TECHNICAL NOTE: Use of mixed model nonlinear functions to adjust pig growth and measurement data

Current methods to adjust pig performance data to a constant BW or age assume linear relationships between the growth of the measurement and either age or BW. Serial BW and ultrasonic measurements from a set of gilts were used to provide an example of the development of alternative methods to adjust pig performance data. Each pig was weighed at approximately 125, 138 153, 167, and 174 d of age. Ultrasonic backfat depth and loin depth measurements were collected at each weigh day on approximately one-half of the gilts. The BW data were fitted to a mixed model generalized Michaelis-Menten (GMM) function with one pig-specific random effect for mature BW. By including each pig's random effect, the pig's predicted age to achieve target BW was estimated. An allometric equation Y = (1 + ai) A BWB with a pig-specific random effect (ai) provided the best fit of the ultrasonic data to BW. An alternative approach to adjust pig growth or measurement data using mixed model nonlinear equations was suggested. The pig's actual data were used to solve for its specific random effect. Then, using the nonlinear equation and the pig's random effect, the pig's measurement at the target age or BW can be estimated. This alternative approach may better predict the mean and variances of each pig's growth and better adjust pig performance data.