Ecological Artificial Life - contents
The initial description of this project was the following:
"Simulation of the population dynamics of at least two real species in a realistic ecosystem with a temporally varying environment by processes which include both genetic and environmental influences. This will include modelling the relationships between predator, prey and scavenger species, between competing species, and between each species and the environment. Two examples would be leopards, gazelles and hyenas in the African Savannah, or a colony of ants. A graphical display of the progress of the simulation should be provided, together with a play-back facility."
To be more specific, I have chosen to model honey bees. There are different varieties within the species apis mellifera, specifically Africanised and European. However, at my level of abstraction the differences between these bees are negligible. I will use the British climate as my baseline for "normal" conditions, i.e. summer in July and winter in January.
Other researchers (notably Richard Dawkins and John Maynard Smith) have done a lot of work done in the past on natural selection, and the search for evolutionarily stable strategies (ESS). In other words, given a particular environment, which organisms will do best? By contrast, the aim of my project is to say "given a set of creatures, what effect will the environment have?"
More generally, I have represented the bees and food sources within a biodome. This provides a controlled environment, and also prevents the bees from flying too far away, since there is effectively an invisible barrier at a 300m radius from the centre (where the main hive is positioned). This simplifies the coding, and will also make it easier to understand the graphical display.
My specific objectives are to find out what overall population sizes are stable, and also what proportions of drones/workers are stable. There are three areas that this applies to: confirming known behaviour, confirming other predictions, and making new predictions.
The graphics are quite crude, just coloured circles for the bees (colour coding for worker/drone/queen).
Every bee will have a "hive number" attribute. In reality, bees can recognise each other by scent, but this abstraction will suffice for me. In most cases, a bee will remain in the same hive for its entire life, but there are some situations where it may move to a new hive. This is more common amongst drones than workers, since drones often get lost on their way home, and land at the wrong hive!
Next: 2. Literature review and background for the project
This page was last updated on 2004-09-05 by John C. Kirk