As three resources that are necessary for human survival and production, water, energy and food are increasingly closely linked. In recent years, the water-energy-food nexus has attracted special attention from international organizations and academic circles. However, due to the lack of research on its internal mechanisms, there is still controversy on whether the water-energy-food nexus can be used as a new policy basis. The internal mechanisms of the water-energy-food nexus were analysed from the perspective of industrial linkages in this paper and empirically verified by constructing an SVAR (structuralvectorautoregression) model using China’s data. The results showed that there were two forms of conduction in China’s water-energy-food nexus: the water-energy-food nexus with nuclear power participation and that with natural gas participation. The characteristics of China’s water-energy-food nexus were derived. For the interactions of the water-energy segment in China’s water-energy-food nexus, the conduction from energy to water was consistent for different types of energy, while that from water to energy varied depending on the type of energy. Food production always had a negative impact on energy production, while the conduction from energy to food varied for different types of energy. The conduction between food and the water supply was not as significant as was generally considered. Especially, the impact of the water supply on food production was weak. The order of strength intensity and the duration were also available for reference. Accordingly, a new policy basis was presented under the framework of China’s water-energy-food nexus. Both our research design and research findings are significant in contributing to understanding the internal mechanisms of the water-energy-food nexus, and the policy implications are also helpful for achieving better policy effects.

This study sets out to measure CWU (litre/kg ECM, energy-corrected milk) of typical milk production systems in 60 dairy regions from 49 countries representing 85% of the world׳s milk production. The extended version of TIPI-CAL 5.2 including water model was used for data analysis.The results have shown the CWU/kg ECM ranged between 739L on the Danish farm to 5622 l on the Ugandan farm with a global average of 1833L. When looking at averages per region, the CWU was lowest in Europe (913L) and highest in Africa (3384L) with large intra- and inter-regional differences. Compared with grazing and intensive production system, low yielding cows on small-scale farms have the highest CWU/kg ECM. The key driver for variation in CWU/kg ECM is feed, accounting for 94–99% of the total CWU. Increasing milk productivity might be one of the promising ways to reduce CWU/kg ECM. However, this might also lead to the negative impact into water supply systems if this increase is dependent on land irrigation in water scarce areas. The findings of this study showed the need to address the location of the farm, the feed quality, feeding system and milk production intensity simultaneously when aiming at efficient water resource management which would help to contribute food production and livelihood security of dairy farmers worldwide.