Introduction

Multi-agent models can bear several names depending on the field they were initially developed in (e.g. agent-based model in social science, individual-based model in ecology). Agent- and individual-based models (ABMs and IBMs) are becoming tools of choice to simulate complex social–ecological systems (Gimblett, 2002; Janssen & Ostrom, 2006; Monticino et al., 2007; Bennett & McGinnis, 2008). The recent development of dedicated programming platforms and libraries has also contributed to the expansion of multi-agent models coupled with geographic information systems (GIS) (Railsback et al., 2006). Such models have been applied in a wide variety of natural resource management contexts where heterogeneous actors interact, including rangeland management in arid zones (Gross et al., 2006), management of water use and access in river basins (Schlüter & Pahl-Wostl, 2007), control of irrigation channels (van Oel et al., 2010), agriculture management (Manson, 2005), and forest clearing for agriculture (Moreno et al., 2007). ABMs have also been used to support national parks and recreation areas' managers by simulating visitor movements to predict over-crowded areas along vehicular routes and hiking trails (Itami et al., 2003), or along riverside rest areas and attraction sites for rafting trips on the Colorado River (Roberts et al., 2002).