NeilUnreal

This is kind of off-topic, but I know there are a lot of biologists here…

This weekend, in a desperate attempt to find excuses not to clean the kitchen, I decided to play around with writing some a-life based on genetic algorithms (using Holland classifiers). I started with a simple example using a fixed population of organisms which could move on a bounded 2D surface (an arena). They couldn’t die naturally, but they could breed when they came into contact, and since the population size was fixed, the resulting new organism replaced the oldest current member of the population (i.e. the “Logan’s Run” algorithm).

I decided to supply a gradient which they could sense, and used simple coupling between the gradient sensors of each organism and the movement control – similar to the class of simple robots called photovores. It occurred to me that, even if there was no survival advantage to moving along the gradient, a reproductive advantage would develop along the gradient. This would occur because any organisms randomly initialized to move along the gradient would encounter more of the same, and this effect would be compounded in each subsequent generation. In other words, a “breeding ground” would develop.

It worked so well it was boring. (This effect may be old-hat to a-life folks, I’m not current with that literature.)

Two things, however, were surprising. I thought I’d ask if whether any of the biologists here were familiar with field research which shows similarities:

1) Although the breeding grounds can take some time to begin forming, once the effect starts the collapse is rapid. The transition from a non-breeding ground population to a breeding ground population takes only a few generations.

2) Populations in restricted arenas are much less prone to develop breeding grounds than populations in large arenas (think whales and migratory birds). To a certain extent, this effect appears independent of the population density (this was the surprising part). Also, the collapse between the loose breeding grounds in small arenas and tight breeding grounds in large arenas may be non-linear with respect to arena size.

Anyway, that’s what I did this weekend instead of cleaning the kitchen.

-Neil

Blinn

I just did a quick Google search and found this: <a href="http://www.nyu.edu/projects/fitch/resources/student_papers/dastagir.pdf" target="_blank">Evolution of Leks</a>. No time to read beyond the abstract, and I'm not sure if I grasped exactly what you did with your program, but the "hotspot" hypothesis is similar to your model, no?

NeilUnreal

Thanks, Zetek! Slightly different than what my simulation was looking at, but it appears to be an instance of the same principle. A great starting point for further reading.

It's interesting that the paper examines the effect in relation to sexual selection in general. I had been thinking of terms of spatial gradients and hadn't considered the more general case of any gradient which affects the success of courtship.

-Neil