Maize Genetics Cooperation Newsletter vol 81 2007
Sher-e-Kashmir
University of Agricultural Sciences and Technology of Kashmir
A number of studies in maize have been conducted to elucidate the nature of the association between yield and its components which identified traits like ear length, ear diameter, kernels/row, ears/plant, 100-seed weight and rows/ear as potential selection criteria in breeding programmes aiming at higher yield (Debnath and Khan, Pakistan J. Sci. 2nd Res. 34:391-394, 1991; Agrama, Plant Breed. 115:343-346, 1996; Mohan et al., Natl. J. Plant Improv. 4:75-76, 2002; Tollenaar et al., Crop Sci. 44:2086-2094, 2004). The present study was undertaken to elucidate such character association in local and CIMMYT inbred line crosses of maize in the temperate valley of Kashmir. The present study was carried out in 2004-05. The materials were generated by crossing 15 diverse white maize inbred lines (4 local and 11 exotic) to three phenotypically diverse testers (W3, W5 and W3 x W5) in a line x tester design, at trhe winter maize nursery in Amberpet, Hyderabad (India). The parental lines and test crosses were evaluated at two diverse locations of the Kashmir valley, namely Larnoo and Wadura, representing distinct climatic zones. Each genotype was replicated thrice at each location in randomised block design. Each entry was grown in two rows of 2 m length with row to row and plant to plant spacing maintained at 60 and 25 cm, respectively. The recommended practices were followed to ensure a good crop. Data were recorded from 10 randomly selected competitive plants from each replication for 11 quantitative traits, and the data were statistically analysed for correlation coefficients and path analysis as per the methods of Al-Jibouri et al. (Agron. J. 50:633-637, 1958) and Dewey and Lu (Agron. J. 51:515-518, 1959). There was substantial variability for all traits. Grain yield, ear length, ear height, 100-seed weight and ear diameter had high GCV estimates, with high heritability. The genetic advance was higher for plant height, ear length, grain/row and grain yield. The genotypic correlation coefficient revealed that ear diameter, 100-seed weight, ear length, kernel rows/ear and kernels/row had the highest significant correlation with grain yield. The path analysis revealed that the highest direct effect on grain yield was exhibited by 100-seed weight, followed by kernels/row, kernel rows/ear, ear length and ear diameter. Most of the traits exerted their positive indirect effects through 100-seed weight, kernel rows/ear and kernels/row.
Genotypic relationships among traits affecting grain yield elucidate true association as they exclude environmental influences. In the present study, the highest significant positive correlation with grain yield was shown by ear diameter, followed by 100-seed weight. Days to 50% silking and days to pollen shed had significant negative correlation with grain yield. Similar results have been reported in maize by Mohan et al. (2002), Vasic et al. (Acta Agron. Flung. 49:337-342, 2001), Mohammadia et al. (Crop Sci. 43:1690-1697, 2004), Neto and Miranda (Sci. Agric. 58:99016-9018, 2001).
Using the path coefficient analysis revealed positive direct effects on grain yield, with the highest direct effect exhibited by 100-seed weight followed by kernels/row, kernel rows/ear, ear length and ear diameter. Days to pollen shed, days to 50 percent silking and ear height showed negative direct effects on grain yield even though ear height had a positive correlation with grain yield. These traits also shared positive indirect effects on grain yield through other yield traits such as ear length and ear diameter. Ear diameter had the highest indirect effect on grain yield through kernels/row (0.362), followed by ear height (0.316) through rows/ear. In fact, the bulk of the indirect effects on grain yield was exerted by the traits studied through these two traits. Similar results in maize have been reported by Wang et al. (Field Crops Res. 61:211-222, 1998), Vasic et al. (2001), Broccoli and Burak (MNL 74:43-44, 2000), Abdmishani et al. (Maize Genet. Conf. Abst. 46:1-2, 2004) and Mohammadia et al. (2003). Thus, in light of the results obtained in the present study, it can be suggested that the traits such as kernels/row, 100-seed weight, kernel rows/ear, ear length and ear diameter should be used as target traits for improvement of grain yield in maize. Thus, it can be emphasized that the ideal plant type should have higher values of the traits described above, whereas the traits showing negative effects on grain yield should be selected for lower values. In fact, Vasic et al. (2001) used various indices of selection for improvement of grain yield, and were able to show that even with a simple selection for improvement of grain yield.
The conventional path analysis, or the one carried out in the present study, suffers from the limitation of non-independence of predictor variables, often leading to high multicollinearity. In fact, Samonte et al. (Crop Sci. 38:1130-1136, 1998) proposed a sequential path analysis which is based on minimising multicolinearity due to complex interaction of yield component traits, and which delineates the importance of predictor variables into various orders based on their direct effects. Thus multiple regression based path analysis can be improved by stepwise regression analysis by sequentially removing the non-significant predictor variables from analysis. Besides, more and more traits can be included in the path analysis in order to reduce the residual effects.
Please Note: Notes
submitted to the Maize Genetics Cooperation Newsletter may be cited only with
consent of authors.