Analyses of rapeseed varieties with regard to quantitative and qualitative traits

Análises de variedades de colza com relação a características quantitativas e qualitativas

Abstract

Rapeseed (Brassica napus L.) is one of the most important oil crops in terms of economics, ecology, and nutrition. For the purpose of selecting the most suitable canola genotypes for quantitative and qualitative traits, an experiment was conducted in Damavand region with the presence of nine genotypes and the examination of nine traits with three replications. The results of analysis of variance showed that the effect of genotype in terms of all traits had significant differences at the level of 0.01 and 0.05. Additionally, the results of the average comparison indicated that Zargol and Hyola 401 genotypes were more favorable than other cultivars in terms of all traits. According to the three analyses related to the examination of the traits, it was concluded that the grain yield trait was positively correlated with the harvest index trait, the biological yield trait was positively correlated with the oil percentage trait, and the leaf width trait was positively correlated with the number of days to 50% flowering. Using principal component analysis (PCA), the first three components explained more than 81 percent of the variance in the data, and the first and second components had positive coefficients for Zargol and Hyola 401 genotypes. On the basis of the graphical analysis, the Zargol and Sunday genotypes were selected as the best genotypes. In comparison with the cluster analysis and heat map drawn on the data, the genotypes were grouped into two main groups based on traits. Accordingly, Zargol genotypes are considered stable genotypes in terms of their traits and are suitable for cultivation and agricultural research.

Keywords:
rapeseed; correlation; graphic analysis; cluster analysis; heat map

Resumo

A colza (Brassica napus L.) é uma das culturas oleaginosas mais importantes em termos de economia, ecologia e nutrição. Com o propósito de selecionar os genótipos de canola mais adequados para características quantitativas e qualitativas, um experimento foi conduzido na região de Damavand com a presença de nove genótipos e o exame de nove características com três repetições. Os resultados da análise de variância mostraram que o efeito do genótipo em relação a todas as características teve diferenças significativas no nível de 0,01 e 0,05. Além disso, os resultados da comparação média indicaram que os genótipos Zargol e Hyola 401 foram mais favoráveis ​​do que outras cultivares em relação a todas as características. De acordo com as três análises relacionadas ao exame das características, concluiu-se que a característica de rendimento de grãos foi positivamente correlacionada com a característica de índice de colheita, a característica de rendimento biológico foi positivamente correlacionada com a característica de porcentagem de óleo e a característica de largura da folha foi positivamente correlacionada com o número de dias para 50% de floração. Usando a análise de componentes principais (PCA), os três primeiros componentes explicaram mais de 81% da variância nos dados, e o primeiro e o segundo componentes tiveram coeficientes positivos para os genótipos Zargol e Hyola 401. Com base na análise gráfica, os genótipos Zargol e Sunday foram selecionados como os melhores genótipos. Em comparação com a análise de cluster e o mapa de calor desenhado nos dados, os genótipos foram agrupados em dois grupos principais com base nas características. Consequentemente, os genótipos Zargol são considerados genótipos estáveis ​​em termos de suas características e são adequados para cultivo e pesquisa agrícola.

Palavras-chave:
colza; correlação; análise gráfica; análise de cluster; mapa de calor

1. Introduction

This plant has a very high economic (Arun and Dalai, 2020), ecological (Zeng et al., 2020) and nutritional (Bocianowski et al., 2020) value. For this reason, it is very useful to develop rapeseed (Brassica napus subsp. napus) cultivation and conduct research in order to supply the oil needed by humans (Azizinia and Mortazavian, 2015). Countries such as Canada, India, France, Austria, United Kingdom, Germany, Poland, Denmark, Slovakia, Czech Republic, United States of America and Russia are reported as the largest rapeseed producers in the world (Vinnichek and Melnik, 2009). Different researchers in order to select the appropriate genotypes in terms of physiological and agricultural traits with biplot multivariate methods (Yan and Tinker, 2006) and using the data obtained from different traits and analyzing them by the method of principal components analysis graphically, genotypes have been investigated (Hongyu et al., 2014; Shojaei et al., 2023a). Various techniques are used in crop performance modeling, including correlation analysis, regression, path analysis, factor analysis, principal component analysis, and cluster analysis (Aytac et al., 2008; Leilah and Al-Khateeb, 2005). Analyzing correlation coefficients is one of the most valuable statistical methods to analyze the results of research projects to obtain high yield and also to examine the direct and indirect contribution of variables related to seed yield (Shojaei et al., 2023b; Belete, 2011; Khan et al., 2006; Sadat et al., 2010). The correlation coefficient expresses the intensity or weakness and the direction of changes of two variables of plant traits, each of which can be influenced by the environment (Malik et al., 2010). Also, analyzing the correlation between different traits with grain yield helps a lot to decide on the relative importance of these traits and their value as selection criteria (Hamzehpour et al., 2018; Mousavi et al, 2023, Illes et al., 2021). Principal components analysis along with factors analysis and clustering also play a significant role in evaluating different traits. The main application of the factor analysis technique is to reduce the number of variables and identify the structure in the relationship between variables (Rameeh, 2015). Also, GT biplot method has been used to evaluate the relationships between traits on rapeseed genotypes (Dehghani et al., 2008). GT biplot is one of the GGE biplot methods that has been used as a valuable tool for studying multi-trait data (Kang et al., 2003). However, even though an applied criterion was able to select for yield and stability, Spearman's rank correlations were weak and variable for all types of indicators, except for simultaneous selection index and grain yield. The ES Alonso, Kamilo and OKapi genotypes were identified as stable and high-yielding genotypes by several stability methods (Motthari et al, 2018). This method is very useful in investigating the relationships between genotypes and traits and it can be used to evaluate and select some desirable traits in order to identify the most suitable genotype (Swelam, 2012; Al-Naggar et al., 2020). In a research that was conducted in order to investigate the phenological and physiological traits effective on increasing the yield of rapeseed, 20 rapeseed cultivars were compared in terms of traits. The direct and positive effects of harvest index and biological performance, as well as the positive indirect effect of biological performance indicated that these traits are considered reliable components for selecting high-yield cultivars (Frooghi et al., 2017). In another experiment that was conducted in order to determine the most important phenological traits that are effective in increasing grain yield in rapeseed, the correlation of traits such as biological yield, harvest index, number of pods per plant and number of seeds per pod with grain yield were very significant. Many other researchers, considering the importance of increasing the performance of rapeseed and also in order to investigate the relationships between traits, compared different cultivars of rapeseed and used GT biplot methods and multivariate graphic methods to investigate genotypes (Shojaei et al., 2023c; Qasemi et al., 2022; Ghasemi et al., 2023; El-Nenny et al., 2022). I would like to present the results of this research in the form of a study aimed at selecting the best canola cultivars for the study and grouping them in relation to the physiological and quantitative traits tested as well as determining the relationship amongst these traits and to evaluate what each trait contributes to the grain yield.

2. Materials and Methods

In this research, 9 rapeseed genotypes were tested in terms of 9 physiological and agronomic traits in the form of a randomized complete block design (RCBD) in three replications. The experiment was conducted in Damavand region. Figure 1 shows the geographical and climatic characteristics of Damavand region. Table 1 also shows the soil characteristics of the cultivated area. In this study, each experimental plot had 4 rows with a length of 2 meters, and the distance between each row was 50 cm. The seeds were planted with a distance of 10 cm from each other and during the cropping season, all the operations were carried out precisely. In order to eliminate the marginal effects, sampling was done from the middle two rows. Table 2 shows the code and characteristics of the studied canola genotypes. The evaluated traits include grain yield (GYD), plant height (Ph), pod length (PL), number of days to 50% flowering (Day50%), number of days to physiological maturity (DayM), leaf length (LL), weight thousand seeds (WTS), harvest index (HI), biological yield (BW), oil percentage (Oil), leaf width (LW) and grain weight per plant (GPP). Plant height, number of days to 50% flowering, number of days to physiological maturity, leaf length and leaf width were measured before harvest and the rest of the traits were measured after harvest. Sampling was done by selecting 5 random plants to obtain data. In order to analyze data from analysis of variance, mean comparison by Duncan's method, correlation coefficients, principal components analysis, factor analysis, cluster analysis and graphic analysis (including polygon graph, stability diagram of genotypes in terms of traits, the diagram of genotype selection in terms of ideal genotype and the grouping of genotypes in terms of traits) were used.

Figure 1
Geographical and climatic characteristics of the experimental region.

Table 1
‏Soil characteristics of the Damavand region where the experiment are carried out .

EC(ds/m)	Acidity	Lime (%)	Organic carbon (%)	Organic materials (%)	Clay (%)	Silt (%)	Sand (%)	ph
0.27	7.33	7	40	42	26	22	14	6.2

Table 2
Code, name and origin of rapeseed genotypes investigated in the experiment.

Genotype Code.	Genotype Name	Origin	Genotype Code.	Genotype Name	Origin
G1	Mehr 2	Iran	G6	Option	Germany
G2	Tassilo	Iran	G7	Sunday	Denmark
G3	Zargol	Iran	G8	Opera	Sweden
G4	Hyola 401	Canada	G9	Okapi	France
G5	Hyola 308	Canada

In order to investigate the interaction between genotype and trait (GT):

The following method was used: (Yan & Rajcan, 2002)

$\frac{{\text{T}}_{\text{ij}}-{\overline{T}}_j}{{\text{s}}_{\text{j}}}={\text{λ}}_1{\zeta }_{i1}{\tau }_{j1}+{\text{λ}}_2{\zeta }_{i2}{\tau }_{j2}+{\text{ε}}_{\text{ij}}$ (1)

Where T_ij is the average value of genotype i for trait j, ${\overline{T}}_j$ is the average value of trait j over genotype, s_j is the standard deviation of trait j among the genotype averages; ${\zeta }_{i1}$ and ${\zeta }_{i2}$ are the PC1 and PC2 scores, respectively, for genotype i; ${\tau }_{j1}$ and ${\tau }_{j2}$ are the PC1 and PC2 scores, respectively, for trait j; and ${\epsilon }_{\text{ij}}$ is the residual of the model associated with the genotype i in trait j.

Also, due to the difference between the units in the evaluated traits, standardization was used to eliminate the units, which was done by the following Formula 2:

$Z=\frac{X-\mu }{\sigma }$ (2)

In this equation, Z: standard score, X: initial data of the trait, μ: mean of the trait, σ: standard deviation of the trait.

Excel, SASv9.1, Genstat.v12.1 and Excelstat2019 software were used to analyze the data obtained from the experiment.

3. Results and Discussion

3.1. Analysis of variance and mean comparison

The results of the analysis of variance obtained from the experimental data indicated that the block effect had a significant difference in terms of the number of days to 50% of flowering, leaf width, grain yield and grain yield per plant at the probability level of 0.01 and 0.05%. The effect of genotype also showed a significant difference in terms of all traits. The highest percentage of the coefficient of variation was related to the thousand seed weight trait (22.3) and the lowest was related to the number of days to 50% flowering trait (2.6). Also, in the examination of the R-Square value, the highest value of this index was related to the grain yield trait (0.98) and the lowest value was related to the pod length trait (0.82) (Table 3). The results of mean comparison using Duncans method showed that in plant height trait, genotype G4 has a much higher preference than other genotypes and cultivar G1 had the least superiority in terms of this trait. In terms of the number of days until 50% flowering, G3 and G2 genotypes were the best genotypes and G7 was the least desirable. G9 and G4 genotypes were selected as the best and least desirable genotypes in terms of the number of days to physiological maturity. In terms of leaf length trait, G4 genotype was identified as the most desirable cultivar and G9 genotype as the least desirable cultivar. In terms of leaf width, cultivar G3 was selected as the best genotype and cultivar G8 as the least desirable genotype. Genotypes G8 and G3 were selected as the best and least suitable genotypes, respectively, in terms of pod length. In terms of the weight of 1000 seeds, G9 genotype was chosen as the most desirable variety and G2 genotype was selected as the least desirable genotype. Genotypes G2 and G9 were selected as the best and worst genotypes in terms of grain yield traits, respectively. In terms of grain yield in each plant, genotype G1 was identified as the most suitable genotype and genotype G9 as the least suitable genotype. G2 and G5 genotypes were selected as the most desirable and least desirable cultivars, respectively, in the examination of biological performance traits. In terms of harvest index trait, G4 genotype was identified as the most appropriate genotype and G9 and G8 genotypes were identified as the worst genotypes. Genotypes G7, G8, G9 and G5 were selected as the most favorable cultivars and G2 genotype as the least favorable cultivar in terms of oil percentage trait. With the general examination of the evaluated cultivars, G3 and G4 genotypes can be selected as suitable and desirable genotypes and G9 and G6 genotypes as unfavorable genotypes in terms of all traits (Table 4).

Table 3
Analysis of variance for traits of 9 Rapeseed genotypes.

S.O.V	Df	Ph	Day50%	DayM	LL	LW	LP	WTS	GYD	GPP	BW	HI	Oil
Block	2	13.4ns	81.1**	1.37ns	34.77ns	9.1**	7.16ns	0.25ns	131.7**	248.1**	416.2ns	47.3ns	12.6ns
Genotype	8	74.2*	52.7**	443.7**	27.5*	12.09**	1.9*	0.37**	819048.3**	610.7**	2616.2**	765.3**	28.6**
Error	16	222.9	22.6	0.95	10.8	1.7	2.5	0.7	11.8	12.22	94.2	35.07	12.8
CV%	-	15.5	2.6	14.5	21.7	19.8	18.3	22.3	10.3	10.26	12.7	12.6	16.5
R-square	-	0.84	0.91	0.95	0.89	0.9	0.82	0.83	0.98	0.96	0.93	0.91	0.92

• ^*
  , **, and ns: significant at 5%, 1%, and not significant.

Table 4
Mean comparison between traits in 9 tested Rapeseed genotypes.

Genotypes	Rank	Ph	Day50%	DayM	LL	LW	LP	WTS	GYD	GPP	BW	HI	Oil
Mehr 2	3	86.1c	183.6abc	205e	17.4ab	8ab	8.1bc	3.4bc	3449.1bc	55.8a	118.03a	46.9bc	47.3ab
Tassilo	4	91.3bc	185.3ab	204.6ef	16.9ab	8.1ab	8.6b	3.1bc	4037.5a	50.6ab	112.1a	45.6c	44.4c
Zargol	1	98.3bc	187.3a	205ef	16.2abc	8.6a	7.6c	3.8ab	3537bc	47.4b	88.8b	55.1bc	46.09b
Hyola 401	2	102.9a	185.6ab	203.6f	18.5a	7.1bc	8.7b	3.5bc	3971.6b	34.1c	46.3c	73.4a	46.09b
Hyola 308	6	94.9bc	182.6abc	210d	17.2ab	8.1ab	8.7b	3.8bc	3264.9bcd	23.7d	49.1c	48.3bc	48.5a
Option	8	97.2bc	185ab	213c	15.9abc	7.78bc	8.1bc	3.6bc	3197.9cd	26.6d	54.4c	48.7bc	46.5ab
Sunday	5	98.7bc	175.3c	206e	11.3bc	3.9cd	8.5b	4.03ab	3289bcd	28.5cd	49.2c	57.8b	48.5a
Opera	7	97.2bc	177bc	230.6b	12.2abc	3.7d	10.5a	4.1ab	2748.8de	27.6d	103.3ab	26.6d	48.5a
Okapi	9	99.5b	179.3abc	236.3a	10.2c	4.41c	8.3b	4.2a	2406.4e	12.1e	61.7c	19.8d	48.5a

• (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage).

3.2. Correlation analysis

Based on the correlation coefficients of the traits, the trait of plant height had a positive and significant correlation with the trait of 1000- seed weight and had a negative and significant correlation with the trait of grain yield in each plant. The trait number of days to 50% flowering also had a positive and significant correlation with the traits of leaf width and grain yield, and had a negative and significant correlation with the traits of biological yield and oil percentage. The trait of days to physiological maturity has a positive and significant correlation with the trait of 1000-seed weight, and with the traits of leaf width, grain yield, grain yield per plant and harvest index, it has a negative and significant correlation, and the trait of leaf length has a positive and significant correlation with the trait of 1000-seed weight. The trait of leaf width had a positive and significant correlation with the traits of grain yield, grain yield per plant and 1000-seed weight, and had a negative and significant correlation with the traits of biological weight and oil percentage. The trait of pod length has a positive correlation with the trait of grain yield in each plant, and the trait of oil percentage has a negative and significant correlation. Grain yield in each plant and harvest index had a positive correlation and a negative and significant correlation with the trait of oil percentage. The trait of grain yield in each plant had a negative and significant correlation with the trait of oil percentage, and the trait of biological yield had a positive and significant correlation with the trait of oil percentage (Table 5). In the examination of the correlation map (Figure 2), which was drawn to check the intensity of the correlation between traits, red color showed the maximum correlation, brown color showed moderate correlation, blue color showed low correlation and white color showed no correlation. Based on this graph, traits of 1000- seed weight with oil percentage, traits of grain yield with harvest index, traits of leaf length with grain yield, traits of number of days to 50% of flowering and leaf length with leaf width and leaf length showed the highest intensity of correlation. The correlation map has been used in various experiments on plants to investigate the relationships between traits, which can be researched in plants such as rice (Semeskandi et al., 2023), corn (Ahmady and Mazloom, 2023) and rapeseed (Sadeghizadeh, 2023). In order to check the correlation between traits graphically, the correlation diagram between traits was used (Figure 3). In this biplot diagram, the cosine of the angle between trait vectors indicates the intensity of correlation between traits. If the angle between the vectors is less than 90 degrees, the correlation between the vectors is equal to +1, if the angle between the vectors of the traits is 90 degrees, the correlation between the vectors of the traits is equal to zero, and if the angle between the vectors is 180 degrees, it indicates a correlation of -1. (Yan et al., 2007). Based on the correlation diagram, grain yield traits, harvest index, oil percentage, number of days to 50% flowering, grain yield per plant and leaf width together, traits of leaf length, 1000- seed weight, peduncle length and biological yield together and traits of number of days to Physiological maturity, biological yield and pod length have a positive correlation with regard to the angle of less than 90 degrees between their vectors, the trait of harvest index with the number of days until physiological maturity according to the angle of 180 degrees between their vectors has a negative correlation, as well as the traits of days until physiological maturity, There was no correlation with the length of the leaf according to the 90 degree angle between the two vectors.

Table 5
Correlation coefficients between the traits evaluated in the experiment [[Q38: Q38]].

Traits	Ph	Day50%	DayM	LL	LW	LP	WTS	GYD	GPP	BW	HI
Day50%	-0.17
DayM	0.26	-0.5
LL	0.34	-0.2	0.08
LW	-0.41	0.92**	-0.66*	-0.3
LP	-0.33	0.31	-0.2	-0.3	0.26
WTS	0.61*	-0.5	0.69*	0.54*	-0.6*	-0.4
GYD	-0.19	0.64*	-0.8*	-0.28	0.65*	0.4	-0.7*
GPP	-0.6*	0.54	-0.6*	-0.35	0.6*	0.7*	-0.8**	0.7**
BW	0.18	-0.7*	0.28	0.39	-0.6*	-0.26	0.37	-0.44	-0.4
HI	0.17	0.45	-0.8*	0.1	0.46	-0.01	-0.4	0.8**	0.39	-0.2
Oil	0.16	-0.7*	0.56	0.26	-0.6*	-0.6*	0.55	-0.79*	-0.65*	0.7*	0.4

• ^*
  and **: significant at 5% and 1%.

Figure 2
Correlation map to check the intensity of correlation between the traits evaluated in the experiment. (Red color: maximum correlation, brown color: medium correlation, blue color: low correlation, and white color: no correlation). (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage).

Figure 3
Correlation diagram between the traits evaluated in the experiment. (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage).

According to the analysis of three analyzes carried out in order to find the relationship between the traits and evaluate the correlation between them, it can be concluded that the trait of grain yield with the trait of harvest index, the trait of biological yield with the trait of oil percentage, the trait of leaf width with the trait of number of days to 50% flowering had a high positive correlation and also the harvest index with the number of days to physiological maturity showed a negative correlation in terms of these three analyses. In a research that was conducted in order to investigate rapeseed genotypes in different environments, the correlation between the grain yield trait and the harvest index trait as well as other morphological traits were reported and the correlation diagram was used to evaluate the relationship between the traits (Shojaei et al., 2023c).

3.3. Principal component analysis (PCA)

In the principal component analysis (PCA) obtained from the test data analysis and according to the Eigenvalue obtained from the data analysis, the first three components accounted for more than 81% of the data variance (Table 5). In the examination of the Eigenvalue chart obtained from the evaluation of traits, the red axis of the chart represents the Cumulative and the blue bar charts represent the Eigenvalue, based on which the first three components covered more than 81% of the variance of the total data. 54.6% was related to the first component, 15.76% was related to the second main component, and 11.1% was related to the third component (Figure 4a). Based on the first component, which accounted for more than 54% of the variance of the data, the traits number of days to 50% of flowering, leaf width, peduncle length, grain yield, grain yield per plant and harvest index have a positive effect on this. The components were the most positive effect on this component related to the traits of grain yield (0.35) and grain yield per plant (0.34). In the second component, which accounted for more than 15% of the data variance, the traits of plant height, number of days to 50% flowering, leaf length, leaf width, 1000- seed weight, grain yield, biological yield and harvest index have a positive effect on this component, the most positive effect was related to plant height (0.46) and leaf length (0.41). The third component accounted for 11% of the variance of the total data, and the traits of leaf length, pod length, grain yield, grain yield per plant, biological yield, and oil percentage had a positive effect on this component. Biological yield trait (0.55) had the most positive effect (Table 5,6). Based on the graph obtained, the distribution of genotypes was analyzed in terms of the first and second principal components in terms of positive and negative coefficients. Based on this diagram, G3 and G4 genotypes have a positive effect on the first and second main components, G1 and G2 genotypes have positive coefficients on the first main component and negative coefficients on the second main component, G6, G5 and G7 genotypes have a negative effect. In terms of the first component and positive coefficients in the second component, G8 and G9 genotypes had a negative effect in terms of both components (Figure 4b) Based on the graph drawn on the traits in terms of the first and second main components, grain yield traits, number of days to 50% flowering, leaf width and harvest index have a positive effect on both components, traits grain yield per plant and pod length has a positive effect on the first main component and a negative effect on the second component, plant height, leaf length, biological yield, oil percentage and 1000- seed weight have negative coefficients on the first main component and have positive coefficients on the second main component and the number of days until the physiological maturity had a negative effect on both components (Figure 4c).

Figure 4
Eigenvalue diagram, distribution of genotypes and traits evaluated in the experiment based on the first and second principal components, a: Eigenvalue diagram, b: distribution diagram of genotypes, c: distribution diagram of traits. (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage). (G1: Mehr2, G2: Tassilo, G3: Zargol, G4: Hyola 401, G5:Hyola 308, G6: Option, G7: Sunday, G8: Opera, G9: Okapi).

Table 6
Principal component analysis (PCA) based on the traits evaluated in the experiment.

Trait	PC1	PC2	PC3	PC4	PC5	PC6	PC7	PC8	PC9	PC10	PC11	PC12
Ph	-0.17	0.46	-0.32	0.33	-0.45	0.18	0.32	0.12	-0.01	0.22	-0.37	0.10
Day50%	0.32	0.08	-0.42	-0.11	0.17	0.32	0.21	0.18	0.07	-0.70	0.00	0.00
DayM	-0.32	-0.31	-0.30	0.09	-0.03	0.03	0.03	0.53	0.22	0.17	0.51	0.28
LL	-0.17	0.41	0.11	0.29	0.75	-0.12	-0.01	0.31	0.09	0.08	-0.10	-0.13
LW	0.33	0.04	-0.19	-0.30	0.29	0.52	0.03	-0.15	0.05	0.61	0.04	0.07
LP	0.22	-0.33	0.10	0.70	0.00	0.26	0.20	-0.16	0.03	0.05	0.20	-0.42
WTS	-0.34	0.14	-0.29	0.18	0.17	-0.01	-0.07	-0.71	0.25	-0.14	0.21	0.28
GYD	0.35	0.22	0.14	0.08	-0.23	-0.03	-0.39	0.10	0.76	0.00	0.00	0.00
GPP	0.34	-0.18	0.25	0.16	0.13	-0.25	0.46	-0.03	0.10	0.02	-0.17	0.66
BW	-0.26	0.08	0.55	0.07	-0.04	0.65	-0.16	0.07	-0.09	-0.20	0.03	0.33
HI	0.23	0.55	0.18	-0.05	-0.15	-0.13	0.23	-0.02	-0.22	0.01	0.69	0.00
Oil	-0.33	-0.03	0.26	-0.37	-0.01	0.06	0.60	-0.07	0.47	-0.02	0.00	-0.31
Eigenvalue	6.56	1.89	1.33	0.92	0.7	0.299	0.16	0.11	0.0001	0.00001	0.000001	0.0000001
Proportion	0.54	0.15	0.11	0.07	0.05	0.02	0.01	0.009	0.0009	0.00009	0.00001	0.000001
Cumulative	0.54	0.7	0.81	0.89	0.95	0.97	0.99	0.999	0.9999	0.9999	0.99999	1

• (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage).

3.4. Graphical analysis

The polygon diagram was used to examine the genotypes and select the most appropriate genotype in terms of traits. Based on this diagram, the genotypes that have the greatest distance from the origin of the diagram and are connected by important lines are identified as desirable genotypes compared to other genotypes. Considering that these genotypes create a polygon, other genotypes are placed inside this polygon (Yan and Tinker, 2006). In each section, the genotypes with specific traits are separated by lines. According to the polygon graph, the first component accounted for 47.43%, the second component for 24.92%, and more than 72% of the total data variance. Based on this diagram, genotypes G3, G2, G1, G9, G6 and G7 were identified as desirable cultivars according to the distance from the origin of the diagram. In each section specifically with traits, G3 genotype in terms of harvest index traits, grain yield and leaf width, G4 genotype in terms of number of days to 50% flowering and grain yield per plant and genotypes G7 and G6 had more performance than other genotypes in terms of leaf length trait (Figure 5a). In examining the stability diagram of the genotypes in terms of the traits evaluated in the experiment, the axis that is marked with an arrow (horizontal axis or AEC - abscissa) and is past the mean of the traits (circle), determines the performance of the traits in the cultivars. So that any genotype that is on the right side of this axis will have the performance of desirable traits. The average of traits is obtained by calculating the average value of PC1 and PC2 for all traits, and the AEC-Ordinate axis confirms the stability or instability of rapeseed cultivars. Genotypes that have a greater distance from the origin of this axis are less stable, in other words, cultivars that have a greater vertical distance from the AEC-abscissa axis are less stable (Yan et al., 2006). In terms of examining the stability diagram and the performance of genotypes in terms of all studied traits, G7, G6, G3, G4 and G5 genotypes had good performance and G1 and G8 genotypes were selected as undesirable genotypes. In terms of the stability of genotypes, in terms of all traits, G7, G6, G3, G4, G5 and G8 genotypes can be selected as stable genotypes. Among these genotypes, G5 was selected as the most stable genotype. Considering that G7, G3 and G6 genotypes were identified as genotypes with good performance and high stability in terms of traits, these genotypes can be selected as desirable cultivars (Figure 5b). The ranking diagram of genotypes based on the ideal genotype is drawn in such a way that a line is connected from the origin of the coordinates of the diagram to the average point and continues to both sides. The best genotype is the genotype that tends to the positive end and its vertical distance is less than this line. In this type of diagram, the best point is the center of concentric circles, which is indicated by an arrow, and other genotypes are ranked based on this point. In examining the ranking of genotypes based on the ideal genotype, G5, G7, G3, G6 and G4 genotypes were identified as undesirable genotypes (Figure 5c). The order of genotypes from favorable to unfavorable is as follows:

Figure 5
Graphic analysis based on the first and second main components, a: the polygon diagram of the distribution of genotypes in terms of traits, b: the stability diagram of genotypes in terms of the evaluated traits, c: the diagram of selecting the best genotype based on the ideal genotype. (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage). (G1: Mehr2, G2: Tassilo, G3: Zargol, G4: Hyola 401, G5:Hyola 308, G6: Option, G7: Sunday, G8: Opera, G9: Okapi).

$$\begin{array}{l}G5>G7>G3>G6>G4>\\ G2>G8>G9>G1\end{array}$$ (3)

3.5. Cluster analysis

Based on the cluster analysis performed on the analyzed rapeseed genotypes in terms of all the traits evaluated in the experiment, the genotypes were grouped into two main groups. The first group was divided into two subgroups, the first subgroup included G1, G4, G5 and G6 genotypes, the second subgroup included G2 and G3 genotypes. The second main group was also grouped into two subgroups, the first subgroup included G7 and G8 genotypes and the second subgroup included G9 genotype. In general, based on cluster analysis, genotypes were grouped into two main groups and four subgroups.

Based on the heat map drawn in terms of the investigated traits, the genotypes were grouped into two main groups. The first group consisted of two subgroups, the first subgroup included G9 genotype, the second subgroup included G7 and G8. The second group also included two subgroups. The first subgroup included G1, G2 and G3 genotypes and the second subgroup included G4, G5 and G6 genotypes. Genotypes G8, G7 and G9 showed a similar reaction in terms of harvest index traits, seed yield and leaf width, and G1, G2, G3 and G4 genotypes also in terms of 1000- seed weight, days to physiological maturity, leaf length, oil percentage and biological yield showed a similar response(Figure 6).. Using the mean seed yields, a cluster analysis revealed only one major plant group. Genetypes 2, 3, 8 and 9 were not grouped in the first year. Only genotype 8 failed to classify compared to the rest of the Canadian genotypes. The three European genotypes (2, 3 and 9) were not grouped. In the second year, genotypes 10 and 6 were not included in the major cluster (Kamundia and Mahasi, 2007). Using cluster analysis, four genotype groups were identified, and the genotype with the highest seed yield and the highest mean pod value per plant was determined to be the genotype with the highest seed yield (Rameeh, 2016)

Figure 6
Cluster analysis and heat map drawn based on the attributes evaluated in the experiment, a: cluster analysis, b: heat map. (Ph:Plant height, Day50%: Number of days to 50% flowering, DayM: Number of days to physiological maturity, LL: Leaf length, LW: Leaf width, LP: Pod length, WTS: Weight 1000-seeds, GYD: Grain yield, GPP: Grain weight per plant, BW: Biological yield, HI: Harvest Index, OIL: Oil percentage). (G1: Mehr2, G2: Tassilo, G3: Zargol, G4: Hyola 401, G5:Hyola 308, G6: Option, G7: Sunday, G8: Opera, G9: Okapi).

References

• AHMADY, G. and MAZLOOM, P., 2023. Selection of weeding method and plant density in Maize Zea mays L. based on morphological and physiological characteristics. Asian Journal of Research in Agriculture and Forestry, vol. 9, no. 4, pp. 19-28. http://doi.org/10.9734/ajraf/2023/v9i4231
  » http://doi.org/10.9734/ajraf/2023/v9i4231
• AL-NAGGAR, A.M.M., SHAFIK, M.M. and MUSA, R.Y.M., 2020. Genetic diversity based on morphological traits of 19 maize genotypes using principal component analysis and GT biplot. Annual Research & Review in Biology, vol. 35, no. 2, pp. 68-85. http://doi.org/10.9734/arrb/2020/v35i230191
  » http://doi.org/10.9734/arrb/2020/v35i230191
• ARUN, N. and DALAI, A.K., 2020. Environmental and socioeconomic impact assessment of biofuels from lignocellulosic biomass. In: A. YOUSUF, D. PIROZZI and F. SANNINO, eds. Lignocellulosic biomass to liquid biofuels, pp. 283-299. Cambridge: Academic Press. http://doi.org/10.1016/B978-0-12-815936-1.00009-5
  » http://doi.org/10.1016/B978-0-12-815936-1.00009-5
• AYTAÇ, Z., KINACI, G. and KINACI, E., 2008. Genetic variation, heritability and path analysis of summer rapeseed cultivars. Journal of Applied Biological Sciences, vol. 2, no. 3, pp. 35-39.
• AZIZINIA, S. and MORTAZAVIAN, M.M., 2015. A Yield stability survey in winter type canola using univariate methods and genotypic distribution pattern. Isfahan University of Technology-Journal of Crop Production and Processing, vol. 5, no. 15, pp. 57-68. http://doi.org/10.18869/acadpub.jcpp.5.15.57
  » http://doi.org/10.18869/acadpub.jcpp.5.15.57
• BELETE, Y.S., 2011. Genetic variability, correlation and path analysis studies in Ethiopian mustard Brassica carinata A. Brun genotypes. International Journal of Plant Breeding and Genetics, vol. 5, no. 4, pp. 328-338. http://doi.org/10.3923/ijpbg.2011.328.338
  » http://doi.org/10.3923/ijpbg.2011.328.338
• BOCIANOWSKI, J., LIERSCH, A. and NOWOSAD, K., 2020. Genotype by environment interaction for alkenyl glucosinolates content in winter oilseed rape Brassica napus L. using additive main effects and multiplicative interaction model. Current Plant Biology, vol. 21, pp. 100137. http://doi.org/10.1016/j.cpb.2020.100137
  » http://doi.org/10.1016/j.cpb.2020.100137
• DEHGHANI, H., OMIDI, H. and SABAGHNIA, N., 2008. Graphic analysis of trait relations of rapeseed using the biplot method. Agronomy Journal, vol. 100, no. 5, pp. 1443-1449. http://doi.org/10.2134/agronj2007.0275
  » http://doi.org/10.2134/agronj2007.0275
• EL-NENNY, E.M.M., IBRAHIM, H.E., SHAWKY, A.M. and ABD EL-RAHMAN, R.H., 2022. Graphic analysis of trait relations of canola genotypes using the biplot method. Egyptian Journal of Applied Sciences, vol. 37, no. 9-10, pp. 17-29.‏
• FROOGHI, A., BIYABANI, A., RAHEMI KARIZAKI, A. and RASSAM, G., 2017. Relationships of phenology and physiological traits with the yield of Rapeseed Brassica napus L. in Northern Khorasan. Journal of Crop Ecophysiology, vol. 40, no. 10, pp. 1007-1024.‏
• GHASEMI, A., QAFFARIPOUR, Z., TOURANI, M., SALEKI, K., RAHMANI-KUKIA, N., KHATAMI, S.H. and TAHERI-ANGANEH, M., 2023. The relationship between long non-coding RNAs and Wnt/β-catenin signaling pathway in the pathogenesis of Alzheimer’s disease. Experimental Neurology, vol. 366, pp. 114434. http://doi.org/10.1016/j.expneurol.2023.114434 PMid:37201745.
  » http://doi.org/10.1016/j.expneurol.2023.114434
• HAMZEHPOUR, G., TOBEH, A. and SHEIKHZADEH, P., 2018. Study of correlations and regression analysis between quantitative and qualitative characteristics in different cultivars of rapeseed Brassica napus L. in different planting patterns.‏ Journal of Plant Ecophysiology, vol. 31, no. 9,
• HONGYU, K., GARCÍA-PEÑA, M., DE ARAÚJO, L.B. and DOS SANTOS DIAS, C.T., 2014. Statistical analysis of yield trials by AMMI analysis of genotype× environment interaction. Biometrical Letters, vol. 51, no. 2, pp. 89-102. http://doi.org/10.2478/bile-2014-0007
  » http://doi.org/10.2478/bile-2014-0007
• ILLÉS, Á., BOJTOR, C., MOUSAVI, S.M.N., SZÉLES, A., TÓTH, B., SZABÓ, A. and NAGY, J., 2021. Evaluation of complete fertilizer in the aspect of the antioxidant enzyme system of maize hybrids. Agronomy (Basel), vol. 11, no. 11, pp. 2129. http://doi.org/10.3390/agronomy11112129
  » http://doi.org/10.3390/agronomy11112129
• KAMUNDIA, J.W. and MAHASI, M.J., 2007. Cluster analysis in rapeseed (Brassica napus L.). International Scholars Journals, vol. 10, no. 1, pp. 001-003.
• KANG, B., RIEDEL, C.H. and TAN, C.A., 2003. Free vibration analysis of planar curved beams by wave propagation. Journal of Sound and Vibration, vol. 260, no. 1, pp. 19-44. http://doi.org/10.1016/S0022-460X(02)00898-2
  » http://doi.org/10.1016/S0022-460X(02)00898-2
• KHAN, F.A., SAJID ALI, S.A., AMIR SHAKEEL, A.S., ASIF SAEED, A.S. and GHULAM ABBAS, G.A., 2006. Correlation analysis of some quantitative characters in Brassica napus L. Journal of Agricultural Research, vol. 44, no. 1, pp. 1-8.‏
• LEILAH, A.A. and AL-KHATEEB, S.A., 2005. Statistical analysis of wheat yield under drought conditions. Journal of Arid Environments, vol. 61, no. 3, pp. 483-496. http://doi.org/10.1016/j.jaridenv.2004.10.011
  » http://doi.org/10.1016/j.jaridenv.2004.10.011
• MALIK, S.R., AHMAD BAKHSH, A.B., ASIF, M.A., UMER IQBAL, U.I. and IQBAL, S.M., 2010. Assessment of genetic variability and interrelationship among some agronomic traits in chickpea.‏ International Journal of Agriculture and Biology, vol. 12, no. 1, pp. 81-85.
• MOTTHARI, A.R., HERVAN, E.M., ALIZADEH, B. and KHOSROSHALI, M., 2018. Interaction effect of genotype× environment for seed yield of winter hybrids and open pollinated oilseed rape (Brassica napus L.) genotypes. Iranian Journal of Crop Sciences, vol. 20, no. 3, pp. 251-237.
• MOUSAVI, S.M.N., ILLÉS, A., SZABÓ, A., SHOJAEI, S.H., DEMETER, C., BAKOS, Z., VAD, A., SZÉLES, A., NAGY, J. and BOJTOR, C., 2023. Stability yield indices on different sweet corn hybrids based on AMMI analysis. Brazilian Journal of Biology = Revista Brasileira de Biologia, vol. 84, pp. e270680. http://doi.org/10.1590/1519-6984.270680 PMid:36921158.
  » http://doi.org/10.1590/1519-6984.270680
• QASEMI, S.H., MOSTAFAVI, K., KHOSROSHAHLI, M., BIHAMTA, M.R. and RAMSHINI, H., 2022. Genotype and environment interaction and stability of grain yield and oil content of rapeseed cultivars. Food Science & Nutrition, vol. 10, no. 12, pp. 4308-4318. http://doi.org/10.1002/fsn3.3023 PMid:36514761.
  » http://doi.org/10.1002/fsn3.3023
• RAMEEH, V., 2015. Heritability, genetic variability and correlation analysis of some important agronomic traits in rapeseed advanced lines. Cercetări Agronomice în Moldova, vol. XLVIII, no. 4, pp. 71-80. http://doi.org/10.1515/cerce-2015-0054
  » http://doi.org/10.1515/cerce-2015-0054
• RAMEEH, V., 2016. Multivariate analysis of some important quantitative traits in rapeseed (Brassica napus) advanced lines. Journal of Oilseed Brassica, vol. 1, no. 2, pp. 162-169.
• SADAT, H.A., NEMATZADEH, G.A., JELODAR, N.B. and CHAPI, O.G., 2010. Genetic evaluation of yield and yield components at advanced generations in rapeseed Brassica napus L. African Journal of Agricultural Research, vol. 5, no. 15, pp. 1958-1964.
• SADEGHIZADEH, M., 2023. Selection of suitable genotypes based on physiological and morphological traits in rapeseed Brassica napus L. Asian Journal of Research in Agriculture and Forestry, vol. 9, no. 3, pp. 172-180. http://doi.org/10.9734/ajraf/2023/v9i3221
  » http://doi.org/10.9734/ajraf/2023/v9i3221
• SEMESKANDI, M.N., MAZLOOM, P., ARABZADEH, B., MOGHADAM, M.N. and AHMADI, T., 2023. Application of correlation coefficients and principal components analysis in stability of quantitative and qualitative traits on rice improvement cultivation. Brazilian Journal of Biology = Revista Brasileira de Biologia, vol. 84, pp. e268981. http://doi.org/10.1590/1519-6984.268981 PMid:36921192.
  » http://doi.org/10.1590/1519-6984.268981
• SHOJAEI, S. H., MOSTAFAVI, K., GHASEMI, S. H., BIHAMTA, M. R., ILLÉS, Á., BOJTOR, C., NAGY, J., HARSÁNYI, E., VAD, A., SZÉLES, A and MOUSAVI, S.M.N., 2023a. Sustainability on different canola (Brassica napus L.) cultivars by GGE biplot graphical technique in multi-environment. Sustainability, vol. 15, no. 11, pp. 8945.‏ https://doi.org/10.3390/su15118945
  » https://doi.org/10.3390/su15118945
• ‏SHOJAEI, S.H., MOSTAFAVI, K., BIHAMTA, M., OMRANI, A., BOJTOR, C., ILLES, A., SZABO, A., VAD, A., NAGY, J., HARSÁNYI, E. and MOUSAVI, S.M.N., 2023b. Selection of maize hybrids based on genotype× yield× trait GYT in different environments. Brazilian Journal of Biology = Revista Brasileira de Biologia, vol. 84, pp. e272093. http://doi.org/10.1590/1519-6984.272093. PMid:37283408. ‏
• SHOJAEI, S.H., MOSTAFAVI, K., KHODARAHMI, M. and ZABET, M., 2023c. Response study of canola Brassica napus L. cultivars to multi-environments using genotype plus genotype environment interaction GGE biplot method in Iran. African Journal of Biotechnology, vol. 10, no. 53, pp. 10877-10881. http://doi.org/10.5897/AJB11.1913
  » http://doi.org/10.5897/AJB11.1913
• ‏SWELAM, A.A., 2012. Phenotypic stability, biplot analysis and interrelationship among agronomic characters for some bread wheat genotypes. Egyptian Journal of Plant Breeding, vol. 16, no. 4, pp. 147-165.‏
• VINNICHEK, L. and MELNIK, N., 2009. Development of agricultural production on the base of territorial labour division CABI.
• YAN, L., FU, D., LI, C., BLECHL, A., TRANQUILLI, G., BONAFEDE, M.A.R.C.O.S., SANCHEZ, A., VALARIK, M., YASUDA, S. and DUBCOVSKY, J., 2006. The wheat and barley vernalization gene VRN3 is an orthologue of FT. Proceedings of the National Academy of Sciences of the United States of America, vol. 103, no. 51, pp. 19581-19586. http://doi.org/10.1073/pnas.0607142103 PMid:17158798.
  » http://doi.org/10.1073/pnas.0607142103
• YAN, W. and RAJCAN, I., 2002. Biplot analysis of test sites and trait relations of soybean in Ontario. Crop Science, vol. 42, no. 1, pp. 11-20. http://doi.org/10.2135/cropsci2002.1100 PMid:11756248.
  » http://doi.org/10.2135/cropsci2002.1100
• YAN, W. and TINKER, N.A., 2006. Biplot analysis of multi-environment trial data: principles and applications. Canadian Journal of Plant Science, vol. 86, no. 3, pp. 623-645. http://doi.org/10.4141/P05-169
  » http://doi.org/10.4141/P05-169
• YAN, W., KANG, M.S., MA, B., WOODS, S. and CORNELIUS, P.L., 2007. GGE biplot vs. AMMI analysis of genotype‐by‐environment data. Crop Science, vol. 47, no. 2, pp. 643-653. http://doi.org/10.2135/cropsci2006.06.0374
  » http://doi.org/10.2135/cropsci2006.06.0374
• ZENG, X., ZOU, D., WANG, A., ZHOU, Y., LIU, Y., LI, Z., LIU, F., WANG, H., ZENG, Q. and XIAO, Z., 2020. Remediation of cadmium-contaminated soils using Brassica napus: effect of nitrogen fertilizers. Journal of Environmental Management, vol. 255, pp. 109885. http://doi.org/10.1016/j.jenvman.2019.109885 PMid:31765948.
  » http://doi.org/10.1016/j.jenvman.2019.109885

Publication Dates

• Publication in this collection
  14 Oct 2024
• Date of issue
  2024

History

• Received
  28 May 2024
• Accepted
  05 Aug 2024