Social life is not only present in various organisms but also in all kinds of forms: simple aggregations of individuals, facultative or obligatory parental care, cooperatively breeding groups and super-organismal eusocial societies (Bourke, 2011). However, research on animal sociality has long focused exclusively on the fascinating eusocial organisation of certain hymenopteran and isopteran insects, hampering our general understanding of the evolution of all forms of social life in nature (Elgar, 2015; Meunier and Steiger, 2018). Somewhat surprisingly, the results of our call for this Research Topic suggest that this bias is being corrected: 7 of the 16 published articles are based on non-eusocial systems (Table 1). All else being equal, this indicates that the research community studying the link between microbes and social evolution has moved away from exclusive eusocial model systems. However, the results of our call also point out existing partitions between communities, as we only managed to receive two non-insect articles and unfortunately none on a non-eukaryotic system (Table 1). This bias may come from the fact that the editors of this special issue are mostly working on insect models, and have failed to reach out and/or to attract members of the other communities. This bias may also stem from the multiple definitions of sociality present in the literature (Rubenstein and Abbot, 2017), which are often very specific to each community, and which may have—for instance—excluded vast communities of researchers studying family life and aggregation behaviours much more frequent in non-insects species.

A long-standing question about the link between microbial pathogens and social evolution is whether access to additional defences against microbial pathogens (i.e., social immunity) necessarily comes with reduced investment in individual immunity. Five manuscripts of this Research topic illustrate the complexity of this question and the diversity of answers that can be given. First, Meusemann et al. used genomic and transcriptomic data from eight termite species, representing wood-dwelling and foraging species, plus 14 other winged insects (Pterygota) and found that differences in the types of sociality do not reflect differences in the intensity of natural selection on immune genes. Instead, they found evidence for a genome-wide pattern of relaxed selection on these genes in termites. Second, Baeuerle et al. used experimental approaches on 14 bumblebee species to show that the investment in an external defence against pathogens (in the venom) does not trade-off against the investment in internal immune defence present in the hemolymph. Third, Esparza-Mora et al. revealed in a termite that inhibiting an external enzyme capable of degrading entomopathogenic fungi does not trigger collective defences such as allogrooming, but instead reduce defensive cannibalistic behaviours. This suggests that the individual immune system is linked to certain collective immune behaviours in this termite. Fourth, Cole and Rosengaus emphasised the importance of social environment on pathogen resistance, as they showed that the presence of a king may help to mitigate the negative effects of a queen's infection during colony foundation in a dampwood termite. Finally, Pull and McMahon propose a comprehensive review of social immunity and emphasise that “superorganism immunity” may fulfil an analogous function to the immune system of Metazoa.

The recent proposal that social immunity is not specific to eusocial systems (Cotter and Kilner, 2010; Van Meyel et al., 2018) opened numerous questions about the diversity of its forms, the type of immune benefits provided by the social environment and the mechanisms mediating the recognition of pathogenic microbes in social species. Two manuscripts of this Research Topic offer an overview of these questions. First, Trienens and Rohlfs investigated forms of social immunity in groups of fruit fly larvae. They showed that larvae suppress the invasion of a harmful fungus by the summative effect of individuals at high densities and that larger groups of larvae at the same density can control fungal growth more efficiently. This indicates a potential collective defence against habitat invasion by pathogenic fungi in insects that exhibit mere aggregation behaviour. Second, Goes et al. reviewed the literature on leaf-cutting ants' social immunity to investigate how workers protect their fungal garden against harmful microbes. They reveal that workers discriminate against harmful microbes via chemical cues originating from the antagonistic microbe and/or semiochemicals released by the fungus-garden during harmful interactions, as well as via associative learning when workers connect the microbe cues with damage in the fungus garden.

Several articles dealt with the issue that habitat can shape interactions of animals with microbes, which in turn affects the animal's social behaviour. Three of these articles specifically looked at wood as a substrate for insects, which is acknowledged to be very favourable for the evolution of sociality because of its structural resistance and longevity compared to the life of an insect (Hamilton, 1978; Kirkendall et al., 2015). In a review article, Dillard and Benbow argue that two additional factors may facilitate prolonged parent-offspring contact and cooperative behaviours in insects. First, the low nutritional value of wood selects for associations with nutrient-enriching microbes that need to be transferred to offspring, which is often through social contact. Furthermore, insects compete with many antagonistic microbes in this habitat and collective defence is often better to keep them in cheque. A research study by Biedermann found evidence for collective pathogen defence in a cooperatively breeding, fungus-farming ambrosia beetle. Delayed dispersing female offspring showed more social hygienic behaviour at places within the nest with higher abundances of antagonistic fungi. Nuotclà et al. showed in a closely related ambrosia beetle species with a similar social system that the two nutritionally important fungal mutualists of this species vary in their relative abundance depending on the dryness of the wood substrate. Interestingly this fed back on the social behaviour and the delayed dispersal periods of daughters in this facultatively eusocial beetle. Finally, a study by Tragust et al. on founding Lasius niger ant queens showed that queens exposed to pathogens invest simultaneously in formic acid defence and higher worker production. Surprisingly there was no measurable trade-off between this individual immune defence and reproduction at an early nest stage, but this may have effects on later fitness.

The growing awareness that microbes residing on and in a host can provide it with major benefits has recently stimulated a great number of experimental and theoretical research on the impacts of symbionts in the social evolution of hosts. This is illustrated by five articles in this Research Topic. First, Bratburd et al. examined the general role of defensive microbial symbionts in host protection against pathogens in terms of behavioural and immune responses and discussed why insects are good models to study issues relating to human health and agriculture. Second, Dunn et al. used a comparative approach to study how microbiomes of hominins have changed over evolutionary time, questioning their impact on the evolution of several host functions and discussing the possibility that prosocial microbes promoted hominin social behaviour. Third, Nalepa discussed why symbiont transmission via proctodeal trophallaxis and sociality are likely to have entangled evolutionary histories and conclude that the vertical transmission of gut microbes (flagellates) and the origin of host subsociality are two sides of the same coin in termites. Fourth, Sinotte et al. explored the link between the division of labour and symbiosis in social insects. Their review suggests that structured microbiomes have evolved in parallel to social complexity, and predicts that mature social insect colonies with the most extreme division of labour shows the strongest distinction between caste microbiomes. This suggests that caste-specific microbiomes may enhance symbiotic benefits and the efficiency of division of labour. Finally, Figueiredo and Kramer took another perspective and focused on the microbes themselves. Their review describes cooperation and conflict within the microbiota. They discuss how these parameters can affect animal hosts and conclude that an explicit consideration of social dynamics within symbiont communities is crucial to advance our understanding of how microbes shape animal function and evolution.