Crown-of-Thorn Outbreak

Authors: Alex Tee Neng Heng, David G. Green

Crown-of-Thorns Outbreak

The Crown of Thorns (Acanthaster planci) is a starfish that lives on Australia’s Great Barrier Reef. It has several times caused alarm for the tourism industy because it feeds on coral and outbreaks of the starfish can eat out entire reefs, which then take several years to recover.

What the model shows

This model shows starfish (COTs) as dots moving around on a reef, eating coral as they go. The reef map is a digitized satellite image. Different parts of the reef have different levels of coral cover.

Technically the model is a cellular automaton (coral) with individual agents (COTs) overlaid and interacting with it.

How to use the simulation

There are two views:

  • Seascape is a map with different areas in different colours. This view reveals where the COTs are moving.
  • Coral is a map of the level of coral cover. This view shows how much coral is left.

The controls allow you to set several kinds of parameters:

  1. Time and motion – number of time steps for which the model will run; starfish speed is the number of cells that starfish can cross in a single time step (in effect this represents the size of each time step).
  2. Coral change – how fast the coral grows; how fast the starfish eat them
  3. Coral cover – Max coral limits the maximum cover possible on each part of the reef
  4. Starfish movement – This sets up one of two scenarios about starfish behaviour: either they move around the reef at random, or they actively move towrds cells with the highest coral cover within their neighbourhood.

Note that three parameters – coral growth rate, eating rate and starfish speed – are related to the size of the timestep. The default values are all meant to be realistic settings for a typical reef, with a timestep of one day.


  1. Press START to run the model with the default settings. Use the STEP selection to experiment with viewing speeds. Do the COTs congregate in particular areas?
  2. Switch from SEASCAPE to CORAL view and run the model again. Can you see the coral disappearing? Run the model with different numbers of starfish. How many does it take to eat out the reef completely?
  3. You should see more evidence of starfish clustering in areas of high coral when their numbers are small. This is because higher numbers eat out most of the coral, so no areas have high coral cover. Experiment with settings to see where the clustering is most obvious. Then change the behaviour to RANDOM to see the difference.

Links and references

Demo screenshot