Division of labor and brain evolution in insect societies: Neurobiology of extreme specialization in the turtle ant Cephalotes varians

Bumblebee cognitive abilities are robust to changes in colony size

Abstract 

Eusocial insect colonies act as a superorganism, which can improve their ability to buffer the negative impact of some anthropogenic stressors. However, this buffering effect can be affected by anthropogenic factors that reduce their colony size. A reduction in colony size is known to negatively affect several parameters like brood maintenance or thermoregulation, but the effects on behaviour and cognition have been largely overlooked. It remains unclear how a sudden change in group size, such as that which might be caused by anthropogenic stressors, affects individual behaviour within a colony. In this study, the bumblebee Bombus terrestris was used to study the effect of social group size on behaviour by comparing the associative learning capabilities of individuals from colonies that were unmanipulated, reduced to a normal size (a colony of 100 workers) or reduced to a critically low but functional size (a colony of 20 workers). The results demonstrated that workers from the different treatments performed equally well in associative learning tasks, which also included no significant differences in the learning capacity of workers that had fully developed after the colony size manipulation. Furthermore, we found that the size of workers had no impact on associative learning ability. The learning abilities of bumblebee workers were thus resilient to the colony reduction they encountered. Our study is a first step towards understanding how eusocial insect cognition can be impacted by drastic reductions in colony size.

Significance statement

While anthropogenic stressors can reduce the colony size of eusocial insects, the impact of this reduction is poorly studied, particularly among bumblebees. We hypothesised that colony size reduction would affect the cognitive capacity of worker bumblebees as a result of fewer social interactions or potential undernourishment. Using differential conditioning, we showed that drastic reductions in colony size have no effect on the associative learning capabilities of the bumblebee Bombus terrestris and that this was the same for individuals that were tested just after the colony reduction and individuals that fully developed under the colony size reduction. We also showed that body size did not affect learning capabilities. This resilience could be an efficient buffer against the ongoing impacts of global change.

What is really social about social insect cognition?

It is often assumed that social life imposes specific cognitive demands for animals to communicate, cooperate and compete, ultimately requiring larger brains. The “social brain” hypothesis is supported by data in primates and some other vertebrates, but doubts have been raised over its applicability to other taxa, and in particular insects. Here, we review recent advances in insect cognition research and ask whether we can identify cognitive capacities that are specific to social species. One difficulty involved in testing the social brain hypothesis in insects is that many of the model species used in cognition studies are highly social (eusocial), and comparatively little work has been done in insects that live in less integrated social structures or that are solitary. As more species are studied, it is becoming clear that insects share a rich cognitive repertoire and that these abilities are not directly related to their level of social complexity. Moreover, some of the cognitive mechanisms involved in many social interactions may not differ from those involved in non-social behaviors. We discuss the need for a more comparative and neurobiologically grounded research agenda to better understand the evolution of insect brains and cognition.

Deterministic Response Threshold Models of Reproductive Division of Labor Are More Robust Than Probabilistic Models in Artificial Ants

We implement an agent-based simulation of the response threshold model of reproductive division of labor. Ants in our simulation must perform two tasks in their environment: forage and reproduce. The colony is capable of allocating ant resources to these roles using different division of labor strategies via genetic architectures and plasticity mechanisms. We find that the deterministic allocation strategy of the response threshold model is more robust than the probabilistic allocation strategy. The deterministic allocation strategy is also capable of evolving complex solutions to colony problems like niche construction and recovery from the loss of the breeding caste. In addition, plasticity mechanisms had both positive and negative influence on the emergence of reproductive division of labor. The combination of plasticity mechanisms has an additive and sometimes emergent impact.

AnimalTraits - a curated animal trait database for body mass, metabolic rate and brain size

Trait databases have become important resources for large-scale comparative studies in ecology and evolution. Here we introduce the AnimalTraits database, a curated database of body mass, metabolic rate and brain size, in standardised units, for terrestrial animals. The database has broad taxonomic breadth, including tetrapods, arthropods, molluscs and annelids from almost 2000 species and 1000 genera. All data recorded in the database are sourced from their original empirical publication, and the original metrics and measurements are included with each record. This allows for subsequent data transformations as required. We have included rich metadata to allow users to filter the dataset. The additional R scripts we provide will assist researchers with aggregating standardised observations into species-level trait values. Our goals are to provide this resource without restrictions, to keep the AnimalTraits database current, and to grow the number of relevant traits in the future.

Measurement(s)	metabolic rate quantification • body mass • brain size
Technology Type(s)	metabolic rate measurement • body mass quantification • brain mass brain volume

Social Brain Energetics: Ergonomic Efficiency, Neurometabolic Scaling, and Metabolic Polyphenism in Ants

Metabolism, a metric of the energy cost of behavior, plays a significant role in social evolution. Body size and metabolic scaling are coupled, and a socioecological pattern of increased body size is associated with dietary change and the formation of larger and more complex groups. These consequences of the adaptive radiation of animal societies beg questions concerning energy expenses, a substantial portion of which may involve the metabolic rates of brains that process social information. Brain size scales with body size, but little is understood about brain metabolic scaling. Social insects such as ants show wide variation in worker body size and morphology that correlates with brain size, structure, and worker task performance, which is dependent on sensory inputs and information-processing ability to generate behavior. Elevated production and maintenance costs in workers may impose energetic constraints on body size and brain size that are reflected in patterns of metabolic scaling. Models of brain evolution do not clearly predict patterns of brain metabolic scaling, nor do they specify its relationship to task performance and worker ergonomic efficiency, two key elements of social evolution in ants. Brain metabolic rate is rarely recorded and therefore the conditions under which brain metabolism influences the evolution of brain size are unclear. We propose that studies of morphological evolution, colony social organization, and worker ergonomic efficiency should be integrated with analyses of species-specific patterns of brain metabolic scaling to advance our understanding of brain evolution in ants.

Brain size and behavioral specialization in the jataí stingless bee (Tetragonisca angustula)

Social insects are instructive models for understanding the association between investment in brain size and behavioral variability because they show a relatively simple nervous system associated with a large set of complex behaviors. In the jataí stingless bee (Tetragonisca angustula), division of labor relies both on age and body size differences among workers. When young, both minors and soldiers engage in intranidal tasks and move to extranidal tasks as they age. Minors switch to foraging activities, while soldiers take over defensive roles. Nest defense performed by soldiers includes two different tasks: (1) hovering around the nest entrance for the detection and interception of heterospecific bees (a task relying mostly on vision) and (2) standing at the nest entrance tube for inspection of returning foragers and discrimination against conspecific non-nestmates based on olfactory cues. Here, using different-sized individuals (minors and soldiers) as well as same-sized individuals (hovering and standing soldiers) performing distinct tasks, we investigated the effects of both morphological and behavioral variability on brain size. We found a negative allometric growth between brain size and body size across jataí workers, meaning that minors had relatively larger brains than soldiers. Between soldier types, we found that hovering soldiers had larger brain compartments related to visual processing (the optic lobes) and learning (the mushroom bodies). Brain size differences between jataí soldiers thus correspond to behavioral specialization in defense (i.e., vision for hovering soldiers) and illustrate a functional neuroplasticity underpinning division of labor.

Statistical Atlases and Automatic Labeling Strategies to Accelerate the Analysis of Social Insect Brain Evolution

Current methods used to quantify brain size and compartmental scaling relationships in studies of social insect brain evolution involve manual annotations of images from histological samples, confocal microscopy or other sources. This process is susceptible to human bias and error and requires time-consuming effort by expert annotators. Standardized brain atlases, constructed through 3D registration and automatic segmentation, surmount these issues while increasing throughput to robustly sample diverse morphological and behavioral phenotypes. Here we design and evaluate three strategies to construct statistical brain atlases, or templates, using ants as a model taxon. The first technique creates a template by registering multiple brains of the same species. Brain regions are manually annotated on the template, and the labels are transformed back to each individual brain to obtain an automatic annotation, or to any other brain aligned with the template. The second strategy also creates a template from multiple brain images but obtains labels as a consensus from multiple manual annotations of individual brains comprising the template. The third technique is based on a template comprising brains from multiple species and the consensus of their labels. We used volume similarity as a metric to evaluate the automatic segmentation produced by each method against the inter- and intra-individual variability of human expert annotators. We found that automatic and manual methods are equivalent in volume accuracy, making the template technique an extraordinary tool to accelerate data collection and reduce human bias in the study of the evolutionary neurobiology of ants and other insects.

Caste, Sex, and Parasitism Influence Brain Plasticity in a Social Wasp

Brain plasticity is widespread in nature, as it enables adaptive responses to sensory demands associated with novel stimuli, environmental changes and social conditions. Social Hymenoptera are particularly well-suited to study neuroplasticity, because the division of labor amongst females and the different life histories of males and females are associated with specific sensory needs. Here, we take advantage of the social wasp Polistes dominula to explore if brain plasticity is influenced by caste and sex, and the exploitation by the strepsipteran parasite Xenos vesparum. Within sexes, male wasps had proportionally larger optic lobes, while females had larger antennal lobes, which is consistent with the sensory needs of sex-specific life histories. Within castes, reproductive females had larger mushroom body calyces, as predicted by their sensory needs for extensive within-colony interactions and winter aggregations, than workers who frequently forage for nest material and prey. Parasites had different effects on female and male hosts. Contrary to our predictions, female workers were castrated and behaviorally manipulated by female or male parasites, but only showed moderate differences in brain tissue allocation compared to non-parasitized workers. Parasitized males maintained their reproductive apparatus and sexual behavior. However, they had smaller brains and larger sensory brain regions than non-parasitized males. Our findings confirm that caste and sex mediate brain plasticity in P. dominula, and that parasitic manipulation drives differential allocation of brain regions depending on host sex.

Soldier neural architecture is temporarily modality-specialized but poorly predicted by repertoire size in the stingless bee Tetragonisca angustula

Individual heterogeneity within societies provides opportunities to test hypotheses about adaptive neural investment in the context of group cooperation. Here we explore neural investment in defense specialist soldiers of the eusocial stingless bee (Tetragonisca angustula) which are age sub-specialized on distinct defense tasks and have an overall higher lifetime task repertoire than other sterile workers within the colony. Consistent with predicted behavioral demands, soldiers had higher relative visual (optic lobe) investment than non-soldiers but only during the period when they were performing the most visually demanding defense task (hovering guarding). As soldiers aged into the less visually demanding task of standing guarding this difference disappeared. Neural investment was otherwise similar across all colony members. Despite having larger task repertoires, soldiers had similar absolute brain size and smaller relative brain size compared to other workers, meaning that lifetime task repertoire size was a poor predictor of brain size. Both high behavioral specialization in stable environmental conditions and reassignment across task groups during a crisis occur in T. angustula. The differences in neurobiology we report here are consistent with these specialized but flexible defense strategies. This work broadens our understanding of how neurobiology mediates age and morphological task specialization in highly cooperative societies. This article is protected by copyright. All rights reserved.

When and Why Did Human Brains Decrease in Size? A New Change-Point Analysis and Insights From Brain Evolution in Ants

Human brain size nearly quadrupled in the six million years since Homo last shared a common ancestor with chimpanzees, but human brains are thought to have decreased in volume since the end of the last Ice Age. The timing and reason for this decrease is enigmatic. Here we use change-point analysis to estimate the timing of changes in the rate of hominin brain evolution. We find that hominin brains experienced positive rate changes at 2.1 and 1.5 million years ago, coincident with the early evolution of Homo and technological innovations evident in the archeological record. But we also find that human brain size reduction was surprisingly recent, occurring in the last 3,000 years. Our dating does not support hypotheses concerning brain size reduction as a by-product of body size reduction, a result of a shift to an agricultural diet, or a consequence of self-domestication. We suggest our analysis supports the hypothesis that the recent decrease in brain size may instead result from the externalization of knowledge and advantages of group-level decision-making due in part to the advent of social systems of distributed cognition and the storage and sharing of information. Humans live in social groups in which multiple brains contribute to the emergence of collective intelligence. Although difficult to study in the deep history of Homo, the impacts of group size, social organization, collective intelligence and other potential selective forces on brain evolution can be elucidated using ants as models. The remarkable ecological diversity of ants and their species richness encompasses forms convergent in aspects of human sociality, including large group size, agrarian life histories, division of labor, and collective cognition. Ants provide a wide range of social systems to generate and test hypotheses concerning brain size enlargement or reduction and aid in interpreting patterns of brain evolution identified in humans. Although humans and ants represent very different routes in social and cognitive evolution, the insights ants offer can broadly inform us of the selective forces that influence brain size.

Active Inferants: An Active Inference Framework for Ant Colony Behavior

In this paper, we introduce an active inference model of ant colony foraging behavior, and implement the model in a series of in silico experiments. Active inference is a multiscale approach to behavioral modeling that is being applied across settings in theoretical biology and ethology. The ant colony is a classic case system in the function of distributed systems in terms of stigmergic decision-making and information sharing. Here we specify and simulate a Markov decision process (MDP) model for ant colony foraging. We investigate a well-known paradigm from laboratory ant colony behavioral experiments, the alternating T-maze paradigm, to illustrate the ability of the model to recover basic colony phenomena such as trail formation after food location discovery. We conclude by outlining how the active inference ant colony foraging behavioral model can be extended and situated within a nested multiscale framework and systems approaches to biology more generally.

The neuroplasticity of division of labor: worker polymorphism, compound eye structure and brain organization in the leafcutter ant Atta cephalotes

Our understanding of how sensory structure design is coupled with neural processing capacity to adaptively support division of labor is limited. Workers of the remarkably polymorphic fungus-growing ant Atta cephalotes are behaviorally specialized by size: the smallest workers (minims) tend fungi in dark subterranean chambers while larger workers perform tasks outside the nest. Strong differences in worksite light conditions are predicted to influence sensory and processing requirements for vision. Analyzing confocal scans of worker eyes and brains, we found that eye structure and visual neuropils appear to have been selected to maximize task performance according to light availability. Minim eyes had few ommatidia, large interommatidial angles and eye parameter values, suggesting selection for visual sensitivity over acuity. Large workers had larger eyes with disproportionally more and larger ommatidia, and smaller interommatidial angles and eye parameter values, indicating peripheral sensory adaptation to ambient rainforest light. Optic lobes and mushroom body collars were disproportionately small in minims. Within the optic lobe, lamina and lobula relative volumes increased with worker size whereas medulla volume decreased. Visual system phenotypes thus correspond to task specializations in dark or light environments and illustrate a functional neuroplasticity underpinning division of labor in this socially complex agricultural ant.