Article

Identification of Salinity Tolerant Stable Sugarcane Cultivars

Using AMMI, GGE and Some Other Stability Parameters
Under Multi Environments of Salinity Stress
Ravinder Kumar 1,*, Pooja Dhansu 1,*, Neeraj Kulshreshtha 1, Mintu Ram Meena 1,
Mahadevaswamy Huskur Kumaraswamy 2, Chinnaswamy Appunu 2, Manohar Lal Chhabra 1 and S. K. Pandey 1

1 ICAR-Sugarcane Breeding Institute Regional Centre, Karnal 132001, India

2 ICAR-Sugarcane Breeding Institute, Coimbatore 641007, India
* Correspondence: [email protected] (R.K.); [email protected] (P.D.)

Abstract: Additive main effects and multiplicative interaction (AMMI), as well as various AMMI-
derived statistics, Genotype, and Genotype × Environment Interaction (GGE) models were em-
ployed on 24 sugarcane genotypes planted during two seasons (2017–18, 2018–19) under different
induced salinity stress environments using saline water irrigation (iw) viz., E1 (Normal iw during
crop season 2017–18), E2 (Normal iw during crop season 2018–19), E3 (4 dsm−1 ECiw during crop sea-
son 2017–18), E4 (4 dsm−1 ECiw during crop season 2018–19), E5 (8 dsm−1 ECiw during crop season
2017–18), E6 (8 dsm−1 ECiw during crop season 2018–19), E7 (12 dsm−1 ECiw during crop season 2017–
18) and E8 (12 dsm−1 ECiw during crop season 2018–19) to assess the genotype by the environment
interaction for the cane yield, commercial cane sugar (CCS) yield, number of millable cane (NMC),
single cane weight (SCW), and pol % in juice. Individual and interactive effects of the genotype and
environment for all the traits were significant. In the expression of total variability, the environmen-
tal contribution was higher for the cane yield (66.98%), CCS yield (67.60%), NMC (65.78%), and SCW
(43.27%), whereas genotypic contribution was higher in the expression of pol% (82.48%). As per
Citation: Kumar, R.; Dhansu, P.;
AMMI Stability Value (ASV), G14 (Co 13033), G23 (Co 15026), G7 (Co 05009), G17 (Co 13036), and
Kulshreshtha, N.; Meena, M.R.;
G2 (Co 15025) were the most stable genotypes for the cane yield. Whereas as per GSI (genotype
Kumaraswamy, M.H.; Appunu, C.;
selection index), genotypes G24 (Co 15027), G21 (Co 15023), G23 (Co 15026), and G17 (Co 13036)
Chhabra, M.L.; Pandey, S.K.
Identification of Salinity Tolerant
were found most stable. The Sustainability Index (SI) of the cane yield (CY) and its contributing and
Stable Sugarcane Cultivars Using CY-based computed traits were low for most of the genotypes, which indicates the negative impact
AMMI, GGE and Some Other of increased levels of irrigation-induced salinity in the expression of these traits. In the mean vs
Stability Parameters Under Multi stability biplot analysis, G21 (Co 15023), G24 (Co 15027), G16 (Co 13036), G6 (Co 0238), and G20 (Co
Environments of Salinity Stress. 14036) were found to be highly productive and stable genotypes for the cane yield. The superior
Sustainability 2023, 15, 1119. and stable performance of early maturing notified varieties G21 (Co 15023) and G6 (Co 0238) for CY
https://doi.org/10.3390/su15021119 and CCS yield indicates that they will help the farmers to obtain sustainable income in saline soil
Academic Editor: Jun-Ichi Sakagami conditions.

Received: 26 October 2022

Keywords: genotype selection index; salt tolerant varieties; sugarcane; sustainability index;
Revised: 13 December 2022
stability index
Accepted: 28 December 2022
Published: 6 January 2023

1. Introduction
Copyright: © 2023 by the authors. Li-
censee MDPI, Basel, Switzerland. Sugarcane, being a C4 crop, thrives well in tropical environments but is cultivated
This article is an open access article globally under varied climatic conditions of tropical and sub-tropical regions [1]. In the
distributed under the terms and con- recent past, Indian sugarcane agriculture had witnessed a swift improvement in cane and
ditions of the Creative Commons At- sugar production. India is the second largest sugarcane producer after Brazil, with around
tribution (CC BY) license (https://cre- 4.79 m ha area, 370.5 mt production, and 77.35 t ha−1 productivity [2]. Sugarcane has the
ativecommons.org/licenses/by/4.0/). ability to survive against various environmental-related limiting factors [3–6]. The

Sustainability 2023, 15, 1119. https://doi.org/10.3390/su15021119 www.mdpi.com/journal/sustainability

Sustainability 2023, 15, 1119 2 of 19

theoretical maximum cane yield for sugarcane is 472 t ha−1 [7], but various environmental
(biotic and abiotic) and agronomic factors (crop geometry, planting time, crop duration,
rotation, water, weeds, diseases, pests and nutrients management practices, crop lodging
etc.) limit the crop yield. Soil salinity is significantly coercing crop productivity, predom-
inantly in arid and semiarid regions. In India, the great Indogangetic plane (Punjab, Har-
yana, Uttar Pradesh, Bihar, and some parts of Rajasthan), arid tracts of Gujarat and Raja-
sthan and semi-arid tracts of Gujarat, Madhya Pradesh, Maharashtra, Karnataka, and An-
dhra Pradesh are largely affected by soil salinity [8]. Five states [Gujarat (2.23 m ha),
Uttar Pradesh (1.37 m ha), Maharashtra (0.61 m ha), West Bengal (0.44 m ha) and Raja-
sthan (0.38 m ha)] together account for almost 75% of the saline and sodic soils; further-
more, the area under salt-affected soils is estimated to be tripled (20 million ha) in India
by 2050 [9]. The problem of fringe quality water also befalls the northwestern arid part of
India, especially in the states of Rajasthan, Haryana, and Punjab [10]. Salinity inhibits
plant growth through ion toxicity, nutritional imbalances, osmotic effect, and oxidative
stress [11–13]. Although sugarcane is pondered as moderately sensitive against salinity
stress [14,15], the prolonged stay of the crop (one plant + one or two ratoon crops) in the
salt-affected area greatly limits the sugarcane yield [15,16]. The ill effect of salinity on the
growth, development, and yield contributing traits of sugarcane have been reported by
several researchers [17–21]-. Rao et al. [22] in their study identified the threshold value for
sugarcane under salinity as 1.7 dSm−1 and suggested that yield decreases by 5.9% with a
per unit increase in salinity. Irrigation-induced salinity or sodicity in sugarcane is reported
in several countries viz., Australia, Egypt, Iraq, the United States, India, Pakistan, Swazi-
land, South Africa, and Zimbabwe [23–25]. Increased soil salinity or sodicity are the most
substantial soil chemical processes which lead to soil degradation under irrigated sugar-
cane [26]. The resistant and susceptible sugarcane genotypes express differential re-
sponses for various growth and yield traits against dissimilar salinity levels [27–31]. It
indicates that there is a need to study the interaction effects of genotype, environment,
and G × E thoroughly against different levels of salinity [5,32]. Plant breeders overcome
the GEI (genotype by environment interaction) challenges by evaluating genotypes in a
variety of distinct environments to ensure that the specific genotype with a high yield and
stable performance is selected [4]. Stable performance and better adaptability are the main
criteria for the selection of genotypes in any breeding program [33]. Environment plays a
dominant role in the expression of the cane yield and attributed traits in sugarcane, and
the importance of G × E interaction is widely recognized [34–37]. The two widely used
biplot models are the AMMI biplot (the additive main effects and multiplicative interac-
tion and the GGE biplot (genotype + genotype × environment). The analysis of variance
of the genotype and environment main effects with the PCA of the GEI generates the
AMMI model, whereas the AMMI2 or GEI biplot is performed based on the singular value
decomposition (SVD) of a double centered G×E table [38]. The GGE biplots are based on
the environment-centered SVD, provided by Yan et al. [39] and graphically represent both
the genotype and genotype-by-environment based on the primary sources of variation
associated with the genotype assessment. The AMMI technique is effective in identifying
the discriminating genotypes that had stable performance across diverse environmental
conditions [40,41]. During yield trials, an additive main effect and a multiplicative inter-
action (AMMI) model are widely used to analyze G × E interaction [42]. AMMI is capable
of detecting GEI in a multi-dimensional environment and displaying it using a biplot.
GGE biplots-based multi-environment trail, genotype evaluation, and environmental val-
uation have been successfully executed for varietal stability analysis by several research-
ers [43-48]. The GGE biplot approach for decomposing genotype plus genotype-by-envi-
ronment (G + G × E) is more effective in a biplot graph compared with AMMI analysis. In
addition, the GGE biplot analysis has been used by several researchers to classify the mega
environment, assess genotype rankings, and decide the discriminative and representative
among the tested environments [42]. The GGE biplot will aid researchers in better under-
standing complicated GE interactions in multi-environment breeding line trials and
Sustainability 2023, 15, 1119 3 of 19

agronomic investigations [45]. The GGE biplot was used to determine the performance of
crop cultivars in a variety of stress conditions, ideal cultivars, mega-environment, and
core testing sites [49]. The direct presentation of genotype effects is not feasible in AMMI2
biplots since this approach only decomposes G × E interaction effects in the PCA. GGE
biplot analysis, on the other hand, is regarded as a useful statistical technique for produc-
ing phenotypically stable and superior cultivars, identifying stable genotypes across sev-
eral environments and achieving crop yield stability across multiple locations.
The AMMI model does not provide a quantitative stability measure, which is essen-
tial to quantify and rank the genotypes in terms of yield stability [50]. The AMMI stability
value (ASV) is used as a selection criterion to quantify and rank the genotypes according
to their yield stability [51]. Further stability per se should not be the only selection criteria,
because the most stable genotypes may not necessarily give the best yield performance
[34, 35]. Therefore, various authors have introduced different selection criteria for the sim-
ulation selection of yield and stability [52-58]. Therefore, the present study aims to identify
superior genotypes with stable yield performance over different salinity stress environ-
ments by evaluating the efficacy of various stability analysis methodologies.

2. Materials and Methods

2.1. Research Materials
The experiment with 24 sugarcane genotypes consisting of recently released com-
mercial varieties viz., Co 98014 (G3), Co 0118 (G4), Co 0237 (G5), Co 0238 (G6), Co 05009
(G7), Co 05011 (G8), Co 06034 (G9) and Co 12029 (G13), newly evolved “Co” canes viz.,
Co 11027 (G10), Co 12026 (G11), Co 12027 (G12), Co 13033 (G14), Co 13034 (G15), Co 13035
(G16), Co 13036 (G17), Co 14034 (G18), Co 14035 (G19), Co 14036 (G20), Co 15023 (G21),
Co 15025 (G22), Co 15026 (G23), Co 15027 (G24), and old varieties which had wider adap-
tation in past (Co 1148 (G1) and CoS 767 (G2) were chosen for evaluating at different levels
of salt concentrations under controlled field conditions at ICAR-SBI, RC, Karnal (29.68 N
latitude and 76.99 E longitude at 243 m above mean sea level), Haryana, India during crop
seasons 2017–18 and 2018–2019. All the studied material except CoS 767 (variety devel-
oped by UPCSR, Shahjahanpur) was developed at ICAR-SBI, RC, Karnal under the eco-
nomic breeding program. The information on parentage and key characters of the studied
material is given in Supplementary Table S1.

2.2. Experimental Design

To evaluate the sugarcane genotypes under salinity stress, the pit method of planting
was followed. The experiment was planted in round pits of size 60 cm × 45 cm using a
factorial Randomized Block Design with three replications. Black durable polysheets were
properly spread in each pit before refilling them with a nursery mixture containing a 1:1:1
ratio of soil, sand, and farmyard manure. The polysheets around the pits help to stop the
leaching of salts due to irrigation or rain.

2.3. Salt Treatments

The salinity (NaCl) stress was imposed to create four production environments of
different ECiw (Electrical conductivity of irrigation water) salinity levels using three lev-
els of saline irrigation water 4 ECiw (S1), 8 ECiw (S2), and 12 ECiw (S3), along with control
(normal water). The pH and EC of normal water used for irrigation were 7.7 and 0.01
dSm−1, respectively. A total of 12 two-budded setts of each genotype were planted in each
pit during the second fortnight of March 2017 and March 2018. The buds were allowed to
germinate by providing normal irrigation water at an interval of 7–10 days during the first
month. After 30 days of planting, the normal EC irrigation water (Niw), saline water of 4
EC dsm−1ECiw, 8 EC dsm−1ECiw, and 12 EC dsm−1ECiw was used for irrigation at 10-day in-
tervals until the onset of monsoon. During monsoon season, the crop was irrigated when
required with the different salinity levels of irrigation water and with the recommended
Sustainability 2023, 15, 1119 4 of 19

package and practices. In the two consequent years, a total of eight induced production
environments of irrigation water salinity viz., E1 (Niw during crop season 2017–18), E2
(Niw during crop season 2018–19), E3 (4 dsm−1ECiw during crop season 2017–18), E4 (4
dsm−1ECiw during crop season 2018–19), E5 (8 dsm−1ECiw during crop season 2017–18), E6
(8 dsm−1ECiw during crop season 2018–19), E7 (12 dsm−1ECiw during crop season 2017–18),
and E8 (12 dsm−1ECiw during crop season 2018–19) were developed in the ring pits.

2.4. Determination of Salinity Stress on Genotype Performance

The crop was harvested after 240 days, and data on the cane yield (t ha−1), NMC
(numbers) and SCW (kg) was recorded. At the same time, juice quality traits viz., brix%,
pol%, purity% and CCS% were estimated in a laboratory as per ICUMSA methods [59].
CCS (t ha−1) was worked out with the following equation:
𝐶𝑎𝑛𝑒 𝑦𝑖𝑒𝑙𝑑 × 𝐶𝐶𝑆%
𝐶𝐶𝑆 (𝑡 ℎ𝑎 −1 ) =
100
Different salinity levels were considered as an environment, and G × E interaction
was analyzed using additive main effects and multiplicative interaction (AMMI) analysis,
as per Gauch and Zobel [50], and GGE biplot analysis as per Yan et al. [60] was performed
graphically, based on AMMI and GGE biplot using R studio (a simplified version of R
statistical software) developed by the R Core Team [61]. The metan package of R studio
was used for AMMI and GGE biplots analysis [62]. The GGE biplots and AMMI are graph-
ical images to exemplify G × E interaction and genotype ranking based on mean and sta-
bility. The graph generated is based on multi-environment evaluation (which-won-where
pattern), genotype evaluation (mean versus stability), and tested environment ranking
(discriminative versus representative). The ranking of genotypes was allocated in increas-
ing order of each stability parameter.

2.5. AMMI Stability Value (ASV)

The AMMI stability value (ASV), used to compare the stability of genotypes as de-
scribed by Purchase et al. [51], was calculated as follows:

𝑆𝑆𝐼𝑃𝐶𝐴1
𝐴𝑆𝑉 = √[( ) × (𝐼𝑃𝐶𝐴1𝑠𝑐𝑜𝑟𝑒 )] + (𝐼𝑃𝐶𝐴2𝑠𝑐𝑜𝑟𝑒 )2
𝑆𝑆𝐼𝑃𝐶𝐴2

where SSIPCA1/SSIPCA2 is the weight given to the IPCA1-value by dividing the IPCA1 sum of
squares by the IPCA2 sum of squares. The larger the IPCA score, either negative or posi-
tive, the more specifically adapted a genotype is to certain environments. Smaller ASV
scores indicate a more stable genotype across environments.

2.6. Genotype Selection Index (GSI)

The genotype selection index (GSI) was calculated for each genotype, which incor-
porates both the mean of the trait and the ASV index in a single criterion [52].
𝐺𝑆𝐼𝑖 = 𝑅𝑀𝑖 + 𝑅𝐴𝑆𝑉𝑖
where GSIi is the genotype selection index for the ith genotype, RMi is the rank of the trait
mean of the ith genotype, and RASVi is the rank of the AMMI stability value for the ith
genotype. GSI incorporate both the mean of the trait and stability in a single criterion. The
low value of this parameter shows desirable genotypes with high mean and stability.

2.7. Sustainability Index (SI)

The sustainability index was estimated by the following formula as described by
Manjoor et al. [53].
Sustainability 2023, 15, 1119 5 of 19

𝑌 − 𝜎𝑛
𝑆𝐼 = [ ] × 100
𝑌𝑀
where Y = Average performance of a genotype, σn = Standard deviation, and YM = Best
performance of a genotype in any environment. The values of the sustainability index
were divided arbitrarily into five groups viz. very low (up to 20%), low (21–40%), moder-
ate (41–60%), high (61–80%), and very high (above 80%).

3. Results
3.1. Effect of Salinity on Sugarcane Genotypes—AMMI Analysis
The analysis of variance (ANOVA) showed that individual and interactive effects of
Genotypes (G), Environments (E), and genotype × environment interaction (G × E) were
significant for all the studied traits viz., the cane yield, CCS yield, NMC population, SCW
and Pol%. The environmental main effect represented 66.98%, the genotype explained
25.41%, and the G × E interaction explained 7.61% of the total variation for the cane yield.
Similarly for the CCS yield (t ha−1), the contribution of environment, genotype, and inter-
action effects on the total variation was 67.60%, 24.88%, and 8.12%, respectively. In addi-
tion, significant variations for NMC (represented 65.78% environmental variation, 21.81%
genotypic variation and 12.41 % G × E variation) were also observed. The analysis of var-
iance revealed 43.27% environmental variation, 42.74% genotypic variation, and 13.99 %
G × E variation for SCW (Table 1). For pol% (sucrose% in juice), the analysis of variance
revealed the presence of a highly significant variation for genotype (82.48%), environment
main effect (9.45%), and moderate interaction effects (8.07%).

Table 1. Analysis of variance of main effects and interactions for Cane yield, CCS t ha−1, NMC, SCW
and Pol % and variability explained in percent.

Cane Yield ha−1 CCS t ha−1 NMC ‘000 no. ha−1 SCW (kg) Pol%
Source df
M.S. VE (%) M.S. VE (%) M.S. VE (%) M.S. VE (%) M.S. VE (%)
Genotypes 23 4453.61 ** 25.41 68.57 ** 24.88 1413.3 ** 21.81 0.75 ** 42.74 45.60 ** 82.48
Environments 7 38,571.59 ** 66.98 627.28 ** 67.60 14,003.6 ** 65.78 2.48 ** 43.27 17.17 ** 9.45
Interactions 161 190.48 ** 7.61 8.12 ** 8.12 114.9 ** 12.41 0.03 ** 13.99 0.64 * 8.07
IPCA1 29 605.88 ** 57.29 11.84 ** 65.10 320.5 ** 50.25 0.11 ** 54.51 1.88 ** 53.02
IPCA2 27 224.25 ** 19.74 3.23 ** 16.52 157.0 ** 22.91 0.06 ** 27.55 0.87 ** 22.02
IPCA3 25 148.54 ** 12.10 2.00 ** 9.48 125.1 ** 16.91 0.03 ** 12.15 0.46 ns 11.25
IPCA4 23 113.50 ** 8.51 1.46 ** 6.34 65.2 ** 8.10 0.01 * 3.77 0.22 ns 5.97
Residuals 384 22.16 ns 00 0.46 ns 00 15.18 ns 00 0.002 ns 00 0.49 ns 00
Note: df = degree of freedom, M.S = Mean sum of square, VE% = variability explained %, *, ** and
*** indicates 5%, 1% and 0.1% level of significance, ns = non significance.

The difference between the principal components (IPCA) values was also highly sig-
nificant, and the first two components accounted for 77.03% of the whole effect on the
variation of the cane yield, 81.62% for CCS t ha−1, 73.16% for NMC (thousand ha−1), 82.06%
for SCW (kg), and 75.04% for Pol% (Table 1).

3.1.1. Biplot Analysis for Determination of Main Effect and Environment Influence
PC1 (principle component 1) Vs CY (Cane yield) environments E5, E6, E7, and E8
[higher ECiw 8 (E5, E6) and ECiw 12 (E7, E8)] demonstrated lower average main effects,
whereas environments E1, E2 (Normal ECiw), expressed the highest main effects and were
gainful for most of the genotypes. The PC1 scores for environments E3 and E4 (4 EC iw) are
near zero. Genotypes G20, G24, G21, G6, G4, G13, G8, G13, and G15, expressed higher
main effects; for the cane yield, on the contrary, genotypes G12, G2, G10, G19, G1, and
G18 exhibited lower main effects. (Figure 1a). The AMMI biplot using PC1 and PC2 scores
Sustainability 2023, 15, 1119 6 of 19

for the cane yield is displayed in Figure 1b. Genotypes G23, G24, G6, G14, G3, G21, and
G20 are stable near the origin, hence they have better adaptation across the environments.

(a)

(b)
Figure 1. (a). AMMI biplot PC1 Vs Cane Yield (CY); (b). AMMI Biplot PC1 Vs PC2 for Cane yield (t
ha−1); Where G = Genotypes (1–24) and E = Environments (1–8) of different ECiw levels viz., Normal
(E1, E2), 4EC (E3, E4), 8EC (E5, E6) and 12 EC (E7, E8).

The AMMI biplot for CCS yield drawn using PC Vs CCSY (Figure 2a) and PC1 Vs
PC2 (Figure 2b) indicates E1, E2, E3, and E4 as the favorable environments, whereas E5,
E6, E7, and E8 as poor environments. Similarly, G21, G24, G4, G6, G13, and G20 are rated
as the better performer genotypes for the trait, while G1, G2, G10, and G12 are rated as
the poor performer genotypes. Genotypes G24, G21, and G16 are located near the origin
in the PC1 Vs PC2 biplot and are stable performers (Figure 2b).
Sustainability 2023, 15, 1119 7 of 19

(a)

(b)
Figure 2. (a). AMMI biplot PC1 Vs CCS Yield (CCSY); (b). AMMI Biplot PC1 Vs PC2 for CCS yield
(t ha−1); Where G = Genotypes (1–24) and E = Environments (1–8) of different ECiw levels viz., Normal
(E1, E2), 4EC (E3, E4), 8EC (E5, E6) and 12 EC (E7, E8).

3.1.2. AMMI Stability Value (ASV)

As per ASV, a stable variety is one with close to zero value. The ASV explained G14
(Co 13033), G23 (Co 15026), G7 (Co 05009), G17 (Co 13036), and G2 (Co 15025) as the top
ranked genotypes for the cane yield (Table 2) due to a lower ASV, hence they are classified
as the most stable genotypes for the cane yield. For CCS yield, G6 (Co 0238), G16 (Co
13035), G14 (Co 13033), G3(Co 98014) and G22 (Co 15025) found most stable genotypes.
G19 (Co 14035), G17 (Co 13036), G15 (Co 13034), G21 (Co 15023) and G10 (Co 11027) for
NMC; G17 (Co 13036), G11 (Co 12026), G21 (Co 15023), G9 (Co 06034), G24 (Co 15027) for
SCW; whereas, G21 (Co 15023), G12 (Co 12027), G5 (Co 0237), G9 (Co 06034), and G16 (Co
Sustainability 2023, 15, 1119 8 of 19

13035) for Pol%, were the most stable genotypes for respective traits. On the contrary, G12
(Co 12027), G2 (CoS 767), G4 (Co 0118) for cane yield; G9(Co 06034), G11 (Co 12026), G8
(Co 05011) for CCS yield; G9 (Co 06034), G11 (Co 12026), G8 (Co 05011) for NMC; G5 (Co
0237), G22 (Co 15025), G1 (Co 1148) for SCW and G6 (Co 0238), G5 (Co 0237), and G4 (Co
0118) for pol% due to highest ASV ranks found most unstable for the respective traits.

3.1.3. Genotype Stability Index (GSI) and Sustainability Index (SI)

For the cane yield the GSI of genotypes G24 (Co 15027), G21 (Co 15023), G23 (Co
15026) and G17 (Co 13036) was lowest (Table 2). Similarly for CCS yield G16 (Co 13035),
G6 (Co 0238), G24 (Co 15027), G21 (Co 15023), and G3 (Co 98014) were the best performer
genotypes as per GSI values. G21 (Co 15023) and G24 (Co 15027) also had better GSI for
rest of the traits viz., NMC, SCW and Pol%.
The SI (supplementary Table S2), of genotypes G18, G1, G10, G2, G16, G20, and G21,
was moderate for the cane yield, whereas rest of the genotypes represents low to very low
category. For CCS yield, the SI, was moderate for G18, while rest of the genotypes catego-
rized into low to very low category. G18, G22, G1, G21, G24 and G20 had high, while rest
of the genotypes had moderate values of SI for NMC. For SCW, G2, G3, G4, G6, G15, G16,
G18, G19, G20, G21, and G24 represents high SI, while rest of the genotypes reflects mod-
erate category. All the genotypes, represents very high category of SI for pol% in juice.
Sustainability 2023, 15, 1119 9 of 19

Table 2. Average values of observed traits for genotypes, AMMI stability value (ASV), rank of the ASV (RA), rank of trait mean (RM) and genotype
selection index (GSI).

Clo Cane Yield ha−1 CCS t ha−1 NMC Thousands ha−1 SCW (kg) Pol%
nes Mean ASV RA RM GSI Mean ASV RA RM GSI Mean ASV RA RM GSI Mean ASV RA RM GSI Mean ASV RA RM GSI
G1 57.67 7.78 18 10 28 7.58 4.08 21 12 33 71.57 2.93 13 1 14 0.80 0.66 22 16 38 15.41 0.69 15 21 36
G2 53.55 8.48 23 20 43 5.73 4.31 22 22 44 54.74 2.15 9 16 25 0.75 0.49 18 18 36 15.88 0.22 4 19 23
G3 71.23 2.21 7 14 21 7.97 0.59 4 11 15 65.74 1.97 8 5 13 0.79 0.23 7 17 24 17.16 0.33 7 12 19
G4 37.33 8.24 22 3 25 10.80 5.33 24 3 27 59.68 3.50 16 12 28 1.16 0.23 6 3 9 17.30 1.20 22 9 31
G5 70.60 4.88 14 23 37 6.34 3.72 19 18 37 46.40 4.55 21 23 44 0.73 0.88 24 19 43 18.85 2.29 23 3 26
G6 44.69 4.45 13 5 18 9.83 0.20 1 6 7 60.09 3.49 15 10 25 1.11 0.32 12 4 16 17.00 2.40 24 13 37
G7 61.76 1.18 3 16 19 6.37 0.70 6 17 23 59.94 3.85 18 11 29 0.72 0.31 11 20 31 16.10 1.18 21 17 38
G8 40.33 7.91 19 9 28 8.64 3.26 16 9 25 69.80 5.49 22 2 24 0.83 0.45 16 14 30 16.07 0.88 20 18 38
G9 42.27 3.21 10 21 31 6.32 1.43 11 19 30 54.09 8.34 24 17 41 0.68 0.21 4 23 27 18.13 0.56 10 4 14
G10 39.58 7.47 17 18 35 5.32 4.83 23 23 46 50.91 1.18 5 20 25 0.80 0.44 15 15 30 14.85 0.84 17 22 39
G11 31.54 2.17 6 22 28 5.95 1.18 9 21 30 52.46 6.63 23 18 41 0.69 0.15 2 22 24 17.19 0.58 11 11 22
G12 63.98 8.75 24 24 48 5.27 3.72 18 24 42 46.51 2.31 10 22 32 0.66 0.63 21 24 45 19.04 0.29 6 2 8
G13 43.28 8.01 21 8 29 9.55 3.78 20 8 28 60.23 3.62 17 8 25 1.01 0.44 14 8 22 17.32 0.15 1 8 9
G14 56.60 0.72 1 17 18 6.40 0.50 3 16 19 45.40 3.16 14 24 38 0.89 0.58 19 13 33 16.93 0.20 2 14 16
G15 66.49 5.74 16 11 27 8.59 3.52 17 10 27 55.90 0.80 3 14 17 0.95 0.42 13 12 25 17.29 0.22 5 10 15
G16 55.15 3.28 11 7 18 10.38 0.33 2 4 6 67.25 1.47 11 4 15 0.97 0.60 20 9 29 18.00 0.88 19 5 24
G17 67.08 1.38 4 12 16 7.44 0.88 8 14 22 55.69 0.78 2 15 17 0.95 0.14 1 11 12 15.63 0.21 3 20 23
G18 42.09 7.93 20 6 26 10.20 3.09 15 5 20 65.16 4.31 20 6 26 1.02 0.46 17 6 23 17.78 0.67 14 6 20
G19 42.03 2.87 9 19 28 6.09 1.61 13 20 33 56.12 0.77 1 13 14 0.71 0.28 8 21 29 16.66 0.60 12 16 28
G20 78.93 4.96 15 1 16 9.62 1.54 12 7 19 63.13 2.41 11 7 18 1.22 0.30 10 2 12 14.36 0.84 18 24 42
G21 70.87 2.66 8 4 12 12.06 2.64 14 1 15 68.44 0.87 4 3 7 1.01 0.16 3 7 9 19.74 0.80 16 1 17
G22 51.23 1.97 5 15 20 7.47 0.66 5 13 18 51.53 3.90 19 19 38 0.97 0.67 23 10 33 16.82 0.53 9 15 24
G23 54.61 0.82 2 13 15 6.85 0.78 7 15 22 48.45 2.88 12 21 33 1.08 0.30 9 5 14 14.56 0.67 13 23 36
G24 76.22 3.44 12 2 14 11.42 1.26 10 2 12 60.21 1.46 6 9 15 1.23 0.23 5 1 6 17.60 0.43 8 7 15
Sustainability 2023, 15, 1119 10 of 19

3.1.4. GGE Biplots for Cane Yield and CCS Yield

For the cane yield and CCS yield, environments E1 and E2 have the lengthiest envi-
ronmental vector with narrow angles to AEC, have better discriminative power, and are
deliberated as ideal environments for the expression of the respective traits, i.e., cane yield
and CCS yield (Figures 3 and 4). The shortest environmental vector observed for E7 and
E8 for these traits, indicating lesser differentiation power of the environments in relation
to the genotypes, i.e., all the genotypes had the average or similar performance in those
environments.

Figure 3. Discriminativeness vs representativeness for Cane yield (t ha−1).

Figure 4. Discriminativeness vs representativeness for CCS yield (t ha−1).

Sustainability 2023, 15, 1119 11 of 19

The mean vs stability biplots envisage the mean performance of the genotypes
through the environments. The line in Figures 5 and 6 passing through the origin repre-
sents the “average-environment coordinates”, and the right-hand side of the vertical line
symbolizes a higher mean yield for the genotypes. The second axis denotes stability. For
the cane yield, G21, G24, G16, G6, and G20 are positioned closer to the AEC, and the right-
hand side indicates that these are highly productive and stable genotypes for the cane
yield. Though genotypes G4, G18, G13 etc., are positioned towards a higher mean yield,
they are also unstable due to their far away position from the AEC line. For the CCS yield,
G20, G24, G21, G5, and G16 are located near the AEC towards the arrow, away from the
origin, which indicates their higher productiveness and stable performance for the trait.

Figure 5. Mean Vs Stability biplot for Cane yield (tha −1).

Figure 6. Mean Vs Stability biplot for CCS yield (t ha−1).

Sustainability 2023, 15, 1119 12 of 19

The ranking of genotypes (Figures 7 and 8) revealed the best productiveness of gen-
otypes G21 and G24 for the cane yield, whereas G20, G24, G6, and G21 for the CCS yield
due to occupying the first concentric circles for these traits.

Figure 7. GGE biplot ranking of genotypes for cane yield (t ha−1).

Figure 8. GGE biplot ranking of genotypes for CCS yield (t ha−1).

The GGE biplot polygons for the cane yield and CCS yield are displayed in Figures
9 and 10. Genotypes G4, G21, G18, G1, G2, G10, G12, and G5 were the vertex genotypes
for cane yield, whereas G13, G4, G20, G18, G2, G12, and G5 were the vertex genotypes for
the CCS yield, confirming the biggest distance from the origin. For the cane yield, the eight
environments were grouped into two mega-environments, i.e., the mega environment (i)
formed by E2 and the mega environment (ii) formed by E1, E3, E4, E5, E6, E7, and E8.
Genotypes G4, G21, and G18 were the vertexes of the mega environment (2) indicating
that they are better performers at all the studied levels of irrigation water salinity. For CCS
Sustainability 2023, 15, 1119 13 of 19

t ha−1 as well, the eight environments formed two mega groups (Figure 10) viz., the mega
environment (i) represented by E1, E2, E3, E4, E5, and E6 environments, whereas E7 and
E8 positioned in Mega environment (ii). G4 and G20 were the vertices genotypes in the
mega environment (i) whereas G18 was the vertices genotype in the mega environment
(ii), indicating their superior performance in these mega environments. There were also
vertexes of genotypes which were located in the regions with no environment at all, viz.,
G5, G12, G10, G2 for cane yield while G5, G12, G2, for CCS t ha−1, indicates their poor
performance in all the environments.

Figure 9. ‘Which won where’ polygon for Cane yield (t ha−1)..

Figure 10. ‘Which won where’ polygon for CCS yield (t ha−1)..
Sustainability 2023, 15, 1119 14 of 19

4. Discussions
Identifying the salt-tolerant genotype with stable performance across different salin-
ity levels is a challenging task for any breeder due to very high environmental and GE
interactions. AMMI incorporates ANOVA and PCA into a single model and enables a
simple visual interpretation of the GE interaction. It enables the clustering of genotypes
based on the similarity of response characteristics and identifying potential trends across
environments [63]. In our study, a large sum of the square for environments indicated that
the environments were diverse, with large differences among environmental means, caus-
ing most of the variation in the cane yield, CCS yield, and NMC. The significant G × E
interaction indicated the differential response of the genotypes at different salinity levels.
A lot of findings are in support of the result that the environment’s main effect used to be
higher for the yield and related traits [64-66]. In our study, the AMMI model with only
two PCA interactions was the best predictive model, which is in agreement with many
researchers [65-70].

4.1. Biplot Analysis for Determination of Main Effect and Environment Influence
For the AMMI1 biplot, exposing main effects means on the abscissa and PC1 values
as the ordinates, genotypes, or environments that locate almost on a vertical line have
interrelated means and those that locate nearly on a horizontal line have similar interac-
tion patterns [64]. These authors proved that the genotypes/environments with great PC1
scores (positive or negative) have higher interactions, whereas genotypes/environments
with PC1 scores near zero have slight interactions [64]. In general, environments with
scores near zero have little interaction across environments and provide low discrimina-
tion among genotypes [68]. In our study, PC1 Vs CY (Cane yield), environments E5, E6,
E7, and E8 [higher ECiw 8 (E5, E6) and ECiw 12 (E7, E8)] demonstrated lower average main
effects and, thus, were rated as deprived environments. Environments E1 and E2 (Normal
ECiw) expressed the highest main effects and were gainful for most of the genotypes. The
PC1 scores for environments E3 and E4 (4 ECiw) are near zero, which indicates that ECiw 4
ds m−1 shows the least interaction among the studied environments in the expression of
the cane yield. Genotypes G20, G24, G21, G6, G4, G13, G8, and G15, expressed higher
main effects for the cane yield, which indicates their better adaptation under favorable
environments. Contrary, genotypes G12, G2, G10, G19, G1, and G18 exhibited lower main
effects for cane yield, indicating their specific adaptation under environments of higher
salinity [ECiw 8 (E5, E6) and ECiw 12 (E7, E8)]. The genotypes near the origin are indifferent
to the environmental interaction and vice versa [69]. The ideal environment and genotype
are those which are near the origin [34]. In our study, the cane yield genotypes G23, G24,
G6, G14, G3, G21, and G20, due to their position near the origin, are comparatively stable
under the various salinity environments (Figure 2a).

4.2. Genotype Stability Index (GSI) and Sustainability Index (SI)

Since ASV takes into account both IPCA1 and IPCA2, most of the variation in the GE
is justified. The rank of ASV and trait mean is such that the lowest ASV is rank one, while
the highest trait mean is rank one; the ranks are then summed in a single simultaneous
selection index of trait, and its stability, called the GSI, incorporates both the mean of the
trait and the ASV index in a single criterion [52]. This is also known as the yield stability
index (YSI) [70]. A low value for this parameter represents stable genotypes with a high
mean yield. In our study, for the cane yield, the GSI of genotypes G24 (Co 15027), G21 (Co
15023), G23 (Co 15026), and G17 (Co 13036) were the lowest (Table 2), which indicates
they are high yield performers coupled with stability under salinity conditions. Similarly,
for the CCS yield, G16 (Co 13035), G6 (Co 0238), G24 (Co 15027), G21 (Co 15023), and G3
(Co 98014) were the best performer genotypes, as per the GSI values. G21 (Co 15023) and
G24 (Co 15027) also had better GSI for the rest of the traits viz., NMC, SCW, and Pol%. In
Sustainability 2023, 15, 1119 15 of 19

our earlier study, these entries were identified as tolerant or moderately tolerant against
salinity [5].
The sustainability index (SI) is the ratio of the average performance of the genotype
across the environments (excluding its standard deviation) to its best performance in a
particular environment [53]. It categorizes the genotypes into five groups viz., very low
(up to 20%), low (21–40%), moderate (41–60%), high (61–80%), and very high (above 80%).
This index is being used by various researchers in the selection of stable genotypes [70-
73]. In our study, except for a few (G18, G1, G10, G2, G16, G20 and G21 had moderate SI
for CY; G18 had moderate SI for CCS yield), most of the genotypes represent lower groups
of SI for the cane yield and CCS yield. This indicates that these traits were greatly affected
by the increased level of irrigation-induced salinity. On the contrary, NMC and SCW rep-
resent a moderate-to-high category of SI, which indicates a moderate reduction in the ex-
pression of these traits, with an increased level of irrigation-induced salinity stress and
the least impact on the juice quality as compared with yield and contributing traits.

4.3. GGE Biplot Analysis

The GGE biplot analysis is a very useful statistical tool in examining the aggression
of the genotype by environment (GE) interaction as well as identifying mega environ-
ments and superior genotypes. It is a scatter plot which graphically shows both the entries
(e.g., genotypes) and the tester’s (e.g., environments) of two-way data for the picturing of
the mega environments, the ranking of the genotypes, and the identification of stable en-
vironments [74]. The GGE Biplot analysis is being extensively used by researchers in the
delineation of Genotype and Genotype by Environment interaction in various crops, in-
cluding sugarcane [75,76]. The discriminativeness vs. representative explanations of GGE
Biplots recognizes the best environments with the highest discriminative power to dis-
criminate the genotypes. The average environmental coordinates (AEC) and test environ-
ments are proficient in envisaging the three types of environments viz., type-1 (short vec-
tors with ordinary distinguish power showing the average performance of genotypes),
type-2 (longest vectors with the utmost discriminative power, proficient to discriminate
the performance of the genotypes), and type-3 (longest vector with great angles, suitable
to harsh effects of environments) [77]. In our study, for the cane yield and CCS yield en-
vironments, E1 and E2 have the lengthiest environmental vector with narrow angles to
AEC and better discriminative power, and they are deliberated as ideal environments for
the expression of the respective traits, i.e., the cane yield and CCS yield (Figures 3 and 4).
The shortest environmental vector was observed for E7 and E8 for these traits, indicating
lesser differentiation power of the environments in relation to the genotypes, i.e., all the
genotypes had the average or similar performance in those environments.
The mean vs. stability biplots envisage the mean performance of the genotypes
through the environments. The line in Figures 4 and 5 passing through the origin repre-
sents the “average-environment coordinates” and the right-hand side of the vertical line
symbolizes higher mean yield for the genotypes. The second axis denotes stability; geno-
types nearer to the origin are more stable [78–80] and the genotypes situated closer and in
the track of AEC are considered perfect and best performers [77]. In our study, for the
cane yield G21, G24, G16, G6, and G20 are positioned closer to the AEC and on the right-
hand side, which indicates that these are highly productive and stable genotypes for the
cane yield. Though genotypes G4, G18, and G13 are also positioned towards higher mean
yield but are unstable due to far away position from the AEC line. For CCS yield G20,
G24, G21, G5, and G16 are located near the AEC towards the arrow, away from the origin,
indicating their higher productiveness and stable performance for the trait.
The ranking of genotypes on biplots comforts to pinpoint the best genotypes based
on their positions in the concentric circle [77]. A line passing through the biplot origin
from the lower right to the upper left is the AEC as defined by the first two PCs of the
environment scores. In our study, genotypes G21 and G24 were shown for the cane yield,
Sustainability 2023, 15, 1119 16 of 19

whereas G20, G24, G6, and G21 were shown for the CCS yield due to occupying the first
concentric circles for these traits are the best productive genotypes.
The GGE biplot polygon recognized the pictorial grouping of environments based on
crossed GEI between the highest yield genotypes [81]. It is constructed by an asymmetrical
polygon. The number of lines starting at the bi-plot’s origin and intercepting the polygon
vertically is equal to the sides in the polygon. The genotypes that are out of the origin in
all directions polygons denote the vertices; hence all genotypes are inside the polygon
[78]. An imaginary environment is denoted by a line that vertically crosses a side of the
polygon, if both genotypes that formed that side, have better productivity, the genotype’s
comparative rank would be reversed in environments at the line’s opposite extreme
(crossed GE). Therefore, the lines radiating from the origin split the bi-plot into sections,
and each section had a vertex (genotype) that represents the best yield performance in
environments contained in that section, which is named a mega-environment. The vertex
genotypes are either the best or the poorest performers in some or all environments [60].
Genotypes G4, G21 and G18 were the vertexes of better performing mega environment
for cane yield, indicating that they are a better performer at all the studied levels of irri-
gation water salinity. For CCS t ha−1, the eight environments formed two mega groups
viz., mega environment (i) represented by E1, E2, E3, E4, E5, and E6 environments,
whereas E7 and E8 positioned in the mega environment (ii). G4 and G20 were the vertices
genotypes in the mega environment (i) whereas G18 was the vertices genotype in mega
environment (ii), indicating their superior performance in these mega environments.
There were also vertexes of genotypes which were located in the regions with no environ-
ment at all, viz., G5, G12, G10, G2 for cane yield while G5, G12, G2, for CCS t ha−1, indicates
their poor performance in all environments.

5. Conclusions
The genotype and environment main effects as well as G × E interaction effects were
significant for all the studied traits. The genotype G21 (Co 15023) was found as the most
stable genotype under different saline water environments, as it was among the best-
ranked genotypes for AMMI analysis, GSI, SI, and GGE biplot analysis for the CY and
CCS yield. The AMMI and GGE biplot analysis reveals the superiority of G6 (Co 0238) for
CY and CCS yield. Hence, the cultivation of these genotypes would help farmers to obtain
sustainable income in harsh environmental conditions.

Supplementary Materials: The following supporting information can be downloaded at:

https://www.mdpi.com/article/10.3390/su15021119/s1, Table S1: List of the studied genotypes along
with information on their parentage and key characters, Table S2: Sustainability Index of different
sugarcane genotypes against Irrigation water salinity
Author Contributions: R.K.: Conceptualization, investigation, data visualization, original draft
preparation; P.D.: investigation, data visualization and editing; N.K.: Conceptualization, investiga-
tion and final draft editing; M.R.M.: investigation; C.A.: Data visualization, Draft Editing; M.H.K.:
data analysis; M.L.C.: Investigation, S.K.P.: Investigation and supervision. All authors have read
and agreed to the published version of the manuscript.
Funding: The research was self funded by the institute.
Data Availability Statement: The data presented in the study are available on request from the
corresponding author. The data are not publicly available due to restrictions.
Acknowledgments: The authors are grateful to Director, ICAR-SBI, Coimbatore, Head ICAR-SBI
RC, Karnal and Director, ICAR-CSSRI, Karnal for providing necessary facilities to conduct this ex-
periment. Thanks are also due to B.N. Manjhi and Vijay Kumar in managing the field experiments.
Conflicts of Interest: The authors declare no conflict of interest.
Sustainability 2023, 15, 1119 17 of 19

References
1. Zhao, D.; Li, Y.-R. Climate Change and Sugarcane Production: Potential Impact and Mitigation Strategies. Int. J. Agron. 2015,
2015, 547386. https://doi.org/10.1155/2015/547386.
2. FAOSTAT data for crops and production for year 2020. Available online: https://www.fao.org/faostat/en/#data/QCL (accessed
on 9 September 2022).
3. Kumar, R.; Meena, M.R.; Dhansu, P.; Karuppaiyan, R.; Appunu, C.; Kulshreshtha, N.; Kaushik, P.; Ram, B. Winter Tolerance
Potential of Genetically Diverse Sugarcane Clones under Subtropical Climate of Northern India. Sustainability 2022, 14, 11757.
https://doi.org/10.3390/su141811757.
4. Kumar, A.; Sharma, P. Climate Change and Sugarcane Productivity in India: An Econometric Analysis. J. Soc. Dev. Sci. 2014, 5,
111–122.
5. Kumar, R.; Meena, M.R.; Kulshreshtha, N.; Kumar, A.; Ram, B. Genotypic Response of Recently Evolved Sugarcane “Co” Clones
Under Different Levels of Saline Irrigation Water. J. Sugarcane Res. 2017, 7, 159–168.
6. Dhansu, P.; Kulshreshtha, N.; Kumar, R.; Raja, A.K.; Pandey, S.K.; Goel, V.; Ram, B. Identification of Drought-Tolerant Co-canes
Based on Physiological Traits, Yield Attributes and Drought Tolerance Indices. Sugar Tech 2021, 23, 747–761.
https://doi.org/10.1007/s12355-021-00967-7.
7. Moore, P. Sugarcane Biology, Yield and Potential for Improvement. Presentation at the BIOEN Workshop on Sugarcane Im-
provement, Sao Paulo. 2009. Available online: http://www.fapesp.br/materia/5064/bioen/workshop-bioen-on-sugarcaneim-
provement-18-e-19-3-2009-.html (accessed on 10.10.2022).
8. Shrivastava, P.; Kumar, R. Soil salinity: A serious environmental issue and plant growth promoting bacteria as one of the tools
for its alleviation. Saudi J. Biol. Sci. 2014, 22, 123–131. https://doi.org/10.1016/j.sjbs.2014.12.001.
9. Sharma, D.K.; Singh, A. Salinity research in india achievements, challenges and future prospects. Water Energy Int. 2015, 58, 35–
45. https://www.researchgate.net/publication/304395327_Salinity_Research_in_India-Achievements_Challenges_and_Fu-
ture_Prospects (assessed on 10.10.2022)
10. Singh, G. Salinity-related desertification and management strategies: Indian experience. Land Degrad. Dev. 2009, 20, 367–385.
https://doi.org/10.1002/ldr.933.
11. Leigh, R.A.; Wyn Jones, R.G. A Hypothesis Relating Critical Potassium Concentrations for Growth to the Distribution and
Functions of this Ion in the Plant Cell. New Phytol. 1984, 97, 1–13. https://doi.org/10.1111/j.1469-8137.1984.tb04103.x.
12. Zhu, J.K. Plant salt tolerance. Trends Plant Sci. 2001, 6, 66–71.
13. Chinnusamy, V.; Jagendorf, A.; Zhu, J.K. Understanding and Improving Salt Tolerance in Plants. Crop Sci. 2005, 45, 437–448.
https://doi.org/10.2135/cropsci2005.0437.
14. Workman, M.; Scott, P.M.; Nixon, D.J. A review of the management and amelioration of saline/sodic soils at Mhlume (Swaziland)
sugar company. Proc. South Afr. Sugar Technol. Assoc. 1986, 60, 162–167.
15. Nelson, P.; Ham, G. Exploring the response of sugar cane to sodic and saline conditions through natural variation in the field.
Field Crop. Res. 2000, 66, 245–255. https://doi.org/10.1016/s0378-4290(00)00077-0.
16. Rozeff, N. Irrigation water salinity and macro yields of sugarcane in South Texas. Sugar Cane 1998, 2, 3–6.
17. Bernstein, L.; Francois, L.E.; Clark, R.A. Salt tolerance of N.Co varieties of sugar cane. I: Sprouting, growth, and yield. Agr. J.
1966, 58, 489–493.
18. Dev, G.; Bajwa, M.S. Studies on salt tolerance of sugarcane. Indian Sugar 1972, 22, 723–726.
19. Rozeff, N. Sugarcane and salinity—A review paper. Sugar Cane 1995, 5, 8–19.
20. Chowdhury, M.K.A.; Miah, M.A.S.; Ali, S.; Hossain, M.A.; Alam, Z. Influence of sodium chloride salinity on germination and
growth of sugarcane (Saccharum officinarum L.). Sugarcane Int. 2001, 7, 15–16.
21. Rietz, D.N.; Haynes, R.J. Effect of irrigation-induced salinity and sodicity on sugarcane yield. Proc. South Afr. Sug. Technol. Assoc.
2002, 76, 173–185.
22. Rao, V.P.; Sengar, R.S.; Singh, S.; Sharma, V. Molecular and metabolic perspectives of sugarcane under salinity stress pressure.
Progress. Agric. 2015, 15, 77–84.
23. Hussein, J. Management and irrigation of Vertisols derived from basalts in Zimbabwe. In Fourth Zimbabwe Sugar Seminar. 20
August 1998; Clowes, M.S.J., Ed.; Zimbabwe Sugar Association Experiment Station: Chiredzi, Zimbabwe, 1998.
24. Haynes, R.J.; Hamilton, C.S. Effects of sugarcane production on soil quality: A synthesis of world literature. Proc. South Afr.
Sugar Technol. Assoc. 1999, 73, 45–51.
25. Dhansu, P.; Kumar, R.; Kumar, A.; Vengavasi, K.; Raja, A.K.; Vasantha, S.; Meena, M.R.; Kulshreshtha, N.; Pandey, S.K. Differ-
ential Physiological Traits, Ion Homeostasis and Cane Yield of Sub-Tropical Sugarcane Varieties in Response to Long-Term
Salinity Stress. Sustainability 2022, 14, 13246. https://doi.org/10.3390/su142013246.
26. Van Antwerpen, R.; Meyer, J.H. Soil degradation under sugarcane cultivation in northern KwaZulu-Natal. Proc. South Afr. Sugar
Technol. Assoc. 1996, 70, 29–33.
27. Gomathi, R.; Thandapani, P. Influence of Salinity Stress on Growth Parameters and Yield of Sugarcane. IOSR J. Pharm. Biol. Sci.
2014, 9, 28–32. Available online: www.iosrjournals.org (accessed on 10.12.2022).
28. Lingle, S.E.; Weigand, C.L. Soil salinity and sugarcane juice quality. Field Crops Res. 1997, 54, 259–268.
29. Akhtar, S.A.; Wahid, A.; Akram, M.; Rasul, E. Some growth, photosynthetic and anatomical attributes of sugarcane genotypes
under NaCI salinity. Int. J. Agri. Biol. 2001, 4, 439–443.
Sustainability 2023, 15, 1119 18 of 19

30. Husain, S.; Von Caemmerer, S.; Munns, R. Control of salt transport from roots to shoots of wheat in saline soil. Funct. Plant Biol.
2004, 31, 1115–1126. https://doi.org/10.1071/fp04078.
31. Singh, R.; Sengar, R. Effect of salinity stress on morphological and yield attributes of sugarcane (Saccharum of ficinarum L.)
genotypes. Int. J. Chem. Stud. 2020, 8, 2312–2316. https://doi.org/10.22271/chemi.2020.v8.i5af.10648.
32. Simões, W.L.; Calgaro, M.; Coelho, D.S.; Dos Santos, D.B.; De Souza, M.A. Growth of sugar cane varieties under salinity. Rev.
Ceres 2016, 63, 265–271. https://doi.org/10.1590/0034-737x201663020019.
33. Wolde, L.; Keno, T.; Tadesse, B.; Bogale, G.; Abebe, B. Megaenvironment Targeting of Maize Varieties Using AMMI and GGE-
Biplot Analysis in Ethiopia. Ethiop. J. Agric. Sci. 2018, 28, 65–84.
34. Alarmelu, S.; Balakrishnan, R.; Hemaprabha, G. G × E Interaction Studies in Multi-location Trials of Sugarcane Using GGE
Biplot and ANOM Analysis. J. Sugarcane Res. 2015, 5, 12–23.
35. Otieno, O.V.; Owuor, N.O. Multivariate Genotype and Genotype by Environment Interaction Biplot Analysis of Sugarcane
Breeding Data Using R. Int. J. Stat. Distrib. Appl. 2019, 5, 22–31.
36. Tiwari, D.K.; Pandey, P.; Singh, R.K.; Singh, S.P.; Singh, S.B. Genotypes x Environment Interaction and Stability Analysis in
Elite Clones of Sugarcane (Sachharum officinarum L.). Int. J. Plant Breed. Genet. 2011, 5, 93–98.
37. Mehareb, E.M.; Osman, M.A.M.; Attia, A.E.; Bekheet, M.A.; Elenen, F.F.M.A. Stability assessment for selection of elite sugarcane
clones across multi-environment based on AMMI and GGE-biplot models. Euphytica 2022, 218, 95.
https://doi.org/10.1007/s10681-022-03025-9.
38. Gauch, H.G. Statistical Analysis of Regional Yield Trials: AMMI Analysis of Factorial Designs 278; Elsevier: Amsterdam, The Neth-
erlands, 1992.
39. Yan, W.K.; Hunt, L.A.; Sheng, Q.L.; Szlavnics, Z. Cultivar Evaluation and Mega-Environment Investigation Based on the GGE
Biplot. Crop Sci. 2000, 40, 597–605. https://doi.org/10.2135/cropsci2000.403597x.
40. Kaya, Y.; Akcura, M.; Ayaranci, R.; Taner, S. Pattern analysis of multienvironment trials in bread wheat. Commun. Biometry Crop
Sci. 2006, 1, 63–71.
41. Admassu, S.; Nigussie, M.; Zelleke, H. Genotype-Environment Interaction and Stability Analysis for Grain Yield of Maize (Zea
mays L.) in Ethiopia. Asian J. Plant Sci. 2008, 7, 163–169. https://doi.org/10.3923/ajps.2008.163.169.
42. Oladosu, Y.; Rafii, M.Y.; Abdullah, N.; Magaji, U.; Miah, G.; Hussin, G.; Ramli, A. Genotype × Environment interaction and
stability analyses of yield and yield components of established and mutant rice genotypes tested in multiple locations in Ma-
laysia. Acta Agric. Scand. Sect. B Soil Plant Sci. 2017, 67, 590–606. https://doi.org/10.1080/09064710.2017.1321138.
43. Kumar, A.; Verulkar, S.; Mandal, N.; Variar, M.; Shukla, V.; Dwivedi, J.; Singh, B.; Swain, P.; Mall, A.; Robin, S.; et al. High-
yielding, drought-tolerant, stable rice genotypes for the shallow rainfed lowland drought-prone ecosystem. Field Crop. Res. 2012,
133, 37–47. https://doi.org/10.1016/j.fcr.2012.03.007.
44. Mohammadi, R.; Haghparast, R.; Amri, A.; Ceccarelli, S. Yield stability of rainfed durum wheat and GGE biplot analysis of
multi-environment trials. Crop. Pasture Sci. 2010, 61, 92–101. https://doi.org/10.1071/cp09151.
45. Luo, J.; Pan, Y.-B.; Que, Y.; Zhang, H.; Grisham, M.P.; Xu, L. Biplot evaluation of test environments and identification of mega-
environment for sugarcane cultivars in China. Sci. Rep. 2015, 5, 15505. https://doi.org/10.1038/srep15505.
46. Al-Naggar, A.M.M.; Shafk, M.M.; Musa, R.Y.M. AMMI and GGE Biplot Analyses for Yield Stability of Nineteen Maize Geno-
types Under Different Nitrogen and Irrigation levels. Plant Arch 2020, 20, 4431–4443.
47. Mohammadi, R.; Abdulahi, A.; Haghparast, R.; Armion, M. Interpreting genotype- environment interactions for durum wheat
grain yields using non-parametric methods. Euphytica 2007, 157, 239–251.
48. Mohammadi, R.; Amri, A. Comparison of parametric and non-parametric methods for selecting stable and adapted durum
wheat genotypes in variable environments. Euphytica 2008, 159, 419–432. https://doi.org/10.1007/s10681-007-9600-6.
49. Karimizadeh, R.; Mohammadi, M.; Sabaghni, N.; Mahmoodi, A.A.; Roustami, B.; Seyyedi, F.; Akbari, F. GGE Biplot Analysis of
Yield Stability in Multi-environment Trials of Lentil Genotypes under Rainfed Condition. Not. Sci. Biol. 2013, 5, 256–262.
https://doi.org/10.15835/nsb529067.
50. Gauch, H.G.; Zobel, R.W. Imputing missing yield trial data. Theor. Appl. Genet. 1990, 79, 753–761.
51. Purchase, J.L.; Hatting, H.; van Deventer, C.S. Genotype × environment interaction of winter wheat (Triticum aestivum L.) in
South Africa: II. Stability analysis of yield performance. South Afr. J. Plant Soil 2000, 17, 101–107.
https://doi.org/10.1080/02571862.2000.10634878.
52. Farshadfar, E.; Mahmodi, N.; Yaghotipoor, A. AMMI Sability Value and Simultaneous Estimation of Yield and Yield Stability
in Bread Wheat (Triticumaestivum L.). Austral J. Crop Sci. 2011, 5, 1837–1844.
53. Manzoor, B.A.; Tariq, M.S.; Ghulam, A.; Muhammad, A. Genotype × environment interaction for seed yield in Kabuli Chickpea
(Cicer arietinum L.) genotypes developed through mutation breeding. Pak. J. Bot. 2009, 41, 1883–1890.
54. Eskridge, K.M. Selection of Stable Cultivars Using a Safety-First Rule. Crop. Sci. 1990, 30, 369–374. https://doi.org/10.2135/crop-
sci1990.0011183x003000020025x.
55. Kang, M.S. Simultaneous Selection for Yield and Stability in Crop Performance Trials: Consequences for Growers. Agron. J. 1993,
85, 754–757. https://doi.org/10.2134/agronj1993.00021962008500030042x.
56. Dashiell, K.E.; Ariyo, O.J.; Ojo, K. Genotype X environment interaction and simultaneous selection for high yield and stability
in soybeans (Glycine max (L.) Merr.). Ann. Appl. Biol. 1994, 124, 133–139. https://doi.org/10.1111/j.1744-7348.1994.tb04121.x.
57. Bajpai, P.K.; Prabhakaran, V.T. A new procedure of simultaneous selection for high yielding and stable crop genotypes. Indian
J. Genet. Plant Breed. 2000, 60, 141–146.
Sustainability 2023, 15, 1119 19 of 19

58. Rao, A.R.; Prabhakaran, V.T. Use of AMMI in simultaneous selection of genotypes for yield and stability. J. Indian Soc. Agric.
Stat. 2005, 59, 76–82.
59. Chen, J.C.P.; Chi, C.C. Cane Sugar Handbook: A Manual for Cane Sugar Manufacturers and Their Chemists; John Wiley and Sons:
Hoboken, NJ, USA, 1993; p. 1090.
60. Yan, W.; Kang, M.S. GGE Biplot Analysis: A Graphical Tool for Breeders, Geneticists, and Agronomists, 1st ed.; CRC Press: Boca Raton,
FL, USA, 2002. https://doi.org/10.1201/9781420040371.
61. R Studio Team. RStudio: Integrated Development for R; RStudio, PBC: Boston, MA, USA, 2020. Available online: http://www.rstu-
dio.com/ (accessed on 8.10.2022).
62. Olivoto, T.; Lúcio, A.D. metan: An R package for multi‐environment trial analysis. Methods Ecol. Evol. 2020, 11, 783–789.
https://doi.org/10.1111/2041-210x.13384.
63. Bocianowski, J.; Niemann, J.; Nowosad, K. Genotype-by-environment interaction for seed quality traits in interspecific cross-
derived Brassica lines using additive main effects and multiplicative interaction model. Euphytica 2019, 215, 7.
https://doi.org/10.1007/s10681-018-2328-7.
64. Crossa, J.; Gauch, H.G.; Zobel, R.W. Additive Main Effects and Multiplicative Interaction Analysis of Two International Maize
Cultivar Trials. Crop. Sci. 1990, 30, 493–500. https://doi.org/10.2135/cropsci1990.0011183x003000030003x.
65. Rea, R.A.; De Sousa-Vieira, O.; Lucena, A.D.; Ramón, M.; Cárdenas, R.B. Genotype by environment interaction and yield sta-
bility in sugarcane. Rev. Fac. Nac. De Agron. Medellín 2017, 70, 8129–8138. https://doi.org/10.15446/rfna.v70n2.61790.
66. Tena, E.; Goshu, F.; Mohamad, H.; Tesfa, M.; Tesfaye, D.; Seife, A. Genotype × environment interaction by AMMI and GGE-
biplot analysis for sugar yield in three crop cycles of sugarcane (Saccharum officinirum L.) clones in Ethiopia. Cogent Food Agric.
2019, 5, 1651925. https://doi.org/10.1080/23311932.2019.1651925.
67. Meena, M.R.; Karuppiayan, R..; Ram, B.; Kumar, R.; Kulshreshtha, N. Genotypes × Environment Interactions and Stability Anal-
ysis of Sugarcane Clones (Saccharum spp.) by AMMI Model in Sub-tropical Regions of India. Indian J. Genet. Plant Breed. 2017,
77, 540–546.
68. Sheelamary, S.; Karthigeyan, S. Evaluation of promising commercial sugarcane genotypes for stability by AMMI analysis. Elec-
tron. J. Plant Breed. 2021, 12, 371–378. https://doi.org/10.37992/2021.1202.055.
69. Anandan, A.; Sabesan, T.; Eswaran, R.; Rajiv, G.; Muthalagan, N.; Suresh, R. Appraisal of environmental interaction on quality
traits of rice by additive main effects and multiplicative interaction analysis. Cereal Res. Commun. 2009, 37, 131–140.
https://doi.org/10.1556/crc.37.2009.1.15.
70. Bose, L.K.; Jambhulkar, N.N.; Pande, K.; Singh, O.N. Use of AMMI and other stability statistics in the simultaneous selection of
rice genotypes for yield and stability under direct-seeded conditions. Chil. J. Agric. Res. 2014, 74, 3–9.
https://doi.org/10.4067/s0718-58392014000100001.
71. Singh, P.; Agarwa, D.K. Sustainability index as an aid for determining the genotypic stability in diploid cotton (Gossypiumar-
boretum). J. Cotton Res. 2003, 17, 90–92.
72. Tuteja, O.P. Comparative studies on stability parameters and sustainability index for selecting stable genotypes in upland cotton
(Gossypiumhirsutum L.). Indian J. Genet. Plant Breed. 2006, 66, 221–224.
73. Gangwar, B.; Katyal, V.; Anand, K.V. Stability and efficiency of cropping systems in Chatisgarh and Madhya Pradesh. Indian J.
Agric. Sci. 2004, 74, 521–528.
74. Yan, W. Singular-Value Partitioning in Biplot Analysis of Multienvironment Trial Data. Agron. J. 2002, 94, 990–996.
75. Mahadevaiah, C.; Hapase, P.; Sreenivasa, V.; Hapase, R.; Swamy, H.K.M.; Anilkumar, C.; Mohanraj, K.; Hemaprabha, G.; Ram,
B. Delineation of genotype × environment interaction for identification of stable genotypes for tillering phase drought stress
tolerance in sugarcane. Sci. Rep. 2021, 11, 18649. https://doi.org/10.1038/s41598-021-98002-y.
76. Zubair, M.; Ahmad, S.; Rasool, A.; Farooq, M.A.; Khalil, I.A. Evaluation of Sugarcane Genotypes for Different Ecologies of
Pakistan by Employing Gge-Biplot Technique. Pak. J. Agric. Res. 2019, 33, 579–588. https://doi.org/10.17582/jour-
nal.pjar/2019/32.4.579.588.
77. Yan, W.; Kang, M.S.; Ma, B.; Woods, S.; Cornelius, P.L. GGE Biplot vs. AMMI Analysis of Genotype-by-Environment Data. Crop
Sci. 2007, 47, 643–653. https://doi.org/10.2135/cropsci2006.06.0374.
78. Yan, W.K. GGE Biplot vs. AMMI Graphs for Genotype-by Environment Data Analysis. Indian Soc. Agric. Stat. 2011, 65, 181–193.
Available online: https://www.cabdirect.org/cabdirect/abstract/20113313412 (accessed on 10.10.2022).
79. Hongyu, K.; Penña, M.G.; Araújo, L.B.; Dias, C.T.S. Statistical analysis of yield trials by AMMI analysis of genotype × environ-
ment interaction. Biom. Lett. 2014, 51, 89–102.
80. Hongyu, K.; Silva, F.L.; Oliveira, A.C.S.; Sarti, D.A.; Araújo, L.C.; Dias, C.T.S. Comparação Entre OsModelos AMMI e GGE
Biplot Para os Dados de Ensaiosmulti-Ambientais. Rev. Bras. Biom. 2015, 33, 139–155.
81. Neisse, A.C.; Kirch, J.L.; Hongyu, K. AMMI and GGE Biplot for genotype × environment interaction: A medoid–based hierar-
chical cluster analysis approach for high–dimensional data. Biom. Lett. 2018, 55, 97–121. https://doi.org/10.2478/bile-2018-0008.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual au-
thor(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.