A SIMULTANEOUS EQUATIONS MODEL OF THE CAUSE-EFFECT RELATIONSHIPS AMONG BIOLOGICAL VARIABLES: AN ANALYSIS USING POSTPARTUM BEEF COWS
A path diagram, as the basic model of a biological system, was used to explore the possible relationships generating observed correlations among physiological and body-trait responses in beef cows. The implications for statistical model specification and estimation were discussed. The model specified was a simultaneous system of equations with the physiological and body-trait responses constituting the set of endogenous variables; length of postpartum anestrous interval (PPI) was chosen as the physiological response of primary interest. The set of predetermined variables representing basic determinants of the biological system included plane of nutrition and obstetrical assistance experiments. From the generalized least-squares parameter estimates it was concluded that basic determinants in common explained much of the correlation among observed values of the physiological and body-trait responses. The cause-effect relationships included a recursive dependency of PPI upon postpartum body condition score and parturition date. In addition, it was concluded from the results that management decisions and events before the experimental period were more important in explaining variation in PPI than were the experimental treatments. Leading to the further conclusion that ignoring or attempting to randomize over such influences in the experimental design may jeopardize experimental results. The system of equations approach to analysis of biological data makes clear the importance of completely thinking through the problem before the experiment, the limitations of ordinary least-squares procedures and should have applications in other biological systems. Key words: Simultaneous equations analysis, anestrus, beef cows, body condition, return interval