Abstract: Soybean is an alternative stapple food which use by all people in Indonesia. However, domestic soybean production does not meet the needs of the national soybean. To meet the needs done, it important to assemble the soybean breeding to improved the noble varieties. Plant breeding is largely determined by the ability of breeders to select. Selection can be done by estimating the correlation coefficient between the components of the results and outcomes. To determine the causal relationship between yield and yield components used path analysis. Path analysis of correlation coefficient will break into the direct and indirect influence. The purpose of this study are to determine (1) the correlation between the components of the results and outcomes; (2) direct and indirect influence between the components of the results and outcomes. The experiment was conducted from March to July 2014 in field trials of POLINELA (Lampung State Polytechnic) and in Seed Laboratory of University of Lampung. Soybean seed used is the result of crossing Wilis F6 generation x B3570. Treatment arranged in a perfect randomized group design with two replications. The results of correlation analysis showed correlation great and positive value indicated by the component total number of pods per plant and grain weight per plant (0.80 *), hundred grain weight and grain weight per plant (0.72 *), and the number of productive branches and grain weight per plant (0.57 *). Results of the analysis showed cross large direct effect on the character of the total number of pods per plant to grain weight per plant (0.89), which is almost as large as the correlation (0.80 *). Then the correlation coefficient was actually measure the degree of closeness of their relationship. Therefore, the selection is based on the total number of pods will be very effective to get genotype expectations that have high yields.
Penulis: Tibor Eka Saputra, Maimun Barmawi, Ermawati Ermawati, Nyimas Sa`diyah
Kode Jurnal: jppertaniandd160244