Accounting for water use in Australian red meat production
Artikel i vetenskaplig tidskrift, 2010
Life cycle assessment (LCA) and life cycle inventory (LCI) practice needs to engage with the debate on water use in agriculture and industry. In the case of the red meat sector, some of the methodologies proposed or in use cannot easily inform the debate because either the results are not denominated in units that are meaningful to the public or the results do not reflect environmental outcomes. This study aims to solve these problems by classifying water use LCI data in the Australian red meat sector in a manner consistent with contemporary definitions of sustainability. We intend to quantify water that is removed from the course it would take in the absence of production or degraded in quality by the production system. The water used by three red meat supply systems in southern Australia was estimated using hybrid LCA. Detailed process data incorporating actual growth rates and productivity achieved in two calendar years were complemented by an input-output analysis of goods and services purchased by the properties. Detailed hydrological modelling using a standard agricultural software package was carried out using actual weather data. The model results demonstrated that the major hydrological flows in the system are rainfall and evapotranspiration. Transferred water flows and funds represent small components of the total water inputs to the agricultural enterprise, and the proportion of water degraded is also small relative to the water returned pure to the atmosphere. The results of this study indicate that water used to produce red meat in southern Australia is 18-540 L/kg HSCW, depending on the system, reference year and whether we focus on source or discharge characteristics. Two key factors cause the considerable differences between the water use data presented by different authors: the treatment of rain and the feed production process. Including rain and evapotranspiration in LCI data used in simple environmental discussions is the main cause of disagreement between authors and is questionable from an environmental impact perspective because in the case of some native pastoral systems, these flows may not have changed substantially since the arrival of Europeans. Regarding the second factor, most of the grain and fodder crops used in the three red meat supply chains we studied in Australia are produced by dryland cropping. In other locations where surface water supplies are more readily available, such as the USA, irrigation of cattle fodder is more common. So whereas the treatment of rain is a methodological issue relevant to all studies relating water use to the production of red meat, the availability of irrigation water can be characterised as a fundamental difference between the infrastructure of red meat production systems in different locations. Our results are consistent with other published work when the methodological diversity of their work and the approaches we have used are taken into account. We show that for media claims that tens or hundreds of thousands of litres of water are used in the production of red meat to be true, analysts have to ignore the environmental consequences of water use. Such results may nevertheless be interesting if the purpose of their calculations is to focus on calorific or financial gain rather than environmental optimisation. Our approach can be applied to other agricultural systems. We would not suggest that our results can be used as industry averages. In particular, we have not examined primary data for northern Australian beef production systems, where the majority of Australia's export beef is produced.