Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.

3
ISSN 0975-928X

Research Article
AMMI biplot analysis for stability in basmati rice (Oryza sativa L.) in different
production systems
Bharat Taindu Jain1*, A. K. Sarial2, R. P. Saharan1, HariKesh1 and Hirdayesh Anuragi1
1
Department of Genetics and Plant Breeding, CCS HAU, Hisar, Haryana, India
2
Vice-Chancellor, CSKHPKV, Palampur, Himachal Pradesh, India
E-Mail: [email protected]

(Received: 14 Sep 2017; Revised: 25 May 2018; Accepted: 26 Jun 2018)

Abstract
The study of G×E Interaction (GEI) is critical for evaluating the mean performance and stability of cultivars across wide range of
environmental conditions. An experiment was conducted during kharif 2014-2015 for discriminating 22 basmati rice genotypes
for days to 50% flowering and days to 75% maturity using AMMI and GGE stability models under four different environments
viz. direct (DSR-wet) and (DSR-dry) and indirect seeding transplanted rice (TPR) and system of rice intensification (SRI)
conditions at CCS HAU farm, Kaul. Estimates of G×E interaction following Eberhart and Russell (1966) and AMMI biplot
analysis as per Gauch and Zobel model (1989) were computed. For days to 50% flowering, genotypes like Pusa Basmati 6, Pusa
Sugandh 3, Haryana Basmati-1 and Pusa RH 10 were identified under SRI, CSR-30 under DSR (dry) and DSR (wet) and HKR
98-476, Pusa Sugandh 2 and Pusa Sugandh 5 under TPR conditions. However, for days to 75% maturity, Pusa Sugandh 3, Pusa
Basmati 1121, Pusa Basmati 1 and HKR 06-434 were adapted best to SRI, Traori Basmati, Basmati-370, HKR 98-476 and HKR
06-443 to TPR and DSR (dry) and HKR 06-487 and Pusa RH 10 to DSR (wet).

Key words
G×E interaction, AMMI biplot, GGE stability, days to 50% flowering and basmati rice.

Introduction
Rice (Oryza sativa L.) is the staple food for a large and easier planting, ensure proper plant population, and
proportion of the world’s population (Zhang, 2007). reduce labour, 10-12 days earlier crop maturity, more
India is the second largest rice-growing country in the efficient water use, higher tolerance to water-deficit
world; however, its productivity per unit area is low. In and often high profit in areas with assured water supply
India, rice is cultivated on 44.01 million hectares with (Datta et al., 1986). To date, no specific varieties have
a production of 105.31 MT and productivity of 2.23 been developed for the above non-conventional
T/ha. Although more than 900 rice varieties have been technologies. Released varieties for TPR do not
released in India, many of them were no longer perform well under non-conventional techniques.
cultivated within a few years due to inconsistent Varieties differed in their genetic potential and all
performance in diverse environments and only a few varieties are not promising for diversification. The
varieties with stable performance continue under varietal response to different production systems is
cultivation after 15 to 20 years of their release. The rice wide (Ghritlahre et al., 2011). To overcome such
production areas in the country are very diverse in challenges estimate G×E interaction and adaptability.
hydrology and combined to other soil and climatic The Additive Main effects and Multiplicative
factors make a difference in rice yield (Singh et al., Interaction (AMMI) model is a hybrid model involving
1997). Rice is primarily grown by transplanting of both additive and multiplicative components. Using
seedling in puddled field which is very cumbersome AMMI analysis and biplot facility, the promising rice
and labour intensive as it requires 30 man days ha -1 quantitative data were analyzed for determine G×E
(Parsad et.al, 2014). Due to conventional transplanting interaction effects in different production systems of
method, it has been causing a sharp decline in water rice to identify stable genotype and to determine which
table. It is imperative to shift from conventional genotype stable for specific environment.
namely transplanting method (TPR) to non-
conventional cultivation techniques namely direct Materials and Methods
seeded rice (DSR). The direct seeding technique offers The experimental materials consisted of twenty two
a useful option to reduce the limitations of transplanted released basmati varieties including elite lines. The
paddy. Direct-seeded rice offers the advantage of faster experiment was conducted in a randomized block

502
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

design with three replications in four environments Basmati and CSR-30 having regression coefficient
created agronomically termed production systems of significantly less than one, non-significant deviation
rice during (kharif) rainy season of 2014-2015 at CCS from regression and mean greater than population
Haryana Agricultural University, Rice Research mean (late flowering) while genotype Pusa Basmati
Station, Kaul farm (District: Kaithal). They are 1121, Pusa Basmati 1509, HKR 06-443, HKR 08-417
conventional transplanted rice (TPR) and non- and Pusa RH 10 with mean less than population (early
conventional viz; system of rice intensification (SRI), flowering) were found suitable for poor environment.
direct seeded dry (DSR, dry) and direct seeded wet
(DSR, wet). Plot size consisted of 5 rows of 2m length On the other hand, biplot analysis is possibly the most
and 0.20m breadth. The production systems have been powerful interpretive tool for AMMI models. There are
described in Table 1. The data were recorded on five two basic AMMI biplots, the AMMI 1 biplot where the
randomly selected plants per genotype per replication main effects (genotype mean and environment mean)
for days to 50% flowering and days to75% maturity. and IPCA 1 scores for both genotypes and
The G×E interaction was analyzed following Eberhart environments are plotted against each other. In the
and Russell (1966) model and AMMI biplot (Gauch, second AMMI 2 biplot scores for IPCA 1 and IPCA 2
1989). are plotted. The biplot technique was used to identify
appropriate genotype adapted to specific
Results and Discussion locations/environments (Gauch, 1996). For days to
Pooled analysis of variance computed as per Eberhart 50% flowering presence of GEI was clearly
and Russell (1966) model for days to 50% flowering demonstrated by the AMMI model had 22.38% of the
and days to 75% maturity (Table 2) showed that the total sum of squares attributed to environmental
variance due to genotypes were significant (p<0.05) for effects, 74.78% to genotypic and 2.38% to G×E
both the characters. This revealed the presence of interaction effects (Table 4). The G×E interaction mean
considerable genotypic variability among the sum of squares was very meager than that of
genotypes for traits under study. The mean sum of genotypes, so predominant difference was due to
squares due to genotype x environment interaction genotypic effect. The PCA 1 explained 61.79% of the
when tested against pooled error was significant for G×E interaction, PCA 2, 25.75% and PCA 3, 12.77%.
both the traits. Further partitioning of combined The cumulative captured by first two PCA axis was
environment and genotype x environment variance into 87.54% of total GEI using 44 DF.
linear and non-linear components showed that
environment linear was highly significant, G×E (linear) The AMMI analysis of variance revealed that G×E
was non significant while, pooled deviation (nonlinear interaction mean sum of squares was very less than that
component) when tested against pooled error was for genotypes, so major difference was due to
significant for both the characters. The Eberhart and genotypic effect. The PCA1 and PCA 2 cumulatively
Russell (1966) model used two parameters (bi and S- captured > 85 % of total G×E interaction. This implied
2
di) to define stability. S-2d is largely used to rank the that the interaction of the 22 rice genotypes with four
relative stability of cultivars. For days to 50% environments was predicted by the first two PCA
flowering (Table 3) stability analysis revealed all components. The findings are in confirmatory to that of
genotypes had significant regression coefficient (bi) Padmavati et al., 2013; Grihtlahre and Sarial, 2011,
and non-significant S-2di for days to 50% flowering. and Gauch and Zobel, 1996. Pusa Basmati 1509 had
Significant G×E interaction for number of days to 50% IPAC1 score near zero, hence had small interaction
flowering was reported by many workers like Grihtlhre effects indicating that this variety was less influenced
and Sarial, 2011; Sinha and Biswas, 1987, Kulkami et by the environment thus found stable. The interaction
al., 1988; and Ramya and Senthil kumar, 2008. None of environments was high and varied. Normal
of the genotypes was found stable. Genotype Pusa transplanting and SRI had positive interaction while
Basmati 6, Pusa Basmati 1, Imp Pusa Basmati 1 and DSR (wet) and DSR (dry) negative. Environments
HKR 98-476 having regression coefficient DSR (wet) and DSR (dry) were always on the right
significantly greater than one, non-significant deviation hand side of the midpoint of the main effect axis,
from regression and mean greater than population seemed to be unfavorable and SRI and TPR were
mean (late flowering) while genotype Pusa Sugandh 2, favorable environments. Genotypes HKR 06-487,
Pusa Sugandh 3, Pusa Sugandh 5, Haryana Basmati 1 CSR-30, HKR 08-417, HKR 08-425 and Super
and Basmati 370 with mean less than population mean Basmati with high mean and negative interaction were
(early flowering) were found suitable for better found adapted to unfavorable environment i.e. DSR
environment. Genotype HKR 3-408, HKR 06-434, (wet) and DSR (dry). The genotype Pusa Sugandh 5,
HKR 06-487, HKR 08-425, Traori Basmati, Super Pusa Sugandh 3 and Pusa Sugandh 2 with low mean

503
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

but high interaction were found adapted to TPR and all the environments. Similar findings were also
SRI. As per AMMI2 biplot, the environment DSR reported by Das et al. (2009) and Kulsum et al. (2013).
(dry) had short spokes and it did not exert strong In AMMI 2 biplot TPR and DSR (dry) had short
interactive force while environment SRI, TPR and spokes and they did not exert strong interactive force
DSR (wet) having long spoke exert strong interaction. while environment SRI and DSR (wet) having long
Genotypes near the origin were HKR 06-487, Super spoke exert strong interaction. Genotypes HKR 06-
Basmati, HKR06-443, Pusa Basmati 1509, Basmati 434, Haryana Basmati-1 and Super Basmati had high
370, HKR 3-408, HKR06-434, Haryana Mehak-1 and PCA score and away from origin were most responsive
Imp Pusa Basmati -1 hence they are non-sensitive to genotypes. Pusa Sugandh 3, Pusa Basmati 1121, Pusa
environmental interactive forces. Genotypes Pusa Basmati 1 and HKR 06-434 were adapted to SRI.
Basmati 6 and HKR 98-476 were most responsive Genotypes Traori Basmati, Basmati-370, HKR 98-476
genotype. Accordingly, suitable genotypes identified and HKR 06-443 to TPR and DSR (dry) while
with respect to site SRI were Pusa Basmati 6, Pusa genotype HKR 06-487 and Pusa RH 10 were adapted
Sugandh 3, Haryana Basmati-1 and Pusa RH 10, while to DSR (wet).
CSR-30 for DSR (dry) and DSR (wet) and HKR 98-
476, Pusa Sugandh 2 and Pusa Sugandh 5 for TPR. The present study provided an evaluation of genotypic
and environmental performance of twenty-two rice
The AMMI analysis of variance showed that G×E genotypes over a range of environments. According to
interaction sum of squares was very less than that for Eberhart and Russel (1966) regarding days to 50%
genotypes, so major difference was due to genotypic flowering, TPR and SRI had positive interaction
effect. The variance of G×E interaction was partitioned categorizing these as favorable environments while
into three significant interaction principal component DSR (wet) and DSR (dry) negative identified as
axes (IPCAs). Of which IPCA1 exhibited 56.22% of unfavorable. Genotype Pusa Sugandh 5, Pusa Sugandh
the total G×E interaction the cumulative captured by 3 and Pusa Sugandh 2 flowered early had positive
the 1st two component was IPCA 87.73%. This implied interaction with TPR and SRI thus adapted to favorable
that the interaction of the 22 rice genotypes with four environments while genotype CSR-30 to unfavorable
environments was predicted by the first two environments. With regards to days to 75% maturity
components of PCAI and PCAII. These findings were genotypes had positive interaction with SRI while
in conformity to those of Islam et al. (2014) and Gauch negative with TPR, DSR (wet) and DSR (dry).
et al. (1996). AMMI1 biplot revealed that interactions Genotypes Traori Basmati, HKR 06-487, Haryana
of environments were high and varied. DSR (wet) had Mehak-1, Pusa Basmati 6 and HKR 08-425 with
positive interaction while DSR (dry), TPR and SRI had positive interaction suited to SRI while genotypes
negative. Environments TPR, SRI and DSR (wet) Pusa Sugandh 2, Pusa Sugandh 3 and Pusa Sugandh 5
being on the right hand side of the midpoint of the to DSR(wet and dry ).
main effect axis, seemed to be favourable
environments for test grain weight, while DSR (dry) References
away from mid-point were generally less favourable
Das, S., Misra, R.C., Patnaik, M.C. 2009. G×E interaction of
environments. Genotypes Pusa RH 10 and Pusa
mid-late rice genotypes in LR and AMMI model
Basmati 6 had high mean and positive interaction were and evaluation of adaptability and yield stability.
adapted to DSR (wet). Conversely, the genotypes HKR Environ Ecol. 27: 529-535
08-425, Basmati-370 and HKR 98-476 with low mean
and negative interaction were adapted to DSR (dry). De Datta, S. K. 1986. Technology development and the
The genotypes HKR 06-434, Pusa Sugandh 3, and Pusa spread of direct-seeded flooded rice in South-East
Sugandh 5 with low mean but high interaction were Asia. Exp Agri. 22: 417-26.
adapted to TPR and SRI. Genotypes that grouped
together have similar adaptation while environments Eberhart, S.A. and Russel, W.A. 1966. Stability parameters
for comparing varieties. Crop Sci, 6:36-40.
which grouped together influences the genotypes in the
same way (Kemptom,1984). Genotypes HKR 06-487, Gauch, H.G. and Zobel, R.W. 1996. AMMI analysis of yield
HKR 3-408, HKR 08-417, Haryana Mehak-1, Pusa trials genotype x environment interaction, Kang,
Sugandh 2, Pusa Basmati 1509, Pusa Basmari 1, Imp M.S. and H.G. Gauch (Eds.). CRC Press, Boca
Pusa Basmati 1, CSR-30, and HKR 06-443 had IPAC1 raton, FL., USA: 85-122.
score near zero, hence had small interaction effects
indicating that these varieties were less influenced by Gauch, H.G. and Zobel, R.W. 1989. Accuracy and selection
the environment. Among them Pusa Sugandh 2 had success in yield trials analysis. Theoret Appl
high mean hence, found stable and recommended for Analysis, 77: 443-481.

504
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Ghritlahre, S.K., and Sarial, A.K. 2011. G×E Interaction and Parsad, P.S. 2014. AMMI analysis of yield performance and
Adaptability of Rice Cultivars in SRI and Normal stability of rice genotypes across different haor
Production Systems. Cereal Res Commun, 39(4): areas. Eco-friendly Agri J, 7(02): 20-24.
589-597.
Ramya, K. and Senthilkumar, N. 2008. Genotype ×
Islam, M.R., Anisuzzaman, M., Khatun, H.,Sharma, N., environment interaction for yield and its
Islam, M.Z., Akter, A., and Biswas, P.S. 2014. component traits in rice (Oryza sativa L.). Genetics
AMMI analysis of yield performance and stability and Plant Breeding, Department of Agriculture
of rice genotypes across different haor areas. Eco- Botany, Faculty of Agriculture, Annamalai
friendly Agri J, 7(02): 20-24. University, Annamalainagar India. Crop
Improvement, 35(1): 11-15.
Kempton, R.A. 1984. The use of biplots in interpreting
variety by environment interactions. J Agri Sci, S.K. and Biswas, S. 1987. Adaptability of rice varieties
103: 123-135. during boro season of west Bengal. Indian Agri,
31(1):15-18.
Khandhola, S.S. and Panwar, D.V.S. 1999. Studies on
genetic divergence, stability heterosis and Singh, B.N., Fagade, S., Ukwungwu, M.N., Williams, C.,
combining ability in rice (Oryza sativa L.). Jagtap, S.S. and Oladimeji, O. 1997. Rice growing
environment and biophysical constraint in rice
Kulkami, R.S. and Shashaidal, H.E. 1989. Genotype × agroecological Zones of Nigeria. Meteorol J, 2:35-
environment interaction in low land rice genotypes 44.
of hill areas of Kamataka. Ind J Genet Pl Breeding,
61(4): 350-352. Zhang, Z., Lu, C. and Xiang, Z.H. 1998. Analysis of variety
stability based on AMMI model. Acta Agronom
Kulsum, M.U., Hasan, M.J., Akter, A.K. and Biswas, H.R.P. Sinica, 24:304-309.
2013. Genotype × Environment interaction and
stability analysis in hybrid rice: an application of
Additive main effects and multiplicative
interaction. Bangladesh J Botony, 42(1): 73-81.

Padmavathi, P.V., Satyanarayana, P.V., and Ahamed, M.L.

2013. Stability analysis of Rice (Oryza sativa L.)
hybrids utilizing Regression and AMMI models.
Society Pl Res. 26 (2):148-153.

505
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Table 1. Description of four different environments

Environment E1 E2 E3 E4
Description TPR SRI DSR (wet) DSR (dry)
Seed rate (Kg/ha) 20 5 20 20
Seedling age (Days) 25 15 Direct sowing at 5 cm depth Direct sowing at 5 cm depth
2
Spacing (cm ) 15×15 25×25 20 (R-R) 20 (R-R)
Seedling /Hill 2 1 2 2
No. of Irrigation 30-33 18-20 18-20 16-18
Weeding Spray Spray Hand 30-35 (DAS) Hand 30-35(DAS)
Source: Jain and Sarial, 2015.

Table 2. Pooled Analysis of variance over 4 environments (production systems) for days to 50% flowering
and days to 75% maturity per plant in rice. (Eberhart and Russell, 1966 model)

Source Days to 50% flowering Days to 75% maturity

Genotype 226.49* 495.79*

Environment 474.66* 399.39*

Gen X Env 2.86* 30.82*

Env+Gen X Env 24.31* 47.57*

Env (Linear) 1,423.99 1,198.20

Env X Gen (Lin) 5.12* 15.959

Pooled Deviation 1.65* 36.51*

Pooled Error 2.11 4.63

* Significant at 5% level of significance

506
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Table 3. Stability parameters for days to 50% flowering and days to 75% maturity of rice genotypes tested
over 4 environments (production systems)

Days to 75% maturity Days to 50% flowering

S. No Genotpes
Mean bi S-2di Mean bi S-2di

1. Pusa Basmati 1121 130.91 1.09* 13.72 99 0.70* -0.65

2. Pusa Basmati 1509 104.83 1.05* 3.17 85.08 0.97* 0.42

3. Pusa Sugandh 2 110.91 1.73* 8.54 90.08 1.17* 1.24

4. Pusa Sugandh 3 116.33 1.97* 12.49 93 1.36* 2.12

5. Pusa Sugandh 5 117.08 2.20* 4.98 93.83 1.44* 4.26

6. Pusa Basmati 6 137.91 1.19* 16.67* 107.16 1.41* 6.39

7. Pusa Basmati 1 141.16 0.83* 4.46 103.66 1.05* -0.58

8. Imp Pusa Basmati 1 141.33 0.95* 0.78 102.5 1.33* 2.83

9. HKR 98-476 141.5 1.33* 1.9 107.16 1.57* 0.58

10. HKR 3-408 141.08 1.35* 0.7 109.33 0.99* -0.6

11. HKR 06-434 142.25 0.85* 0.05 110.66 0.83* -0.28

12. HKR 06-443 132.16 0.77 64.99* 98.83 0.72* 0.62

13. HKR 06-487 139.58 1.14 64.57* 110.53 0.65* 0.39

14. HKR 08-417 134.33 0.75 24.09* 101.16 0.78* 1.81

15. HKR 08-425 132.41 0.85* -0.53 101.91 0.75* 2.21

16. Haryana Mahek-1 143 1.35 87.30* 114.25 0.99* 0.48

17. Haryana Basmati-1 126.83 0.12 198.94* 98.58 1.11* 0.17

18. Traoari Basmati 131.66 0.92 46.47* 104 0.93* 0.19

19. Super Basmati 134.16 0.31* -0.46 102.33 0.71* -0.55

20. CSR-30 138.16 0.01 0.73 106.5 0.68* 1.75

21. BASMATI-370 132.91 0.49 5.25 98.91 1.05* 0.54

22 PUSA RH-10 117 0.64 210.48* 89.75 0.73* 0.97

Mean 131.25 101.28

Standard error 0.81 0.16

* Significant at 5% level of significance

507
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Table 4. AMMI analysis of days to 50% flowering and days to 75% maturity in rice across 4 production
systems

Source Days to 50% flowering Days to 75% maturity

MSS % explained MSS % explained

Trials 73.11 155.76

Genotypes 226.49* 74.78 495.78* 76.82
Environments 474.67* 22.38 399.39* 8.84
G*E Interaction 2.86* 2.83 30.82* 14.32
PCA I 4.83** 61.79 71.57** 84.8
PCA II 2.20** 25.75 11.45** 12.39
PCA III 1.20* 12.77 2.89** 2.83
Error 0.71 1.51
*, ** Significant at 5% & 1% level of significance, respectively

508
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Mean days to 50% flowering

Fig. 1. AMMI biplot of days to 50% flowering showing main effects and G x E interaction of rice genotypes in
four environments

Legend : 1:Pusa Basmati 1121, 2:Pusa Basmati 1509, 3:Pusa Sugandh 2, 4:Pusa Sugandh 3, 5:Pusa Sugandh 5, 6:Pusa Basmati
6, 7:Pusa Basmati 1, 8: Improved Pusa Basmati 1, 9:HKR 98-476, 10:HKR 3-408, 11:HKR06-434, 12:HKR 06-443, 13:HKR 06-
487, 14:HKR 08-417, 15:HKR 08-425, 16:Haryana Mehak-1, 17:Haryana Basmati-1, 18:Traori Basmati, 19:Super Basmati,
20:CSR-30, 21: Basmati 370, 22:Pusa RH 10, E1:Normal production system, E2:SRI, E3:DSR (wet), E4:DSR (dry)

IPAC2

IPAC1
Fig. 2. AMMI 2 biplot of days to 50% flowering showing IPCA scores of rice genotype (G) ploted across
environment (E)

509
 Electronic Journal of Plant Breeding, 9 (2) : 502-510 (June 2018) DOI: 10.5958/0975-928X.2018.00061.3
ISSN 0975-928X

Mean days of 75% maturity

Fig. 3. AMMI biplot of days to 75% maturity showing main effects and G×E interaction of rice genotypes in
four environments
Legend : 1:Pusa Basmati 1121, 2:Pusa Basmati 1509, 3:Pusa Sugandh 2, 4:Pusa Sugandh 3, 5:Pusa Sugandh 5, 6:Pusa Basmati
6, 7:Pusa Basmati 1, 8: Improved Pusa Basmati 1, 9:HKR 98-476, 10:HKR 3-408, 11:HKR06-434, 12:HKR 06-443, 13:HKR 06-
487, 14:HKR 08-417, 15:HKR 08-425, 16:Haryana Mehak-1, 17:Haryana Basmati-1, 18:Traori Basmati, 19:Super Basmati,
20:CSR-30, 21: Basmati 370, 22:Pusa RH 10, E1:Normal production system, E2:SRI, E3:DSR (wet), E4:DSR (dry)

IPCA2

IPCA1

Fig. 4. AMMI-2 biplot of days to 75 % maturity showing IPCA scores of rice genotype (G) ploted across
environment (E)

510