Water footprinting – A comparison of methods using New Zealand dairy farming as a case study

A case study is presented to (1) assess the water footprint of New Zealand (NZ) dairy farming in two contrasting regions of Waikato (North Island, non-irrigated moderate rainfall) and Canterbury (South Island, irrigated low rainfall), (2) illustrate differences in water footprint methods and (3) evaluate the suitability of indicators derived from each water footprint method. The water footprint methods (1) water footprint following the Water Footprint Network (WF-WFN), (2) stress-weighted water footprint (WF-Ridoutt), following Ridoutt and Pfister (2010) and Ridoutt et al. (2010), (3) environmental impacts of freshwater consumption expressed in damage to resources (ΔR), damage to ecosystem quality (ΔEQ), and damage to human health (ΔHH) following Pfister et al. (2009), and (4) freshwater ecosystem impacts (FEIs) and freshwater depletion (FD) following and were applied to two average dairy systems in the different regions. Total WF-WFN was 945 and 1084 L H2O/kg fat-and-protein-corrected milk (FPCM) for the average Waikato and Canterbury dairy farm systems, respectively. The Waikato farm system had a higher green WF-WFN, whereas the Canterbury farm system had the highest blue WF-WFN impact, grey WF-WFN, WF-Ridoutt, ΔEQ, FEI, and FD mostly due to groundwater and surface water used for irrigation of pasture. For both dairy farm systems, ΔR and ΔHH were zero. Data collection in the inventory showed similarities between the water footprint methods. The methods, however, all resulted in different total estimates due to (1) the inclusion of how land conversion from native vegetation to agriculture affects freshwater availability in FEI, (2) application of different characterisation factors, and (3) inclusion of a normalisation procedure for WF-Ridoutt. For example, WFc-Ridoutt was 0.011 L H2O-equivalents (eq)/kg FPCM for Waikato and 7.1 L H2O-eq/kg FPCM for Canterbury dairy farm systems whereas corresponding values for FEI were −1.8 L ecosystem-eq H2O/kg FPCM and 14.6 L ecosystem-eq H2O/kg FPCM respectively. Use of catchment-specific characterisation factors are preferred over characterisation factors based on globally-spatial data for WF studies. For FD, we recommend incorporating the concept of sustainable yield for aquifers into the characterisation factor. Summing the different coloured waters into one WF-WFN reduces the relevance and indicators related to regional water scarcity have more meaning for end-users compared to volumetric estimates. The blue WF-WFN impact, WF-Ridoutt, and FEI are all useful indicators but only relate to freshwater availability. In respect to dairy farming, these indicators should be complemented by FD and ΔR and other impact categories which assess water degradation impacts on ecosystems and human health while avoiding double-counting: e.g., ΔEQ, ΔHH, and grey WF-WFN impact or eutrophication potential (EP).