Partitioning of Additive Main Effects and Multiplicative Interaction from Sugarcane Varieties on Multi-environmental Trial

Abstract

Additive main effects and multiplication interaction ( AMMI ) is a statistical technique of multienvironment yield trial. It is used to serve three main purposes; (i) model diagnosis in which the initial model is a part of AMMI (ii) to clearify the complicated genotype by environment interactions, and (iii) improving accuracy of yield estimates (Crossa, 1990). The objective of this study was to analyse and estimate cane and sugar yield and cane CCS value using AMMI analysis. Data from 18 experiments conducted during 1984 to 1989 in which three sugarcane varieties, namely F140, UT1 and CN1, were common in all environments. Results of the analyses indicated that AMMI model can separate a larger porting from the total variation. Percent of sum of square for cane yield were 93.62, 94.01, 92.39 and 93.87 for CCS were 99.01, 89.45, 92.66 and 96.39 for sugar yield were 93.58, 95.02, 89.11 and 92.73 in plant cane, first and second ratoon, and all of three crops respectively. PCA1 score of UT1 was 0.59, 0.17 and 1.03 in cane yield, CCS and sugar yield, respectively. UT1 was rather stable in cane yield and CCS value, but sensitive to environments in sugar yield. F140 had the lowest PCA1 score (0.27) in sugar yield, which was considered stable, but it gave lower yield when compared with CN1 and UT1, respectively.