Evolution of a kin recognition system in social insects – EVOKIN

A range of approaches, including neuroanatomy, genomics, chemical analysis, and behavior, were used under a phylogenetic framework.

1. Field sampling: more than 700 individuals, ranging from solitary to eusocial species and spanning the entire phylogeny of Hymenoptera, were collected in France, the United Kingdom, Panama, and Malaysia

2. Neuroanatomy: using confocal microscopy (with anterograde tracing and immunohistochemistry), we performed 3D reconstructions of antennal lobe (AL) glomeruli from high-resolution images, enabling volumetric quantification and the counting of neuroanatomical units (including glomeruli associated with basiconic sensilla). Additionally, scanning electron microscopy (SEM) was used to analyze the density, morphology, and distribution of basiconic sensilla on the antennae.

3. Genomic and bioinformatics: in addition to data available in public databases, genomic data with high coverage were generated for eight key species in the Hymenopteran phylogeny. OR genes were annotated on a broad scale and comparative genomics assessed expansions/contractions of 9-exon OR genes (putatively linked to CHC detection) across Hymenoptera. In addition, RNA-seq was used to link nestmate recognition in some species to molecular-level sensory adaptations.

4. Chemical analysis: cuticular hydrocarbons (CHCs) were extracted from a range of Hymenopteran species, both solitary and eusocial. Cuticular extracts were analyzed using Gas‐Chromatography coupled with Mass‐Spectrometry (GC‐MS) and individual compound identities determined by comparing mass spectra to standards and the literature. Peak area data were used to obtain CHC profile complexity indices, also considering CHC types (linear alkanes vs. methyl-branched/unsaturated hydrocarbons).

5. Behavior: to test CHC-mediated recognition in some ant species, aggression experiments were performed, in which focal individuals were presented with nestmate vs non-nestmate extracts, or dead specimens (washed/recoated with CHCs), ensuring that chemical cues were sufficient for eliciting the behaviors. Behaviors were video-recorded and analyzed using dedicated software to produce an aggression index integrating duration and intensity of aggressive behaviors.

6. Comparative phylogenetic analyses: to understand the evolution of the BaS-subsystem in Hymenoptera, ancestral state reconstructions were performed, and neuroanatomical traits (glomeruli numbers in particular) were mapped onto Hymenopteran phylogenies. To integrate neuroanatomical, genomic, and chemical data, scaling relationships were used, with glomeruli number/volume being regressed against CHC complexity and 9-exon OR gene repertoire to identify adaptive patterns.

In Formicidae, the BaS subsystem and its TB glomerular cluster are universal, with phylogenetic reconstructions indicating an early origin and complexity predating eusociality (Fig 1; Marty et al., 2025). TB complexity scales with overall AL complexity, revealing a conserved allometric coupling. In contrast, serotonergic innervation is labile, with repeated gains and losses among closely related genera, showing that neuromodulatory wiring evolves independently of the TB. Variation in BaS investment is not explained by social or broad ecological traits. Instead, integrating CHC data shows covariation with species‑specific CHC properties: species with low‑diversity CHC blends tend to invest more in the TB, consistent with a compensatory enhancement of discrimination when chemical signals are impoverished. In ants, social structure still influences the olfactory system, as we showed caste-specific adaptations in a polymorphic ant, where small workers—specialized in nestmate recognition—exhibit a higher density in basiconic sensilla and upregulated 9-exon OR genes.

Comparative data from vespids and other apoid wasps reveals a distinct pattern (Fig 2). Although the BaS subsystem is widespread, its relative investment is often decoupled from the rest of the AL. Several solitary clades, (a.g. Eumeninae, Ampulicidae, Crabronidae) show exceptionally high TB glomerular numbers, sometimes exceeding those of sympatric social species. Genomic analyses parallel this pattern: TB glomerular number is strongly correlated with the size of the 9-exon OR repertoire, while the rest of the AL covaries with non-9-exon OR families. These results indicate that the BaS subsystem is an ancestral and evolutionarily labile sensory module whose expansion and contraction are driven primarily by ecological and chemical constraints rather than eusocial complexity.

Patterns of reduction were also informative: convergent shrinkage of TB cluster and 9-exon repertoires occurs in pollen-collecting lineages (e.g. in Anthophila, Masarinae), consistent with a shift toward floral cues. This supports the hypothesis that the ancestral BaS function lay in host/prey detection, with nestmate recognition representing a derived deployment in social taxa. CHC complexity likewise does not track social complexity. We are finalizing phylogenetically controlled multivariate models integrating CHC composition, BaS investment, and OR counts to test whether the sensory‑compensation pattern observed in ants holds across Hymenoptera: do species with less chemically diverse CHC profiles tend to exhibit higher BaS investment as sensory compensation that enhances discrimination when the peripheral signal is impoverished?

Overall, EVOKIN positions the BaS/TB as an ancient, modular olfactory subsystem repeatedly tuned by ecological and phylogenetic constraints, providing a preadapted sensory toolkit that could be co‑opted for kin recognition where selective conditions favoured it.

These findings are significant because they challenge a long-standing assumption: the transition to eusociality was not the primary force shaping the olfactory system of Hymenoptera. Instead, our data support an alternative hypothesis: a pre-existing, highly developed subsystem, originally evolved for detecting and discriminating complex CHC profiles in the context of parasitoid lifestyles, was later co-opted for nestmate recognition.

The discoveries of the EVOKIN project and the collaborative discussions it generated have led to three innovative projects: one on cognitive variability and personality in ants, another on the influence of social and ecological factors in brain evolution among Hymenoptera, and a third on the study of the evolutionary arms race for prey/host detection between cuckoo wasps and their hosts. These projects will deepen our understanding of the evolution of sensory systems and brain function in insects.

In evolutionary terms, life is about reproduction. Yet, in some species, individuals forgo their own reproduction to support the reproductive efforts of others. Bees, ants and wasps (Hymenoptera) are well known for their ‘eusocial’ lifestyle where queens regulate the reproductive output of up to a million workers. This sacrifice of workers to their colony was a major challenge to the theory of natural selection, acknowledged by Darwin himself. The realisation that helping related individuals indirectly promotes the helper’s overall reproductive success solved part of this evolutionary conundrum. However, this strategy is vulnerable to exploitation by unrelated ‘parasites’. An ability to distinguish kin from non-kin is therefore critical for eusociality to evolve. Given that eusociality is virtually absent in most insect lineages, a major unanswered question is how it was able to evolve nine times in Hymenoptera. Recent studies suggest that eusocial ants possess an olfactory subsystem that is specialized to function in kin recognition, providing a mechanism to detect social parasites. It consists of one type of sensory structure on the antennae, called basiconic sensillae (BaS), which detect cuticular hydrocarbons (CHCs), a blend of chemical cues that can reveal individual’s colony origin. The BaS are connected to a segregated region of the primary olfactory centre in the brain, and are thought to express a group of olfactory receptor (OR) genes with a distinctive 9-exon structure. Strikingly, a similar suite of features has recently been observed in Vespid wasps, which independently evolved a eusocial lifestyle. This suggests that the BaS-subsystem may have played a critical role in the convergent evolution of eusociality by enabling kin recognition. However, we do not know if the BaS-subsystem was present in ancestral solitary species, if it was pre-adapted to support kin-recognition, or how it may have been refined for this task. Answering these questions is central to understanding the striking diversity of social behaviours in Hymenoptera, and to explain why eusociality, normally so rare, has evolved so many times in this clade.

The EVOKIN project proposes an unprecedented endeavour to answer these questions. We will collect data from species representing all major Hymenopteran lineages, a diverse range of social organisations, and multiple independent origins of eusociality. By combining neuroanatomical, chemical and molecular data, together with behavioural and neurophysiological assays, we will test the evolutionary link between the BaS-specific subsystem and eusociality. Specifically, we will:
Aim 1: Reveal how the anatomical organisation of the BaS-specific subsystem varies across Hymenoptera. We will test whether this variation is associated with social organisation, and reconstruct the evolutionary history of key traits across the phylogeny, enabling us to place behavioural and neuronal changes in chronological order.
Aim 2: Use an extensive dataset of OR repertoires across Hymenoptera to test whether 9-exon OR gene number co-evolves with each species’ CHC profile complexity and the number of anatomical units in the BaS-specific subsystem, and examine how transitions in social organisation shape the selection pressures acting on OR genes.
Aim 3: Confirm the function of the BaS-specific subsystem in kin recognition in eusocial lineages, and determine its possible ancestral function in solitary species, by performing behavioural experiments (e.g. dyad encounters) and using in vivo optical imaging to record the neural activity of the BaS-specific subsystem in response to kin, conspecifics, and environmental odours.
Combining detailed neuroanatomical, molecular and functional characterisation of the sensory structures underpinning kin recognition, the EVOKIN project will uncover the neural adaptations that support social behaviour and the remarkable diversity of Hymenopteran social systems.