Withdraw
Loading…
Development and validation of prediction equations for the metabolizable energy content of distillers dried grains with solubles from different sources for pigs
Mendoza, Omarh
Loading…
Permalink
https://hdl.handle.net/2142/46610
Description
- Title
- Development and validation of prediction equations for the metabolizable energy content of distillers dried grains with solubles from different sources for pigs
- Author(s)
- Mendoza, Omarh
- Issue Date
- 2014-01-16T17:56:07Z
- Director of Research (if dissertation) or Advisor (if thesis)
- Ellis, Michael
- Doctoral Committee Chair(s)
- Ellis, Michael
- Committee Member(s)
- Stein, Hans H.
- Pettigrew, James E.
- Gaines, Aaron M.
- Department of Study
- Animal Sciences
- Discipline
- Animal Sciences
- Degree Granting Institution
- University of Illinois at Urbana-Champaign
- Degree Name
- Ph.D.
- Degree Level
- Dissertation
- Keyword(s)
- Pigs
- Dried Distiller's Grains with Solubles (DDGS)
- Metabolizable Energy
- Prediction Equations
- Validation
- Abstract
- Three experiments were conducted to determine the apparent DE and ME of samples for distillers dried grains with solubles (DDGS) from 17 different sources, either unground (665.8 ± 284.4 µm) or ground to a common particle size (337.5 ± 39.0 µm). The experiments were conducted simultaneously using an incomplete block design to determine the apparent DE and ME of samples of DDGS as follows: Exp. 1 used 18 dietary treatments, a corn-based control diet (common to all experiments) and 17 diets composed of each of 17 DDGS samples unground. Exp. 2 used 17 dietary treatments, using 15 DDGS samples ground and two unground (from Exp. 1); Exp. 3 used 5 dietary treatments using one source with 5 different particle sizes (1,557, 1,180, 890, 560, and 351 µm). All results are expressed on a DM basis unless otherwise noted. Results for Exp. 1 showed that the mean values for DE and ME of unground DDGS samples were 3,842 ± 116.3, and 3,596 ± 108.4 kcal/kg, respectively. For Exp. 2, mean values for DE and ME of ground DDGS samples were 3,954 ± 117.7, and 3,719 ± 122.5 kcal/kg, respectively. In addition, data from Exp. 1 and 2 were combined to evaluate the effect of DDGS source and particle size, and the two-way interaction. There were no important interactions, suggesting that the effect of particle size reduction was constant across DDGS samples. There was an effect (P < 0.01) of DDGS source on DE and ME, in addition to an effect of particle size, with the ground DDGS samples having 134 and 144 kcal/kg greater (P < 0.01) DE and ME, respectively, compared to the unground DDGS samples. In Exp. 3, reducing particle size in a single sample of DDGS resulted in no difference in DE, however, grinding the sample to the lowest particle size (351 μm) resulted in a 234 kcal/kg increase (P < 0.05) in ME, compared to particle sizes of 560, 890, 1,180, and 1,557 μm. The data generated in these experiments was used, along with the chemical composition (CP, crude fat, crude fiber, ADF, NDF, ash, and starch) of each DDGS sample (analyzed by 2 laboratories), and GE and particle size to develop regression equations to predict the ME of DDGS based on chemical composition and particle size. Regression equations to predict the ME of DDGS were developed using the PROC REG procedure of SAS. A series of equations were developed with those producing the greatest R ̅2 values being selected. For Laboratories 1 and 2, R ̅2 values were maximized using a 4-variable equation, however, different chemical components were included in the equation for each of the laboratories (crude fiber, ADF, NDF, and GE for the equation based on Laboratory 1; CP, crude fat, NDF, and starch for Laboratory 2; with R ̅2 values 0.79 and 0.75, respectively). For validation purposes a separate experiment was conducted to determine the apparent ME of DDGS samples from 4 sources to check the accuracy of the selected equations. The root mean square error of prediction (RMSEP) and mean percent bias were used as criteria to evaluate the accuracy of the equations. The major finding was that the most accurate prediction of ME of DGGS was achieved when the same analytical laboratory was used both for the chemical analysis of the original samples used to develop the prediction equations and also for the analysis of the samples being evaluated (i.e., the samples for which ME was being predicted). This research, also, highlighted the need to develop standard procedures for the development and validation of equations to predict the energy concentration of DDGS and other ingredients, which is essential if users of equations are to have accurate predictions of energy value of feedstuffs.
- Graduation Semester
- 2013-12
- Permalink
- http://hdl.handle.net/2142/46610
- Copyright and License Information
- Copyright 2013 Omarh Mendoza
Owning Collections
Graduate Dissertations and Theses at Illinois PRIMARY
Graduate Theses and Dissertations at IllinoisManage Files
Loading…
Edit Collection Membership
Loading…
Edit Metadata
Loading…
Edit Properties
Loading…
Embargoes
Loading…