Social signal learning of the waggle dance in honey bees

Differential response by tandem leaders and followers to landmark-rich and landmark-poor environments

When animals use the same route repeatedly, they have the opportunity to update information that might help them to navigate more quickly and more accurately. Here we analyse ants involved in tandem running, in which the leader has evaluated a new nest and decided to recruit to it while the follower has chosen to be led and shown the route. We used a motorised gantry equipped with a camera to track the movements of tandem members on their tandem and return trips in a landmark-rich and a landmark-poor environment. Although the amount of visual navigational information did not affect the movements of leaders or followers on their tandem trip, the paths of followers were significantly more tortuous and their speeds significantly slower than those of leaders on their return trips in the landmark-poor environment. By contrast, there were no such differences between the followers and leaders on their return trips in the landmark-rich environment even though the return paths of followers in the landmark-rich environment were significantly more tortuous than that of leaders in the landmark-poor environment. Indeed, in the landmark-rich environment, the majority of the leaders' return paths had loops while most were straight in the landmark-poor environment. Thus, the availability of more information when many landmarks are present may induce tandem leaders to make the loops, typically associated with the paths of tandem followers. This suggests knowledgeable individuals slow down to update navigational information and has implications for the formation of leader oligarchies in tandem running.

A marine predator relies on both social cues and frequently updated memory to search for prey

Finding scattered resources is a challenge for marine predators, their search strategies shaping in turn their response to global changes. Three types of search strategies are generally considered: random opportunistic search, reliance on conspicuous social cues or long-term individual fidelity to predictably productive areas. More complex strategies, namely intermediate, composite and/or those varying across time and environmental conditions are likely to prevail but are rarely explored. We investigated memory strategies in Cape gannets, a marine predator heavily relying on social cues, and closely related to the northern gannet which shows long-term individual foraging fidelity. Repeat GPS tracks within seasons revealed that two different individuals leaving the colony more than 3 days apart showed random chances to forage in a similar direction, whereas within-individuals, birds persisted in a same direction for longer timescales, reaching random levels only after 10 days, or 9 consecutive trips. This strategy of transient individual preferences within a breeding season was observed in most individuals every year. These results suggest a consistent and complex search strategy mixing social cues and private information updated every few days. Our simple approach can readily be applied to other central place foragers to better understand the evolutionary ecology of search strategies.

Evolutionary biology of social expertise

There is increasing evidence that competent handling of social interactions among conspecifics has positive effects on individual fitness. While individual variation in social competence has been appreciated, the role of long-term experience in the acquisition of superior social skills has received less attention. With the goal of promoting further research, we integrate knowledge across disciplines to assess social expertise, defined as the characteristics, skills and knowledge allowing individuals with extensive social experience to perform significantly better than novices on a given social task. We focus on three categories of social behaviour. First, animals can gain from adjusting social behaviour towards individually recognised conspecifics that they interact with on a regular basis. For example, there is evidence that some territorial animals individually recognise their neighbours and modify their social interactions based on experience with each neighbour. Similarly, individuals in group-living species learn to associate with specific group members based on their expected benefits from such social connections. Individuals have also been found to devote considerable time and effort to learning about the spatial location and timing of sexual receptivity of opposite-sex neighbours to optimise reproduction. Second, signallers can enhance their signals, and receivers can refine their response to signals with experience. In many birds and insects, individuals can produce more consistent signals with experience, and females across a wide taxonomic range can adaptively adjust mating preferences after perceiving distinct male signals. Third, in many species, individuals that succeed in reproducing encounter the novel, complex task of caring for vulnerable offspring. Evidence from a few species of mammals indicates that mothers improve in providing for and protecting their young over successive broods. Finally, for social expertise to evolve, heritable variation in social expertise has to be positively associated with fitness. Heritable variation has been shown in traits contributing to social expertise including social attention, empathy, individual recognition and maternal care. There are currently limited data associating social expertise with fitness, most likely owing to sparse research effort. Exceptions include maternal care, signal refinement, and familiarity with neighbours and group members. Overall, there is evidence that individuals in many species keep refining their social skills with experience throughout life. Hence we propose promising lines of research that can quantify more thoroughly the development of social expertise and its effects on fitness.

Connecting and integrating cooperation within and between species

There has long been a fundamental divide in the study of cooperation: researchers focus either on cooperation within species, including but not limited to sociality, or else on cooperation between species, commonly termed mutualism. Here, we explore the ecologically and evolutionarily significant ways in which within- and between-species cooperation interact. We highlight two primary cross-linkages. First, cooperation of one type can change the context in which cooperation of the other type functions, and thus potentially its outcome. We delineate three possibilities: (i) within-species cooperation modulates benefits for a heterospecific partner; (ii) between-species cooperation affects the dynamics of within-species cooperation; and (iii) both processes take place interactively. The second type of cross-linkage emerges when resources or services that cooperation makes available are obtainable either from members of the same species or from different species. This brings cooperation at the two levels into direct interaction, to some extent obscuring the distinction between them. We expand on these intersections between within- and between-species cooperation in a diversity of taxa and interaction types. These interactions have the potential to weave together social networks and trophic dynamics, contributing to the structure and functioning of ecological communities in ways that are just beginning to be explored. This article is part of the theme issue 'Connected interactions: enriching food web research by spatial and social interactions'.

A framework for the emergence and analysis of language in social learning agents

Neural systems have evolved not only to solve environmental challenges through internal representations but also, under social constraints, to communicate these to conspecifics. In this work, we aim to understand the structure of these internal representations and how they may be optimized to transmit pertinent information from one individual to another. Thus, we build on previous teacher-student communication protocols to analyze the formation of individual and shared abstractions and their impact on task performance. We use reinforcement learning in grid-world mazes where a teacher network passes a message to a student to improve task performance. This framework allows us to relate environmental variables with individual and shared representations. We compress high-dimensional task information within a low-dimensional representational space to mimic natural language features. In coherence with previous results, we find that providing teacher information to the student leads to a higher task completion rate and an ability to generalize tasks it has not seen before. Further, optimizing message content to maximize student reward improves information encoding, suggesting that an accurate representation in the space of messages requires bi-directional input. These results highlight the role of language as a common representation among agents and its implications on generalization capabilities.

Tropical and montane Apis cerana show distinct dance-distance calibration curves

Social bees have evolved sophisticated communication systems to recruit nestmates to newly found food sources. As foraging ranges can vary from a few hundred meters to several kilometers depending on the environment or season, populations of social bee species living in different climate zones likely show specific adaptations in their recruitment communication. Accordingly, studies in the western honey bee, Apis mellifera, demonstrated that temperate populations exhibit shallower dance-calibration curves compared with tropical populations. Here, we report the first comparison of calibration curves for three Indian Apis cerana lineages: the tropical Apis indica, and the two montane Himalayan populations Apis cerana cerana (Himachal Pradesh) and Apis cerana kashmirensis (Jammu and Kashmir). We found that the colonies of the two montane A. cerana populations show dance-distance calibration curves with significantly shallower slopes than those of the tropical A. indica. Next, we transferred A. c. cerana colonies to Bangalore (∼ 2600 km away) to obtain calibration curves in the same location as A. indica. The common garden experiment confirmed this difference in slopes, implying that the lineages exhibit genetically fixed differences in dance-distance coding. However, the slopes of the calibration curves of the transferred A. c. cerana colonies were also significantly higher than those of the colonies tested in their original habitat, indicating an important effect of the environment. The differences in dance-distance coding between temperate and tropical A. cerana lineages resemble those described for Apis mellifera, suggesting that populations of both species independently evolved similar adaptations.

ARBUR, a machine learning-based analysis system for relating behaviors and ultrasonic vocalizations of rats

Deciphering how different behaviors and ultrasonic vocalizations (USVs) of rats interact can yield insights into the neural basis of social interaction. However, the behavior-vocalization interplay of rats remains elusive because of the challenges of relating the two communication media in complex social contexts. Here, we propose a machine learning-based analysis system (ARBUR) that can cluster without bias both non-step (continuous) and step USVs, hierarchically detect eight types of behavior of two freely behaving rats with high accuracy, and locate the vocal rat in 3-D space. ARBUR reveals that rats communicate via distinct USVs during different behaviors. Moreover, we show that ARBUR can indicate findings that are long neglected by former manual analysis, especially regarding the non-continuous USVs during easy-to-confuse social behaviors. This work could help mechanistically understand the behavior-vocalization interplay of rats and highlights the potential of machine learning algorithms in automatic animal behavioral and acoustic analysis.

Learning in the honey bee waggle dance

The waggle dance of honey bees is a classic example of complex behavior and communication in animals. Despite long being considered a completely fixed and innate behavior, recent work is showing a role for social learning in tuning components of the waggle dance in naïve bees.

The Honey Bee Colony's Criterion for Candidate Selection: "Ongoing" or "One-Shot"?

In the honey bee, the queen's death severely threatens the survival of the colony. In an emergency, new queens are reared from young worker larvae, where nepotism is thought to influence the choice of queen candidates by the workers. This article simulates the emergency queen-rearing process in a colony under natural conditions and records the results of colony selection (without nepotism). In queenless colonies, worker larvae aged three days or younger were preferred for queen rearing, and 1-day-old larvae were the first to be selected for the queen-cell cups. In the capping stage, the number of capped queen cells selected from the 1-day-old larvae was much higher than the 3-day-old larvae. On the first day, the number of emerging queens reared from 1-day-old larvae was significantly higher than the queens reared from 2-day-old and 3-day-old larvae. However, there was no significant difference in the birth weights of queens reared from 1-day-old, 2-day-old, or 3-day-old larvae. When the newly emerged queens were introduced into the original queenless colony, 1-day-old larval queens triggered more worker followers than 2-day-old larval queens. The expression of ovarian development-related genes (vg, hex110, and Jh) was higher in queens reared from 1-day-old larvae than those reared from 2-day-old and 3-day-old larvae, indicating that the quality of the queens reared from 1-day-old larvae is superior. This study shows that in the absence of nepotism, the colony selection of queen candidates at the larval stage, capping stage, and emerging stage is not final, but is gradually optimized to maximize colony development through a "quality control" process.

New comparative genomic evidence supporting the proteomic diversification role of A-to-I RNA editing in insects

Adenosine-to-inosine (A-to-I) RNA editing, resembling A-to-G mutation, confers adaptiveness by increasing proteomic diversity in a temporal-spatial manner. This evolutionary theory named "proteomic diversifying hypothesis" has only partially been tested in very few organisms like Drosophila melanogaster, mainly by observing the positive selection on nonsynonymous editing events. To find additional genome-wide evidences supporting this interesting assumption, we retrieved the genomes of four Drosophila species and collected 20 deep-sequenced transcriptomes of different developmental stages and neuron populations of D. melanogaster. We systematically profiled the RNA editomes in these samples and performed meticulous comparative genomic analyses. Further evidences were found to support the diversifying hypothesis. (1) None of the nonsynonymous editing sites in D. melanogaster had ancestral G-alleles, while the silent editing sites had an unignorable fraction of ancestral G-alleles; (2) Only very few nonsynonymous editing sites in D. melanogaster had corresponding G-alleles derived in the genomes of sibling species, and the fraction of such situation was significantly lower than that of silent editing sites; (3) The few nonsynonymous editing with corresponding G-alleles had significantly more variable editing levels (across samples) than other nonsynonymous editing sites in D. melanogaster. The proteomic diversifying nature of RNA editing in Drosophila excludes the restorative role which favors an ancestral G-allele. The few fixed G-alleles in sibling species might facilitate the adaptation to particular environment and the corresponding nonsynonymous editing in D. melanogaster would introduce stronger advantage of flexible proteomic diversification. With multi-Omics data, our study consolidates the nature of evolutionary significance of A-to-I RNA editing sites in model insects.

Termites can learn

It is generally believed that termites can't learn and are not "intelligent". This study aimed to test whether termites could have any form of memory. A Y-shaped test device with one release chamber and two identical test chambers was designed and constructed by 3D printing. A colony of damp wood termites was harvested from the wild. Worker termites were randomly selected for experiment. Repellent odors that could mimic the alarm pheromone for termites were first identified. Among all substances tested, a tea tree oil and lemon juice were found to contain repellent odors for the tested termites, as they significantly reduced the time that termites spent in the chamber treated with these substances. As control, a trail pheromone was found to be attractive. Subsequently, a second cohort of termites were operant conditioned by punishment using both tea tree oil and lemon juice, and then tested for their ability to remember the path that could lead to the repellant odors. The test device was thoroughly cleaned between trials. It was found that conditioned termites displayed a reduced tendency to choose the path that led to expectant punishment as compared with naïve termites. Thus, it is concluded that damp wood termites are capable of learning and forming "fear memory", indicative of "intelligence" in termites. This result challenges established presumption about termites' intelligence.

Replicator-mutator dynamics of the rock-paper-scissors game: Learning through mistakes

We generalize the Bush-Mosteller learning, the Roth-Erev learning, and the social learning to include mistakes, such that the nonlinear replicator-mutator equation with either additive or multiplicative mutation is generated in an asymptotic limit. Subsequently, we exhaustively investigate the ubiquitous rock-paper-scissors game for some analytically tractable motifs of mutation pattern for which the replicator-mutator flow is seen to exhibit rich dynamics that include limit cycles and chaotic orbits. The main result of this paper is that in both symmetric and asymmetric game interactions, mistakes can sometimes help the players learn; in fact, mistakes can even control chaos to lead to rational Nash-equilibrium outcomes. Furthermore, we report a hitherto-unknown Hamiltonian structure of the replicator-mutator equation.

Bumblebees socially learn behaviour too complex to innovate alone

Culture refers to behaviours that are socially learned and persist within a population over time. Increasing evidence suggests that animal culture can, like human culture, be cumulative: characterized by sequential innovations that build on previous ones1. However, human cumulative culture involves behaviours so complex that they lie beyond the capacity of any individual to independently discover during their lifetime1-3. To our knowledge, no study has so far demonstrated this phenomenon in an invertebrate. Here we show that bumblebees can learn from trained demonstrator bees to open a novel two-step puzzle box to obtain food rewards, even though they fail to do so independently. Experimenters were unable to train demonstrator bees to perform the unrewarded first step without providing a temporary reward linked to this action, which was removed during later stages of training. However, a third of naive observer bees learned to open the two-step box from these demonstrators, without ever being rewarded after the first step. This suggests that social learning might permit the acquisition of behaviours too complex to 're-innovate' through individual learning. Furthermore, naive bees failed to open the box despite extended exposure for up to 24 days. This finding challenges a common opinion in the field: that the capacity to socially learn behaviours that cannot be innovated through individual trial and error is unique to humans.

Honeybee comb-inspired stiffness gradient-amplified catapult for solid particle repellency

Natural surfaces that repel foreign matter are ubiquitous and crucial for living organisms. Despite remarkable liquid repellency driven by surface energy in many organisms, repelling tiny solid particles from surfaces is rare. The main challenge lies in the unfavourable scaling of inertia versus adhesion in the microscale and the inability of solids to release surface energy. Here we report a previously unexplored solid repellency on a honeybee's comb: a catapult-like effect to immediately eject pollen after grooming dirty antennae for self-cleaning. Nanoindentation tests revealed the 38-μm-long comb features a stiffness gradient spanning nearly two orders of magnitude from ~25 MPa at the tip to ~645 MPa at the base. This significantly augments the elastic energy storage and accelerates the subsequent conversion into kinetic energy. The reinforcement in energy storage and conversion allows the particle's otherwise weak inertia to outweigh its adhesion, thereby suppressing the unfavourable scaling effect and realizing solid repellency that is impossible in conventional uniform designs. We capitalize on this to build an elastomeric bioinspired stiffness-gradient catapult and demonstrate its generality and practicality. Our findings advance the fundamental understanding of natural catapult phenomena with the potential to develop bioinspired stiffness-gradient materials, catapult-based actuators and robotic cleaners.

Development of Sustainable Insecticide Candidates for Protecting Pollinators: Insight into the Bioactivities, Selective Mechanism of Action and QSAR of Natural Coumarin Derivatives against Aphids

Plants employ abundant toxic secondary metabolites to withstand insect attack, while pollinators can tolerate some natural defensive compounds. Coumarins, as promising green alternatives to chemical insecticides, possess wide application prospects in the crop protection field. Herein, the bioactivities of 30 natural coumarin derivatives against Aphis gossypii were assessed and revealed that 6-methylcoumarin exhibited potent aphicidal activity against aphids but displayed no toxicity to honeybees. Additionally, using biochemical, bioinformatic, and molecular assays, we confirmed that the action mode of 6-methylcoumarin against aphids was by inhibiting acetylcholinesterase (AChE). Meanwhile, functional assays revealed that the difference in action site, which located in Lys585 in aphid AChE (equivalent to Val548 in honeybee AChE), was the principal reason for 6-methylcoumarin being toxic to aphids but safe to pollinators. This action site was further validated by mutagenesis data, which uncovered how 6-methylcoumarin was unique selective to the aphid over honeybee or mammalian AChE. Furthermore, a 2D-QSAR model was established, revealing that the central structural feature was H3m, which offers guidance for the future design of more potent coumarin compounds. This work provides a sustainable strategy to take advantage of coumarin analogues for pest management while protecting nontarget pollinators.

A Method for Studying Social Signal Learning of the Waggle Dance in Honey Bees

Honey bees use a complex form of spatial referential communication. Their waggle dance communicates to nestmates the direction, distance, and quality of a resource by encoding celestial cues, retinal optic flow, and relative food value into motion and sound within the nest. This protocol was developed to investigate the potential for social learning of this waggle dance. Using this protocol, we showed that correct waggle dancing requires social learning. Bees (Apis mellifera) that did not follow any dances before they first danced produced significantly more disordered dances, with larger waggle angle divergence errors, and encoded distance incorrectly. The former deficits improved with experience, but distance encoding was set for life. The first dances of bees that could follow other dancers had none of these impairments. Social learning, therefore, shapes honey bee signaling, as it does communication in human infants, birds, and multiple other vertebrate species. However, much remains to be learned about insects' social learning, and this protocol will help to address knowledge gaps in the understanding of sophisticated social signal learning, particularly in understanding the molecular bases for such learning. Key features It was unclear if honey bees (Apis mellifera) could improve their waggle dance by following experienced dancers before they first waggle dance. Honey bees perform their first waggle dances with more errors if they cannot follow experienced waggle dancers first. Directional and disorder errors improved over time, but distance error was maintained. Bees in experimental colonies continued to communicate longer distances than control bees. Dancing correctly, with less directional error and disorder, requires social learning. Distance encoding in the honey bee dance is largely genetic but may also include a component of cultural transmission.

In search of behavioral and brain processes involved in honey bee dance communication

Honey bees represent an iconic model animal for studying the underlying mechanisms affecting advanced sensory and cognitive abilities during communication among colony mates. After von Frisch discovered the functional value of the waggle dance, this complex motor pattern led ethologists and neuroscientists to study its neural mechanism, behavioral significance, and implications for a collective organization. Recent studies have revealed some of the mechanisms involved in this symbolic form of communication by using conventional behavioral and pharmacological assays, neurobiological studies, comprehensive molecular and connectome analyses, and computational models. This review summarizes several critical behavioral and brain processes and mechanisms involved in waggle dance communication. We focus on the role of neuromodulators in the dancer and the recruited follower, the interneurons and their related processing in the first mechano-processing, and the computational navigation centers of insect brains.