Blog challenge: Sheep cyclone

In relation to multiscale modelling, a category of problem that receives a good amount of attention is understanding how collective (group) movement arises from movement rules defined at the individual level. For example, the direction of each individual’s next move might depend on where that individual’s nearest neighbours are. What type of rules are necessary for the group to stay together? If the group stays together, what does the collective movement look like (directed or meandering)?

Some nice examples are:

and more generally see here, here, here and here. So this is my question:

What type of individual movement rules are needed to produce a Sheep cyclone?

Bonus question: How close together do the parallel walls (relative to the car width) need to be?

This entry was posted in Fun, Questions to readers by Amy Hurford. Bookmark the permalink.

I am a theoretical biologist. I became aware of mathematical biology as an undergraduate when I conducted an internet search to learn about the topic. Now, twelve years later, I want to know, what is it that makes great models great? This blog is the chronology of my thoughts as I explore this topic.

9 thoughts on “Blog challenge: Sheep cyclone”

1. Presumably you want movement rules that would also explain sheep movement in other contexts. As opposed to rules like “always turn left”. 😉

2. The sheep cyclone is one of my favourite youtube clips! I wonder if sheep herds have a pecking order. Do “alpha” sheep exist? If so, I bet this alpha sheep determines the direction of the cyclone. There’s really only one way to find out though … we need two dozen sheep, spray paint and a car in an alley!

• Attention everyone: Jennifer is the one who put me onto Sheep cyclone!

That’s brilliant: what determines the direction of the sheep cyclone! There are some papers on how leaders influence collective movement. The irony in all this is that my Dad was a sheep farmer up until a few years ago. Why didn’t we think of this experiment earlier!

3. I gave this a try in Netlogo with an individual based model. I thought with some flocking and really simple avoidance rules I could generate the behavior. I didn’t generate really clean cyclones, just the occasional loop around a car by a few sheep. I wrote up a quick post with links to a couple videos I made and the Netlogo code with a some thoughts on the kinds of rules that I think will work. A neat diversion, thanks!
http://currentecology.blogspot.com/2012/02/sheep-cyclone-problema-start.html

• This. Is. Awesome. And there’s actually a car (and not a rectangle) and the sheep are not just white dots they actually have noses.

I feel like the cyclone arises from a super strong desire to minimize the distance to the nearest sheep ahead of you – where ahead is defined within some fixed arc, and the car moves just fast enough that the nearest sheep is always ‘in front’ and not ‘to the side’.

In addition to being a sheep farmer, at one time my dad farmed goats, and so here’s what I think is interesting: if you run a dog (simulating a predator?) at sheep, the sheep flock together. If you run a dog at goats they just break apart. Why? Why the different behavioural adaptations to predators? Do the sheep try and hide the weak/sick/elderly in the middle of the flock, but goats are every goat for themselves? Is this related to average heard size/social group size. Are there more sick sheep then there are sick goats? Why?!

• Thanks!