Evaluation of Performance and Yield Stability Analysis Based on AMMI and GGE-Biplot in Promising Pigeonpea [Cajanus cajan (L.) Millspaugh] Genotypes

Pigeonpea [Cajanus cajan (L.) Millspaugh], a multipurpose food legume occupies a prominent place in Indian rainfed agriculture. It is an integral component of various agro ecologies of the country mainly inter cropped with cereals, pulses, oilseeds and millets. It is the second most important pulse crop next to chickpea, covering an area of around 4.42 m ha (occupying about 14.5% of area under pulses).  India ranked first in area and production which accounts for 36% of the area and 23% of production, respectively of the worlds share. However, country’s productivity at 853 kg/ha is far below the world average productivity of 1023 kg/ha and production of 2.86 MT (contributing to 16% of total pulse production) and in Telangana state, pigeonpea is being cultivated in 2.51 lakh ha. with productivity of 785 kg/ha (Anonymous, 2017). Pigeonpea is an often-cross-pollinated crop (20-70%) with diploid (2n=2X) chromosome number of 22 serving as life line for resource poor farmers and commonly known in India, as redgram or arhar or tur or thogari. Protein calorie malnutrition is a global concern especially in infants, young children and nursing mothers. The current per capita availability of pulses in India is 43.8 g/capital/day as against 52.0 g/capital/day of ICMR recommendation (2015-16). The protein content in pigeonpea is substantially higher (21.7 g for 100 g) compared to major cereals (6.0-15.0 for 100 g). Keeping in view the trends in population growth rate and that several other options besides pulses are now available for meeting protein requirements of the people and also their changing food habits, the pulse requirement in the country is projected at 32 million tons by the year 2030 and 39 million tons by the year 2050. This necessitates an annual growth rate of 2.2% (Singh and Paharaj, 2020).
       
Pigeonpea is affected by various abiotic stresses during its life cycle such as moisture, temperature, photoperiod and mineral stresses. Among stresses, moisture stress is common because pigeonpea is grown as a rainfed crop (Chaudhary et al., 2011). However, shifts in rainfall patterns and seasons due to climatic change require the development of varieties that have wider adaptability. Such new varieties must show high performance for yield and other essential agronomic traits and their superiority should be consistent (stable) over a wide range of environmental conditions (Kumar et al., 2020).
       
The genotypes exhibit different levels of phenotypic expression under different environmental conditions resulting in varied performances, in such case, the interaction of the genotype location is an important factor in plant breeding strategies. Therefore, analysis of any variance combined can measure GEI and identify main effectiveness, however it is not enough to declare the GEI effectiveness. A convenient analytic pattern like the additive main effects and multiplicative interaction (AMMI) can cure both the additive main effect and multiplicative interaction constituent utilize the ANOVA (Analysis of Variance) and IPCA (Interaction Principal Components), respectively (Kilic, 2014; Mohammadi et al., 2018). The GGE biplot exploits the PCA approach to investigate the multi environment data and allow the visual presentation of the relations and this can be defined as genotype × environment (G × E) interaction.
       
The aim of the present study was to (i) analyses the effect of GEI on grain yield of 15 pigeonpea genotypes by two biplot models (AMMI and GGE) (ii) identifying high yielding and stable pigeonpea genotype (s) across locations.

The present experimental material comprised of 15 genotypes of pigeonpea including two checks, Asha (ICPL-87119) and WRG-65. The trials were conducted in randomized block design with three replications at five locations during Kharif, 2018 grown under rainfed conditions (Table 2). The plot size of four rows each with 4 m length was followed with spacing of 120 cm between rows and 20 cm between the plants. Standard package of practices were followed to raise the crops. Genotypes across the locations were sown during 1^st week of July and harvested from each plot separately. Observations were recorded for grain yield on plot basis.
 
Statistical Analysis
 
The performance of pigeonpea genotypes was tested over five locations and was assessed using stability models viz, (1) Additive Main effects and Multiplicative Interaction (AMMI) (Gauch and Zobel, 1996) and (2) GGE Biplot or Site Regression model (Yan and Kang, 2003). These models were used to interpret and visualize the stability and GEI patterns. In the AMMI model, only the GEI term is absorbed in the multiplicative component, whereas in the GGE model, the main effects of genotypes (G) plus the GEI are absorbed into the multiplicative component. The AMMI model (Gauch, 1988) was used in analyzing the stability and interaction for yield traits. The AMMI model is a combination of analysis of variance (ANOVA) and principal component analysis (PCA). The G × E interaction was evaluated with the AMMI model by considering the first two principal components. ANOVA model was used to analyze the trait data with main effects of genotype and environment without the interaction, then, a principal component analysis was integrated using the standardized residuals. These residuals include the experimental error and the effect of the GEI. The analytical model can be written as
 
Where,
Y_ij = Yield of i^th-genotypes in j^th-environment.
μ = Overall mean.
g_i = Effect of the i^th genotype.
e_j = Effect of the j^th environment.
λk = Eigen value of the PCA for axis k.
a_ik and y_jk = Genotype and environment principal components scores for axis k, respectively.
R_ij = Residual term.
       
Environment and genotype PCA scores are expressed as unit vector times the square root of λk.
       
The GGE biplot graphically represents G and GEI effect present in the multi-location trial data using environment centered data. This methodology uses a biplot to show the factors (G and GE) that are important in genotype evaluation and that are also sources of variation in GEI analysis of multilocation trial data (Yan et al., 2000; Yan, 2001). GGE biplots were used to evaluate (1) mega environment analysis (which-won-where pattern), where genotypes can be recommended to specific mega environments. (2) Genotype evaluation, where stable specific genotypes can be recommended across all locations and (3) location evaluation, explains discriminative power of target locations for genotypes under study. The data was subjected to IRRI P.B. tools 1.4 version to get AMMI and GGE biplots.

The pooled analysis of variance for grain yield (Table 3) showed significant differences among genotypes, environments and genotype × environment interaction. The genotype × environment interaction effect was significant emphasizing the influence of environment on grain yield of pigeonpea genotypes under study.  In the current study, as depicted by (Table 1) the mean grain yield of 15 genotypes ranged 1673 kg/ha (G3) to 1078 kg/ha (G11). The genotypes G3 (WRG-327) and G2 (WRG-330) with grain yield of 1673 kg/ha and 1608 kg/ha were two high yielding genotypes, respectively compared to the standard check G4 (Asha) with mean yield of 1478 kg/ha. Among the five environments, the highest mean grain yield was obtained from environment E5 (Jagtial; 1680 kg/ha) and the lowest from E2 (Palem; 1256 kg/ha).
 

 
AMMI stability analysis
 
The AMMI analysis of variance for grain yield (kg ha^-1) of 15 pigeonpea genotypes evaluated across the five environments revealed that the main effects of genotypes (G) and environments (E) accounted for 17.73% and 13.39% of the total sum of squares respectively (Table 3). The G × E interaction also accounted for 55.55% of the total sum of squares indicating that the differences in the response of the genotypes across the environment were substantial and the presence of G × E interaction and it was clearly demonstrated by the AMMI model, when the interaction was partitioned among the first four interaction principal component axis (IPCA) as they were significant in predictive assessment. All the interaction PCA were highly significant capturing 54.4%, 25.6%, 12.4% and 7.6% of the total variation in the G × E interaction sum of squares, respectively. The first two interaction PCA axes jointly accounted for 80.0% of the G × E interaction sum of squares. Thus, the GEI of the 15 pigeonpea genotypes tested in five diverse environments was mostly explained by the first two principal components of genotypes and environments. Previous reports confirmed that in most of the cases the maximum GEI could be explained through using the first two PCAs (Fikere et al., 2014; Biswas et al., 2019). Therefore, IPCA1 and IPCA2 were used for construction of AMMI1 and AMMI2 biplots.
 
Biplot analysis
 
The results of AMMI analysis further enlightened the relative contribution of the first two IPCA axes to the interaction effects by plotting with genotype and environment means as presented in Fig 1 and 2. The mean performance and PCA1 scores for both the varieties and environments used to construct the biplots are presented in Table 2 and 3. In the biplot, environments are designated by the letter ‘E’ followed by numbers 1 to 5 as suffix (Table 2, Fig 1), while genotypes represented by numbers from 1 to 15 (Table 3, Fig 1). The quadrants in the graph represent: (QI & QII) higher mean, (QIII & QIV) lower mean, (QI & QIV) +ve IPCA1 and (QII & QIII) -ve IPCA1 scores (Fig 1). When a variety and environment have the same sign on PCA1 axis, their interaction is positive and if opposite, their interaction is negative. Thus, if a variety has a PCA1 score near to zero, it has small interaction effect and was considered as stable over wide environments. Conversely, varieties with high mean yield and large PCA scores were considered as explicitly adapted to specific environments (Abdi and Williams, 2010; Askari et al., 2017; Mustapha and Bakari, 2014).
 

 

 

 

       
Accordingly, in the present study, the pigeonpea genotypes, G3, G5 and G7 exhibited high yield of positive IPCA 1 score, out of which G3 and G5 had high IPCA 1 score in which G3 is being the overall best genotype. On the other hand, G1, G9 and G4 were high yielding genotypes with negative IPCA 1 scores, While IPCA 1 for G1 and G14 were near to zero score and hence have less interaction with the environments out of which only G1 had above average yield performance.
 
AMMI-2 relationships among genotypes and environments
 
In AMMI 2 biplot (Fig 2). The biplot 2 provides on the G×E interaction only and not like AMMI 1 as the AMMI biplot 1 included main effect also. From AMMI 2 biplot analysis (Fig 2), it was observed that the genotypes with less interaction in both axes are positioned near the origin and vice-versa. Hence, the genotypes nearer to the origin were considered as stable when compared to others. Those genotypes falling apart form the origin those with long spokes were termed as highly interacting genotypes. Hence, environments E1, E2 and E5 exerted strong interaction forces while, the rest two E3 and E4 did less. In the present study, G12, G11, G2, G4, G8, G6, G9 and G13 had more responsive since they were away from the origin whereas the genotypes G14, G1, G5, G10, G3, G7 and G15 were close to origin and hence they were less sensitive to environmental forces. In overall, G14 exhibited very less Genotype × environmental interaction showing high stability with poor yield.
 
GGE biplot analysis
 
GGE biplot of environment-view for yield
 
Environment centered GGE biplot used to estimate the pattern of environments (Fig 3). To compare the relationship between environments, some lines are drawn to connect the test environments to the biplot origin as environment vectors. The angle cosine between two environments is used to extent of the correlation between them (Dehghani et al., 2010).  Environments E5, E3 and E2 are positively correlated (an acute angle). The presence of wide obtuse angle among environments is an indication of high cross over genotype × environment interaction (Yan and Tinker, 2006). In the present study, the environments E2 and E1 are negatively correlated (an obtuse angle).
 

 
GGE biplot of genotype view for grain yield
 
Vector of GGE biplot in the genotype focused scaling also measures their dissimilarity in discriminating the genotypes. Genotypes G1, G2, G3 and G4 showed same group position (Fig 4). Genotypes G11, G12, G13 and G14 were in different group. Likewise, genotypes can be discriminated based on dissimilarity.
 

 
GGE biplot on environment for comparing environments with ideal environment
 
Discriminating ability and representativeness of the testing environments are an important measure in the GGE biplot. The concentric circles in Fig 5 can help us to visualize the length of the environment vectors, which are a measure of the discriminating ability of the environments as well as standard deviation within the respective environments. (Kang-Bo-Shim et al., 2015). The environments E4 and E1 are most discriminating. The average environment which is drawn as small circle at the end of arrow (Fig 6) has the average coordinates of all test environments and average environment axis (AEA) is the line passing through the average environment and the biplot origin. A test environment showing a smaller angle with the AEA is more representative than test environments (Yan and Rajcan, 2002). Accordingly, the environments E4 and E5 are most representative whereas the environments E1 and E3 are least representative. Test environments with both discriminating and representative are good test environments for selecting adaptable genotypes. Discriminating but non representative test environment like E1 is useful for selecting adaptable genotypes.
 

 

 
Biplot of stability and mean performance of genotypes across average environments
 
The line that passes through the biplot origin and the average environment with single arrow is the average environment axis (AEA). Projections of genotype markers to the average environment axis show the mean yield of genotypes (Fig 6). Genotypes are ranked along the ordinate. The genotype G3 was high yielding while G11 was the lowest. The AEA ordinate is the double arrowed line that passes through the biplot origin and is perpendicular to the AEA abscissa. Greater projection onto AEA ordinate regardless of the direction means greater stability. Accordingly, the genotypes G6 and G12 are unstable. The genotypes G14, G1, G13 and G2 with shorter projections are stable over environments.
 
“What-Won-Where” pattern analysis
 
Genotypes having specific adaptative ability for specific environment or group of environments were identified using “What-Won-Where pattern analysis” and “ranking of genotypes in individual environments” using GGE Biplot tools. GGE bi-plot can best identify G × E interaction pattern of data and clearly shows which genotypes perform best in which environments and thus facilitates mega-environment identification than AMMI. Otherwise, both GGE and AMMI models are equivalent as far as their accuracy is concerned. The studied environments were divided into two mega environments i.e., E4 and E1. In mega environment E1, the winning genotype is G6, while the genotype G3 is the winner in mega environment is E4, whereas, E2, E3 and E5 are closely related and fall under the mega environment E4. The polygon view of the GGE bi-plot (shown in Fig 3) indicates the best genotype(s) in each environment. The vertex genotypes (G3, G6, G8, G14, G11 and G12) have the longest vectors, in their respective direction, which is a measure of responsiveness to environments (Fig 7). The vertex genotypes for each sector are the ones that gave the highest yield for the environments that fall within that sector. The genotype with the high yield in E4, E5, E3 and E2 is G3 followed by G2 and G5. In E1 the best genotype was G6. The genotypes G2, G5, G1, G7 and G5 performed better in E4, E5, E3 and E2. The other vertex genotypes G10, G15, G13, G14, G11 and G12 are the poorest in all environments because there is no environment in their sectors.
 

The genotype × environment interaction (GEI) has been an important and challenging issue among plant breeders engaged in performance testing. The GEI reduces association between phenotypic and genotypic values and leads to bias in the estimates of gene effects and combining ability for various characters that are sensitive to environmental fluctuations. Such traits are less amenable to selection. In this context, the present study describes an AMMI based simultaneous selection model based on the yield performance and stability of genotypes in a multi environmental evaluation trial. The model works well in identifying pigeonpea genotypes with higher yield as well as stable performance over locations. The results of the investigation proved that yield performance of pigeonpea genotypes were influenced by genotypes, the environments and GE interaction effect, as well. The mean grain yield value of genotypes averaged over environments indicated that WRG-327 had the highest (1673 kg/ha), followed by WRG-330 (1608 kg/ha). As revealed by AMMI and GGE bi plots, the genotype G3 (WRG-327) identified as most adapted line and stable performer with negligible G×E interaction and high yield (1673 kg/ha), could be used directly as variety. Therefore, WRG-327 could be popularized through large scale demonstrations in farmer’s fields which can improve the pigeonpea productivity and ensure livelihood of the resource poor farming communities.

The authors are highly thankful to Dr. N. Lingaiah, Associate Professor (GPBR), Agricultural College, Warangal for statistical analysis.