March 13, 2005

Artificial evolution

In some biology simulations, you don't even need to store the location of each critter. Surprisingly enough, evolution simulations are often of this type.
Each critter is either male or female.
Each critter has it's own ratio of
male to female offspring.
Each timestep:
Select a male critter C1 at random
Select a female critter C2 at random
delete 2 random critters
create 2 offspring from C1 and C2
When creating offspring from C1 and C2, one offspring inherits C1's male/female probability gene, the other inherits C2's, and they are male or female with their respective probabilities.

To provide a bit of variation, occasionally randomize a critter's male/female gene.

When you run this extremely simple simulation, the ratio of males to females in the population quickly approaches 50/50. The genepool also changes significantly, from a set of random ratios to a pool entirely filled with genes near 50/50.

This happens because of negative feedback. When one sex becomes more numerous, genes that produce more of the other sex gain an advantage at reproducing themselves, thus rebalancing the population.

So, sorry guys, all those sci-fi stories with races of space-babes whose populations are 7/8ths female just aren't possible. Unless, of course, they've got the genetics of a hive insect, in which case most of them won't be interested.


At March 15, 2005, Blogger Fraxas said...

That simulation is assuming that it takes 1 of each sex to reproduce, and that males and females are equally fit (i.e. neither has a longer life expectancy than the other).

You could have a 7/8ths space-babe society if space babes died really quickly and were born more often...

At March 15, 2005, Blogger JeremyHussell said...

If that happened, genes that increased the number of males at the expense of their longevity would be selected for too. In other words, you could have such a situation, but it wouldn't be in the least bit stable in the long term.

Of course, for a story, a temporary situation may be enough.

At March 15, 2005, Blogger Fraxas said...

Whoops, I missed a part. If space dudes had 7 times the lifespan of space babes, *and* it took 7 space babes and one space dude to make an offspring, then populations would stabilize at 1:7 ratios...wouldn't they?

At March 15, 2005, Blogger JeremyHussell said...

I think if a species required 7 of one gender and 1 of the other to reproduce, then it would stabilize at a 7:1 ratio with equal lifespans.

The need for 8 parents would imply that they're octoploid, i.e. they have 8 copies of each chromosome. As of 1999, there is only 1 known mammal that isn't diploid: a rat from Argentina that's tetraploid. Even then, each rat parent donates two copies of each chromosome, so that they only need 2 parents to reproduce.

So our hypothetical 7:1 species would have a lot of trouble reproducing and wouldn't evolve naturally. Even if someone genetically engineered them, it would make more sense for them to have 7 males per female. Otherwise, which of the 7 females has the kid, and how, exactly, did the remaining 6 contribute?

At March 15, 2005, Blogger JeremyHussell said...

There are no known species that produce offspring from more than 2 parents.

It is possible to have a species with much higher male mortality, so that by the time they reach reproductive maturity, there aren't many left. If you think about it, this doesn't affect the ratio at birth, because although most males die before breeding, they also get to breed more often, and although most females survive to breed, they don't get to breed as often. So, for the parents, it still makes sense to produce a 1:1 ratio of males and females.


Post a Comment

<< Home