Fecal sample preparation methods for gas chromatography analysis of fatty acids of ruminants fed different amounts of rumen protected conjugated linoleic acids (CLA)

Aim of this study was to compare three methods for determining fatty acid (FA) profiles in ruminant feces by GC. The first method (J) was based on a mild acid\u2013base treatment directly performed on the dry fecal samples, completing in one step hydrolysis, extraction and methylation of FA. The second method (JEE) was based on acid hydrolysis followed by an accelerated solvent extraction (ASE) of ether extract (EE) and by a mild acid\u2013base catalyzed methylation of FA. The third method (CEE) was based on an acid hydrolysis followed by ASE of EE and by an acid catalyzed methylation of FA (CEE). The experimental design involved the fecal samples of 9 bulls fed a total mixed ration supplemented with 0, 8 or 80 g/d of rumen protected CLA (rpCLA; 3 bulls/dose). Feces collected from these bulls were analyzed, by GC, in triplicates by each method expressing FA contents as mg/g DM. The repeatability of FA and CLA measurements of each method was determined. For the content of CLA isomers, the methods presented heteroscedastic residual variances and, thus, were compared by linear regression. Within method, fecal contents of CLA were regressed against the rpCLA dose. The F-test was employed to test the significance of any slope that deviated from unity and any intercept that was different from zero. There were no differences among methods for the total amount of FA extracted, which averaged 24.55 mg/g DM. The J method was the most repeatable method for most single FA, and for the sums of SFA, MUFA and PUFA. The two EE-methods evidenced for C18:2c9,t11 CLA and C18:2t10,c12 CLA linear relationships with slopes and intercepts close to 1 and 0, respectively, whereas the relationships of JEE and CEE with J had slopes lower than unity. With increasing rpCLA dosage the EE-based methods provided lower increase of fecal contents of C18:2c9,t11 CLA and of C18:2t10,c12 CLA isomers and a higher C18:2t9,t11 CLA content than J, probably as a result of a modification of cis-trans isomerism caused by acid hydrolysis and ASE. The J method should be preferred as it provides more repeatable measures of the fecal FA profiles and because it causes a lower shift in CLA isomer composition with respect to procedures based on acid hydrolysis and ASE