Background: In the present study, a set of non-linear prediction equation was developed using records of body measurement traits, namely heart girth, punch girth, height at shoulder, height at back, height at fore leg, height at back leg and body length of pigs. Repeated measurement of traits at birth, 15 d, 30 d and weaning (42 d) of 394 piglets were used. Methods: The regression of body weight on body measurement traits showed non-linear relationship based on equation, Y = aXb. Correlation of heart girth with body weight (r2 = 0.968 for male and 0.969 for females; P less than 0.01) was highest among the traits studied. Result: Heart girth was found to be most suited among the traits to predict body weight in an exponential equation, Y = 0.00058 H2.49 and Y = 0.00061 H2.47 for male and females, respectively, explaining 96 percent of variation.
Pile groups under axial loading: an appraisal of simplified non-linear prediction models
