12: Individual-based Models

MCED Chapter 12

Title

Individual-Based Models

Authors

Hauke Reuter, Broder Breckling, and Fred Jopp

Abstract

This chapter will describe the category of models that represent the behaviour and interaction of distinct individuals with specific properties. Models of this type can become very complex, but have the advantage that model structures operate on a low level of abstraction and represent ecological relations in a form similar to empirical assessment. Individual-based models facilitate studies of emergent properties, where characteristics of higher level entities like populations or communities can be generated on the basis of single actions of particular individuals. They allow to simultaneously investigate energetic and physiological aspects, behaviour, and relations to other organisms and heterogeneous environmental structures. As a technical background, object-oriented programming is frequently used for this model approach. This chapter introduces the conceptual background and describes two case studies, one that investigates spatial aspects of a predator–prey interaction, and a second one which depicts community interactions of Northern Scandinavian small mammals with oscillating population dynamics.

Additional material

  • Short description of the Individual-based predator-prey model
    Link to IPP software (download material)
  • Short description of the Fish-school model
    Link to Fish-school software (download material)

Further Readings

  • Breckling B (ed.) (2005) Emergent properties in individual-based models – case studies from the Bornhöved Project (Northern Germany). Ecol Model 186(4):375–510

  • Breckling B, Middelhoff U, Reuter H (2006) Individual-based models as tools for ecological theory and application: understanding the emergence of organisational properties in ecological system. Ecol Modell 94:102–113

  • DeAngelis DL, Gross L (1992) Individual-based models and approaches in ecology: populations, Communities and ecosystems. Chapman & Hall, New York

  • Grimm V, Railsback S (2005) Individual-based modelling and ecology. Princeton University Press, Princeton

  • Hill DRC (1996) Object oriented analysis and simulation. Addison Wesley, Harlow

  • Huston M, DeAngelis DL & Post W (1988) New computer models unify ecological theory, BioScience 38:682 – 69

Links

Individual-based models, in particular the more complex ones, are frequently programmed “from scratch” using a higher programming language. There is a number of specific software approaches meant to make things easier. Frequently, however, these approaches narrow the potential of IBM in that they focus on certain types of relations and make it more difficult to capture others. General purpose languages which provide object-oriented programming options are, among others,

  • JAVA (tbg)
  • C++ (tbg)
  • SMALLTALK (ObjectWorks) (tbg)
  • DELPHI (tbg)
  • SIMULA (tbg)

The following software facilitates specific frameworks to develop individual-based models

  • MASON (cs.gmu.edu/~eclab/projects/mason/)
    Mason is a discrete-event multi-agent simulation library programmed in Java which provides functionality for object oriented simulations and allows detailed customisation for complex programming issues. It is optimised for fast program execution with a high number of concurrent objects.
  • Cormas (cormas.cirad.fr/indexeng.htm)
    Cormas is a simulation platform aiming at the representation of interactions between human actors and their use of natural resources. It is based on the VisualWorks programming environment which allows the development of applications in SmallTalk. Cormas has a lot of pre-defined entities which the users can apply to create their own models.
  • NetLogo (ccl.northwestern.edu/netlogo/)
    NetLogo is a multi-agent programmable modelling environment developed in JAVA. As it was originally designed for teaching, it is very easy to use but less flexible with regard to event scheduling. It provides a Model Library containing a large collection of simulation examples and many built-in features for model development and visualisation.
  • Swarm (www.swarm.org/index.php/Swarm_main_page)
    Swarm is a modelling platform originally developed at the Santa Fe Institute (
    www.santafe.edu). It enables simulations of agent based models to be written in Objective-C or Java. Following the Swarm’s conceptual framework users write their own model by using the set of code libraries for implementing, observing, and conducting experiments on ABMs.
  • Repast (repast.sourceforge.net)
    The ABM-platform Repast is built in Java and contains a comfortable model development and execution environment that includes many built-in features such as results logging and graphing tools as well as automated connections to external tools for statistics, evaluation and visualisation. It also contains a flexible hierarchically nested definition of space and GIS support.

General Webpages on
Individual-based and Agent-based models (in ecology)

Project web pages linked to
IBM/ABM:

  • Everglades Across Trophic Level System Simulation
    ATLSS.org

Tailored models

  • WATOR (tbg)
  • Eco-Beaker (simbio.com)