Multivariate Genotype and Genotype by Environment Interaction Biplot Analysis of Sugarcane Breeding Data Using R

Abstract

Complexity of Genotype by environment interaction (GxEI) in sugarcane multi-environmental trial (MET) requires further evaluation for genotypes performance determination. Genotype and genotype by environment (GGE) is one of the many statistical techniques for evaluating the interaction with emphasis on genotypes. Many statistical analysis tools for GGE exists with usage depending on cost and knowhow. R open source analytical software ensures availability and the knowledge on the necessary packages is required thus the objective of the paper on utilization of GGE using R software in the evaluation of genotypes with presence GxEI. The application used secondary data of Kenyan Mtwapa series of 96 and 97 preliminary varietal trial stage 4 established under randomized complete block design (RCBD), consisting of 15 test genotypes and three controls in the environments of SONYsugar, Mumias and KibosF9 with the plant crop and ratoon crop cycles as seasons. The 2-way GEI data was handled using singular value decomposition (SVD) through the R package; GGEbiplot programmed scripts and graphical user interface (GUI) were used in ranking genotypes and environments, determining genotypes performance overall and in each environment, determining stabilities and adaptability of the genotypes and identifying mega trial environments. GGEbiplot unpacked the GEI through the principle components (PC) 1 and 2 that sufficiently explained 85.37% of the variations.