Two on-farm data collection methods to determine dynamics of swine compositional growth and estimates of dietary lysine requirements

An experiment was conducted to evaluate the use of two real-time ultrasound data-collection methods to develop a dynamic assessment of live weight growth, protein and lipid accretions, and lysine requirement curves on two commercial swine operations. For the first method, pigs (40 barrows and 40 gilts) were weighed (ranging from 18 to 121 kg) and scanned ultrasonically to collect backfat depth and longissimus muscle area measurements every 3 wk in the finishing facility on two farms (serial method). For the second method, pigs (200 gilts and 200 barrows) of similar corresponding ages on the same two farms were weighed and scanned on 1 d (mass scan) at three different times (February, April, and May). Thirty-two pigs/sex were measured at approximately the same ages as with the serial scans. Pigs on farm 1 grew faster and had smaller backfat depths and larger longissimus muscle areas (P < 0.01) than those on farm 2, irrespective of method. These measurements were used to predict empty-body protein and lipid contents using nonlinear functions, which then were converted to accretion rates and lysine requirements at each BW. Protein accretion (g/d) and daily lysine requirements increased and then decreased for each sex on each farm and were higher on farm 1 than on farm 2. Data from the individual mass scans had larger standard errors for modeled live weight growth than data from the serial scans. Combining data from the three mass scans yielded growth curves with standard errors similar to those for the curves from the serial scans. For the protein accretion curves, the standard errors of the combined mass scans were approximately 20% lower than the standard errors of the serial scans. The standard errors for the modeled lysine:calorie ratio requirement from the serial scans were approximately 1% of the requirement at each BW. These results indicated that either the serial or mass scan data-collection method is a practical means of determining on-farm growth and daily protein and lipid accretion rates, which can be used to determine the farm-specific lysine requirements of growing-finishing pigs.