EMPIRICAL LINEAR REGRESSION MODEL USED TO ESTIMATE METHANE EMISSION IN DAIRY CATTLE

Abstract

Objective: To develop an empirical linear regression model from published data to estimate the CH₄ production in ruminants based on dry matter intake. 

Design/methodology/approach: The present work consisted in conducting a search of scientific articles in SCOPUS and ScienceDirect using the keywords: model, intake, dry matter, methane, estimation.

Results: Two linear regression models are shown, the first with R²=0.073 that explained 73% of the variability of data and a second with a R²=0.95 that explained 95% of the variability to estimate the CH₄ enteric emissions from dry matter intake.

Study limitations/implications: Because the models generated in this study were obtained from data published in the scientific literature, future research is required to validate the estimates in vivo of the proposed empirical linear regression models.

Findings/conclusions: The models presented could be applied to estimate the emissions of CH₄ per animal per day in dairy and beef cattle, although it is possible to estimate the emission of CH₄ in sheep and goats as well, especially the model with a coefficient of determination that explained 73 percent of the variability of the data.