A framework to apply water footprinting for sustainable agricultural water management : a case study on the Steenkoppies Aquifer

Abstract

This study was conducted to better understand the usefulness of water footprint (WF) information for vegetable crops produced on a water-stressed aquifer in South Africa. Different methodologies were investigated in a literature review and the methodologies proposed by the Water Footprint Network (WFN), the Life Cycle Assessment (LCA) communities, and the hydrological-based WF community were selected for a case study on the cultivation of carrots (Daucus carota), beetroot (Beta vulgaris), cabbage and broccoli (Brassica oleracea), lettuce (Lactuca sativa), maize (Zea mays) and wheat (Triticum aestivum) on the Steenkoppies Aquifer, Gauteng, South Africa. A key aim was to identify one or more simple yet effective WF method(s) that can be applied in South Africa to improve water resource management and for raising consumer awareness. The case study on the Steenkoppies Aquifer indicated that WF metrics from the three methodologies differed notably. For example, an annual two-crop rotation sequence (carrots in summer and cabbage in winter) had an average blue WF of 57 m3 tonne-1 according to the WFN methodology, compared to 32 and 44 m3 tonne-1 according to the hydrological-based method and the LCA methodology, respectively. Estimated WFs differed notably between different crops and growing seasons, for example, according to the WFN methodology the blue plus green WFs of lettuce in summer was 56 m3 tonne-1, compared to 327 m3 tonne-1 for the blue plus green WF of broccoli in winter. The WFN methodology consists of goals and scoping phase, an accounting phase, determining the volume of water required to produce a product, followed by a sustainability assessment, which interprets the impact of the WF in the local context of water availability. Finally, a suitable response is formulated. The WFs according to the WFN methodology is not considered to be suitable for awareness raising if it is communicated without the sustainability assessment or outside the local environmental context of the water use. As an alternative, the LCA-based WF methodologies have been developed to incorporate the environmental sustainability in the WF assessments, and represent a stress weighted index of water use, which can be used together with other LCA metrics. Also, a hydrological-based methodology has been proposed and developed to incorporate all the water flows and other aspects of the hydrological system, so the WFs can be reported as a stand-alone value. While the latter two methodologies aimed to develop single WF values that indicate the sustainability of a water use, due to the vast number of variables, complexities and trade-offs involved in sustainable water use, obtaining such a number still does not seem possible at this stage. Following this comparison, the WFN method was selected as the key methodology for this research project for reasons that include the following: • The methodology is simple and well-developed. • The WFs are based on actual water volumes used for a product, a process or by an entity, which can potentially be used in different information systems, such as water use licensing, up-scaling to a catchment level, and quantifying water consumed by different users for allocation purposes. • By using different functional units, such as nutritional value and economic gain, the volume of water can be directly linked to certain benefits derived from the product. • The volumetric WFs can reveal impacts on water resources in different seasons of a hydrological or calendar year. • It can indicate high WFs of certain crop species, such as broccoli, or certain growing regions, such as those which experience relatively high vapour pressure deficits or with poor soils. • It allows for local geographic contextualisation if there is suitable information to conduct the sustainability assessment. Despite the relative simplicity of the WFN methodology, some complexities were encountered in its application for quantification of WFs of selected vegetable crops on the Steenkoppies Aquifer. In this study, we assessed that WF outcomes are influenced by several factors, including natural variations in weather conditions between growing seasons and between different years. Therefore, WFs should be site specific and calculated for a particular season or year. Water footprints are also directly dependent on crop simulation model outputs, which are in turn affected by the quality of parameterisation and input data used. Whether solar radiation (Rs) data was measured or estimated were found to have a notable impact on estimates of crop reference evapotranspiration (ETo) and crop yields in summer season. It was recommended that if estimated data for a specific weather variable is used for crop parameterisation, the same type of data be used when simulations are executed with those crop parameters (assuming that the variation in the error in Rs will be consistent for a crop calibrated in summer or winter and simulated in the same season). The error in Rs estimates was, however, not consistent over different seasons and parameters generated for crop in a particular season should be used cautiously for other seasons. Variations in water content between different crops can impact the WFs, which are most commonly expressed as a volume of water (e.g. in m3) used per unit of yield (e.g. in tonnes) in fresh mass. This resulted in relatively higher WFs for grain crops with low moisture contents, when using yields in fresh mass. For example, alternative functional units, such as nutritional content (such as zinc or iron) or economic gain are proposed to be used to link the WFs to a more specific potential benefit, which makes comparisons possible. Packhouse WFs were calculated to quantify the volume of water used in cleaning and/or packaging a unit yield of carrots, cabbage and lettuce in a packhouse on the Steenkoppies Aquifer according to the WFN methodology. As observed in previous studies, packhouse blue WFs were relatively low compared to the WFs linked to the cultivation phase (ET) (2.2% of the total field to farm gate WF for carrots, 0.5% for cabbage and 1.5% for lettuce). This highlights the importance of water use during cultivation, as compared to the rest of the supply chain, when considering measures to reduce water use impacts on the aquifer. Using phosphorus (P) as the critical pollutant, packhouse grey WFs were estimated to be considerably larger than the packhouse blue WFs. For carrots, cabbage and lettuce, packhouse grey WFs were 44, 12 and 16%, respectively, of the grey WF linked to the cultivation of these crops. The inclusion of recycling and filtration systems, final fate of the disposed water and associated pollutants, and assimilation capacity of the natural environment make the estimation and interpretation of grey WFs challenging! Grey WF assessments, which were not validated by the Steenkoppies Aquifer water quality measurements, require further research and refinement.