Life cycle inventory modelling of land use induced by crop consumption

Abstract

Background, aims and scope Most life cycle inventory data for crops do not include the ultimate (marginal) land use induced by crop consumption. The aims of this study were to present, document and discuss a method which can solve this problem and, furthermore, to present concrete examples for wheat consumption in Brazil, China, Denmark and the USA. A global scope is applied and the simulated adaptation to increased wheat demand corresponds to a long-term temporal scope under present market conditions with present technology. Materials and methods The economic general equilibrium model, Global Trade Analysis Project (GTAP) is modified and applied. Agricultural statistics and a number of global land use and land cover datasets are used in the modification and the processing of the model output. Some of the land use datasets are processed by use of a geographic information system tool. Results The net expansion of the global agricultural area per tonne of wheat consumed in Brazil, China, Denmark and the USA is estimated at 2,000, 260, 1,700, and 3,200 m 2 , respectively. For Brazil, the expansion mainly affects tropical evergreen forest. For China and the USA, the expansion mainly affects boreal deciduous forest, savanna, open shrubland and tropical evergreen forest, and for Denmark, it mainly affects savanna, tropical evergreen forest and dense shrubland. The areas affected are quantified in the land use life cycle inventory (LCI) produced for the four countries. Discussion The method applied allows for an even more detailed land use LCI than the one presented in this study. Results are influenced by existing global trade patterns and their inertia. Such aspects should be acknowledged in life cycle assessment (LCA). The method takes its starting point in consumption rather than production. Conclusions The method presented makes it possible to simulate the main mechanisms of the global agricultural system and thereby construct an LCI containing the land use induced by crop consumption in a given region and the nature types (biomes) affected. The results are sensitive to changes in the so-called Armington elasticities representing the inertia of global trade patterns. It is considered reasonable to double the standard elasticities in the GTAP model for the construction of LCI data. Wheat consumption in different countries result in different land use consequences due to differences in trade patterns, which are governed by transport and trade costs, among other factors. Recommendations and perspectives The modelling could be improved by incorporating a mechanism simulating