DETERMINATION OF YIELD CONTRIBUTING CHARACTERS IN WHEAT (Triticum aestivum L.) USING DIFFERENT STATISTICAL TOOLS

Abstract

An experiment was conducted at the Research Farm of the Department of Genetics and Plant Breeding of Hajee Mohammad Danesh Science and Technology University, Dinajpur, Bangladesh during 2016-2017 to find out the yield contributing traits in wheat using different statistical methods, to establish the relationship between yield and yield contributing characters and to compare among them. Twelve wheat varieties were used in the experiment and they were collected from Wheat Research Institute, Dinajpur. The statistical tools used were path analysis, multiple regression and stepwise regression. The experimental result demonstrated that the highly preferable traits are no. of tillers/plant, internode distance, spike length, spike weight and seed weight/spike, the moderately preferable traits in wheat yield are awn length, no. of spikelet/spike and 1000 seed weight. Plant height, length of flag leaf and length of grain are also important characters in high yield of plants. Flag leaf angle, breadth of flag leaf and breadth of grain excreted a lower acceptability. Path analysis revealed that the traits seed weight/spike (0.63) exerted highest positive direct effect on grain yield per plant followed by spike weight (0.31), no. of tillers/plant (0.26), internode distance (0.13), spike length (0.09), length of grain (0.07), plant height (0.06), 1000 seed weight (0.04) and length of flag leaf (0.02). Among them only four characters showed significance mark in correlation study, those are no. of tillers/plant (0.31), spike weight (0.78), seed weight/spike (0.82) and 1000 grain weight (0.36). This indicated that simultaneous selection of all these characters was important for yield improvement. The multiple regression revealed that no. of tillers/plant, internode distance, awn length, spike length, no. of spikelet/spike, spike weight and seed weight/spike are the most yield contributing characters. With a unit increase in no. of tillers/plant, internode distance, awn length, spike length, no. of spikelet/spike, spike weight and seed weight/spike, there was an increase in the yield/plant to the extent of 1.0120, 0.3944, -0.6159, 0.9644, -0.5169, 2.1580 and 6.0596, respectively. Therefore, selection would be effective with these characters. Stepwise regression for plant yield indicated that 8 variables, including no. of tillers/plant (1.01375), internode distance (0.41303), awn length (- 0.59656), spike length (0.90622), no. of spikelet/spike (- 0.46203), spike weight (2.21780), seed weight/spike (5.94674), flag leaf angle (- 0.09686) and 1000 seed weight (0.08031) are the best selective traits contributing to wheat yield. Among 15 characters only 5 characters showed reliable findings in all the statistical analyses. So, the highly selectable traits are no. of tillers/plant, internode distance, spike length, spike weight and seed weight/spike and stepwise regression was found to be the most variable selective tool among them.