| |
Parameter Sliders
These
allow one to set the values for various parameters. The application checks
them at the beginning of a run only — changing their values while
a run is in progress has no effect. Change the values by clicking and
dragging with the mouse. Alternatively, for finer control click the slider
with the mouse to select it, and use the right and left arrow keys change
the values.
|
8 |
NUMBER
OF TIMESTEPS and NUMBER OF FAUNAS are straightforward. These determine
for how many timesteps each fauna is followed, and for how many
faunas this process is repeated, the whole constituting one run.
NUMBER OF SPECIES
represents the initial number of species in each fauna. Because
species sometimes become extinct and are not replaced right away,
this is a maximum value and the average throughout the course of
the timesteps is somewhat less than this, depending on the other
parameters.
The number
of species has a marked effect on the speed at which the simulation
runs. Also, note that the simulation tends to become unstable
if you specify fewer than about 20 species. |
|
|
|
|
9 |
Each iteration
within a run starts with a new fauna, for which a slope has been
specified (or randomly generated) for the fauna's regression of
population size on body mass. The species are randomly assigned
body masses, then are assigned population sizes based on this slope.
The population sizes are then randomly dispersed about the line
of the slope, by means of a Gaussian distribution using the INITIAL
Y-VARIANCE. This makes it possible to start the simulations with
realistic-looking faunas rather than ones where every species lines
up along the slope. (A value of about 0.25 produces faunas that
look reasonably like empirical ones.) After the first timestep,
the actual variation about the regression slope depends on the simulation
processes and is no longer constrained by the initial y-variance.
(See POOL VARIANCE, 11).
The MAXIMUM
and MINIMUM SLOPE values set the range of slopes that will be used
to start new faunas. Slope values will be chosen from a uniform
distribution within the bounds set here. In the case where both
sliders are set to the same value, all faunas in a run will start
with the same slope.
NOTE: Certain
assumptions of the model cause the simulation to tend to be unstable
if the slope becomes very shallow or positive, particularly in combination
with some values for TOTAL ENERGY (10). Although it is possible
to set a positive slope value for the initial slope, this can increase
the probability that the application will exceed some of its numerical
limits. Try it anyway — it can be informative. |
| |
|
|
|
10 |
SPECIES ADVANCING
— The expected percentage of species that will make an evolutionary
advance in a given timestep.
ENERGY INCREMENT
— For a species making an evolutionary advance, the percentage
by which each individual increases its energy. This can be given
a small random component using the E value program state
(11).
KLEIBER SLOPE
— This allows one to explore the behavior of the simulation
under alternate exponents for the scaling of individual metabolism.
The most widely accepted empirical value is approximately 0.75 .
TOTAL ENERGY
— The total energy available to the fauna is 5 X 10x,
where x is the value specified here. Much of the time this
value is arbitrary and will have no effect on the simulation. However,
it does interact with the number of species and can affect the behavior
of the simulation. It is possible, for example, to have it set too
low, such that within the first few timesteps most of the species
go extinct, leading to unpredictable behavior. Likewise, if it is
set too high and some species become energetically dominant, their
energy use may exceed numerical limits and cause an error. Many
times what is happening will be apparent from the visual output.
If you are having trouble with instability
even though the other parameters seem reasonable,
try varying the Total Energy. |
| |
|
|
|
|
