Effect of summer grazing on carbon footprint of milk in Italian Alps : a sensitivity approach

During the last decades, the Italian Alps were characterized by a very high rate of agricultural abandonment that mainly affected small farms. The remaining farms, especially in the dairy cattle sector, moved towards increasing size, intensified production and abandonment of traditional summer grazing of the herds in high altitude pastures. The aim of the study was to estimate the carbon footprint (CF) of cow milk production in the Italian Alps, comparing the traditional farming system based on summer grazing in the highlands with the emerging more intensive systems. Moreover, a sensitivity analysis was performed to investigate the variations in the outcomes of different allocation methods and of different scenarios of land use change (LUC) emissions for soybean production. A group of 32 dairy farms was analyzed in a “cradle to farm gate” life cycle assessment (LCA); 9 of them transferred the whole herd including lactating cows to the high altitude pastures for three months in the summer season (Summer Grazing ¼ SG) while the remaining 23 (no Summer Grazing ¼ noSG) maintained their lactating cows in the valley barns all over the year. On average the farms had a small herd size (54 61 lactating cows) but a high stocking rate (3.7 2.0 Livestock Unit ha 1). The average milk production was 6206 1892 kg FPCM cow 1 year 1 with a huge difference between the two groups (7017 1445 kg FPCM cow 1 for noSG and 4132 1184 for SG). NoSG farms had also higher feed efficiency than SG ones. The CF values obtained in the baseline scenario were 1.55 0.21 and 1.72 0.37 kg CO2-eq. kg 1 FPCM for, respectively, noSG and SG farms. Considering different allocation methods, no significant differences were observed between the carbon footprint of the two systems. When LUC emissions were accounted for, CF increased especially in noSG farms due to the high amount of concentrate feed purchased. Traditional activity of summer grazing in high pastureland did not show any mitigation effect on the carbon footprint of milk, mainly as a consequence of low milk yield and low feed efficiency. The sensitivity analysis for different allocation methods and for different LUC emissions underlined the fact that one of LCA’s limits is that a change of key assumptions can determine a change of the overall results. In particular including the emissions related to LUC for soybean production significantly affects the final outcome, the amount of feed purchased by the farm becoming one of the major driver of milk CF.