Coupled information networks drive honeybee (Apis mellifera) collective foraging

Collective behaviour by eusocial insect colonies is typically achieved through multiple communication networks that produce complex behaviour at the group level but often appear to provide redundant or even competing information. A classic example occurs in honeybee (Apis mellifera) colonies, where both the dance communication system and robust scent-based mechanisms contribute to the allocation of a colony's workforce by regulating the flow of experienced foragers among known food sources. Here we analysed social connectivity patterns during the reactivation of experienced foragers to familiar feeding sites to show that these social information pathways are not simply multiple means to achieve the same end but intersect to play complementary roles in guiding forager behaviour. Using artificial feeding stations, we mimicked a natural scenario in which two forager groups were simultaneously collecting from distinct patches containing different flowering species. We then observed the reactivation of these groups at their familiar feeding sites after interrupting their foraging. Social network analysis revealed that temporarily unemployed individuals interacted more often and for longer with foragers that advertised a familiar versus unfamiliar foraging site. Due to such resource-based assortative mixing, network-based diffusion analysis estimated that reactivation events primarily resulted from interactions among bees that had been trained to the same feeding station and less so from different-feeder interactions. Both scent- and dance-based interactions strongly contributed to reactivation decisions. However, each bout of dance-following had an especially strong effect on a follower's likelihood of reactivation, particularly when dances indicated locations familiar to followers. Our findings illustrate how honeybee foragers can alter their social connectivity in ways that are likely to enhance collective outcomes by enabling foragers to rapidly access up-to-date information about familiar foraging sites. In addition, our results highlight how reliance on multiple communication mechanisms enables social insect workers to utilise flexible information-use strategies that are robust to variation in the availability of social information.

Biased cultural transmission of a social custom in chimpanzees

Cultural transmission studies in animals have predominantly focused on identifying between-group variation in tool-use techniques, while immaterial cultures remain understudied despite their potential for highlighting similarities between human and animal culture. Here, using long-term data from two chimpanzee communities, we tested whether one of chimpanzees' most enigmatic social customs-the grooming handclasp-is culturally transmitted by investigating the influence of well-documented human transmission biases on their variational preferences. After identifying differences in style preferences between the communities, we show that older and dominant individuals exert more influence over their partners' handclasp styles. Mothers were equally likely to influence their offspring's preferences as nonkin, indicating that styles are transmitted both vertically and obliquely. Last, individuals gradually converged on the group style, suggesting that conformity guides chimpanzees' handclasp preferences. Our findings show that chimpanzees' social lives are influenced by cultural transmission biases that hitherto were thought to be uniquely human.

Social transmission in the wild can reduce predation pressure on novel prey signals

Social transmission of information is taxonomically widespread and could have profound effects on the ecological and evolutionary dynamics of animal communities. Demonstrating this in the wild, however, has been challenging. Here we show by field experiment that social transmission among predators can shape how selection acts on prey defences. Using artificial prey and a novel approach in statistical analyses of social networks, we find that blue tit (Cyanistes caeruleus) and great tit (Parus major) predators learn about prey defences by watching others. This shifts population preferences rapidly to match changes in prey profitability, and reduces predation pressure from naïve predators. Our results may help resolve how costly prey defences are maintained despite influxes of naïve juvenile predators, and suggest that accounting for social transmission is essential if we are to understand coevolutionary processes.

The modularity of a social group does not affect the transmission speed of a novel, socially learned behaviour, or the formation of local variants

The structure of a group is critical in determining how a socially learnt behaviour will spread. Predictions from theoretical models indicate that specific parameters of social structure differentially influence social transmission. Modularity describes how the structure of a group or network is divided into distinct subgroups or clusters. Theoretical modelling indicates that the modularity of a network will predict the rate of behavioural spread within a group, with higher modularity slowing the rate of spread and facilitating the establishment of local behavioural variants which can prelude local cultures. Despite prolific modelling approaches, empirical tests via manipulations of group structure remain scarce. We experimentally manipulated the modularity of populations of domestic fowl chicks, Gallus gallus domesticus, to affect the transmission of a novel foraging behaviour. We compared the spread of behaviour in populations with networks of high or low modularity against control populations where social transmission was prevented. We found the foraging behaviour to spread socially between individuals when the social transmission was permitted; however, modularity did not increase the speed of behavioural spread nor lead to the initial establishments of shared behavioural variants. This result suggests that factors in the social transmission process additional to the network structure may influence behavioural spread.

Detecting and quantifying social transmission using network-based diffusion analysis

Although social learning capabilities are taxonomically widespread, demonstrating that freely interacting animals (whether wild or captive) rely on social learning has proved remarkably challenging. Network-based diffusion analysis (NBDA) offers a means for detecting social learning using observational data on freely interacting groups. Its core assumption is that if a target behaviour is socially transmitted, then its spread should follow the connections in a social network that reflects social learning opportunities. Here, we provide a comprehensive guide for using NBDA. We first introduce its underlying mathematical framework and present the types of questions that NBDA can address. We then guide researchers through the process of selecting an appropriate social network for their research question; determining which NBDA variant should be used; and incorporating other variables that may impact asocial and social learning. Finally, we discuss how to interpret an NBDA model's output and provide practical recommendations for model selection. Throughout, we highlight extensions to the basic NBDA framework, including incorporation of dynamic networks to capture changes in social relationships during a diffusion and using a multi-network NBDA to estimate information flow across multiple types of social relationship. Alongside this information, we provide worked examples and tutorials demonstrating how to perform analyses using the newly developed nbda package written in the R programming language.

Learning strategies and long-term memory in Asian short-clawed otters (Aonyx cinereus)

Social learning, where information is acquired from others, is taxonomically widespread. There is growing evidence that animals selectively employ 'social learning strategies', which determine e.g. when to copy others instead of learning asocially and whom to copy. Furthermore, once animals have acquired new information, e.g. regarding profitable resources, it is beneficial for them to commit it to long-term memory (LTM), especially if it allows access to profitable resources in the future. Research into social learning strategies and LTM has covered a wide range of taxa. However, otters (subfamily Lutrinae), popular in zoos due to their social nature and playfulness, remained neglected until a recent study provided evidence of social learning in captive smooth-coated otters (Lutrogale perspicillata), but not in Asian short-clawed otters (Aonyx cinereus). We investigated Asian short-clawed otters' learning strategies and LTM performance in a foraging context. We presented novel extractive foraging tasks twice to captive family groups and used network-based diffusion analysis to provide evidence of a capacity for social learning and LTM in this species. A major cause of wild Asian short-clawed otter declines is prey scarcity. Furthering our understanding of how they learn about and remember novel food sources could inform key conservation strategies.

Personality composition determines social learning pathways within shoaling fish

In shaping how individuals explore their environment and interact with others, personality may mediate both individual and social learning. Yet increasing evidence indicates that personality expression is contingent on social context, suggesting that group personality composition may be key in determining how individuals learn about their environment. Here, we used recovery latency following simulated predator attacks to identify Trinidadian guppies (Poecilia reticulata) that acted in a consistently bold or shy manner. We then employed network-based diffusion analysis to track the spread of a novel foraging behaviour through groups containing different proportions of bold and shy fish. Informed associates promoted learning to a greater extent in bold individuals, but only within groups composed predominately of bold fish. As the proportion of shy fish within groups increased, bold individuals instead emerged as especially effective demonstrators that facilitated learning in others. Individuals were also more likely to learn overall within shy-dominated groups than in bold-dominated ones. We demonstrate that whether and how individuals learn is conditional on group personality composition, indicating that selection may favour traits enabling individuals to better match their behavioural phenotype to their social environment.

Detecting and quantifying social transmission using network-based diffusion analysis

Although social learning capabilities are taxonomically widespread, demonstrating that freely interacting animals (whether wild or captive) rely on social learning has proved remarkably challenging. Network-based diffusion analysis (NBDA) offers a means for detecting social learning using observational data on freely interacting groups. Its core assumption is that if a target behaviour is socially transmitted, then its spread should follow the connections in a social network that reflects social learning opportunities. Here, we provide a comprehensive guide for using NBDA. We first introduce its underlying mathematical framework and present the types of questions that NBDA can address. We then guide researchers through the process of selecting an appropriate social network for their research question; determining which NBDA variant should be used; and incorporating other variables that may impact asocial and social learning. Finally, we discuss how to interpret an NBDA model's output and provide practical recommendations for model selection. Throughout, we highlight extensions to the basic NBDA framework, including incorporation of dynamic networks to capture changes in social relationships during a diffusion and using a multi-network NBDA to estimate information flow across multiple types of social relationship. Alongside this information, we provide worked examples and tutorials demonstrating how to perform analyses using the newly developed nbda package written in the R programming language.

Network-based diffusion analysis reveals context-specific dominance of dance communication in foraging honeybees

The honeybee (Apis mellifera) dance communication system is a marvel of collective behaviour, but the added value it brings to colony foraging efficiency is poorly understood. In temperate environments, preventing communication of foraging locations rarely decreases colony food intake, potentially because simultaneous transmission of olfactory information also plays a major role in foraging. Here, we employ social network analyses that quantify information flow across multiple temporally varying networks (each representing a different interaction type) to evaluate the relative contributions of dance communication and hive-based olfactory information transfer to honeybee recruitment events. We show that virtually all successful recruits to novel locations rely upon dance information rather than olfactory cues that could otherwise guide them to the same resource. Conversely, during reactivation to known sites, dances are relatively less important, as foragers are primarily guided by olfactory information. By disentangling the contributions of multiple information networks, the contexts in which dance communication truly matters amid a complex system full of redundancy can now be identified.

Honeybees have a sophisticated system to communicate foraging locations through a “dance”, but they also share food-related olfactory cues. Here, Hasenjager and colleagues use social network analysis to disentangle how foraging information is transmitted through these systems in different contexts.

Wild primates copy higher-ranked individuals in a social transmission experiment

Little is known about how multiple social learning strategies interact and how organisms integrate both individual and social information. Here we combine, in a wild primate, an open diffusion experiment with a modeling approach: Network-Based Diffusion Analysis using a dynamic observation network. The vervet monkeys we study were not provided with a trained model; instead they had access to eight foraging boxes that could be opened in either of two ways. We report that individuals socially learn the techniques they observe in others. After having learnt one option, individuals are 31x more likely to subsequently asocially learn the other option than individuals naïve to both options. We discover evidence of a rank transmission bias favoring learning from higher-ranked individuals, with no evidence for age, sex or kin bias. This fine-grained analysis highlights a rank transmission bias in a field experiment mimicking the diffusion of a behavioral innovation.

Social learning in otters

The use of information provided by others to tackle life's challenges is widespread, but should not be employed indiscriminately if it is to be adaptive. Evidence is accumulating that animals are indeed selective and adopt 'social learning strategies'. However, studies have generally focused on fish, bird and primate species. Here we extend research on social learning strategies to a taxonomic group that has been neglected until now: otters (subfamily Lutrinae). We collected social association data on captive groups of two gregarious species: smooth-coated otters (Lutrogale perspicillata), known to hunt fish cooperatively in the wild, and Asian short-clawed otters (Aonyx cinereus), which feed individually on prey requiring extractive foraging behaviours. We then presented otter groups with a series of novel foraging tasks, and inferred social transmission of task solutions with network-based diffusion analysis. We show that smooth-coated otters can socially learn how to exploit novel food sources and may adopt a 'copy when young' strategy. We found no evidence for social learning in the Asian short-clawed otters. Otters are thus a promising model system for comparative research into social learning strategies, while conservation reintroduction programmes may benefit from facilitating the social transmission of survival skills in these vulnerable species.

Social networks predict selective observation and information spread in ravens

Animals are predicted to selectively observe and learn from the conspecifics with whom they share social connections. Yet, hardly anything is known about the role of different connections in observation and learning. To address the relationships between social connections, observation and learning, we investigated transmission of information in two raven (Corvus corax) groups. First, we quantified social connections in each group by constructing networks on affiliative interactions, aggressive interactions and proximity. We then seeded novel information by training one group member on a novel task and allowing others to observe. In each group, an observation network based on who observed whose task-solving behaviour was strongly correlated with networks based on affiliative interactions and proximity. Ravens with high social centrality (strength, eigenvector, information centrality) in the affiliative interaction network were also central in the observation network, possibly as a result of solving the task sooner. Network-based diffusion analysis revealed that the order that ravens first solved the task was best predicted by connections in the affiliative interaction network in a group of subadult ravens, and by social rank and kinship (which influenced affiliative interactions) in a group of juvenile ravens. Our results demonstrate that not all social connections are equally effective at predicting the patterns of selective observation and information transmission.

Bayesian Spatial NBDA for Diffusion Data with Home-Base Coordinates

Network-based diffusion analysis (NBDA) is a statistical method that allows the researcher to identify and quantify a social influence on the spread of behaviour through a population. Hitherto, NBDA analyses have not directly modelled spatial population structure. Here we present a spatial extension of NBDA, applicable to diffusion data where the spatial locations of individuals in the population, or of their home bases or nest sites, are available. The method is based on the estimation of inter-individual associations (for association matrix construction) from the mean inter-point distances as represented on a spatial point pattern of individuals, nests or home bases. We illustrate the method using a simulated dataset, and show how environmental covariates (such as that obtained from a satellite image, or from direct observations in the study area) can also be included in the analysis. The analysis is conducted in a Bayesian framework, which has the advantage that prior knowledge of the rate at which the individuals acquire a given task can be incorporated into the analysis. This method is especially valuable for studies for which detailed spatially structured data, but no other association data, is available. Technological advances are making the collection of such data in the wild more feasible: for example, bio-logging facilitates the collection of a wide range of variables from animal populations in the wild. We provide an R package, spatialnbda, which is hosted on the Comprehensive R Archive Network (CRAN). This package facilitates the construction of association matrices with the spatial x and y coordinates as the input arguments, and spatial NBDA analyses.

Familiarity affects social network structure and discovery of prey patch locations in foraging stickleback shoals

Numerous factors affect the fine-scale social structure of animal groups, but it is unclear how important such factors are in determining how individuals encounter resources. Familiarity affects shoal choice and structure in many social fishes. Here, we show that familiarity between shoal members of sticklebacks (Gasterosteus aculeatus) affects both fine-scale social organization and the discovery of resources. Social network analysis revealed that sticklebacks remained closer to familiar than to unfamiliar individuals within the same shoal. Network-based diffusion analysis revealed that there was a strong untransmitted social effect on patch discovery, with individuals tending to discover a task sooner if a familiar individual from their group had previously done so than if an unfamiliar fish had done so. However, in contrast to the effect of familiarity, the frequency with which individuals had previously associated with one another had no effect upon the likelihood of prey patch discovery. This may have been due to the influence of fish on one another's movements; the effect of familiarity on discovery of an empty ‘control’ patch was as strong as for discovery of an actual prey patch. Our results demonstrate that factors affecting fine-scale social interactions can also influence how individuals encounter and exploit resources.

The spread of a novel behavior in wild chimpanzees: New insights into the ape cultural mind

For years, the animal culture debate has been dominated by the puzzling absence of direct evidence for social transmission of behavioral innovations in the flagship species of animal culture, the common chimpanzee. Although social learning of novel behaviors has been documented in captivity, critics argue that these findings lack ecological validity and therefore may not be relevant for understanding the evolution of culture. For the wild, it is possible that group-specific behavioral differences emerge because group members respond individually to unspecified environmental differences, rather than learning from each other. In a recent paper, we used social network analyses in wild chimpanzees (Pan troglodytes schweinfurthii) to provide direct evidence for social transmission of a behavioral innovation, moss-sponging, to extract water from a tree hole. Here, we discuss the implications of our findings and how our new methodological approach could help future studies of social learning and culture in wild apes.

Interspecific social networks promote information transmission in wild songbirds

Understanding the functional links between social structure and population processes is a central aim of evolutionary ecology. Multiple types of interactions can be represented by networks drawn for the same population, such as kinship, dominance or affiliative networks, but the relative importance of alternative networks in modulating population processes may not be clear. We illustrate this problem, and a solution, by developing a framework for testing the importance of different types of association in facilitating the transmission of information. We apply this framework to experimental data from wild songbirds that form mixed-species flocks, recording the arrival (patch discovery) of individuals to novel foraging sites. We tested whether intraspecific and interspecific social networks predicted the spread of information about novel food sites, and found that both contributed to transmission. The likelihood of acquiring information per unit of connection to knowledgeable individuals increased 22-fold for conspecifics, and 12-fold for heterospecifics. We also found that species varied in how much information they produced, suggesting that some species play a keystone role in winter foraging flocks. More generally, these analyses demonstrate that this method provides a powerful approach, using social networks to quantify the relative transmission rates across different social relationships.

Social network analysis shows direct evidence for social transmission of tool use in wild chimpanzees

Network-based diffusion analysis demonstrates that a novel tool-use behavior, “moss-sponging”, spread via social learning in a wild East-African chimpanzee community.

Social network analysis methods have made it possible to test whether novel behaviors in animals spread through individual or social learning. To date, however, social network analysis of wild populations has been limited to static models that cannot precisely reflect the dynamics of learning, for instance, the impact of multiple observations across time. Here, we present a novel dynamic version of network analysis that is capable of capturing temporal aspects of acquisition—that is, how successive observations by an individual influence its acquisition of the novel behavior. We apply this model to studying the spread of two novel tool-use variants, “moss-sponging” and “leaf-sponge re-use,” in the Sonso chimpanzee community of Budongo Forest, Uganda. Chimpanzees are widely considered the most “cultural” of all animal species, with 39 behaviors suspected as socially acquired, most of them in the domain of tool-use. The cultural hypothesis is supported by experimental data from captive chimpanzees and a range of observational data. However, for wild groups, there is still no direct experimental evidence for social learning, nor has there been any direct observation of social diffusion of behavioral innovations. Here, we tested both a static and a dynamic network model and found strong evidence that diffusion patterns of moss-sponging, but not leaf-sponge re-use, were significantly better explained by social than individual learning. The most conservative estimate of social transmission accounted for 85% of observed events, with an estimated 15-fold increase in learning rate for each time a novice observed an informed individual moss-sponging. We conclude that group-specific behavioral variants in wild chimpanzees can be socially learned, adding to the evidence that this prerequisite for culture originated in a common ancestor of great apes and humans, long before the advent of modern humans.

Chimpanzees are widely considered as the most “cultural” of all animals, despite the lack of direct evidence for the spread of novel behaviors through social learning in the wild. Here, we present a novel, dynamic network-based diffusion analysis to describe the acquisition patterns of novel tool-use behavior in the Sonso chimpanzee community of Budongo Forest, Uganda. We find strong evidence for social transmission of “moss-sponging” (the production of a sponge consisting of moss) along the innovators' social network, demonstrating that wild chimpanzees learn novel tool-use behaviors from each other and supporting the more general claim that some of the observed behavioral diversity in wild chimpanzees should be interpreted as “cultural.” Our model also estimated that, for each new observation, naïve individuals enhanced their chances of developing moss-sponging by a factor of 15. We conclude that group-specific behavioral variants can be socially learned in wild chimpanzees, addressing an important critique of the claim of culture in our closest relatives.

Perching but not foraging networks predict the spread of novel foraging skills in starlings

The directed social learning hypothesis suggests that information does not spread evenly through animal groups, but rather individual characteristics and patterns of physical proximity guide the social transmission of information along specific pathways. Network-based diffusion analysis (NBDA) allows researchers to test whether information spreads following a social network. However, the explanatory power of different social networks has rarely been compared, and current models do not easily accommodate random effects (e.g. allowing for individuals within groups to correlate in their asocial solving rates). We tested whether the spread of two novel foraging skills through captive starling groups was affected by individual- and group-level random and fixed effects (i.e. sex, age, body condition, dominance rank and demonstrator status) and perching or foraging networks. We extended NBDA to include random effects and conducted model discrimination in a Bayesian context. We found that social learning increased the rate at which birds acquired the novel foraging task solutions by 6.67 times, and acquiring one of the two novel foraging task solutions facilitated the asocial acquisition of the other. Surprisingly, the spread of task solutions followed the perching rather than the foraging social network. Upon acquiring a task solution, foraging performance was facilitated by the presence of group mates. Our results highlight the importance of considering more than one social network when predicting the spread of information through animal groups. This article is part of a Special Issue entitled: Cognition in the wild.

Diffusion dynamics of socially learned foraging techniques in squirrel monkeys

Social network analyses and experimental studies of social learning have each become important domains of animal behavior research in recent years yet have remained largely separate. Here we bring them together, providing the first demonstration of how social networks may shape the diffusion of socially learned foraging techniques. One technique for opening an artificial fruit was seeded in the dominant male of a group of squirrel monkeys and an alternative technique in the dominant male of a second group. We show that the two techniques spread preferentially in the groups in which they were initially seeded and that this process was influenced by monkeys' association patterns. Eigenvector centrality predicted both the speed with which an individual would first succeed in opening the artificial fruit and the probability that they would acquire the cultural variant seeded in their group. These findings demonstrate a positive role of social networks in determining how a new foraging technique diffuses through a population.

More on how and why: a response to commentaries

We are grateful to the commentators for taking the time to respond to our article. Too many interesting and important points have been raised for us to tackle them all in this response, and so in the below we have sought to draw out the major themes. These include problems with both the term 'ultimate causation' and the proximate-ultimate causation dichotomy more generally, clarification of the meaning of reciprocal causation, discussion of issues related to the nature of development and phenotypic plasticity and their roles in evolution, and consideration of the need for an extended evolutionary synthesis.

More on how and why: a response to commentaries

We are grateful to the commentators for taking the time to respond to our article. Too many interesting and important points have been raised for us to tackle them all in this response, and so in the below we have sought to draw out the major themes. These include problems with both the term 'ultimate causation' and the proximate-ultimate causation dichotomy more generally, clarification of the meaning of reciprocal causation, discussion of issues related to the nature of development and phenotypic plasticity and their roles in evolution, and consideration of the need for an extended evolutionary synthesis.

Information flow through threespine stickleback networks without social transmission

Social networks can result in directed social transmission of learned information, thus influencing how innovations spread through populations. Here we presented shoals of threespine sticklebacks (Gasterosteous aculeatus) with two identical foraging tasks and applied network-based diffusion analysis (NBDA) to determine whether the order in which individuals in a social group contacted and solved the tasks was affected by the group's network structure. We found strong evidence for a social effect on discovery of the foraging tasks with individuals tending to discover a task sooner when others in their group had previously done so, and with the spread of discovery of the foraging tasks influenced by groups' social networks. However, the same patterns of association did not reliably predict spread of solution to the tasks, suggesting that social interactions affected the time at which the tasks were discovered, but not the latency to its solution following discovery. The present analysis, one of the first applications of NBDA to a natural animal system, illustrates how NBDA can lead to insight into the mechanisms supporting behaviour acquisition that more conventional statistical approaches might miss. Importantly, we provide the first compelling evidence that the spread of novel behaviours can result from social learning in the absence of social transmission, a phenomenon that we refer to as an untransmitted social effect on learning.

Sex ratio affects sex-specific innovation and learning in captive ruffed lemurs (Varecia variegata and Varecia rubra)

Recent years have witnessed extensive research into problem solving and innovation in primates, yet lemurs have not been subjected to the same level of attention as apes and monkeys, and the social context in which novel behavior appears has rarely been considered. We gave novel foraging puzzlebox devices to seven groups of ruffed lemurs (Varecia variegata and Varecia rubra) to examine the factors affecting rates of innovation and social learning. We found, across a range of group sex ratios, that animals of the less-represented sex were more likely to contact and solve the puzzlebox sooner than those of the more-represented sex. We established that while some individuals were able to solve the puzzleboxes there was no evidence of social learning. Our findings are consistent with previously reported male deference as a sexual strategy, but we conclude that the need for male deference diminishes when, within a group, males are rare.

The evolutionary basis of human social learning

Humans are characterized by an extreme dependence on culturally transmitted information. Such dependence requires the complex integration of social and asocial information to generate effective learning and decision making. Recent formal theory predicts that natural selection should favour adaptive learning strategies, but relevant empirical work is scarce and rarely examines multiple strategies or tasks. We tested nine hypotheses derived from theoretical models, running a series of experiments investigating factors affecting when and how humans use social information, and whether such behaviour is adaptive, across several computer-based tasks. The number of demonstrators, consensus among demonstrators, confidence of subjects, task difficulty, number of sessions, cost of asocial learning, subject performance and demonstrator performance all influenced subjects' use of social information, and did so adaptively. Our analysis provides strong support for the hypothesis that human social learning is regulated by adaptive learning rules.