Properly defining animal movement for animal agents within the game will be a critical development process. While there is a great deal of published literature regarding animal movement, very seldom does it include the type of information that would be useful for defining agent-based behavior. Describing the movement of an animal (ie. in a set of rules) is not a typical research goal of ecologists; instead investigations will typically focus on a single characteristic of their movement, and how that characteristic relates to an ecological process.

There are two primary lines of investigation in ecological research that I envision as being useful to define the animal agent rules;
  1. Habitat utilization/requirements
  2. Foraging Strategies

Existing research often uses incompatible temporal and/or spatial scales that must be negotiated to produce useful rules of agent behavior. Spatial discrepancies are most likely because habitat requirements are typically defined by correlating observed counts (often regional-scale data) with existing landscape features, while foraging strategies are based on more detailed observations of individual animals. The temporal resolution of observing the foraging behavior of an individual is also going to be much finer than a census-style bird count, typically performed once per year.

Movement Ecology Paradigm

The movement ecology paradigm is a conceptual framework for the study of all organismal movement linked to four basic mechanistic components (Nathan et al. 2008). It could potentially be a robust methodology for decomposing movement processes and translating them into agent-based rulesets. Potentially applicable to both human and people agents.


  • movement methods intended for application in TrailsForward (first two included in PATCH model - Schumaker 1998):
    • random walk (Gaussian/Brownian + others)
      • PATCH allows you to bias random walk towards good habitat
    • intelligent search (individual moves directly to the best available site within search radius)
    • vector summation approach

  • Big question - how to (properly) combine definition of suitable habitat with proper utilization by agents
    • habitat could be pre-processed as a species-specific binary suitability measure
    • then need some agent-based algorithm preventing overuse of suitable habitat (minimum spacing, ownership by individual agents?)
      • could potentially use avoidance algorithm to cram as many individuals as possible into suitable habitat; external stimulus in the form of a reduction in forage return could trigger response in the form of movement mode switching from 'forage' to 'territory finding'

  • best sources of behavioural data appear to be for highly visible (from the perspective of the public) species, ie. game species (deer, bear) or high conservation priorties (wolves, lynx?).
  • (via van Deelen) - may be able to create ‘generic predator’ and can resize to fit different species according to principles of allometric scaling (ie. a badger is roughly equivalent behavior-wise to a small bear)

Related Topics

Analyzing Agent Behavior
Foraging Strategies

Object Avoidance


Bennett, DA; Tang, W; Modelling adaptive, spatially aware, and mobile agents: Elk migration in Yellowstone. International Journal of Geographical Information Science 20(9) p.1039.

Tang, W; Bennett, DA. 2010. Agent-based Modeling of Animal Movement: A Review. Geography Compass

Turchin, P. 1998. Quantitative Analysis of Movement: measuring and modeling population redistribution in plants and animals. Sinauer Associates, Sunderland, MA.