Agent-Based Models and Optimal Control in Biology: A Discrete Approach - Mathematical Concepts and Methods in Modern Biology - page 145

is present, and black when it is absent. Notice the sliders on the left-hand side: these
values can be changed as the simulation proceeds, or prior to starting a simulation.
What do you expect to happen to the rabbit population if the grass growth is set to a
very low rate? What if the grass grows quickly? What effect would altering the birth
threshold have on the amount of grass present at each time step? Do you think it is
possible for the rabbit population and grass level to stabilize over time? Play around
with the sliders to determine if your predictions are correct. Note that you may need
to let the simulation run for several hundred time steps (or “ticks” in Netlogo) in order
to observe consistent dynamics. Do the populations stabilize? Do they oscillate?
5.3 NETLOGO: AN INTRODUCTION
Hopefully, the example in Section 5.2 has given you an idea of what an agent-based
model is: a computer simulation wherein agents interact with their local environment
(possibly including other agents) based on a set of rules. In this section we guide you
through several exercises involving agent-based models using Netlogo [ 6 ], a software
tool developed to work with agent-based models. Netlogo can be downloaded for free
at http://ccl.northwestern.edu/netlogo/ . There you will find detailed instructions on
how to install it. Netlogo is its own programming language, so named because it is a
variation of the Logo language. While many key features are unique to Netlogo, users
familiar with Logo will likely note similarities. Note too that Netlogo is continually
updated and newer versions released. As of the time of this writing, the latest version
is Netlogo 5.0. All files and exercises associated with this chapter will be conducted
using this version; some adjustment may be necessary for newer (or older) versions.
While there are many other software platforms available and in use for agent-based
modeling, for the remainder of this chapter we will use Netlogo to introduce and
examine agent-based models. It has the convenient advantage of providing the user
with an intuitive graphical interface, which we will use to aid our understanding of the
models we will examine. In addition, the standard installation comes with a library of
models, all of which are open source. Thus we may build models from scratch or we
may choose to alter an existing model. You may wish to go through the tutorials found
at http://ccl.northwestern.edu/netlogo/docs/ . A complete dictionary of programming
terms can be found at http://ccl.northwestern.edu/netlogo/docs/dictionary.html . All
exercises in this section refer to the “Rabbits Grass Weeds” model introduced in
Section 5.2 . It should be noted that for Exercises 5.1 and 5.2 , the web-based version of
the model can be used (via http://ccl.northwestern.edu/netlogo/models/RabbitsGrass
Weeds ), but Exercises 5.3 and 5.4 require installation of the Netlogo software.
Exercise 5.1. Note the two sliders weeds-grow-rate and weed-energy .
By setting weeds-grow-rate to be any nonzero value, you will notice that the
landscape of the model has been altered, as there are now patches containing weeds
interspersed with the grass (represented as purple patches). What effect does this have
on the rabbit population? How about on the grass population? What happens when
you increase weed-energy ? Set grass-grow-rate and weeds-grow-rate