PATH ANALYSIS OF SEED YIELD COMPONENTS USING DIFFERENT CORRELATION COEFFICIENTS IN SAFFLOWER (Carthamus tinctorius L.)
M. Topal, E. Ozturk* and T. Polat*
Department of Biometry and Genetics, and *Department of Field Crops, Faculty of Agriculture, Ataturk University, 25240, Erzurum, TURKEY
Corresponding author e-mail address: firstname.lastname@example.org
Coefficients of Pearson correlation (r) and Spearman’s rho (rs) among plant height, head diameter, number of seeds/head, oil yield, 1000-seed weight on seed yield in safflower (Carthamus tinctorius L.) was investigated by breaking up direct and indirect effects in this study during 2001 and 2002. Direct and indirect effects estimated with parametric and nonparametric path analysis exhibited similarity but parametric path analysis was preferred to nonparametric path analysis. Because, residual effects of parametric path analysis are lower than that of nonparametric path analysis. Furthermore, determination coefficients of parametric path analysis were higher than that of nonparametric path analysis in both years. The direct effect of oil yield and indirect effect of number of seed/head on seed yield via oil yield were found large by parametric and nonparametric path analysis in 2001/02. It was concluded that oil content, number of seed/head and plant height were important selection characters for seed yield of safflower under drought conditions.
Key Words: parametric and non parametric path analysis, safflower; spearman’s rho