Association Analysis and Structured Equation Modelling in Groundnut (Arachis hypogaea L.)
R. Sangeetha Vishnuprabha *
Department of Oilseeds, Centre for Plant Breeding and Genetics, Tamil Nadu Agricultural University, Coimbatore –641003, Tamil Nadu, India.
PL. Viswanathan
Department of Oilseeds, Centre for Plant Breeding and Genetics, Tamil Nadu Agricultural University, Coimbatore –641003, Tamil Nadu, India.
S. Manonmani
Department of Oilseeds, Centre for Plant Breeding and Genetics, Tamil Nadu Agricultural University, Coimbatore –641003, Tamil Nadu, India.
L. Rajendran
Department of Oilseeds, Centre for Plant Breeding and Genetics, Tamil Nadu Agricultural University, Coimbatore –641003, Tamil Nadu, India.
T. Selvakumar
Department of Oilseeds, Centre for Plant Breeding and Genetics, Tamil Nadu Agricultural University, Coimbatore –641003, Tamil Nadu, India.
*Author to whom correspondence should be addressed.
Abstract
India is one among the top countries producing groundnut in the world ranking. The breeding programmes of groundnut aim at developing early maturing and high yielding varieties. Correlation and Path analysis is a tool which provides information on magnitude and intensity of association among yield components and yield. The present study analyses effects of the yield components on yield in the segregating population of groundnut. The traits days to maturity, number of matured pods, kernel yield, hundred pod and kernel weight are positively correlated with pod yield while the vegetative traits height of main axis and number of branches are negatively correlated with pod yield. Path analysis revealed that days to accumulation of 25 flowers, days to maturity, number of matured pods, kernel yield, hundred pod and kernel weight exerted low to moderate direct effects on pod yield. Structural Equation Model (SEM) is a powerful statistical technique for Path analysis that allows for more complex models than multiple regressions with its single dependent variable. In groundnut segregating population, structural equation modelling splits the yield component into traits before harvest and after harvest. Though both the variables are correlated, the traits before harvest have higher influence on pod yield.
Keywords: Legume, correlation, path analysis, structural equation modelling