Eliminating prediction bias in CO2 emission models for lactating cows by incorporating feed intake: accurate quantification of methane-reducing effects using a CO2-based method, demonstrated by a case study on 3-nitrooxypropanol.

objectiveThe methane (CH4) emission prediction method, using predicted CO2 emissions and the CH4:CO2 concentration ratio, faces challenges in evaluating the efficacy of CH4-reducing feed additives due to CO2 prediction bias associated with energy utilization efficiency. We hypothesized that incorporating dry matter intake (DMI), along with metabolic body weight (MBW) and energy-corrected milk (ECM) as explanatory variables, would reduce this bias. The primary objective was to compare the performance of CO2 emission models with and without including DMI. The secondary objective was to assess the CO2-based method's applicability for quantifying CH4-reducing effects, through a case study of 3-nitrooxypropanol (3-NOP).

methodsPrediction models for CO2 emissions were developed including DMI, MBW, and ECM as explanatory variables, based on 219 records obtained from previous experiments with Holstein cows using respiration chambers or headboxes. The model performance was evaluated using cross-validation. Bias associated with energy utilization efficiency was assessed. The applicability of the CO2-based method to quantify the CH4-reducing effect of 3-NOP was assessed using data obtained from the literature, including 10 studies with 22 treatment and control mean comparisons. The agreement between the observed and predicted CH4 reductions was assessed.

resultsCombining DMI along with MBW and ECM improved the predictive performance of CO2 emissions. While the models without DMI showed bias associated with energy utilization efficiency, it was eliminated when DMI was incorporated. Applicability assessment demonstrated that the models without DMI systematically underestimated the CH4-reducing effect of 3-NOP. In contrast, the models with DMI showed smaller discrepancies between observed and predicted CH4 reductions.

conclusionThis study highlights the importance of incorporating DMI as an explanatory variable to achieve accurate and unbiased predictions of CO2 emissions. These findings would contribute to the appropriate application of the CO2-based method for evaluating the CH4-reducing effects of feed additives.