VALIDATION OF PORTABLE NEAR INFRARED REFLECTANCE SPECTROSCOPY (NIRS) FOR THE EVALUATION OF FEEDS AND FODDERS IN THE DEVELOPMENT OF RUMINANT PRODUCTION SYSTEMS
Loading...
Date
2015-04
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
SRI VENKATESWARA VETERINARY UNIVERSITY, TIRUPATI – 517 502, ANDHRA PRADESH
Abstract
ABSTRACT:
The present study was conducted to develop the database on nutritive values of conventional and unconventional feeds and fodders for ruminants in India using source information. Nutrient composition of feeds is an essential pre-requisite for scientific feeding of livestock. Knowledge on feed resources including unconventional and lesser known ones in combination with the use of innovative and appropriate technologies like Near Infrared Reflectance Spectroscopy (NIRS) strengthens the production systems. The database created in this study can be a readyreckoner for end users.
In this study, a portable Near Infrared Reflectance Spectroscopy instrument was used for the rapid prediction of nutritive values of Sorghum fodder and Sorghum silage by developing calibration models. The NIR spectra of 200 samples of fresh and dried Sorghum fodder grown in Rayalaseema region of Andhra Pradesh were obtained by collecting spectra from 350-2500 nm range by using LabSpec 4 NIRS instrument. Further, a total of 200 small quantities of Sorghum silage were prepared in the laboratory by using the corresponding Sorghum fodders and spectra were collected between 1100-2500 nm for both fresh and dried samples.The raw spectra were imported to Unscrambler 10.3 X(Camo software) from Indico pro and were organized in four data matrices (fresh Sorghum, dried Sorghum, fresh Sorghum
silage, and dried Sorghum silage) and subjected to data pre-processing methods like Spectroscopic, Standard Normal Variate (SNV), Detrending and Savitzky Golay to explore the spectra quality and control of noise and random errors. The Kennard- Stone (KS) sample selection technique through Principal Component Analysis (PCA) was performed to select 50 calibration samples each for fodder and silage and were analyzed by standard laboratory procedures.The pre-treated spectra along with the laboratory values were analysed by Partial Least Square (PLS) regression technique by using Unscrambler 10.3 X(Camo software) and calibration models were developed for the nutritional attributes.
The R2 coefficients of 0.57, 0.92, 0.77, 0.85, 0.88, 0.43, 0.67, 0.60 and 0.60 whereas the 1-VR values were 0.40, 0.83, 0.04, 0.28, 0.07, 0.04, 0.18, 0.05 and 0.13 for DM, CP, TA, NDF, ADF, ADL, cellulose, silica and hemicellulose, respectively for fresh Sorghum fodder samples. Among the calibration models developed for different chemical constituents in fresh Sorghum samples a good and ‘Satisfactory’ model was obtained for prediction of CP content with R2 (0.92) and 1-VR (0.83) in unknown samples. Next to the prediction of CP content, calibration models pertaining to ADF (R2=0.88), NDF (R2=0.85), TA (R2=0.77), cellulose (R2=0.67), silica (R2=0.60) and hemicellulose (R2=0.60) were in the order of their predictive ability.
The R2 coefficients of 0.80, 0.97, 0.77, 0.82, 0.75, 0.49, 0.55, 0.69 and 0.32 whereas the 1-VR values were 0.75, 0.91, 0.48, 0.70, 0.25, 0.04, 0.09, 0.13 and 0.10 for DM, CP, TA, NDF, ADF, ADL, cellulose, silica and hemicellulose, respectively for dried Sorghum fodder samples. Among the calibration models developed for different chemical constituents in dried Sorghum fodder samples a ‘robust’ and ‘Satisfactory’ model was obtained for prediction of CP content with R2 (0.97) and 1- VR (0.91) in unknown samples. Next to the prediction of CP content, calibration models pertaining to NDF (R2=0.82), DM (R2=0.80), TA (R2=0.77), ADF (R2 =0.75) and silica (R2=0.69) were in the order of their predictive ability.
The R2 coefficients of 0.48, 0.81, 0.88, 0.75, 0.63, 0.36, 0.66, 0.46 and 0.75 whereas the 1-VR values were 0.21, 0.44, 0.72, 0.40, 0.34, 0.04, 0.33, 0.08 and 0.46 for DM, CP, TA, NDF, ADF, ADL, cellulose, silica and hemicellulose, respectively for fresh Sorghum silage samples.Among the calibration models developed for different chemical constituents in fresh Sorghum silage samples a ‘Satisfactory’ model was obtained for prediction of TA content with R2 (0.88) and 1-VR (0.72) in unknown samples. Next to the prediction of TA content, calibration models pertaining to CP (R2=0.81), NDF (R2=0.75), hemicellulose (R2=0.75), cellulose (R2=0.66) and ADF (R2=0.63) were in the order of their predictive ability.
The R2 coefficients of 0.31, 0.81, 0.71, 0.43, 0.71, 0.71, 0.82, 0.67 and 0.71 whereas the 1-VR values were 0.12, 0.65, 0.53, 0.13, 0.30, 0.18, 0.55, 0.40 and 0.28 for DM, CP, TA, NDF, ADF, ADL, cellulose, silica and hemicellulose, respectively for dried Sorghum silage samples. Among the calibration models developed for different chemical constituents in dried Sorghum silage samples a ‘Satisfactory’ model was obtained for prediction of cellulose content with R2 (0.82) and 1-VR (0.55) in unknown samples. Next to the prediction of cellulose content, calibration models pertaining to CP (R2=0.81), TA (R2=0.71), ADF (R2 =0.71), ADL (R2=0.71), hemicellulose (0.71) and silica (R2=0.67) were in the order of their predictive ability.
Sorghum based complete rations were formulated for the study area based upon the prediction values obtained for CP content in Sorghum fodder and Sorghum silage to the nearest accuracy by NIRS and by using linear programming method. The complete rations suggested to the farmers are on the basis of CP and TDN requirements and are formulated for lactating cows, buffaloes and sheep.
Hence, portable Near Infrared Reflectance Spectroscopy (NIRS) which is rapid, non labour intensive, relatively inexpensive in the long term and simplified technique can predict the nutritive values of unknown Sorghum fodder and Sorghum silage samples under field conditions quickly than depending on laboratory methods which are laborious and costly. The outcome of the study helps in formulation of Sorghum fodder and or Sorghum silage based complete rations effectively in a quick manner helping the farmers to take quick decisions regarding feeding their livestock and strengthening the ruminant production systems.
Description
THESES
Keywords
INFRARED REFLECTANCE SPECTROSCOPY; RUMINANT; NIRS; FEEDS; FODDERS; VALIDATION