November 14, 2013 | 2
Think about all the decisions you’ve made today. Even if you’re reading this in the morning, you’ve probably already made hundreds or even thousands of decisions, without even thinking consciously about most of them.
We like to think of ourselves as being in control of our own decisions, and making rational, well-informed choices. However, more and more research is showing that this just isn’t the case, as has been popularised in books like Predictably Irrational and Thinking fast and slow
However, this isn’t something that’s uniquely human. Even slime moulds are irrational. Even though these organisms might seem very different to ourselves, the more we can learn about how other animals make decisions, the more it will help us understand how decision-making as a behaviour evolves. Understanding this will help reveal why we make decisions the way we do.
Takao Sasaki and Stephen Pratt from Arizona State University looked into decision-making in ants. Specifically they looked at how individuals weight different attributes when making a decision. For example, if you had to choose between two cars, that were identical except for the price, it wouldn’t be a difficult decision. But, rarely in life are we lucky enough to be presented with problems like this. It’s more likely that you would have to make a decision between a cheap car that looks like this:
and one that looks like this:
And even that would be a pretty easy decision. In reality there will be a multitude of attributes of the two cars you’re choosing between, and you have to decide which ones you want to prioritise. However, how you rank things may also depend on what you’ve experienced.
We weight our decisions all the time according to experience, most of the time without even thinking about it. It seems that ants do the same thing. The rock ant Temnothorax rugatulus lives in rock crevices, so one of their main decisions they have to make is which of many rock crevices is the best place to live.
Sasaki & Pratt brought these ants into their lab, and made artificial rock crevices for the ants to choose between. These ants generally prefer small entrances of crevices to larger ones, and darker interiors over brighter ones. First the researchers gave the ants a choice between two options that posed a trade-off between these two attributes: one nest had either a smaller (better) entrance, but was lighter (worse), than the other one.
The ants were then moved to a ‘standard’ nest, which was neither brilliant nor terrible. Next, the scientists removed the roof of their nest to induce them to move from their current home to a new one. The ants had a choice between two new ‘homes’ to choose between. For one treatment group of ants, they had a choice between another ‘standard’ home, the same as the one they had just left, and one that had the same light level, but worse (larger) entrance. The other treatment group of ants had a choice between the same ‘standard’ nest and a nest that had the same entrance size, but worse lighting (lighter). In both cases, the ants would assess both the new housing options, and opt for the ‘standard’ home that was the better choice. Since ants live and function as a colony, they do this by sending out particular workers to check out the two different options. These ants then come back to the home nest, and get another ant to follow them closely behind (so-called ‘tandem running’). Once enough ants decide that they want to move to a particular nest, they start carrying other ants from the colony straight to their new home to speed up the process (so-called ‘nestmate transport’).
A couple of great videos of these behaviours:
Tandem running in ants
In this video below you can see the ants explore both the ‘good’ and the ‘poor’ nest, then start leading other ants to the good one through tandem-running, and finally start carrying the rest of the colony over through ‘transport’.
The scientists let the colonies choose between these two different nests (‘standard’ and the one that was worse in one attribute) three more times. This meant the ants had to make choices between different nest sites four times in total.
After having made these four decisions, the ants were then given the same choice they had been given at first: between a nest that had either a smaller (better) entrance, but was lighter (worse), or a larger (worse) entrance that was darker (better).
In both the treatments, the ants chose the nest that was ‘better’ in the attribute that they had seen be worse up until now. So, for ants that had been choosing between standard nests and those with brighter lighting, they now chose the nest with lighting they preferred (darker). On the other hand, the ants that had been rejecting lots of nests with large entrance holes now chose the nest with the smaller entrance hole.
This is a bit like you thinking that you don’t really care about the colour of the car you’re trying to buy, more than, say, the gas mileage. However, every single car you look at is just an awful colour. Now when you’re given a choice between a car with poor gas mileage and a great colour, and a bad colour and great gas mileage, you choose the one that’s a better colour. You have changed the weighting of what you care about, based on your experience.
Why might animals tune in to one particular attribute when making a decision? On the surface it might seem that by doing so they are missing out on other, potentially important information. However, it might actually be a clever strategy. If one particular attribute is generally more informative than another, then it makes sense to pay more attention to that attribute. In this case, one attribute of the nest was always fine (of ‘standard’ quality to the ants), and so wasn’t worth paying as much attention to. In our car example, if we see a bunch of cars with fine gas mileage that are just horrible colours, then we might stop paying attention to the gas mileage and instead focus in on the colour. Without even realising it, this kind of bias could help us make faster and more accurate decisions.
First car: Scott Feldstein
Second car: William Warby
Tandem Running & Nestmate transport videos: Stephen Pratt
Nest choice video: Takao Sasaki
rock ant Temnothorax rugatulus photos: Stephen Pratt and James Waters
car montage: mix of car photos, in order:
Sasaki T, Pratt SC. 2013 Antslearn to rely on more informative attributes during decision-making. Biol Lett 2013 0667.http://dx.doi.org/10.1098/rsbl.2013.0667