0 avis
Estimation of individual CH4 emissions using fecal near infrared spectra for young, dairy and beef cattle
Archive ouverte
Edité par CCSD ; Wageningen Academic Publishers -
Session 45. Indicators, hi-tech solutions, and tools to effectively manage cattle. International audience. Last years several indirect methods/proxies have been investigated to predict individual enteric CH4 emissionsfrom cattle without using reference measurement methods which are expensive and time-consuming. However,there is a lack of proxy allowing to take into account individual variabilities for non-lactating animals: youngcattle, dry cows or beef cattle. Faeces analysed by near infrared (NIR) spectrometry are relevant to considerbecause such spectral information has already demonstrated their interest to predict forages dry matter intakeand digestibility. Otherwise, CH4 emissions can also be correlated to these last parameters. A dataset of 851reference values has been constituted including CH4 values measured with Greenfeed system averaged on 14days (274±135 g CH4/day) and corresponding spot fecal NIR spectra (sampled on day 14). 256 data were fromlactating cows (Holstein), 249 data from young cattle (Holstein heifers, young Belgian blue bulls) and 346 datafrom beef cattle (Charolais, dual purpose Belgian blue, Belgian blue cows, beef-on-dairy crossbred bulls). Thesecategories of animals have been used to develop independent and common prediction models (10% of animals percategory discarded for independant validation (P)). The PLS model including all categories of animals showed thebest statistics: RPD=3, R²P=0.87 and SEP=48 g/d, demonstrating the great potential of this approach even if therobustness needs to be improved.
Consulter en ligne
Suggestions
Du même auteur
