Relation of 40K Count and Probe to Carcass Composition Changes in Growing and Finishing Swine

Thirty Yorkshire and seventy Hampshire barrows were slaughtered (n = 20 at each weight) at live weights of 45.4, 68.2, 90.9, 113.6 and 136.4 kg and evaluated for carcass merit using two different techniques; potassium-40 and Duncan Lean Meter live backfat probe. Each pig was taken off feed and 40K counted at each weight. Upon reaching the predetermined slaughter weight, each pig was probed using the Duncan Lean Meter, 40>K counted for the final time, and then slaughtered for carcass evaluation. Standard carcass measurements were obtained, the right half of each carcass was separated into lean, fat and bone and lipid-free lean was determined by ether extraction of the lean. As slaughter weight increased up to 113.6 kg the 40K counted at each weight, each pig was probed using the Duncan Lean Meter, count of the Yorkshires appeared to increase in a linear fashion; however, a non-linear (P<.01) response was found in the Hampshire data across five slaughter weight groups. Simple regression studies across slaughter weights indicated that weight alone accounted for 96.1% of the variance when predicting lipid-free lean in Yorkshires and 93.6% of the variance in the Hampshires. Potassium-40, when used alone in a regression equation, accounted for 91.3 and 85.7% of the variation, while live backfat probe was associated with 73.1 and 76.2% of the variation when predicting lipid-free lean in the Yorkshires and Hampshires, respectively. Multiple regression equations involving live weight, 40K count, and Duncan Lean Meter live backfat probe were developed to estimate kilograms of lipid-free lean and accounted for 98.2 and 96.3% of the variation in this trait with standard errors of estimate of 1.78 and 2.04 kg for Yorkshires and Hampshires, respectively.