Journal of Zoology - 2015 - Leadbeater - What Evolves in The Evolution of Social Learning

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Journal of Zoology

bs_bs_banner

Journal of Zoology. Print ISSN 0952-8369

MINI-SERIES

What evolves in the evolution of social learning?


E. Leadbeater
School of Biological Sciences, Royal Holloway University of London, Surrey, UK

Keywords Abstract
social learning; associative learning; social
information use. Social learning is fundamental to social life across the animal kingdom, but we still
know little about how natural selection has shaped social learning abilities on a
Correspondence proximate level. Sometimes, complex social learning phenomena can be entirely
Ellouise Leadbeater, School of Biological explained by Pavlovian processes that have little to do with the evolution of
Sciences, Royal Holloway University of sociality. This implies that the ability to learn socially could be an exaptation, not
London, Egham Hill, Surrey TW20 0EX, UK. an adaptation, to social life but not that social learning abilities have been left
Email: Elli.Leadbeater@rhul.ac.uk untouched by natural selection. I discuss new empirical evidence for associative
learning in social information use, explain how natural selection might facilitate
Editor: Heike Lutermann the associative learning process and discuss why such studies are changing the way
that we think about social learning.
Received 16 June 2014; revised 21 August
2014; accepted 03 September 2014

doi:10.1111/jzo.12197

Introduction led to the acquisition of this association. Note that I use the
term ‘social information’ to describe any social stimulus that
Across the animal kingdom, from human groups to primate provides information about the state of the environment,
troops, bird flocks and insect colonies, social animals capital- including both social cues that are produced by animals as a
ize upon opportunities to learn from others that face the same co-incidental by-product of routine activities and signals that
challenges. Social learning lies at the heart of animal social life have been shaped by selection specifically to promote trans-
and potentially enhances abilities to assess the environment mission of information from one animal to another (Danchin
under a wide range of ecological circumstances (Rendell et al., et al., 2004) [Note: Signals and social cues are typically con-
2010). Yet, we still have little idea of how natural selection has sidered distinct by social learning researchers, but there is no a
shaped social learning processes – if it has even shaped them at priori reason to expect that social learning mechanisms should
all. Evidence suggests that at least some forms of social learn- differ between the two, especially given that signals most likely
ing fall within the capabilities, if not the typical behavioural evolve from social cues (Tinbergen, 1952)].
repertoire, of species that spend most of their life cycle alone Like most cognitive traits, the link between social learning
(e.g. octopuses: Fiorito & Scotto, 1992; tortoises: Wilkinson and fitness in the wild has rarely been explored empirically
et al., 2010). But although the neural circuitry required for (Thornton et al. 2014; but see Sherman & Visscher, 2002;
social learning may be present in solitary animals, it does not Dornhaus & Chittka, 2004), yet theory predicts that learning
necessarily follow that these abilities have remained unim- from others has fitness benefits under the right circumstances
proved by natural selection during the evolutionary time that (Boyd & Richerson, 1985; Rogers, 1988; Rendell et al., 2010).
separates social species from their solitary ancestors. The focus of this review is not whether social learning is typi-
Social learning is a composite ability involving perceptual cally adaptive (see Giraldeau, Valone & Templeton, 2002;
or motivational responses to social information and resultant Laland, 2004; Kendal, Coolen & Laland, 2009; Rendell et al.,
learning that could guide future decisions made in the absence 2011; Grüter & Leadbeater 2014), but rather whether the
of social stimuli (Heyes, 2012, see also Lotem & Halpern, underlying cognitive mechanisms have been shaped by natural
2012). For example, picture a bee that observes another bee selection specifically to promote learning from others. In other
foraging from a sunflower, abandons its own preferred flower words, the question is whether social learning abilities are
species to join in and continues to forage on sunflowers after- exaptations – features that may enhance fitness but were not
wards (Leadbeater & Chittka, 2007). This individual has built by selection for their current role – or adaptations to social
learnt that sunflowers are a good nectar source (a learnt asso- life (Gould & Vrba, 1982). I will argue that simple, taxonomi-
ciation between two asocial stimuli), but the presence of the cally general associative learning processes can explain many
conspecific on the sunflower provided social information that social learning phenomena (Heyes, 1994, 2012). Such processes

4 Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
E. Leadbeater What evolves in the evolution of social learning?

are almost universal within the animal kingdom and have no Responses to social information
direct link to social life (Walters, Carew & Kandel 1981;
Rescorla & Holland 1982; Roman & Davis 2001). Social learning requires that animals respond to social stimuli
The implication is not that ‘associative learning’ and (Table 1), and these responses may themselves sometimes arise
‘natural selection’ are stark alternatives. Selection could through learnt associations. If a response to a particular social
modify responses to social information to increase the likeli- stimulus, such as conspecific presence, alarm vocalizations or
hood that animals are exposed to particular associations. For residual scent cues, is the result of learning (Fig. 1a), reliance
example, animals might be more likely to learn socially if they on social information should vary according to the rewards
are particularly motivated to approach conspecifics, or if social that such responses have evoked in the past. Evidence that
cues are more likely to capture visual attention than less bio- social information use is modified by experience is beginning
logically relevant stimuli (Heyes, 2012). Selection could also act to emerge from a diverse range of taxonomic groups, includ-
upon the asocial associations that develop as a result, making ing humans (Kendal et al., 2009; Harris & Corriveau, 2011;
some more likely to be learnt than others. For example, rhesus Reader, 2013). For example, juvenile house sparrows that are
monkeys can learn socially to associate snakes, but not flowers, reared with an adult model that consistently visits locations
with fear responses (Cook & Mineka, 1989; see also Chivers & containing food are more likely to later join others at feeders
Smith, 1994; Davies & Welbergen, 2009). I will discuss exam- than individuals reared with a model that visited locations
ples of each type of bias in the text below, arguing that the containing no food (Katsnelson et al., 2008). Direct manipu-
underlying cognitive requirement for social learning is still a lation of the benefits of joining in adults also influences the
simple, taxonomically general ability to form associations likelihood that individuals will join on subsequent occasions
between stimuli that are experienced together (classical condi- (Belmaker et al., 2012). Thus, individuals that have previously
tioning) or motor actions and their consequences (operant been differentially rewarded for using social information differ
conditioning). in their tendencies to join others when foraging in exactly the
In the first section of this review, I discuss why (on a proxi- same environment.
mate level) animals respond to social information. I begin by Such evidence suggests that learning modifies responses to
presenting evidence that animals learn to respond to social social information but does not reveal whether such responses
stimuli (conditioned responses) before discussing how such arise through learning. It is not clear that individuals that have
learning could be shaped by specialized input mechanisms never had the chance to learn would ignore social cues, but
(Heyes, 2012), and describing circumstances in which natural studies of naïve individuals that have never foraged socially
selection may have produced unconditioned responses that do can address this possibility. For example, although wild
not require learning at all. In the second section, I examine bumblebees avoid flowers upon which they detect olfactory
how such responses lead to new associations involving asocial ‘footprints’ left inadvertently by other recent visitors
stimuli, allowing animals to learn socially about the world (Goulson, Hawson & Stout, 1998), foragers that have never
around them. previously encountered these scent marks show no aversion to

Figure 1 Responses to social stimuli. (a) Con-


ditioned responses in which animals learn
to (positively or negatively) associate social
stimuli with reward. (b) Unconditioned
responses do not require learning but may be
subsequently modified by learning. For most
empirical examples of social information use,
there is as yet insufficient ontogenetic
evidence to determine whether responses
to social stimuli are conditioned or
unconditioned.

Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London 5
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
What evolves in the evolution of social learning? E. Leadbeater

them (Leadbeater & Chittka, 2011). If marked flowers are a more adaptive alternative if learning is time consuming or
manipulated to contain reward, such bees even develop an risky (Reader, 2013). For example, social insects respond to
attraction to the marks (Saleh & Chittka, 2006). conspecific alarm pheromones with vigorous nest defence
When animals learn to use social information, are social behaviour (Moritz & Burgin, 1987), despite rarely having
stimuli learnt about as quickly as asocial alternatives? Even if much opportunity to learn that such substances predict threat.
the learning mechanism itself is entirely associative (Fig. 1a), Avian nestling responses to parental alarm signals can arise
selection could facilitate the process by making animals more without learning, although they may be fine-tuned through
likely to approach social stimuli (and thus learn about them). experience (Davies, Madden & Butchart, 2004). It is note-
For example, Drosophila that carry the forS (sitter) allele at the worthy that many examples of unconditioned responses to
foraging locus are more likely to aggregate at food patches and social information relate to signals rather than inadvertently
consequently more likely to rely on social information than produced social cues, which might reflect the reliability of the
those that carry the forR (rover) allele (Foucaud et al., 2013). information provided. Signallers are under selection to
However, increased opportunity to obtain social information produce honest signals (Zahavi, 1975; but see Flower et al.
is but one of several potentially adaptive consequences of 2014 for an example of deceptive signalling between non-
grouping behaviour, so it seems unlikely that aggregation relatives), so the information content of signals may be filtered
evolved solely to promote social information use. Likewise, for quality by the signaller itself (although the same may be
schooling behaviour is key to social information use in true of social cues; Rendell et al., 2010). It seems reasonable to
guppies (Chapman, Ward & Krause, 2008), but it is unlikely hypothesize that although ancestral responses to signals may
that schooling evolved purely for that purpose. have involved learning, such responses may frequently
Alternatively, selection could lead individuals to perceptu- become fixed by natural selection, simply because the appro-
ally prioritize social stimuli. There is evidence that animals priate response is predictable.
pay more visual attention to biologically relevant stimuli (e.g. Thus far, I have discussed conditioned responses to social
fear-relevant stimuli; Öhman, Flykt & Esteves, 2001), but the stimuli that arise through associative learning and uncondi-
perceptual relevance of stimuli that provide social information tioned responses that may potentially arise through natural
has rarely been addressed (but see Kanwisher, 2000). selection (although note that other explanations for uncondi-
However, Dawson & Chittka (2012) found that bumblebees tioned responses exist, including sensory bias). However,
learnt to respond to heterospecific (honeybee) and conspecific simply responding to social information does not in itself
stimuli equally quickly, suggesting that neither perceptual nor qualify as social learning. In the next section, I introduce
motivational systems are adaptively specialized to prioritize associative processes by which such responses can become
conspecific cues in this case. However, because honeybees may conditioned to asocial stimuli, leading animals to learn about
provide biologically useful social information to at least some the environment around them.
bumblebee species, similar results might not be obtained for
less biologically relevant stimuli.
Responses to social information can lead to
As the preceding example demonstrates, there is no reason
social learning
why learnt social information use should involve only
conspecifics. Other species that share similar ecological Iberian green frog tadpoles respond to olfactory cues from
requirements may provide useful (or even, better) informa- injured conspecifics by reducing activity levels. However, such
tion (Seppanen, 2007), and some of the most convincing stimuli do not simply evoke immediate responses but can also
evidence for learnt social information use in the wild involves lead to learning about predator identities. If they are encoun-
interspecific interactions. For example, several bird and tered together with the scent of an unfamiliar predator, indi-
primate species respond to the alarm calls of heterospecifics viduals later produce the same response to the predator cues
when the two species exist in sympatry but not when they alone (Gonzalo, Lopez & Martin, 2007). In other words, the
are allopatric (Ramakrishnan & Coss, 2000; Magrath, Pitcher social ‘danger’ cue is a (probably unconditioned) stimulus that
& Gardner, 2009). This suggests that such responses are elicits a behavioural response (activity reduction) and
learnt rather than inherited genetically, but ontogenetic becomes conditioned to a new asocial stimulus (the predator
studies are necessary to rule out genetic variation on a cue) because the two are experienced together. As a result, the
microgeographic scale. Correspondingly, Haff & Magrath asocial stimulus comes to elicit the behavioural response alone
(2013) found that responses to New Holland honeyeater alarm (Fig. 2a). Similar predator identification phenomena are very
calls by sympatric white-browed scrubwren fledglings devel- common in the aquatic world (e.g. Suboski et al., 1990;
oped much later if honeyeaters were not present in the natal Wisenden, Chivers & Smith, 1997).
territory, arising several weeks after individuals had begun to There are conceptual parallels between these examples and
venture further afield. This suggests that exposure is key to two of the most classic examples of social learning. Firstly,
development. when rats encounter food odours in association with carbon
The ontogeny of most responses to social information has disulphide (a component of rat breath), they develop a pref-
never been studied in detail (Galef, 2013), so it is hard to erence for foods of that type (Galef et al., 1988). It has been
estimate the generality of learning as an explanation for suggested that carbon disulphide acts as an inherently attrac-
responses to social information. However, it seems likely that tive unconditioned stimulus (US), and this response becomes
unconditioned responses (Fig. 1b) to social stimuli might offer conditioned to the food odour stimulus (CS) when the two are

6 Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
E. Leadbeater What evolves in the evolution of social learning?

Figure 2 Learning about asocial stimuli


through social stimuli. Animals might learn
about features of the environment through
either (a) unconditioned or (b) conditioned
responses to social stimuli, which become
associated with new asocial stimuli. In the
latter case, second-order conditioning occurs.
Again, in many cases, there is not yet
sufficient ontogenetic work to distinguish
between the two mechanisms (but see
Dawson et al., 2013 for an exception).

encountered together (Galef & Durlach, 1993; Heyes, 2012). through early licking of rat pups by mothers. In support of
Thus, the food odour later elicits attraction when presented this suggestion, pups that are deprived of maternal licking in
alone. Secondly, young rhesus monkeys learn to respond fear- early life subsequently fail to learn food preferences from
fully to snakes by observing frightened conspecifics (Cook & conspecifics (Lindeyer, Meaney & Reader, 2013).
Mineka, 1989). Cook & Mineka (1989) suggest that the sight When there is clear evidence that the motivational value of
of frightened conspecifics (US) elicits an unconditioned fright social stimuli is learnt, the ‘observational conditioning’
response in observers that becomes conditioned to the snake process is conceptually identical to a Pavlovian process termed
stimulus (CS) – a process termed ‘observational conditioning’. ‘second-order conditioning’ (Rizley & Rescorla, 1972).
In both cases, just as in the Iberian green frog, a presu- Second-order conditioning is a mechanism by which individ-
mably unconditioned response to a social stimulus becomes uals form associations between a conditioned (CS2) and an
conditioned to an asocial stimulus that is encountered concur- unconditioned stimulus (US) through an intermediate condi-
rently and comes to elicit the response when presented alone. tioned stimulus (CS1) that is associated with both. For
Both of these examples (and particularly observational con- example, Pavlov’s trained dogs famously learnt to associate
ditioning) have long been viewed as associative learning pro- the presentation of food (US) with the sound of a metronome
cesses that could be explained without invoking adaptive (CS1, a first-order association). A group of the same dogs later
specialization (Cook & Mineka, 1989; Heyes, 1994, 2012). experienced the sound of the metronome together with pres-
However, this initially seems confusing; surely the mechanism entation of a black square (CS2), in the absence of food, and
has been shaped by selection at least in part because it depends subsequently salivated on encountering the black square alone
on initial unconditioned responses elicited by social informa- (Pavlov, 1927). In other words, the black square became asso-
tion. Carbon disulphide seems to be a uniquely relevant stimu- ciated with food even though the two had never been experi-
lus for rats, and pairing of odour cues with alternative stimuli enced together.
does not achieve the same effects (Galef et al., 1988). Yet the In the context of social learning, a social stimulus could
point is that although the response to social information (i.e. become conditioned to both food (or predators) and to fea-
the motivational properties of carbon disulphide) could be the tures of the environment, leading to social learning (Fig. 2b).
result of selective history, the learning process that connects There are numerous empirical examples whereby animals
this response (attraction) with asocial stimuli (food odours) is learn about asocial stimuli through intermediary social
associative (see later for a discussion of bias in this associa- stimuli. For example, rats that observe freezing responses in
tion). In many cases it is not truly clear whether the social conspecifics (evoked by foot shock) paired with an auditory
stimulus acquires its motivational value through natural selec- tone learn to respond similarly to the tone themselves, even
tion or through learning (Heyes, 2012). For example, carbon though they have never directly experienced the association
disulphide might simply acquire rewarding associations between shock and tone. Such learning is not observed in rats

Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London 7
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
What evolves in the evolution of social learning? E. Leadbeater

with limited social experience (Yusufishaq & Rosenkranz, ously learnt to regard cuckoos as a threat), and a large body
2013), suggesting that individuals may learn to associate of evidence now suggests that social learning is more likely
conspecific fear cues with pain during early development, but to occur when social information complements information
other explanations exist because socially deprived rats display that has been acquired previously alone (Kendal et al., 2009;
a wide range of abnormal social behaviours. Grüter & Leadbeater, 2014). Alternatively, such biases might
To unequivocally demonstrate that second-order condi- reflect genetically inherited predispositions to notice and
tioning is occurring, it is necessary to show that the motiva- prioritize biologically relevant stimuli as mentioned earlier in
tional properties of social information are learnt. the text (Öhman et al., 2001). For example, rhesus monkeys
Accordingly, we recently carried out an experiment in which learn socially to fear snakes, but they cannot learn to fear
bumblebees were initially trained to associate conspecific flowers in the same manner even when they have no previous
presence (social stimulus; CS1) with either a sucrose reward or experience of either stimulus (Cook & Mineka, 1989).
an aversive substance (US; Dawson et al., 2013). Subjects However, it is not clear that there is anything specifically
were then permitted to watch conspecifics choosing a particu- social about such biases. In other words, it seems likely that
lar flower colour (asocial stimulus; CS2) through a screen individuals may be more disposed to develop negative asso-
(Fig. 3; following Worden & Papaj, 2005). When later allowed ciations about snakes than flowers generally, irrespective of
to forage alone, bees that had learnt to associate social cues whether such associations develop through observing others
with sucrose preferred the flower colour that they had seen responding fearfully or through direct experience with floral
demonstrators choose, but the ‘aversive’ group actively or snake stimuli.
avoided the same colours. Importantly, naïve bees (that had
never foraged with conspecifics) were not influenced by
conspecific choices, suggesting that this apparently sophisti-
Conclusions
cated observational learning arises entirely through integra- What evolves in the evolution of social information use? In the
tion of two learnt associations. preceding paragraphs, I have presented explanations for social
In the example above, bees were equally likely to learn learning that are based on associative learning, an evolution-
aversive and attractive associations, even though in nature arily ancient ability to form associations between stimuli, or
conspecifics are unlikely to linger on unrewarding flowers. between motor actions and their consequences. I have also
However, in some cases, biologically relevant associations discussed cases where associative learning might be shaped by
seem to be more easily learnt than others. Reed warblers that genetically inherited traits, such as unconditioned responses to
observe conspecifics mobbing cuckoos become more defen- alarm pheromones, or tendencies to seek out conspecifics. The
sive towards cuckoos themselves, but observation of assumption is not that genetically inherited traits are neces-
conspecifics mobbing parrots (a harmless control) does not sarily adaptations for social learning, because they may derive
change responses to parrots (Davies & Welbergen, 2009). from sensory biases or have evolved for a different function.
This difference might simply reflect reinforcement of pre- Rather, evidence for genetic inheritance of social learning
existing learnt associations (i.e. individuals may have previ- traits simply hints at a possible role for natural selection in
shaping the social learning process.
The examples that I have used span the animal kingdom
rather than focusing on species that are typically considered
cognitively ‘advanced’ (e.g. cetaceans, primates), but there is
little evidence that social learning mechanisms are fundamen-
tally different in these species. For example, although social
learning of feeding techniques in humpback whales might
have exciting consequences for animal cultures (Allen et al.,
2013), it is unlikely that the underlying social learning mecha-
nism is particularly unusual or complex. A notable exception
is the ability to imitate, which has a very limited taxonomic
distribution, and is often put forward as instrumental in the
evolution of culture (Whiten & Ham, 1992). Yet, even imita-
tion may have an associative basis. Cook et al. (2014) propose
that when a human infant performs an action for itself (e.g. a
facial expression) and simultaneously sees an adult do the
same thing, neural representations of the motor action and the
social stimulus may become associated. This process may
explain the development of mirror neurons, which fire when
an action is performed and when the same action is observed,
Figure 3 A bumblebee (Bombus terrestris) observer watches model and are thought to underlie imitation.
‘demonstrator’ bees appear to preferentially forage on orange flowers We still have little idea of the role of experience in the
before the demonstrator is removed and the observer is permitted to development of most empirical examples of social learning.
make its own choices. Much more ontogenetic work is necessary if we are to progress

8 Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
E. Leadbeater What evolves in the evolution of social learning?

Table 1 Examples of responses to social stimuli and social learning that Brown, C. & Laland, K.N. (2002). Social learning of a novel
might occur as a result of such responses avoidance task in the guppy: conformity and social release.
Social learning: asocial stimulus Anim. Behav. 64, 41–47.
acquires motivational or aversive Chapman, B.B., Ward, A.J.W. & Krause, J. (2008). Schooling
Response to social stimulus properties and learning: early social environment predicts social learn-
Bee joins conspecific at Bee learns to associate flowers of ing ability in the guppy, Poecilia reticulata. Anim. Behav.
a flower that species with nectar rewards 76, 923–929.
(Leadbeater & Chittka, 2005, Chivers, D.P. & Smith, R.J.F. (1994). Fathead minnows
2007) Pimephales promelas acquire predator recognition when
Fish follows a demonstrator Fish learns to associate spatial
alarm substance is associated with the sight of unfamiliar
that is escaping from landmarks along the route with
fish. J. Fish Biol. 44, 273–285.
a predator escape (Brown & Laland 2002;
Cook, M. & Mineka, S. (1989). Observational conditioning of
Reader et al. 2003)
Monkey becomes frightened in Monkey learns to associate fear to fear-relevant and fear-irrelevant stimuli in Rhesus
response to frightened predator stimulus with fright monkeys. J. Abnorm. Psychol. 98, 448.
conspecific response (Cook & Mineka, Cook, R., Bird, G., Catmur, C., Press, C. & Heyes, C. (2014).
1989) Mirror neurons: from origin to function. Behav. Brain Sci.
Rat finds conspecific breath Rat learns to associate flavours 37, 177–192.
rewarding encountered on conspecific Danchin, E., Giraldeau, L.A., Valone, T.J. & Wagner, R.H.
breath with reward (Galef et al. (2004). Public information: from nosy neighbours to cul-
1988) tural evolution. Science 305, 487–491.
Chimpanzee performs same Chimpanzee learns to associate
Davies, N.B. & Welbergen, J.A. (2009). Social transmission of
motor action as demonstrator motor action with puzzle solution
a host defense against cuckoo parasitism. Science 324,
when solving puzzle (Whiten, 1998)
1318–1320.
Davies, N.B., Madden, J.R. & Butchart, S.H.M. (2004).
Learning fine-tunes a specific response of nestlings to the
with understanding the evolution of social information use at parental alarm calls of their own species. Proc. Roy. Soc.
the proximate level, rather than simply labelling its different Lond. Ser. B. 271, 2297–2304.
forms (Galef, 2013). For the moment, the available evidence Dawson, E.H. & Chittka, L. (2012). Conspecific and
suggests that although natural selection may have shaped heterospecific information use in bumblebees. PLoS ONE
responses to social stimuli, the learning process itself is asso- 7, e31444.
ciative (Heyes, 2012). On a proximate level, so far, it seems Dawson, E.H., Avargues-Weber, A., Chittka, L. &
that social learning is really just learning. Leadbeater, E. (2013). Learning by observation emerges
from simple associations in an insect model. Curr. Biol. 23,
Acknowledgements 727–730.
Dornhaus, A. & Chittka, L. (2004). Why do honey bees
E.L. is funded by a Leverhulme Trust Early Career Fellowship.
dance? Behav. Ecol. Sociobiol. 55, 395–401.
Thanks to Cecilia Heyes for helpful discussion regarding social
Fiorito, G. & Scotto, P. (1992). Observational learning in
learning in rats and to Christoph Grüter for the suggestion that
Octopus Vulgaris. Science 256, 545–547.
adaptive responses to signals might often be predictable.
Thanks again to Christoph Grüter, Erika Dawson, Nichola Flower, T., Gribble, M. & Ridley, A.R. (2014). Deception by
Raihani and two anonymous reviewers for comments on the flexible alarm mimicry in an African bird. Science 344,
manuscript. 513–516.
Foucaud, J., Philippe, A.-S., Moreno, C. & Mery, F. (2013).
A genetic polymorphism affecting reliance on personal
versus public information in Drosophila melanogaster.
References
Proc. Roy. Soc. Lond. Ser. B. 280, 10.
Allen, J., Weinrich, M., Hoppitt, W. & Rendell, L. (2013). Galef, B.G. (2013). Imitation and local enhancement:
Network-based diffusion analysis reveals cultural transmis- detrimental effects of consensus definitions on analyses
sion of lobtail feeding in Humpback whales. Science 340, of social learning in animals. Behav. Processes 100,
485–488. 123–130.
Belmaker, A., Motro, U., Feldman, M.W. & Lotem, A. Galef, B.G. & Durlach, P.J. (1993). Absence of blocking,
(2012). Learning to choose among social foraging strategies overshadowing and latent inhibition in social enhancement
in adult house sparrows (Passer domesticus). Ethology 118, of food preferences. Anim. Learn. Behav. 21, 214–220.
1111–1121. Galef, B.G., Mason, J.R., Preti, G. & Bean, N.J. (1988).
Boyd, R. & Richerson, P.J. (1985). Culture and the evolution- Carbon-disulfide – a semiochemical mediating socially-
ary process. Chicago: Chicago University Press. induced diet choice in rats. Physiol. Behav. 42, 119.

Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London 9
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
What evolves in the evolution of social learning? E. Leadbeater

Giraldeau, L.A., Valone, T.J. & Templeton, J.J. (2002). language or learning another? Proc. R. Soc. B. 276, 769–
Potential disadvantages of using social acquired informa- 774.
tion. Phil. Trans. Roy. Soc. Lond. B. 352, 1559–1566. Moritz, R.F.A. & Burgin, H. (1987). Group response to alarm
Gonzalo, A., Lopez, P. & Martin, J. (2007). Iberian green pheromones in social wasps and the honeybee. Ethology 76,
frog tadpoles may learn to recognize novel predators 15–26.
from chemical alarm cues of conspecifics. Anim. Behav. 74, Öhman, A., Flykt, A. & Esteves, F. (2001). Emotion drives
447–453. attention: detecting the snake in the grass. J. Exp. Psychol.
Gould, S.J. & Vrba, E.S. (1982). Exaptation – a missing term 130, 466–478.
in the science of form. Paleobiology 8, 4–15. Pavlov, I.P. (1927). Conditioned reflexes: an investigation of
Goulson, D., Hawson, S.A. & Stout, J.C. (1998). Foraging the physiological activity of the cerebral cortex. New York,
bumblebees avoid flowers already visited by conspecifics Dover.
or by other bumblebee species. Anim. Behav. 55, Ramakrishnan, U. & Coss, R.G. (2000). Recognition of
199–206. heterospecific alarm vocalizations by Bonnett Macaques
Grüter, C. & Leadbeater, E. (2014). Insights from insects (Macaca radiata). J. Comp. Psychol. 114, 3–12.
about adaptive social information use. Trends Ecol. Evol. Reader, S.M. (2013). Experiential effects on mirror systems
29, 177–184. and social learning: implications for social intelligence.
Haff, T.M. & Magrath, R.D. (2013). Eavesdropping on the Behav. Brain Sci. 37, 217–218.
neighbours: fledglings learn to respond to heterospecific Reader, S.M., Kendal, J.R. & Laland, K.N. (2003).
alarm calls. Anim. Behav. 85, 411–418. Social learning of foraging sites and escape routes
Harris, P.L. & Corriveau, K.H. (2011). Young children’s in wild Trinidadian guppies. Anim. Behav. 66, 729–
selective trust in informants. Philos. T. R. Soc. B. 366, 739.
1179–1187. Rendell, L., Boyd, R., Cownden, D., Enquist, M., Eriksson,
Heyes, C. (1994). Social learning in animals: categories and K., Feldman, M.W., Fogarty, L., Ghirlanda, S., Lillicrap,
mechanisms. Biol. Rev. 69, 207–231. T. & Laland, K.N. (2010). Why copy others? Insights
Heyes, C. (2012). What’s social about social learning? J. from the social learning strategies tournament. Science
Comp. Psychol. 126, 193–202. 328, 208–213.
Kanwisher, N. (2000). Domain specificity in face perception. Rendell, L., Fogarty, L., Hoppitt, J.E., Morgan, T.J.H.,
Nat. Neurosci. 3, 759–763. Webster, M.M. & Laland, K.N. (2011). Cognitive culture:
Katsnelson, E., Motro, U., Feldman, M.W. & Lotem, A. theoretical and empirical insights into social learning strat-
(2008). Early experience affects producer-scrounger forag- egies. Trends Cogn. Sci. 15, 68–76.
ing tendencies in the house sparrow. Anim. Behav. 75, Rescorla, R.A. & Holland, P.C. (1982). Behavioral studies of
1465–1472. associative learning in animals. Annu. Rev. Psychol. 33,
Kendal, R.L., Coolen, I. & Laland, K.N. (2009). Adaptive 265–308.
trade-offs in the use of social and personal information. In Rizley, R.C. & Rescorla, R.A. (1972). Associations in second-
Cognitive Ecology II: 249–271. Dukas, R. & Ratcliffe, J. order conditioning and sensory preconditioning. J. Comp.
(Eds). Chicago: University of Chicago Press. Physiol. Psychol. 81, 1–11.
Laland, K.N. (2004). Social learning strategies. Learn. Behav. Roman, G. & Davis, R.L. (2001). Molecular biology and
32, 4–14. anatomy of Drosophila olfactory associative learning.
Leadbeater, E. & Chittka, L. (2005). A new mode of informa- Bioessays. 23, 571–581.
tion transfer in foraging bumblebees? Curr. Biol. 15, R447– Rogers, A. (1988). Does biology constrain culture? Am.
R448. Anthropol. 90, 819–831.
Leadbeater, E. & Chittka, L. (2007). The dynamics of social Saleh, N. & Chittka, L. (2006). The importance of experience
learning in an insect model, the bumblebee (Bombus in the interpretation of conspecific chemical signals. Behav.
terrestris). Behav. Ecol. Sociobiol. 61, 1789–1796. Ecol. Sociobiol. 61, 215–220.
Leadbeater, E. & Chittka, L. (2011). Do inexperienced Seppanen, J.T. (2007). Social information use is a process
bumblebee foragers use scent marks as social information? across time, space and ecology, reaching heterospecifics.
Anim. Cogn. 14, 915–919. Ecology 88, 1622–1633.
Lindeyer, C.M., Meaney, M.J. & Reader, S.M. (2013). Sherman, G. & Visscher, P.K. (2002). Honeybee colonies
Early maternal care predicts reliance on social learning achieve fitness through dancing. Nature 419, 920–
about food in adult rats. Dev. Psychobiol. 55, 168–175. 922.
Lotem, A. & Halpern, J.Y. (2012). Coevolution of learning Suboski, M.D., Bain, S., Carty, A.E., McQuoid, L.M., Seelen,
and data-acquisition mechanisms: a model for cognitive M.I. & Seifert, M. (1990). Alarm reaction in acquisition
evolution. Philos. T. R. Soc. B. 367, 2686–2694. and social transmission of simulated predator recognition
Magrath, R.D., Pitcher, B.J. & Gardner, J.L. (2009). Recogni- by zebra danio fish Brachydanio rerio. J. Comp. Psychol.
tion of other species’ aerial alarm calls: speaking the same 104, 101–112.

10 Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London
14697998, 2015, 1, Downloaded from https://zslpublications.onlinelibrary.wiley.com/doi/10.1111/jzo.12197 by Cochrane Chile, Wiley Online Library on [27/04/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
E. Leadbeater What evolves in the evolution of social learning?

Thornton, A., Isden, J. & Madden, J.R. (2014). Toward wild Wilkinson, A., Kuenstner, K., Mueller, J. & Huber, L. (2010).
psychometrics: linking individual cognitive differences to Social learning in a non-social reptile (Geochelone
fitness. Behav. Ecol. (Online DOI: 10.1093/beheco/aru095) . carbonaria). Biol. Lett. 6, 614–616.
Tinbergen, N. (1952). ‘Derived’ activities, their causation, bio- Wisenden, B.D., Chivers, D.P. & Smith, R.J.F. (1997).
logical significance and emancipation during evolution. Q. Learned recognition of predation risk by Enallagma dam-
Rev. Biol. 27, 1–32. selfly larvae (Odonata, Zygoptera) on the basis of chemical
Walters, E.T., Carew T.J. & Kandel, E.R. (1981). Associative cues. J. Chem. Ecol. 23, 137–151.
learning in aplysia – evidence for conditioned fear in an Worden, B.D. & Papaj, D.R. (2005). Flower choice copying
invertebrate. Science. 211, 504–506. in bumblebees. Biol. Lett. 1, 504–507.
Whiten, A. (1998). Imitation of the sequential structure of Yusufishaq, S. & Rosenkranz, J.A. (2013). Post-weaning
actions by chimpanzees (Pan troglodytes). J. Comp. social isolation impairs observational fear conditioning.
Psychol. 112, 270–281. Behav. Brain Res. 242, 142–149.
Whiten, A. & Ham, R. (1992). On the nature and evolution of Zahavi, A. (1975). Mate selection – selection for a handicap.
imitation in the animal kingdom: reappraisal of a century J. Theor. Biol. 53, 205–214.
of research. Adv. Study Behav. 21, 239–283.

Journal of Zoology 295 (2015) 4–11 © 2014 The Zoological Society of London 11

You might also like