Related Experiment Video
Updated: Dec 23, 2025
11:02
The Use of an Automated System GreenFeed to Monitor Enteric Methane and Carbon Dioxide Emissions from Ruminant Animals
Published on: September 7, 2015
22.8K
Application of Meta-Analysis and Machine Learning Methods to the Prediction of Methane Production from In Vitro Mixed
Jennifer L Ellis1, Héctor Alaiz-Moretón2, Alberto Navarro-Villa3,4
1Centre for Nutrition Modelling, Department of Animal Biosciences, University of Guelph, 50 Stone Road East, Guelph, ON N1G 2W1, Canada.
Animals : an Open Access Journal From MDPI
|April 25, 2020
Summary
Researchers developed models to predict in vitro methane (CH4) production from gas and volatile fatty acid (VFA) data. Machine learning models significantly improved prediction accuracy compared to meta-analysis for ruminant feed screening.
Area of Science:
- Ruminant nutrition and environmental science.
- In vitro fermentation and gas production analysis.
- Data science applications in agricultural research.
Background:
- In vitro gas production systems are crucial for screening ruminant feed ingredients.
- Measuring in vitro methane (CH4) production is not standardized across all systems and studies.
- Accurate prediction of CH4 emissions is vital for environmental impact assessment of animal diets.
Purpose of the Study:
- To develop predictive models for in vitro CH4 production using total gas and volatile fatty acid (VFA) data.
- To identify key factors influencing CH4 production in in vitro ruminant systems.
- To compare the predictive power of meta-analysis and machine learning (ML) approaches.
Main Methods:
- A comprehensive database of 354 data points from 11 studies was compiled.
- Meta-analysis and various ML methodologies (artificial neural networks, support vector regression) were employed.
- Models predicted CH4 production based on total gas, apparent DM digestibility (DMD), pH, feed type, and VFA profiles.
Main Results:
- Meta-analysis models incorporating DMD, total VFA, propionate, feed type, and valerate showed good predictability.
- ML models demonstrated significantly higher predictability than meta-analysis.
- Artificial neural networks and support vector regression yielded similar predictive performance.
Conclusions:
- Developed models can effectively estimate in vitro CH4 emissions, aiding feed ingredient screening.
- Machine learning offers a powerful approach for predicting CH4 production in in vitro systems.
- Further validation on external datasets is recommended to assess the generalization of ML models.
