Reptiles have ability to learn from each other

Social learning is a feat that is associated with animals that live in groups, especially those that display extended infant care, such as the common woolly monkey shown on the right (Image from bbc). Social learning means that the animal acquires the knowledge of how to complete a task by observing a conspecific – an animal of the same species, performing that task. For example, it’s how we learn basic skills as small children and how primates learn from their parents (and one of the reasons for being carried around all the time!). Animals will only learn from conspecifics with whom they have no conflict of interest. That means, in most cases only their parents and siblings (r=0.5 in most animals). Of course, there are other scenarios where there is no conflict of interest, such as when there is no direct competition between the animals for food (or anything else), such as in the experiment shown below.

Social learning has also been observed in fish, birds and insects, but always in animals that live in social groups. The conclusion that was drawn was that social learning appears to be an adaptation to living in social groups. I was doing some reading recently on social learning as part of some lectures on conflict of interest and so a paper on “Social learning in a non-social reptile (Geochelone carbonaria)” immediately stood out to me. The paper by some researchers at University of Vienna showed that red-footed tortoises can learn to solve a talk by observing a con-specific complete that task. Red-footed tortoises are non-social animals and generally lead solitary lives. So we would expect them not to be capable of social learning. But, further experiments showed that they are.

The following image shows the experimental set up. A bowl of food (presumably something extra tasty) was placed in a V-shaped fence. The tortoise was placed on the other side. First each of four tortoises was left to complete the task. They were given 12 tries on different days. Then, a tortoise was trained in completing the task – which took a long time! But eventually the tortoise figured out that it had to walk around the the fence to get to the food. It was then allowed to eat the tasty treat. (Note: the animals were trained to walk around the right of the fence). Then, the an observer tortoise was placed, in a little cage, inside the experimental set up and watched the observer complete the trial. Then the observer tortoise was let loose. The experimenters were very careful with this part of the experiment – they made sure the observer was actually watching and that the demonstrator did everything correct. Neat setup.

On to the results section – always the more exiting part. Below shows a graph of how well each tortoise performed. Notice how they all have individual names – cute. I couldn’t find out what the name of the demonstrator tortoise was though. I hope it doesn’t feel neglected. On with SCIENCE though. The first part of the graph shows that none of the tortoises which didn’t have a demonstrator manged to complete the task. None at all. So tortoises do not have an innate ability to somehow eventually manage to learn this – also shown by the 30 sessions with several tries each, that it took the demonstrator tortoise to learn the task.

The second part of the graph shows the performance of the tortoises that did watch a demonstration first. The shaded areas indicate that the tortoise took the same path as the demonstrator, the white ones shows they walked the other way around. Clear difference! They managed to get to a yummy treat! Clearly Aldous is the champion of walking around V-shaped fences, closely followed by Moses. However, both Molly and Quinn didn’t do so well.

So what does this all mean? The researchers concluded that the tortoises learned by observing the demonstrator tortoise complete the task. And I agree, they have very nicely shown that tortoises have the ability to learn from conspecifics.

However, they failed to explain the marked differences in performance between the high-performers Moses and Aldous and the low performers Molly and Quinn. Some simple statistics reveal that there are significant differences between the performance of those two groups. They did conclude that that more experiments are needed, for example to show if a conspecific is needed or a different animal could serve as a demonstrator or perhaps even an animate object.

doi: 10.1098/rsbl.2010.0092