Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 1,741 other subscribers

Tree Skinks Go To School: The Complexities of Social Learning in Lizards

By: Fonti Kar & Julia Riley

“Never study an animal that is smarter than you”

                   – Dr Martin Whiting

An adult female tree skink after performing the discrimination task we used to quantify their learning ability – she successfully removed the blue lid from this dish and accessed the food reward . Photo by Anna Küchler.

Animals learn about their environment and use what they have learnt while foraging, to increase mating success, avoid predators, and overall increase their chances of survival. An animal can learn by themselves, but this can be a long and difficult process. Alternatively, an animal can learn through the use of social information that is observed through “spying” on others. This is often an easier strategy, unless the social information is not accurate. We can relate to this – often when figuring out a problem, it’s easier to get help from a friend or teacher, rather than struggling through it on our own! In particular, social animals are likely to have more of a chance to observe others, and learn from them. We studied if tree skinks (Egernia striolata), a social Australian lizard, can learn from others.

Two tree skinks basking together beside a rock crevice in their natural habitat. Tree skinks are part of the Australian Egernia-group of social lizards, which across species, displays a diverse array of social and mating systems, including family-living and parental care. Photo by Martin Whiting.

In a previous study, we wanted to test whether adult female skinks can learn from another female. We trained a group of lizards to remove a lid to forage for some delicious fruit puree. In research, this type of challenge is called a ‘motor’ task. Once they were trained, this group of lizards demonstrated this action to another group of lizards that were naive to this foraging technique. Lizards in this ‘social learning’ group got to observe a demonstrating lizard before they attempted this motor task themselves. We compared this ‘social learning’ group to another group of lizards (our control) that needed to learn this task individually, without any help from others. We found that both ‘social learners’ and ‘individual learners’ (our control) were able to learn this motor task!

But, we then challenged the lizards with another, more complicated, task where they needed to learn to discriminate between dishes covered by a blue lid or a white lid. This was called the ‘discrimination task’. In this task the delicious food reward was accessible under the blue lid, but not the white lid. ‘Social learners’ made fewer errors and learnt this discrimination task a lot more quickly than ‘individual learners’ (our control). This means that that female tree skinks learn from one another! In complicated tasks, these social lizards likely use social information as a short-cut to learning.

Female tree skinks that were able to observe social information made fewer errors in the discrimination task, and learnt it more quickly than tree skinks that had to learn on their own. This shows that female tree skinks can socially learn! See the entire presentation Fonti Kar made on this study here.

Interestingly, we also investigated if juvenile tree skinks, that grew up in different social environments, used social information from an unfamiliar, adult female. We ran a similar experiment to our first, but also added a ‘reversal task’. In the reversal task, the food was now under the white lid. Lizards had to unlearn the rule from the previous, discrimination task (the food is under the blue lid), and now associate the white lid with a food reward (to see the methods for each of the tasks check out this video). In contrast to our previous study, we found no evidence that juveniles used social information from an unfamiliar, adult female. We were a bit surprised that juveniles did not use social information, because adult females do and it is often a beneficial short-cut to learning! But, the use of social information can vary depending on an animal’s age, and the environment and social context in which an animal develops or resides. In the wild, unfamiliar, unrelated, adult tree skinks can be lethally aggressive to juveniles in the wild. So, in our study, juveniles may not have been motivated to learn from an unknown adult that could, potentially, threaten their survival. A similar phenomenon has been observed in young guppies (Poecilia reticulata) – they don’t use social information from adults until they are large enough to not be impacted by aggressive advances of adults. All in all, we think that social feedback played a role in our finding that juvenile tree skinks did not use social information from unfamiliar, unrelated adult females.

Tree skink mothers often associate with their offspring, and this is thought to give them protection from other adult lizards that could, potentially, attack them. Photo by Julia Riley.

Learning is crucial for survival in the animal kingdom. Although we didn’t find evidence that juveniles would use social information presented by an unfamiliar, adult female, we did find that adult females socially learnt from one another. This is the first evidence that a family-living lizard is capable of learning from others!!! We also show that social learning is not guaranteed to be the default strategy of tree skinks. In fact, the use of social learning depends on a multitude of factors (i.e., the complexity of the task, the relationship between the demonstrator and learner, an individual’s age, etc.), and is likely only used when it benefits the learner.

Article references:

Whiting MJ, Xu F, Kar F, Riley JL, Bryne RW, and Noble DWA. 2018. Evidence for social learning in a family living lizard. Frontiers in Ecology and Evolution, doi: 10.3389/fevo.2018.00070

Riley JL,  Küchler A, Damasio T, Noble DWA, Byrne RW, Whiting MJ. 2018. Learning ability is unaffected by isolation rearing in a family-living lizard. Behavioural Ecology and Sociobiology, 72: 20, doi: 10.1007/s00265-017-2435-9

Brains and Brawn: Dominant lizards are better learners!

By Fonti Kar

Dominant individuals tend to have greater monopoly over food and mates and therefore have more offspring compared to subordinate individuals. Are these successes attributed to greater cognitive ability? Or are dominant individuals just better at freeloading from their clever subordinate counterparts?

We investigated whether dominant and subordinate eastern water skinks differ in their ability to learn from one and other (social learning). Previous work has shown in this species that young skinks tend to learn from older skinks, but age and dominance status and body size are inherently confounded. In other words, the age-dependent pattern may actually reflect a dominance effect, whereby young and therefore subordinate skinks tend to learn from older, dominant lizards.

In order to disassociate these confounding factors, we matched skinks closely in size and therefore age and allowed them to fight to determine their dominance statuses (Fig. 1). Winners of this fight was considered ‘dominant’, while the loser was considered ‘subordinate’. We then divided our skink pairs into two treatment groups, a ‘control’ group, where a skink watched their status counterparts do nothing and a ‘social learning’ group, where a skink was able to watch their status counterpart solve a foraging task e.g. a subordinate skink was able to watch a dominant skink, while a dominant skink was able to watch a subordinate skink (Fig. 2).

Control and social learning group set up. The control group watched their status counterparts do nothing. While the social learning group watched their status counterparts solve a task before receiving the task themselves

We gave the lizards two foraging tasks. In the first task, the lizards had to learn from their status counterparts how to learn to flip a blue lid to access a worm and ignore a white lid (association task). In the second task, the lizards had to unlearn the blue lid-worm association and learn to flip the white lid for the worm (reversal task). We then recorded how many trials it took for skinks to learn these tasks.

In the association task, skinks had to learn to flip the blue lid to access food. They then had to unlearn this association in the reversal task and learn to flip the white lid. See video below!


To our surprise, lizards did not seem to learn from the other lizard but instead relied on their own trial-and-error learning abilities. This was consistent for both dominant and subordinate lizards. We also found that dominant lizards learnt faster compared to subordinate lizards.

These results tell us a few neat things about social learning in the eastern water skink. Firstly, skinks were closely matched in size but they didn’t seem to learn from watching another skink. This seems to suggest that skinks may not want to learn from an individual of similar age and actually this may actually impede learning. Secondly, dominant individuals learnt faster compared to subordinate skinks implies that dominant skinks may be less prone to the stress associated with learning in the presence of another skink or they may indeed have both brains and brawn.

For more information, check out the paper published in Animal Cognition here

Meta-analysis on reptile developmental plasticity

It’s been many years in the making with Lisa Schwanz and Vaughn Stenhouse, but finally our utterly massive meta-analysis on the role of incubation temperatures on phenotypic variability and survival is finally out in Biological Reviews! We collated effect sizes from 92 different species across all major reptile orders taken from 175 different studies that manipulated incubation temperatures! Extracting all these data was no small task, but has lead to some important insights that have implications for our understanding of how climate change will affect reptiles and the role of early thermal environments more generally on developmental plasticity in reptiles.

There are lots of new stuff in the paper but a pretty graph (only one so as not to spoil anything) and the major conclusions are summarised below!

(1) The magnitude of the phenotypic effect of incubation temperature is moderate to large across orders, trait categories and ages. There is no evidence that this effect is substantially larger in any single order of reptiles, although data are sparse for Rhynchocephalia (a species-poor order) and Crocodilia.

(2) Effects of incubation temperature can persist for many months post-hatching. Sampling is poor for ages >1 year, thus more data would be useful in increasing our confidence in the persistence of effects.

(3) The effect of temperature on incubation duration is much stronger than on any other trait category. Survival also stands out with particularly strong effect sizes, while the relative strength of other trait categories varies in ways that compels more detailed comparison of reaction norms.

(4) Temperature fluctuations in the incubation environment potentially decrease the phenotypic effect of different mean temperatures, particularly when the temperature differences between treatments are large (although not significantly). More data are needed from fluctuating temperature regimes to assess more rigorously whether this tendency is real, and to quantify the impact of increasing fluctuation.

(5) On average, increased temperature changes lead to greater phenotypic effects. Despite expectations that the exact impact of warming incubation temperatures will depend on the trait studied and the shape of the reaction norm, we can say that, on average, nest temperatures that increase by 4°C would have a greater impact on nearly all phenotypes than would an increase of 2°C.

(6) The effect of increased incubation temperature depends on the temperatures experienced (mid-temperature), and this dependence varies according to trait type. Survival, morphology and performance were affected more strongly at extreme temperatures compared to intermediate temperatures indicating that increasingly warmer nest temperatures will accelerate change in these traits. Thus, collecting phenotypic data from extreme incubation temperatures is important.

Figure 1 – An example of how the average temperatures of pairwise treatments and their temperature difference impact the magnitude of effect for incubation duration.

(7) Substantial variation in the magnitude and direction of the phenotypic effects of incubation temperature remain unexplained. Future research should quantify the shape of the reaction norm to explore interspecific variation along with how parental and/or ecological effects might mediate responses.

If you want to find out more have a read of the paper!