Agent-Based Modelling is a simulation-driven approach for analysing complex systems. The interplay of the system elements is depicted at the individual level and can be spatially explicit. Thus, agents (in our case, e.g., pastoralists or herds of animals) can be endowed with certain rules on how to interact with other agents and their environment. Often times, agents possess only local knowledge of their environment and, hence, are considered to be boundedly rational. From the individual interactions emerge dynamics at the system level which display complex feedbacks between different parts of the system.
Since these models are usually not analytically solvable, they have to be analysed with the help of computer simulations.