Comparing the environmental performance of mixed and specialised dairy farms: the role of the system level analysed

Mixed crop-livestock systems are often considered more environmental friendly compared to specialised systems, but due to the interactions between different farming activities, it is not trivial to quantify possible benefits. Using life cycle assessment (LCA), we tested different allocation procedures and system expansion through avoided burden to compare the environmental impact of milk from either specialised or mixed dairy production systems (product level). In a second approach, we compared the whole farming systems with additive system expansion, where the functional unit comprised milk, live animals sold for meat production and crops (farm level). On the product level, milk from the mixed farm had higher non-renewable cumulative energy demand, terrestrial ecotoxicity and phosphorus use, but lower aquatic eutrophication N, independently of the allocation method. For all other impact categories, differences were not significant. On the farm level, results were partially reversed. The mixed system had a lower energy demand and potassium use, while phosphorus use was higher. All other differences were not significant on farm level. The different rankings on product and on farm level were caused by the way manure was attributed to the farming activities. In order to avoid allocation, manure management was sub-divided into storage and application processes. Storage was attributed to dairy production, application to dairy production only if applied on grassland or feed crops, and to cash crops when applied to produce these crops. Manure applied on cash crop areas was thus out of the scope of the product approach, and mineral fertilisers that could be saved within the cash crop production were thus not attributed to milk production. We conclude that only system expansion was able to cope with the complexity of mixed farming systems in LCA. Based on our results with modelled farms, mixed farming showed the potential to reduce environmental impacts compared to specialised farming. Nevertheless, due to the complexity of the system regarding farm management and interactions between cropping and livestock activities, only an assessment with real farm data could reveal the actual benefits of such systems.