Evaluation of Alternative Methods to Analyze Pig Body Weight Data from Animal Sorting Technologies

Methods to analyze pig BW data collected by an animal sorting technology scale without individual pig identification were evaluated. Data were obtained from 5 grow-finishing groups of pigs from a commercial producer. The daily BW data were ranked and sorted into 3 alternative numbers of groups: percentile (100 groups), one-half percentile (200 groups), and number of pigs. The BW data were fit to a mixed model Bridges function with a random effect for group. The best fit of the BW data was by sorting the daily BW data into 200 one-half percentile groups. There were substantial differences (P < 0.001) in the mean ADG and pattern of BW growth for the 5 replicates. A simple method of forecasting based on the current BW and overall growth curve will not accurately project the age at which pigs will achieve target market BW. Methods combining the overall BW growth curve with the early BW growth of the current group of pigs could be developed. Directly modeling the growth of BW percentiles greatly facilitates the assessment of marketing strategies for animals that are sold in batches.