Competitive resources in dairy cattle feeding: effect on yields, efficiency and emissions

The use of resources (feed, land) competitive with direct human use has been increasingly questioned in terms of appropriateness and food-system level efficiency. Additionally, livestock systems are pursuing more environmentally friendly configurations. The aim was to assess the effect of the inclusion of competitive feed resources in lactating dairy cow’s diets on milk energy yield (MEY), net gross energy (GE) provision, and greenhouse gas emissions (GHGe). Data were collected from 336 farms, representative of northern Italy systems. Farm visits were performed to obtain data on farm size, milk yield and quality, animal diets as well as management and structure to compile farm inventory. Daily MEY was computed as MY x GE content (NASEM protocol). Lactating cows’ dry matter intake (DMI) was computed based on ratio between net energy requirement and dietary net energy. The GE intake (GEI) was computed as DMI x dietary GE content, considering the whole diet (GEIw), contribution due to concentrates target of land occupation (GEIc) and to human-edible portion of the diet (GEIe). Net provisions were calculated as the difference between MEY and GEIw (NEPw), GEIc (NEPc) or GEIe (NPEe). Life Cycle Assessment was applied to compute GHGe (as 100-y global warming potential, GWP) per kg of fat- and protein-corrected milk (FPCM), considering cradle-to-farm gate boundaries, biophysical milk-meat allocation and IPCC computation protocol. Classes of percentage of GEI covered by GEIc (GEIc_p) and GEIe (GEIe_p) (quintiles, average values of 4(6), 17(19), 24(25), 31(30), and 42(39)%, for quintiles from 1 to 5 of GEIc (GEIe)) were used in two separate mixed models to test their effect on MEY, NEPs and GWP. Quartile of farm size, lactation stage and percentage of specialized breeds as well as use of total mixed rations (yes vs no), housing (tie vs loose stall) and random effect of the dairy system were used as additional effects. Average size was 101 lactating cows (min-max: 3-1009), MY 25.7 kg/cow/d (8.5‒44.5) and MEY 78.2 MJ/cow/d (25.3‒135.5). Net provision (mean±sd) averaged -345±38 (NEPw), -23±48 (NEPc), -21±37 (NPEe). Mean GWP was 1.48±0.28 kg CO2-eq/kg FPCM. The GEIc_p significantly affected MYE, NPEs and GWP. In particular, MYE and GWP in first quintile were significantly lower (MYE) and higher (GWP) than 3rd-5th quintiles, with intermediate values in 2nd quintile. About NEPs, increasing GEIc_p was associated with a linear increase in NPEw and decrease in NPEc, with values above zero in the first two quintiles (i.e. positive net provision of energy). Same results were found for GEIe_p. In conclusion, diets with GEIc_p close to 17% and GEIe_p to 19% resulted in the best combination of high MYE, low GWP and positive energy net provision when feed origins and types were considered. Moreover, although not completely overlapping, concentrates could be used as proxy of human-edible portion of the diets, allowing faster and easier evaluation of the lactating cows’ diets.