0 avis
Dairy cows’ diet discrimination through SIFT-MS-based volatolome of different matrices
Archive ouverte
Edité par CCSD ; Wageningen Academic Publishers -
Session 89. Free communications in animal nutrion. International audience. This study tested using SIFT-MS-based volatome of ruminal liquid, urine, faeces, sweat and milk to discriminatediets fed to dairy cows. Sixteen Holstein cows (1.38 ± 0.52 lactations, 74 ± 11 days in milk, 22.6 ± 2.74 kg kg/d ofmilk yield, 627 ± 44.2 kg LW, 1.4 ± 0.20 BCS, mean ± SD) were enrolled in a longitudinal 10 weeks experimentbased on two contrasting diets: 1) High-fibre and low-starch diet (HFibre, n = 8, 12 % starch DM basis), 2) Highstarch(HStarch, n = 8, increasing starch until reaching 31 % in weeks 9 and 10). Cows were sampled at the endof weeks 9 and 10 and the samples stored at -80°C until the analysis. The volatolome was determined throughselected-ion flow-tube mass spectrometry in full scan mode after thawing the samples at 37°C and heating themin water bath (30 min, 45°C). The data was analysed through partial-least square discriminant analysis for eachsampling period independently. The number of ions ranged from 177 to 253 depending on the type of sample.While ruminal liquid, faeces and sweat did not yield predictive models at any time (Q2 ≤ 0.21 in all cases), urineyielded valid models with moderate-to-high predictive ability to discriminate the diet for both weeks 9 (R2 = 0.91,Q2 = 0.74) and 10 (R2 = 0.66, Q2 = 0.51). The milk volatolome also yielded valid models on both weeks 9 and 10,but its predictive ability was low (Q2 = 0.37 and 0.47 for weeks 9 and 10, respectively). Our results suggest thaturine, which is more related to the internal metabolism of the animal than the other matrices, has a better potentialto be used for developing non-invasive techniques to follow the diet and nutritional status of productive animals.
Consulter en ligne
Suggestions
Du même auteur
