Uji Multilokasi Melalui Analisis AMMI Multirespon (Studi Kasus : Penelitian Galur Tanaman Tembakau Madura)

Abstract

AMMI is used to analyze the data of multi location trials. The purpose of this analysis is to classify some stable or specific genotypes. The limitation of this analysis is only able to analyze for one response. Some methods of merged the responses can overcome this limitation. Several methods were considered good enough are range equalization, weighted based on principal component analysis, and division by mean. This study use the data of madura tobacco crop research in 2010-2012 were obtained from multi location trials involving 10 genotypes of tobacco plants and 9 environments. Measurement of responses includes the number of leaves, index of plant, chopped productivity, and quality index. The result of AMMI analysis for the origins responses shows that all genotypes are not stable at all origins responses. Only two genotypes are stable for two responses. The other genotypes grew most specifically expressed in each location. The Results of adaptability genotypes obtained by the combined response shows that just genotype C is stable in all combined responses and the other genotypes are grown in specific locations. Overall from the classification of genotypes, three combined responses are represented almost the entire of origins responses relatively. But from three methods were used, range equalization methods is the best choice for the data of this research.