Energy evaluation of grass silage

Kridis, Mansour S.F (1989) Energy evaluation of grass silage. PhD thesis, University of Glasgow.

Abstract

The areas covered in the literature review include: a) Metabolisable energy as a measure of the nutritive value of grass silages: b) The importance of digestibility as a useful index of nutritive value; c) Factors affecting the digestibility measurements; d) Laboratory methods for predicting the organic matter digestibility of grass silages. The effect of different washing procedures on the losses of organic matter and nitrogen from samples of hay incubated in polyester bags within the rumen of sheep was investigated. For organic matter, post-incubation detergent washing reduces variability without altering the form of the degradation curve. For nitrogen, post incubation detergent washing might remove contaminating bacteria which could otherwise lead to the underestimation of protein degradability. Washing the bags after rumen incubation with domestic washing powder in the washing machine is both cheap and convenient. One hundred and seventy dried samples of grass silages which had been evaluated in vivo for organic matter digestibility (OMD), were collected from different sources around the UK. These sources include: a) Agricultural Development and Advisory Service - 100 silages; b) Rowett Research Institute - 43 silages; c) School of Agriculture, Aberdeen - 27 silages. All silages were subjected to seven laboratory predictors of in vivo OMD, including those used routinely by advisory services in the UK. To avoid between-laboratory differences, each method was performed by a laboratory which makes routine use of the particular method. Of all the 170 silages, 122 silages were selected to derive prediction equations (calibration silages) and the remaining 48 silages were reserved for subsequent validation purposes. The aim of this work was to investigate the robustness of each method as a predictor of in vivo OMD and then explore the possibility of establishing an improved technique which could be used by all of the advisory services in the UK. 4 The in vivo OMD of 122 calibration silages was not precisely predicted by the MADF (R2 = 0.34 and RSD% = 5.1), LIGA (R2 = 0.52 and RSD% = 4.4), NCOMD (R2 = 0.54 and RSD% = 4.3) and PCOMD methods (R2 = 0.55 and RSD% = 4.2). All gave significant between-population differences in the regression equations obtained. The in vivo OMD was more precisely predicted by the rumen liquor methods (NB48 OMD and IVOMD) [R2 = 0.68, 0.74; RSD% = 3.6, 3.2 respectively]. In each case their application require one single regression equation to describe all silage populations. Of the methods tested in this work, the NIR method was 2 the best predictor of in vivo OMD (R = 0.85 and SEC% = 2.5). It also gave single regression line provided that more than five terms were used in the multiple regression equation. The best prediction of in vivo OMD of a blind test of 48 in vivo silages was obtained by the NIR method using a 2 multiple linear regression involving eight terms (R2 = 0.76, SEP% = 2.6). 5 The effect of external factors in the relationship between in vivo OMD and its predictors was investigated. The IVOMD and NCOMD methods were significantly affected by the year of harvest. Cut number was also found to significantly affect the relationship based on the MADF and LIGA methods. The method of ensiling, wilting time, additive application and nitrogen fertilisation were found not to affect the regression equations of any predictor studied in this work. 6 The use of the NIR method to predict in vivo DOMD was examined. This method can predict directly in vivo DOMD, however with lesser precision than predicting in vivo OMD (R2 = 0.64, SEP% = 2.97 and R2 = 0.76, SEP% = 2.6 respectively). The calculation of in vivo DOMD by NIR prediction of in vivo OMD and then measuring ash content, a parameter useful to indicate soil contamination, gave more precise prediction than the direct prediction of in vivo DOMD by NIR (R2 = 0.78, SEP% = 2.43 and R2 = 0.64, SEP = 2.97 respectively).

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Animal sciences.
• Colleges/Schools: College of Medical Veterinary and Life Sciences
• Supervisor's Name: Barber, Dr. Geoffrey
• Date of Award: 1989
• Depositing User: Enlighten Team
• Unique ID: glathesis:1989-72644
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 11 Jun 2019 11:06
• Last Modified: 15 Jun 2021 13:46
• URI: https://theses.gla.ac.uk/id/eprint/72644