Previous Top Next


From one generation to another

Two unmarried randomly chosen individuals (of different genders!) are mated one to another.
Monogamy is fully respected by the software. This is one of the approximations taxing the model's validity.

Births

Once married the software awards a number of children to the couple.
Each child has its gender randomly chosen. Each parent gives randomly one of its two alleles, thus constructing the egg's genotype very similarly to the real meiotic mechanism.
So meiosis and fertilisation are simulated without approximations.

Number of children.

Fertility rate is the mean children number for a mother (putting it another way it's the mathematical hope of the random variable number of child).
This number randomly fluctuates according to a Poisson's law whose parameter (mean value) is the population's Fertility rate.

Examples with 2 different Fertility rates
Probability for a couple to get n children

Children number
0
1
2
3
4
5
6
7
8
Fertility rate 1,5
0,223
0,335
0,251
0,126
0,047
0,014
0,004
0,001
0,000
Fertility rate 2,1
0,122
0,257
0,270
0,189
0,099
0,042
0,015
0,004
0,001


The actual fertility of a given population can differ sensibly from the expected value. Luck may allow many couples with high fertility for instance. This difference is of course linked to the size of the population and that would be just the same in a real population.

Maximum size of the population.

Once the max size (part of the user's parameters) has been reached while producing a new generation, the software deletes randomly the excess number of individuals.
This emulates crudely the effect of a full environment. It's clearly an approximation and not a too good one.
A more realistic way would be to use a logistic curb, flexing the size increase when getting near the limit.

So the simulation is poor with populations near the size limit, but it remains correct as far as the allelic frequencies are on focus.