Bridging research and practice using digital twin modeling to mitigate methane emissions across dairy production stages.

This study examined dietary determinants of enteric methane (CH4) emissions in high-yielding Holstein Friesian dairy cows across different physiological stages. Emissions were estimated using the IPCC Tier 2 methodology during peak lactation, the full lactation cycle, and the dry period in two commercial groups with distinct productivity. Group A (38-40 kg milk/day) showed higher peak dry matter intake and fiber content than Group B (32-35 kg milk/day), which had greater ether extract (EE) levels. Peak-lactation CH4 emissions were significantly higher in Group A (P < 0.05), while dry-period values did not differ (P > 0.05). Dietary EE was inversely associated with CH4 output, suggesting a potential mitigation pathway. Phase-specific regression models (adjusted R2 = 0.88-0.93) confirmed diet composition and physiological stage as major drivers of emissions. Digital twin simulations based on these models offer a non-invasive, reproducible tool for predicting emission scenarios, which is particularly valuable in farms where direct measurements are impractical. These findings highlight the feasibility of integrating diet optimization and predictive modeling into herd management strategies, enabling substantial reductions in CH4 emissions while sustaining milk yield and overall productivity in intensive dairy systems.