L.F. Wang

A novel approach for a functional group to predict protein in undigested residue and protein digestibility by mid-infrared spectroscopy.

To evaluate nutrient digestibility, we propose the novel approach of functional group digestibility (FGD). The FGD was based on the absorbance of specific Fourier transform infrared (FT-IR) peaks and the ratio of an inorganic indigestible marker in diet and digesta, without calibration. For application, samples of diet and digesta of wheat with predetermined crude protein (CP) digestibility were scanned on an FT-IR spectrometer equipped with a single-reflection attenuated total reflection (ATR) attachment. The FGD in the amide I region (1689–1631 cm−1) of digesta spectra was strongly related (R 2 = 0.99) with CP digestibility. The measured diet CP digestibility ranged from 60.4 to 87.8% with a standard error of prediction of 1.09%. In conclusion, instead of predictions based on calibrations, FGD can be calculated directly from spectra, provided the ratio of marker in diet and undigested residue is known, and then accurately predicts nutrient digestibility.

Evaluation of energy digestibility of canola coproducts by in vitro analyses and characterization of fat digestion using spectroscopy1

In vitro methods did not accurately predict in vivo apparent total tract digestibility of energy for canola coproducts (r = -0.81; P = 0.002). We scanned 153 samples of digesta, feces, and in vitro digestion residues (ivR) on a Fourier transform midinfrared instrument with a single-reflection attenuated total reflectance attachment. The second derivative spectral net intensities of the carbonyl ester peak at 1745 cm(-1) and olefinic hydrocarbon (=C-H) peak at 3008 cm(-1) were both higher in ivR than in feces (3.83 × 10(-04) vs. 3.46 × 10(-05) and 7.92 × 10(-05) vs. 5.17 × 10(-06), respectively; P < 0.001), indicating poor enzymatic digestion of unsaturated fat. In conclusion, fat digestion of in vitro procedure for canola coproducts requires improvement to adequately mimic in vivo digestion in pigs.

Development and validation of a spectroscopic method to predict wheat protein digestibility

The CP digestibility is traditionally measured by chemical analyses of CP and marker concentration in digesta and diets. Potentially, CP digestibility can also be predicted by marker concentrations and spectral analyses of digesta and diet. Spectroscopy is a rapid, nondestructive method to ascertain qualitative and quantitative chemical information. Based on Beers law, a spectroscopic method was developed to predict in vivo CP digestibility. To validate, samples of digesta and diet of wheat grain with predetermined apparent ileal digestibility (AID) of CP were scanned on a Fourier transform midinfrared (FTIR) instrument with a single-reflection attenuated total reflectance attachment. The AID of CP was calculated from peak intensities of spectra and measured marker concentrations in digesta and diet and then compared with in vivo AID of CP. The AID of CP of a wheat-based diet was predicted accurately with a deviation of 0.68 ± 0.86% from in vivo AID of CP ranging from 60.4 to 87.8%. Functional group digestibility based on the peak at 1,643 cm(-1) or the Amide I region was strongly correlated (r ≥ 0.99; P < 0.001) with in vivo AID of CP. In conclusion, instead of predictions based on calibrations, CP digestibility can also be potentially predicted directly from FTIR spectra.

Feed preference of weaned pigs fed diets containing soybean meal, Brassica napus canola meal, or Brassica juncea canola meal

Brassica napus and Brassica juncea canola meal (CM) may replace soybean meal (SBM) in pig diets, but differ in fiber, glucosinolates content and profile. Preference of weaned pigs provided double-choice selections to diets containing 20% SBM, B. napus CM, or B. juncea CM was evaluated in two studies. In experiment 1, 216 pigs (9.4 ± 1.6 kg initial BW) were housed in 27 pens of 8 pigs (four gilts and four barrows). In experiment 2, 144 pigs (8.9 ± 1.1 kg) were housed in 36 pens of 4 pigs (two gilts and two barrows). Pigs were offered three dietary choices: B. napus CM with SBM as reference (B. napus CM [SBM]), B. juncea CM with SBM as reference (B. juncea CM [SBM]), and B. juncea CM with B. napus CM as reference (B. juncea CM [B. napus CM]) in a replicated 3 × 3 Latin square. Diets were formulated to provide 2.4 Mcal NE/kg and 4.5 g standardized ileal digestible Lys/Mcal NE and were balanced using canola oil and crystalline AA. Each pair of diets was offered in two self-feeders per pen as mash (experiment 1) or pellets (experiment 2) during three test-periods of 4-d, followed by a 3-d non-test period when a common diet was offered in both feeders. Feeders with different diets were rotated daily among pens during preference periods for both experiments, and feeder positions (right or left) were switched daily in experiment 2. Prior to the study and between periods, pigs were fed non-test diets containing SBM (experiment 1) or without test feedstuffs (experiment 2). Overall in both experiments, pigs preferred (P < 0.001) SBM over B. napus and B. juncea CM diets, and preferred (P < 0.001) B. napus over B. juncea CM diet. Dietary choice did not affect (P > 0.05) growth performance in both experiments, except for greater G:F (P < 0.05) for pigs fed the B. juncea CM [B. napus CM] diets than pigs fed the B. napus CM [SBM] or B. juncea CM [SBM] diets in experiment 1. In conclusion, weaned pigs preferred SBM over CM diets when given a choice, and preferred B. napus over the B. juncea diet that contained more total glucosinolates especially gluconapin. Weaned pigs fed the B. juncea CM [B. napus CM] diets in the double-choice selection did not reduce feed intake, weight gain, and G:F compared to pigs fed the B. napus CM [SBM] or B. juncea CM [SBM] diets.

A noncalibration spectroscopic method to estimate ether extract and fatty acid digestibility of feed and its validation with flaxseed and field pea in pigs

Digestibility of ether extract (EE) or fatty acids (FA) is traditionally measured by chemical analyses for EE or GLC methods for FA combined with marker concentration in diet and digesta or feces. Digestibility of EE or FA may be predicted by marker concentrations and spectral analyses of diet and digesta or feces. On the basis of Beers law, a noncalibration spectroscopic method, which used functional group digestibility (FGD) determined with marker concentration and peak intensity of spectra of diets and undigested residues (digesta or feces), was developed to predict the apparent ileal digestibility (AID) of total FA and apparent total tract digestibility (ATTD) of EE. To validate, 4 diets containing 30% flaxseed and field pea coextruded with 4 extruder treatments and a wheat and soybean basal diet with predetermined AID of total FA and ATTD of EE were used. Samples of ingredients, diets, and freeze-dried digesta and feces were scanned on a Fourier transform infrared (FT-IR) instrument with a single-reflection attenuated total reflection (ATR) accessory. The intensity of either the methylene (CH2) antisymmetric stretching peak at 2,923 cm(-1) (R(2) = 0.90, P < 0.01) or the symmetric stretching peak at 2,852 cm(-1) (R(2) = 0.86, P < 0.01) of ingredients, diet, and digesta spectra was related strongly to the concentration of total FA. The AID of total FA of diets measured using GLC was predicted by the spectroscopic method using FGD at 2,923 and 2,852 cm(-1) (R(2) = 0.75, P < 0.01) with a bias of 0.54 (SD = 3.78%) and -1.35 (SD = 3.74%), respectively. The accumulated peak intensity in the region between 1,766 and 1,695 cm(-1) of spectra was related to EE concentration in ingredients and diets (R(2) = 0.61, P = 0.01) and feces (R(2) = 0.88, P < 0.01). The relation was improved by using second-derivative spectra of the sum of peak intensities at 1,743 and 1,710 cm(-1) for ingredients and diets (R(2) = 0.90, P = 0.01) and at 1,735 and 1,710 cm(-1) for feces (R(2) = 0.92, P < 0.01). The ATTD of EE of test diets determined with proximate analysis was estimated by the FGD of nonderivative spectra with or without baseline (R(2) = 0.90, P < 0.01) with a bias of 3.15 (SD = 3.14%) and 3.50 (SD = 3.24%), respectively. In conclusion, instead of using GLC methods or predictions based on calibrations, the AID of total FA and ATTD of EE can also be estimated directly from ATR FT-IR spectra, provided the ratio of marker in the diet and undigested residue is known.