Factors influencing the carbon footprint of milk production on dairy farms with different feeding strategies in western Europe

Greenhouse gas emissions from dairy production represent a major source of emissions especially in Western Europe where the sector has grown over the past decade. Different feeding strategies have evolved and there is a need to identify effective mitigation measures. Life cycle assessment (LCA) was used to examine carbon footprints (CFs) of milk production across 71 commercial dairy farms in Ireland, Northern Ireland, England, Spain (Galicia and Basque regions), Portugal and France based on monthly data collection over two years. Emissions up to the farm gate were calculated within a global boundary with both higher tier emission factors (HTEF) applicable in respective countries, and default emission factors (DEF). The global warming potential (GWP) used was the GWP(100) metric, however results were also calculated using GWP(20) for comparison. Functional units were: (i) one tonne fat and protein corrected milk (FPCM); (ii) 1 ha of on-farm agricultural area (FAA); (iii) 1 ha of global agricultural area (GAA). Farms were categorised based on the proportion of time that cows spent grazing. Mean CF per tonne FPCM (FPCM-CF) were 1,129, 1237 and 1519 kg CO(2)e for grazing (>220d grazing; n = 16), mixed (up to 219d grazing; n = 17) and housed farms (0d grazing; n = 38), respectively. housed had the widest range, from 884 to 2494 kg CO(2)e/tonne FPCM, and included the farm with the overall lowest FPCM-CF. housed also had the highest mean CF per ha FAA: 44.1 tonne CO(2)e, followed by mixed (15.2 tonne) and grazing (11.6 tonne). CF (tonne CO(2)e) per ha GAA followed the same ranking: housed (15.1), mixed (9.8) and grazing (9.2). There was no difference in ranking of the feeding strategies using DEF in comparison with HTEF. A stepwise regression analysis identified feed efficiency and age at first calving as important factors in determining FPCM-CF for all farms. Furthermore, N surplus was important for grazing & mixed farms. The proportion of uncovered slurry storage, milk yield per cow and the amount of bought in concentrate per cow were important for housed farms. Wide variation in CFs implies considerable potential for lowering emissions per tonne FPCM and per ha FAA and GAA, but it is imperative that mitigation measures are tailored to feeding strategy.