Machine Learning models identify gene predictors of waggle dance behaviour in honeybees

Learning performance and GABAergic pathway link to deformed wing virus in the mushroom bodies of naturally infected honey bees

Viral infections can be detrimental to the foraging ability of the western honey bee, Apis mellifera. The deformed wing virus (DWV) is the most common honey bee virus and has been proposed as a possible cause of learning and memory impairment. However, evidence for this phenomenon so far has come from artificially infected bees, while less is known about the implications of natural infections with the virus. Using the proboscis extension reflex (PER), we uncovered no significant association between a simple associative learning task and natural DWV load. However, when assessed through a reversal associative learning assay, bees with higher DWV load performed better in the reversal learning phase. DWV is able to replicate in the honey bee mushroom bodies, where the GABAergic signalling pathway has an antagonistic effect on associative learning but is crucial for reversal learning. Hence, we assessed the pattern of expression of several GABA-related genes in bees with different learning responses. Intriguingly, mushroom body expression of selected genes was positively correlated with DWV load, but only for bees with good reversal learning performance. We hypothesise that DWV might improve olfactory learning performance by enhancing the GABAergic inhibition of responses to unrewarded stimuli, which is consistent with the behavioural patterns that we observed. However, at higher disease burdens, which might be induced by an artificial infection or by a severe, natural Varroa infestation, this DWV-associated increase in GABA signalling could impair associative learning as previously reported by other studies.

Transcriptomic responses to location learning by honeybee dancers are partly mirrored in the brains of dance-followers

The waggle dances of honeybees are a strikingly complex form of animal communication that underlie the collective foraging behaviour of colonies. The mechanisms by which bees assess the locations of forage sites that they have visited for representation on the dancefloor are now well-understood, but few studies have considered the remarkable backward translation of such information into flight vectors by dance-followers. Here, we explore whether the gene expression patterns that are induced through individual learning about foraging locations are mirrored when bees learn about those same locations from their nest-mates. We first confirmed that the mushroom bodies of honeybee dancers show a specific transcriptomic response to learning about distance, and then showed that approximately 5% of those genes were also differentially expressed by bees that follow dances for the same foraging sites, but had never visited them. A subset of these genes were also differentially expressed when we manipulated distance perception through an optic flow paradigm, and responses to learning about target direction were also in part mirrored in the brains of dance followers. Our findings show a molecular footprint of the transfer of learnt information from one animal to another through this extraordinary communication system, highlighting the dynamic role of the genome in mediating even very short-term behavioural changes.

Co-expression Gene Networks and Machine-learning Algorithms Unveil a Core Genetic Toolkit for Reproductive Division of Labour in Rudimentary Insect Societies

The evolution of eusociality requires that individuals forgo some or all their own reproduction to assist the reproduction of others in their group, such as a primary egg-laying queen. A major open question is how genes and genetic pathways sculpt the evolution of eusociality, especially in rudimentary forms of sociality-those with smaller cooperative nests when compared with species such as honeybees that possess large societies. We lack comprehensive comparative studies examining shared patterns and processes across multiple social lineages. Here we examine the mechanisms of molecular convergence across two lineages of bees and wasps exhibiting such rudimentary societies. These societies consist of few individuals and their life histories range from facultative to obligately social. Using six species across four independent origins of sociality, we conduct a comparative meta-analysis of publicly available transcriptomes. Standard methods detected little similarity in patterns of differential gene expression in brain transcriptomes among reproductive and non-reproductive individuals across species. By contrast, both supervised machine learning and consensus co-expression network approaches uncovered sets of genes with conserved expression patterns among reproductive and non-reproductive phenotypes across species. These sets overlap substantially, and may comprise a shared genetic "toolkit" for sociality across the distantly related taxa of bees and wasps and independently evolved lineages of sociality. We also found many lineage-specific genes and co-expression modules associated with social phenotypes and possible signatures of shared life-history traits. These results reveal how taxon-specific molecular mechanisms complement a core toolkit of molecular processes in sculpting traits related to the evolution of eusociality.

An artificial neural network explains how bats might use vision for navigation

Animals navigate using various sensory information to guide their movement. Miniature tracking devices now allow documenting animals' routes with high accuracy. Despite this detailed description of animal movement, how animals translate sensory information to movement is poorly understood. Recent machine learning advances now allow addressing this question with unprecedented statistical learning tools. We harnessed this power to address visual-based navigation in fruit bats. We used machine learning and trained a convolutional neural network to navigate along a bat's route using visual information that would have been available to the real bat, which we collected using a drone. We show that a simple feed-forward network can learn to guide the agent towards a goal based on sensory input, and can generalize its learning both in time and in space. Our analysis suggests how animals could potentially use visual input for navigation and which features might be useful for this purpose.

Assessing Anatomical Changes in Male Reproductive Organs in Response to Larval Crowding Using Micro-computed Tomography Imaging

Ecological conditions shape (adaptive) responses at the molecular, anatomical, and behavioral levels. Understanding these responses is key to predict the outcomes of intra- and inter-specific competitions and the evolutionary trajectory of populations. Recent technological advances have enabled large-scale molecular (e.g., RNAseq) and behavioral (e.g., computer vision) studies, but the study of anatomical responses to ecological conditions has lagged behind. Here, we highlight the role of X-ray micro-computed tomography (micro-CT) in generating in vivo and ex vivo 3D imaging of anatomical structures, which can enable insights into adaptive anatomical responses to ecological environments. To demonstrate the application of this method, we manipulated the larval density of Drosophila melanogaster Meigen flies and applied micro-CT to investigate the anatomical responses of the male reproductive organs to varying intraspecific competition levels during development. Our data is suggestive of two classes of anatomical responses which broadly agree with sexual selection theory: increasing larval density led to testes and ejaculatory duct to be overall larger (in volume), while the volume of accessory glands and, to a lesser extent, ejaculatory duct decreased. These two distinct classes of anatomical responses might reflect shared developmental regulation of the structures of the male reproductive system. Overall, we show that micro-CT can be an important tool to advance the study of anatomical (adaptive) responses to ecological environments.