Validation of enteric methane emissions by cattle estimated from mathematical models using data from in vivo experiments

Several authors have developed equations to estimate methane (CH4) emissions by cattle according to variables such as dry matter and nutrient intake, live weight, or weight gain. Mathematical models using these variables show a large variability of results, being necessary to identify those which provide more precise and accurate predictions. For this reason, the objective of this study was to validate enteric CH4 emissions estimated from mathematical models through a comparison with a database of CH4 emissions obtained from cattle experiments carried out in tropical regions. A database of 495 individual cattle CH4 emissions data (g day-1) obtained from 19 studies in three tropical Latin American countries was built for this study. Results showed that mathematical models developed for cattle in tropical production systems overestimated CH4 emissions when they were compared with our database. The mathematical model with higher precision and accuracy was the one that included dry matter intake and organic matter digestibility in the equation (Equation 7. R2=0.34, Cb=0.94, CCC=0.55, RMSE=60.8%, r=0.58), followed by models that included neutral detergent fiber intake data (Equation 5). Our data did not show a relationship between CH4 emissions and gross energy intake or live weight.