Research Article

Evaluation of Sugarcane Genotypes for Different Ecologies of Pakistan by Employing Gge-Biplot Technique

Muhammad Zubair1, Sagheer Ahmad2*, Awais Rasool1, Muhammad Asad Farooq1 and Ibni Amin Khalil3

1National Agricultural Research Center, Islamabad, Pakistan; 2Plant Sciences Division, Pakistan Agricultural Research Council, Islamabad, Pakistan; 3Department of Plant Breeding and Genetics, KPK Agricultural University, Peshawar, Khyber Pakhtunkhwa, Pakistan.

Received | May 30, 2018; Accepted | February 03, 2019; Published | October 24, 2019

*Correspondence | Sagheer Ahmad, Plant Sciences Division, Pakistan Agricultural Research Council, Islamabad, Pakistan; Email: sagheersc@hotmail.com

Citation | Zubair, M., S. Ahmad, A. Rasool, M.A. Farooq and I.A. Khalil. 2019. Evaluation of sugarcane genotypes for different ecologies of Pakistan by employing gge-biplot technique. Pakistan Journal of Agricultural Research, 32(4): 579-588.

DOI | http://dx.doi.org/10.17582/journal.pjar/2019/32.4.579.588

Keywords | GGE Biplot, Sugarcane genotypes, Stability performance, Larkana, Sugarcane yield

Introduction

Sugarcane (Saccharum officinarum L.) is an important cash and sugar crop in Pakistan. It is being grown on an area of about 1.31 million hectares with a total production of 81.44 million tons (Jamshaid, 2017) ranking 5th in the world. However, the per acre sugarcane yield in Pakistan (62.2 ton ha-1) is far less than the varietal potential yield (100-150 ton ha-1) as well as sugarcane yield in significant cane growing countries and world average yield of 70.6 ton ha-1 (FAO, 2018). This is due to lack of proper climate resilient varietal selection along with many other agro-technological reasons.

In Pakistan, sugarcane is grown in the areas ranging from costal hot humid in the southern Sindh to very hot dry in upper Sindh and southern parts of Punjab as well as frosty cool climate of parts of Punjab and KP. Hence sugarcane growing regions of Pakistan vary climatologically and crop has to withstand the severity of the climate in the country. It necessitates developing and identifying specific varieties for each region.

Genotype × Environment interaction (GEI) is an important aspect of plant breeding programs. It may arise when certain genotypes are grown in diverse set of environments. A significant G × E interaction for a quantitative trait such as yield can seriously limit the efforts on selecting superior genotypes for both crop production and improved cultivar development (Kang and Gorman, 1989). By growing cultivars in different environments, the highest yielding and most stable cultivars can be identified (Lu’quez et al., 2002; Farshadfar et al., 2012; Akhter et al., 2015). When selecting genotypes for wide adaptation, plant breeders look for a non-crossover GEI or preferably the absence of GEI (Matus-Ca´diz et al., 2003). Thus, the estimation of stability of performance becomes important to identify consistent performing and high-yielding genotypes (Kang, 1998). Different researchers have used different stability statistics to determine whether or not cultivars evaluated in multi-environment yield trials (MET) are stable (Zubair and Ghafoor, 2001; Sabaghnia et al., 2006; Luo et al., 2012). Because the most stable genotype(s) may not be the highest yielding, the use of methods that integrate yield performance and stability to select superior genotypes becomes important (Kang and Magari, 1996; Luo et al., 2015).

The conventional method of partitioning for variety × environment interaction conveys little information on the individual patterns of response (Kempton, 1984) and are being covered by regression analysis (Gauch, 1988), multivariate analysis (Westcoff, 1987), cluster analysis (Crossa et al., 1991). In recent time additive main effect and multiplicative interaction (AMMI) model is also being used (Rea et al., 2011; Ready et al., 2011). Recently GGE Biplot Analysis is used (Yan 1999; Luo et al., 2012; Akhtar et al., 2015). Yan (1999) and Yan et al. (2000) proposed that GGE biplot allowed visual examination of the GE interaction pattern of MET data emphasizing two concepts viz., genotype by environment interaction (GE) and biplot technique developed by Gabriel (1971) employed to approximate and display the GGE of a MET. In addition, the GGE biplot also has a usage in selecting superior genotypes and test environments for a given mega-environment.

The objectives of this study were to investigate the stability in yield performance of sugarcane genotypes using GGE-biplot technique tested under diverse environmental conditions of Pakistan.

Materials and Methods

Nineteen sugarcane candidate varieties, contributed by eight different institutes were used in this investigation (Table 1). Seven locations were selected in the Punjab, Sindh and Khyber Pakhtunkhwa sugarcane growing regions which were diverse in agro-climatic conditions (Table 2). At each location, a randomized complete block design with three replications was employed for the trial conducted.

Table 1: Sugarcane varieties name, code name and contributing institutes.

Table 2: Latitude, longitude, altitude and soil type of the experimental locations during 2007-08.

The cane yield data was subjected to Analysis of Variance (ANOVA) technique (Steel and Torrie, 1980) to find out whether genotype × environment interaction was significant or not, thereafter the GGE-biplot method (Yan et al., 2000) was employed to study the stability of genotype with different sites for cane yield. The evaluation based on the model: yij - yj = y1εi1pj1 + y2εi2pj2, where: yij is the average yield of ith population in jth environment; yi is the overall mean of population j in environment j; y1εi1pj1 is the first principal component (PC-1); y2εi2pj2 is the second major component (PC-2); y1, y2 are the eigenvalues associated to PC-1 and PC-2, ε1 and ε2 are the scores of 1st and 2nd main component of the ith population and pj1 and pj2 are the scores of the 1st and 2nd principal component for the jth environment, respectively; and εij is the error associated with the model of the ith population and jth environment (Yan and Kang, 2003; Mattos et al., 2013).

Results and Discussion

Results from the analysis of variance showed that both the main effects i.e. genotype (G) and location/environment (E) and their interaction effect (G × E) had a significant influence on cane yield (Table 3). The relative magnitude of the variance explained by G, E and G × E is also shown in Table 3. The variation due to E was responsible for 16.1 % of the variance, while the variation due to G was 48.2 %. The variation due to G × E interaction had a contribution of 34.5 % to the total variation which is an indication of possible existence of differences among the locations in which the sugarcane varieties were evaluated. Zobel et al. (1988) and Mattos et al. (2013) reported that in multi-locational trials, the environment (E) normally explains up to 80% of the variation while G and G × E both usually represent around 10-20 % of each variation. Likewise, Luo et al. (2015) reported about 40% impact on cane yield only due to environment. However, in present study variations due to G and G × E is more suggesting that genotypes may be selected for specific environments as have been reported by other researchers (Xu et al., 2014; Akhter et al., 2015). Analysis of Variance (ANOVA) also represented that G × E interactions were significant for yield response. The results also indicated that the genotypes can be characterized for environmentally induced variations (Mattos et al., 2013; Luo et al., 2015). The mean cane yield of different sugarcane varieties across various locations is shown in Table 5. Among the locations maximum cane yield (119 ton ha-1) was attained at LRK (Larkana) followed by ISL (Islamabad) (97 ton ha-1), while minimum cane yield of 74 ton ha-1 was noticed at DIK (D.I. Khan). Across locations cane variety LRK-2004 (V-9) outclassed all other varieties by producing cane yield of 111 ton ha-1, followed by CPD-01-335 (V-19) with an average cane yield of 108 ton ha-1, while minimum cane yield of 81 ton ha-1 was produced by cane varieties S2002-US-560 (V-3) and GT-11 (V-11).

Table 3: ANOVA for sugarcane yield data (ton ha-1) obtained from sugarcane varietal trial at seven locations/environments.

The adaptability and stability of sugarcane genotypes can be graphically interpreted considering biplots with the first two axis of G × E interaction (Mattos et al., 2013). Hence yield performance of 19 sugarcane varieties is shown in the biplot across seven locations (Figure 1). These seven locations were distributed in four sectors. The first sector contained LRK (Larkana), FSD (Faisalabad) and DIK (D.I. Khan); the second sector had THT (Thatta). The third sector contained ISL and TJM (Tandojam), while the fourth sector had only JNG (Jhang). Figure 1 of GGA biplot represents a polygon view that is important to study the possible existence of mega environments within a growing region (Yan and Rajcon, 2002; Mattos et al., 2013; Xu et al., 2014). In this case two mega environments are Mega-I comprising Larkan, D.I. Khan and Faisalabad and Mega-II consisted of Islamabad and Tandojam. The other two minor environments were Thatta and Jhang. The polygon also indicated some varieties on the vertices while the rest on the inside of the polygon. These vertex varieties are the most responsive varieties since they have the longest distance from the biplot origin and best average performance in one or all environments (Yan and Rajcan, 2002; Luo et al., 2015). In the first sector the vertex varieties V-9 and V-15 were the best at both Larkana and Faisalabad locations. In the same sector sugarcane varieties V-12 and V-16 also performed well at all the three locations i.e. Larkana, Faisalabad and D.I. Khan. In the third sector V-19 was best at Islamabad and Tandojam locations. Sugarcane variety V-3 is supposed to be the best at Jhang as it is the vertex variety in sector four. Since, vertex 2 occupied no sugarcane variety therefore it is concluded that no variety performed the best at Thatta.

Table 4: Overall mean, GGE distance of sugarcane varieties from “ideal” variety and ranking (via GGE-biplot) on both average yield and stability performance across seven environments.

Figure 2 and 3 represents to the average yield and stability performance of 19 sugarcane varieties evaluated at seven locations. The average yield of varieties is approximated by the projection of their markers to the ATC X-axis (the single arrowed line; the direction indicates the positive end) while the stability of the varieties is approximated by the projection of their markers to the ATC Y-axis (double arrowed line) (Yan, 2001). The GGE identified V-15 as highest yielding variety, followed by V-9 and V-19 while V-3 as the poorest yielding variety. For only stability of performance V-15 was the best among all varieties followed by V-16 and V-12. However, varieties with high cane yield and relatively stable performance are important for growers. An “ideal” genotype is one that is the highest yielding (longest projection on ATC X-axis) across test environments and is absolutely stable (Shortest projection on ATC Y-axis) in performance (i.e., one that ranks the highest in all test environments) (Yan and Kang, 2003; Fan et al., 2007). When an “ideal” genotype view was drawn, sugarcane variety V-15 (GGE distance 1.2) was the closest to “ideal” variety

(center of the concentric circle) and supposed to be the best among all on the basis of average yield and stability of performance (Figure 3 and Table 4). Sugarcane varieties can be ranked on the basis of cane

The GGE-biplot also shows the relationship between yield and stability with respect to the vectorial standpoint of the environment and they are connected by vectors with the origin of the biplot (Mattos et al., 2013). The lesser the difference between average yield of genotypes in an environment the less would be G × E interaction (Yan and Kang 2003; Fan et al., 2007). In present study Faisalabad (FSD) and Larkana (LRK) environments were the most unstable indicating greater G × E interaction (Figure 4). Similarly, the difference between the average yields of genotypes was the lowest in Dera Ismail Khan (DIK) and Tandojam (TJM) contributing lesser towards G × E interaction (Figure 4; Table 5). Figure 4 also shows relationship among the locations. The vectors of all seven locations represent the discriminating ability and the linear map at the right side of the graph (in degrees) helps in explaining the relationship among them. Islamabad and Tandojam had smaller angle between their vectors, similarly the angles between DIK and LRK is also smaller as compared to rest of the locations. According to Yan and Kang (2003) and Fan et al. (2007), the cosine of the angle between two vectors of locations represents correlation between them.

Discriminating ability and representativeness of the test locations can be measured as the absolute distance of location from the biplot origin and the length of the projection from the marker of location on to the ATC Y-axis (Yan, 2001). Thus, Larkana was identified as the most discriminating location as it had a larger projection on to ATC X-axis, followed by Thatta (Figure 5) while, D.I. Khan and Tandojam were identified as representative locations as they had smaller projections on to ATC Y-axis. Yan (1999) and Yan et al. (2000) reported that genotype by environment interaction with respect to discriminating ability and representativeness of test environments is a measure of desirability. Therefore, the desirable location must have highly discriminating ability as well as representative of all locations. The center of the concentric circles represents an “ideal” location that is most discriminating for varieties and is representative of all other locations (Yan, 2001; Yan and Kang, 2003).

Table 5: Average cane yield (t ha-1) of 19 sugarcane varieties at seven locations.

Akande, S.R. 2007. Biplot analysis of genotype by environment interaction of cowpea grain yield in the forest and southern guinea savanna agro-ecologies of Nigeria. J. Food Agric. Environ. 5 (3, 4): 464-467.

Akter, A., M.J. Hasan, U. Kulsum, M.H. Rahman, M. Khatun and M.R. Islam. 2015. GGE Biplot analysis for yield stability in multi-environment trials of promising hybrid Rice (Oryza sativa L.). Bangladesh Rice J. 19(1): 1 – 8. https://doi.org/10.3329/brj.v19i1.25213

Crossa, J., P.N. Fox, W.H. Pfeiffer, S. Rajaram and H.G. Gauch. 1991. AMMI adjustment for statistical analysis of an international wheat yield trial. Theor. Appl. Genet. 81: 27–37. https://doi.org/10.1007/BF00226108

Fan, X.M., M.S. Kang, H. Chen, Y. Zhang, J. Tan and C. Xu. 2007. Yield stability of maize hybrids evaluated in multi-environment trials in Yunnan, China. Agron. J. 99: 220-228. https://doi.org/10.2134/agronj2006.0144

FAO. 2018. http://www.fao.org/faostat/en/#data/QC

Farshadfar, E., R. Mohammadi, M. Aghaee and Z. Vaisi. 2012. GGE biplot analysis of genotype × environment interaction in wheat-barley disomic addition lines. Aust. J. Crop Sci. 6(6): 1047-1079.

Gabriel, K.R. 1971. The biplot graphic display of matrices with application to principal component analysis. Biometrika. 58: 453–467. https://doi.org/10.1093/biomet/58.3.453

Gauch, H.G. 1988. Model selection and validation for yield trials with interaction. Biometrics. 44: 705–715. https://doi.org/10.2307/2531585

Jamshaid, M.K. 2017. Minutes of the 9th Meeting of Federal Committee on Agriculture (FCA), Minist. Nat. Food Secur. Res. Islamabad. December 4, 2017.

Kang, M.S. 1998. Using genotype-by-environment interaction for crop cultivar development. Adv. Agron. 62: 199–252. https://doi.org/10.1016/S0065-2113(08)60569-6

Kang, M.S. and D.P. Gorman. 1989. Genotype × environment interaction in maize. Agron. J. 81(4): 662-664. https://doi.org/10.2134/agronj1989.00021962008100040020x

Kang, M.S. and R. Magari. 1996. New developments in selecting for phenotypic stability in crop breeding. pp. 1–14. In M.S. Kang and H.G. Gauch, Jr (ed.) Genotype-by-environment interaction. CRC Press, Boca Raton, FL.

Kempton, R.A. 1984. The use of biplot in interpreting variety by environment interactions. J. Agric. Sci. (Cambridge). 103: 123–135. https://doi.org/10.1017/S0021859600043392

Luo, J., Y. Pan, Y. Que, H. Zhang, M.P. Grisham and L. Xu. 2015. Biplot evaluation of test environments and identification of mega-environment for sugarcane cultivars in China. Sci. Rep. 5: 15505. https://doi.org/10.1038/srep15505

Luo, J., H. Zhang, Z.H. Deng and Y.X. Que. 2012. Trait stability and test site representativeness of sugarcane cultivars based on GGE-biplot analysis. Chin. J. Appl. Ecol. 23: 1319–1325.

Luquez, J.E., L.A.N. Aguirrezabal, M.E. Aguero and V.R. Pereyra. 2002. Stability and adaptability of cultivars in non-balanced yield trials: Comparison of methods for selecting ‘high oleic’ sunflower hybrids for grain yield and quality. J. Agron. Crop Sci. 188: 225-234. https://doi.org/10.1046/j.1439-037X.2002.00562.x

Matus-Cadiz, M.A., P. Hucl, C.E. Perron and R.T. Tyler. 2003. Genotype × environment interaction for grain color in hard white spring wheat. Crop Sci. 43: 219–226. https://doi.org/10.2135/cropsci2003.2190

Mattos, P.H.C., R.A. Oliveira, J.C.B. Filho, E. Daros and M.A.A. Veríssim. 2013. Evaluation of sugarcane genotypes and production environments in Paraná by GGE biplot and AMMI analysis. Crop Breeding Appl. Biotechnol. 13: 83-90. https://doi.org/10.1590/S1984-70332013000100010

Rea, R., O. De Sousa-Vieira, M. Ramon, G. Alejos, A. Diaz and R. Briceno. 2011. Ammi analysis and its application to sugarcane regional trials in Venezuela. Sugar Tech. 13(2): 108-113. https://doi.org/10.1007/s12355-011-0070-8

Reddy, P.S., A. Rathore, B.V.S. Reddy and S. Panwar. 2011. Application GGE biplot and AMMI model to evaluate sweet sorghum (Sorghum bicolor) hybrids for genotype × environment interaction and seasonal adaptation. India. J. Agric. Sci. 81: 438–444.

Sabaghnia, N., H. Dehghani and S.H. Sabaghpour. 2006. Nonparametric methods for interpreting genotype × environment interaction of lentil genotypes. Crop Sci. 46: 1100–1106. https://doi.org/10.2135/cropsci2005.06-0122

Steel, R.G.D. and J.H. Torrie. 1980. Principles and procedures of statistics. 2nd ed. New York: McGraw-Hill.

Westcoff, B. 1987. A method of analysis of the yield stability of crops. J. Agric. Sci. (Cambridge), 108: 267–274. https://doi.org/10.1017/S0021859600079259

Xu, F., F. Tang, Y. Shao, Y. Chen, C. Tong and J. Bao. 2014. Genotype × Environment interaction for agronomic traits of rice revealed by Association Mapping. Rice Sci. 21(3): 133-141. https://doi.org/10.1016/S1672-6308(13)60179-1

Xu, N., M. Fok, G. Zhang, J. Li and Z. Zhou. 2014. The application of GGE biplot analysis for evaluating test locations and Mega-Environment investigation of cotton regional trials. J. Integr. Agric. 13(9): 1921-1933. https://doi.org/10.1016/S2095-3119(13)60656-5

Yan, W. 1999. Methodology of cultivar evaluation based on yield trial data-with special reference to winter wheat in Ontario. Ph.D. Diss. Univ. Guelph. Canada.

Yan, W. 2001. GGE-Biplot a windows application for graphical analysis of multi-environment trial data and other types of two-way data. Agron. J. 93: 1111–1118. https://doi.org/10.2134/agronj2001.9351111x

Yan, W. and I. Rajcan. 2002. Biplot analysis of test sites and trait relations of soybean in Ontario. Crop Sci. 42: 11-20. https://doi.org/10.2135/cropsci2002.0011

Yan, W. and M.S. Kang. 2003. GGE Biplot analysis: A graphical tool for breeders, geneticists and agronomists. CRC Press, Boca Raton, FL.

Yan, W., L.A. Hunt, W.Q. Sheng and Z. Szlavnics. 2000. Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Sci. 40: 597–605. https://doi.org/10.2135/cropsci2000.403597x

Zobel, R.W., M.J. Wright and H.G. Gauch. 1988. Statistical analysis of a yield trial. Agron. J. 80: 388-393. https://doi.org/10.2134/agronj1988.00021962008000030002x

Zubair, M. and A. Ghafoor. 2001. Genotype × environment interaction in mungbean. Pak. J. Bot. 33(2): 187-190.