The goal of this research was to study the policy environment to encourage transition from conventional to organic farming and associated impacts on GHG emissions using a loosely-coupled geographic information system (GIS) and Agent-based Model (ABM, also known as multi-agent models). ABM is a relatively recent modeling technique, and has been defined as the modeling of complex systems composed of interacting, independent “agents” (Macal and North, 2010).
For our work, the agent-based model must be capable of exploring policy environments and evaluating farm conversion and GHG emission changes at farm and watershed scales over time. The ABM, as a spatiotemporal simulation approach, is potentially useful to model an agricultural sub-watershed as a complex system. The integration of GIS and ABM (Birkin et al., 2012, Goncalves et al., 2004, Murphy, 1995) takes advantage of automata theory to study dynamic phenomena in a complex environmental system with many individual entities.
Agricultural watersheds are complex coupled human-natural systems. They have multiple interrelated actors and elements, and multiple relationships with their environment (Pacini et al., 2003). These multidimensional relationships (and their impacts) interact and combine to present watershed level outcomes and behavior that is often difficult to predict from the study of a narrow range of elements.
We approach the loosely coupled GIS-ABM with an environmental and socio-economic sustainability perspective. Greenhouse gas emissions (GHGs) are used as an example to demonstrate and assess the model performance.
Numerous authors, including Pelletier et al. (2008) and Lynch et al. (2011), have argued that the transition from conventional to organic agriculture is especially effective to reduce GHG emissions. Organic systems minimize the use of external inputs such as synthetic fertilizers and pesticides, and demonstrate much greater energy efficiency. For example, Pelletier et al. (2008), in a 12-year study of energy inputs, energy outputs, and energy-use efficiency in conventional and organic crop production systems in Canada, found that energy use was 50% lower with organic compared to conventional management. They also found approximately 50% fewer GHGs produced.
But the transition to organic production is difficult, with many economic, psycho-social and institutional obstacles (MacRae et al., 2009). Other researchers have examined many of these challenges in relation to sustainability and/or agricultural systems. Hannon (1991) studied ecosystem indicators, Halberg (1999) stressed economic variables, Hardaker (1997), Hammond and Goodwin (1997), and Callens and Tyteca (1999) focused on government programs, regulations, technical, social, and political aspects and strategies.
The ABM that we present in this paper draws upon these authors and others to define relationships and parametize the model. The model was applied over time (32 years using one-year increments) to predict not only GHG, but economic outputs such as net income. We were also able to estimate the effects of a number of exogenous factors and forces influencing farmers on the process of transition. We applied the ABM to the Middle Maitland Valley, an agricultural sub-watershed in Ontario, Canada, based on typical crop rotations for both conventional and organic farming. The ABM was designed and constructed using the REPAST platform, which is open source and compatible with spatial data.
Below we introduce Agent-based Modeling. Later we discuss the anatomy of our model and data sources. Finally, we present preliminary results from application of the model with respect to GHG emissions during scenarios of conversion from conventional to organic agriculture in the Middle Maitland Valley sub-watershed (MMVS).
