A return on experience from the application of agent-based simulations coupled with life cycle assessment to model agricultural processes

Agent-Based Models (ABMs) are becoming a widespread approach to model human-environment interactions. They belong to the class of individual-based modelling approaches, which allow a bottom-up representation of the system being modelled, eliciting its macro-level evolution while modelling the micro-level behavior of its individuals. This paper deals with the application of an ABM to simulate future crop patterns in the Grand-Duchy of Luxembourg under a pre-defined scenario. The simulated scenario deals with the introduction of a "green consciousness" component in farmers' decisions, substituting a purely rational approach based only on profit maximization. The results of the ABM are used to perform a life cycle assessment of Luxembourg's agricultural system. The paper first describes the difficulties and the challenges connected with building an ABM for agriculture and then shows the results of the selected case study. The results show that, from a lifecycle perspective, a "greenness" criterion aimed only at reducing greenhouse gases emissions reveals patently a sub-optimal choice and causes burden shifts to other impact categories. Finally the ABM-based (bottom-up) approach is compared with a top-down approach applied in a previous study by the same authors to model the same system. Assets and drawbacks of the two approaches are highlighted. (C) 2016 Elsevier Ltd. All rights reserved.