Factors associated with variation in the fatty acid composition and iodine value of carcass fat in pigs fed increasing levels of dried distillers grains with solubles

The effect of fat sampling location, pig removal program (pigs taken off test at one time [Single-group] vs. pigs taken off test in 6 groups over time [Multiple-groups]), and experimental subsample size (number of pigs/pen selected for IV measurement [2, 4, 6, 8, 10, or 12 pigs]) on fat quality composition was evaluated in a study involving 1,632 pigs. The pigs were reared under commercial conditions from 23.4 ± 1.14 kg to 128.3 ± 2.15 kg BW in groups of 34 animals and were fed diets with a range of DDGS inclusion levels (0, 20, 40, 60%). At the end of the growth period, pigs were harvested at a commercial plant and fat samples were collected from the belly (anterior end), jowl (anterior end), backfat 1 (3rd thoracic vertebra), and backfat 2 (clear plate). Fat composition was measured on subsamples of pigs from each pen. Fatty acid profile was measured using gas chromatography (GC). Iodine value (IV) was either predicted with an equation from AOCS (1998) based on unsaturated fatty acid profile from the GC analysis or measured using near infrared spectroscopy (NIR). Increasing DDGS level increased fat IV linearly for all sampling locations. Removal program had limited to no effect on fat composition and IV. IV measured using either GC or NIR were similar (P > 0.05) for all locations (Pearson correlation = 0.96). Jowl fat IV was higher (P < 0.05) compared to belly, backfat 1 and backfat 2 (5, 2, and 3 g/100g, respectively). The correlations between IV at the 4 sampling locations were relatively high (from 0.86 to 0.92). The development of equations using other sampling locations to predict belly fat IV indicated that using the IV of the fat from jowl, backfat 1, or backfat 2 would give similar precision of prediction of belly fat IV (R2 = 0.78, 0.75, and 0.73, respectively). The DDGS inclusion level and dietary iodine value product (IVP) were highly correlated (i.e., 0.86 to 0.90) with fat IV for all sampling locations. For equations to predict the IV of fat at the sampling locations, DDGS inclusion level was the first variable selected (e.g., jowl fat IV = DDGS level × 0.24 + 72.99; R2 = 0.81). The inclusion of other variables resulted in little or no improvement in R2. DDGS inclusion level was highly correlated with dietary IVP (0.99), so using DDGS inclusion level or dietary IVP to predict pork fat IV resulted in similar R2 values. For Single- and Multiple-groups removal programs, a minimum subsample size of 2 and 4 pigs (closest to the mean pen BW) were needed to represent the mean and standard deviation, respectively, of fat IV. Results of this study can be used to appropriately design studies to evaluate fat composition under commercial conditions