Prediction of enteric methane emission from cattle using linear and non-linear statistical models in tropical production systems: Fid-Bau Portal

Prediction of enteric methane emission from cattle using linear and non-linear statistical models in tropical production systems

Patra, Amlan Kumar

Abstract

Abstract The objective of this study was to develop linear and nonlinear statistical models to predict enteric methane emission (EME) from cattle (Bos) in the tropics based on dietary and animal characteristic variables. A database from 35 publications, which included 142 mean observations of EME measured on 830 cattle, was constructed to develop EME prediction models. Several extant equations of EME developed for North American and European cattle were also evaluated for suitability of those equations in this dataset. The average feed intake and methane production were 7.7 ± 0.34 kg/day and 7.99 ± 0.39 MJ/day, respectively. The simple linear equation that predicted EME with high precision and accuracy was: methane (MJ/day) = 1.$ 29_{(±0.906)} $ + 0.$ 878_{(±0.125)} $ × dry matter intake (DMI, kg/day), [root mean square prediction error (RMSPE) = 31.0 % with 92 % of RMSPE being random error; R2 = 0.70]. Multiple regression equation that predicted methane production slightly better than simple prediction equations was: methane (MJ/day) = 0.$ 910_{(±0.746)} $ + 1.$ 472_{(±0.154)} $ × DMI (kg/day) – 1.$ 388_{(±0.451)} $ × feeding level as a multiple of maintenace energy intake – 0.$ 669_{(±0.338)} $ × acid detergent fiber intake (kg/day), [RMSPE = 22.2 %, with 99.6 % of MSPE from random error; R2 = 0.84]. Among the nonlinear equations developed, Mitscherlich model, i.e., methane (MJ/day) = 71.$ 47_{(±22.14.6)} $ × (1 - exp{−0.$ 0156_{(±0.0051)} $ × DMI (kg/day), [RMSPE = 30.3 %, with 97.6 % of RMSPE from random error; R2 = 0.83] performed better than simple linear and other nonlinear models, but the predictability and goodness of fits of the equation did not improve compared with the multiple regression models. Extant equations overestimated EME, and many extant models had low accuracy and precision. The equations developed in this study will be useful for improved estimates of national methane inventory preparation and for a better understanding of dietary factors influencing EME for tropical cattle feeding systems.