The Effect of Sorting Wheat or Barley, Based on the Predicted Crude Protein Content, on Physical Characteristics, Feed Processing Characteristics and Nutrient Digestibility
MetadataShow full item record
Variability in the physiochemical profile of cereal grains represents a challenge for the livestock industry. Currently, nutrient values are based on sample averages, ignoring the variation between individual seeds. These experiments were designed to determine if: (1) fractions obtained by an instrument calibrated to separate individual kernels based on predicted crude protein (CP) content have different physical characteristics and/or differ in dry matter digestibility (DMD) and amino acid digestibility (AAD); (2) the grinding method and intensity differentially influences digestibility of each fraction; and (3) hydrothermal treatment effects differ for the individual fractions. The BoMill TriQ (TriQ), which employs near infrared transmittance spectroscopy (NIT), was used to separate individual kernels based on predicted CP content. In the first study, the TriQ was used to sort six independent sources of wheat into ten fractions each. A minimum of 100 kernels from each fraction were randomly selected and used to obtain measurements of length, width, height, area, DGM, perimeter, sphericity, colour (HunterLab), and mass. HunterLab was used to determine L (100 white to 0 black) a (-a green to +a red) b (-b blue to +b yellow). Data were analyzed as a complete randomized design (CRD) using Proc Mixed procedure of SAS 9.4 with the fixed effect being fraction. Physical characteristics were not different among fractions (P > 0.10), except that fractions with lower predicted CP content tended to have greater L* (54.12 vs. 50.95) based on the HunterLab calorimetric approach (P < 0.10). In the second study, two fractions [predicted high CP (HCP) and low CP (LCP)] produced from five independent sources of feed grade wheat or barley, were compared to the unsorted (UNS) grain. Each fraction (UNS, HCP, and LCP) was ground through a 0.375-mm (coarse grind) or a 0.188-mm (fine grind) screen using a hammer mill or a roller mill to produce coarse and fine treatments. The UNS fraction was used to adjust the roller mill to produce ground samples with a similar processing index (w/v) relative to the hammer mill. In vitro DMD (using rumen inoculum; %), starch digestibility (%) and total gas production (TGP; mL) were determined after a 12-h incubation. Data were analyzed independently by grain type including the effects of fraction, grinder type and severity of processing, and their interactions. A split plot iii design was used where main plot was fraction and the subplots were grinder type and severity of grinding. Significance was defined as P < 0.05 and a trend was defined as P > 0.05 and P < 0.10. The TGP (mL) and DMD (%) of barley ground using a hammer mill were greater than when processed using a roller mill (P < 0.05; 59.4 mL ± 2.0 mL and 41.8% ± 1.0%, respectively). A similar response was observed for wheat processed using either a hammer mill or a roller mill (P < 0.05; 63.8 mL ± 1.4 mL and 27.8% ± 1.5%, respectively). Increasing the severity of processing increased TGP (47.4 mL ± 1.96 mL vs 35.9 mL ± 1.98 mL), DMD (P < 0.05; 36.2 % ± 1.03 % vs 29.6 % ± 1.04 %) for barley and for wheat (P < 0.05; 48.9 mL ± 1.48 mL vs 42.7 mL ± 1.44 mL) and DMD (P < 0.05; 36.4% ± 0.82% vs 32.2% ± 0.80%). Sorting individual seeds based on predicted CP content did not affect physical characteristics, DMD, or TGP for either wheat or barley. In the final experiment, the response of fractions to hydrothermal treatment on AAD was assessed. Eight wheat-based and eight barley-based treatment diets were used. These treatments followed a 2 2 2 factorial arrangement where the main factors were processing temperature (low vs. high temperature pelleting), fractions (LCP vs. HCP) and grain sources (two independent sources for each of wheat and barley). Additionally, a nitrogen-free diet was fed to estimate endogenous losses. Sixteen ileal cannulated pigs were fed the diets in six blocks, providing an n = 6 per treatment. Digestibility of amino acids for rations composed of barley had a fraction × temperature of processing interaction; the same was observed for wheat grain except for proline (P = 0.27), glycine (P = 0.16) and histidine (P = 0.46), while trends were observed for phenylalanine (P = 0.07), tyrosine (P = 0.10), isoleucine (P = 0.07), methionine (P = 0.08) and glutamic acid (P = 0.08). Most of the amino acids for wheat and barley did not exhibit differences between fractions in digestibility. This lack of difference for the majority of the parameters was attributed to the similarity in chemical composition between the fractions. The similarity between the fractions was attributed to the inability of the TriQ to separate a source into fractions that were different chemically. The key findings of the experiments were that using NIT to sort on an individual seed basis for predicted CP content did not result in chemical (starch or CP content) or physical variation. Additionally, DMD and TGP did not differ between the fractions produced by NIT. iv The current NIT technology was found to have limitations in its ability to differentiate kernels that may affect chemical, physical or processing traits.
DegreeMaster of Science (M.Sc.)
DepartmentAnimal and Poultry Science
SupervisorPenner, Greg; Beaulieu, Denise
CommitteeTabil, Lope; Buchanan, Fiona
Copyright DateDecember 2018
Grain, Feed, NIR, Swine, Beef.