Prediction of the fatty acid composition in backfat of pigs as breeding tool

Pork fat is major constituent of many processed meat products. The quality of these products is largely depending on oxidative stability and consistency of the fat, which in turn are mainly determined by the fatty acid composition of the adipose tissues. It is well known that the fatty acid composition of pig adipose tissue can be influenced considerably by the lipid composition of the feed. Nevertheless, high heritabilities described for the proportion of major fatty acids in pig adipose suggest that genetic factors, controlling e.g. fat synthesis and endogenous desaturation of fatty acids, also may largely influence the fatty acid composition. Therefore breeding for a favorable pig fat composition would basically be feasible. However, estimating genetic parameters and breeding values in order to select those animals with desired traits would require large records of phenotypic data and therefore highly efficient methods to measure the fatty acid composition in high numbers of animals.
It was previously shown that the fatty acid composition of pig backfat can be reliably predicted by using a fibre optic probe to take NIR-scans (NIRFlex N-500, BUCHI, Flawil, Switzerland) at the surface of the subcutaneous fat. As reference method for calibration and validation gas-chromatrographic analysis of the fatty acid composition of 191 animals was performed.
Calibrations were developed using the NIRCal software (NIRCal 5.5, BUCHI) applying the PLS-method. Two thirds of the samples were assigned to the calibration set, one third to the validation set. First derivatives (Buhler Chemical Analytic Package BCAP 5 points, 1st Savitzky-Golay 9 points) and normalization (standard normal variate or full Multiplicative Scatter Correction MSC full ) proved to be the most suitable pretreatments. The coefficient of determination and the standard error of prediction in an independent data set were 0.92, 0.76, 0.92 and 0.62, 0.67, 0.49 %-points for the proportion of saturated (SFA), monounsaturated (MUFA) and polyunsaturated (PUFA) fatty acids, respectively. This rapid method was applied at the backfat of 1964 pigs in the pig performance testing station at Sempach, Switzerland. Although all the pigs received the same feed, the proportion of SFA, MUFA, and PUFA showed a large variation form 31.3-43.3, 43.9-53.3, and 8.7-19.1 % respectively. Based on this data, heritabilities of the proportion of these fatty acid types in backfat were estimated and found to be in a medium to high range with 0.43, 0.42, and 0.49 for the respective fatty acid types (SFA, MUFA, PUFA).
These results showed that the fatty acid composition of pig adipose tissue can be modified by breeding measures. The developed NIR-method to predict pig backfat composition is currently applied as routine method in the pig performance testing station at Sempach in order to estimate breeding values for fat quality of boars employed at the artificial insemination station.