Social insects are a peculiar, yet conspicuous set of about 2% of known species of insects. Within the western tradition, they have been recognized since antiquity as a more or less distinct group of species. At the same time, it has proven much easier to recognize who the social insects are than to say exactly what they are.
What, then, are the core attributes that unite the termites, ants, some wasps, and some bees, as well as a very few other insects, while setting them apart from the great mass of nonsocial insects? To say that they are characterized by living in durable structured groups is a fair first draft, but biologists have long deemed this to be much too imprecise. A key point came with Charles D. Michener’s definition of eusociality as comprising three features of the group: (a) overlap of adult generations, (b) reproductive castes, and (c) cooperative brood care. The first of these simply means that the mother (and sometimes the father) remains with the brood until it matures, so that at least for a time the two generations interact as adults. Reproductive castes are present where there is any significant bias between interacting individuals that reproduce more and those that reproduce less, with the extreme being those very large colonies of thousands of individuals, in which one female lays all the eggs. And cooperative brood care is any breakdown of the primitive pattern in which adults care exclusively for their own offspring .
This concept has since received such wide acceptance that “social insects” is virtually synonymous with “eusocial insects” in the writing of possibly a large majority of biologists. For a long time the only known eusocial animals were termites and several lineages of aculeate hymenoptera. In recent decades, however, species of aphids, thrips, beetles, a shrimp, and even a mammal have been shown to fit the definition.
At this point some further definitions are useful. The vast majority of insect species, like most other animals, show none of the three criteria and are termed solitary. Within this broad array are the many presocial insects, which show some attributes of group living but none of the three criteria of eusociality. Between the two poles of the solitary-eusocial continuum are six possible combinations of either one or two of these features, but not all three. These have names (tables in refs. [2, 11]) and are manifested by various species, at least as a temporary condition. For example, if there is cooperative brood care, but neither of the other two features, the association is termed quasisocial.
However, it is noteworthy that most entomologists have not bothered to learn these six names, and for good reason. While the recognition of eusociality was initially treated largely as an exercise in definition, it increasingly came to be accepted as something real, with this particular suite of three features forming an organic whole [6, 11]. One indication in favor of this view is the relative numbers of species showing one, two, or three of the features. As a simple mathematical proposition, if the three criteria evolved independently of each other, we would expect to find at least some of the six intermediate syndromes represented by more species than the 2% of eusocial species, even if much less so than the solitary species. This is conspicuously not the case. None of the intermediate patterns is very common. Furthermore, they tend to be unstable, so that few are known to represent the permanent condition of any species.
Phylogenetic tests make it plain that (a) solitary life is in the ground plan of insects as a whole, as well as all orders with eusocial species (now that the termites are regarded as a subgroup of the order of cockroaches), and (b) eusociality evolved many times from solitary life, including several times within the Aculeata. The exact number remains uncertain, as new researches continue to alter our estimates up or down. As an example, results of a recent study are consistent with the hypothesis that sociality originated in the social wasps twice, not once, as was previously believed . Added to this is the complicating factor that evolution in some populations appears to have proceeded in the opposite direction, from sociality to at least facultative solitary life.
If eusociality arises from solitary life by the sequential addition of criteria, there are six possible sequences. However, some of these seem plainly unrealistic (e.g., that reproductive castes could arise ahead of cooperative brood care). The question over the evolutionary pathway, while it lasted, was around just two possibilities , with no suggestion that only one could have taken place in all independent origins of eusociality: (a) cooperative brood care → reproductive castes → overlap of generations, and (b) overlap of generations → cooperative brood care → reproductive castes.
However, this controversy did not last but faded before the view of eusociality as a true adaptive syndrome, such that the three criteria arise in concert, possibly even within very few generations . Possibly the first empirical support for this was the finding that some bees in Japan could show a eusocial colony cycle at one altitude and yet can be solitary at another altitude of the same island. This was supported by experimental findings, in which females of normally solitary bees, when forced into persistent interaction, developed cooperative brood care and a marked reproductive bias (caste).
Consistent with this hypothesis is the theoretical concept of a eusociality threshold, which a population crosses in evolutionary time or much more quickly into a new adaptive space. This does not necessarily involve any physical distinction between castes or even in the form of the nest (e.g., sweat bees, whose status as solitary or social usually usually cannot be determined except through direct behavioral observations), but it does open up a new way of life.
Within the traditional view, a distinction is commonly made between two levels in the broad, heterogeneous array of eusocial forms. In primitively social species, reproductive castes are not distinct from the workforce except sometimes by size. In advanced eusocial species, in contrast, they differ consistently and usually conspicuously [6, 11, 12]. Both levels are well represented in social wasps. Queens and workers of the primitively social Polistes and Mischocyttarus, for example, are not distinguishable by external physical structure. In contrast, in most colonies of vespines the queen can be recognized at a glance.
It is widely thought that primitive sociality can revert to solitary life, as is especially in evidence in sweat bees , crossing back over the eusociality threshold if environmental conditions favor it. However, the prevailing view is that advanced sociality is irreversible [1, 4, 12, 13]. That is, through the evolution of physiologically and physically distinct castes those lineages have reached a point of no return. This is readily seen in such groups as honey bees or stingless bees, for example, in which it is inconceivable that the queens could perform most worker tasks. However, at least a semantic complication is presented by those ants that secondarily have lost the physical queen, so that reproduction is by laying workers.
While the use of “social insects” as a synonym for eusocial species or populations enjoys wide acceptance – even near hegemony – and can now be regarded as traditional, it is not universally preferred. Some have advocated narrowing the definition to refer only to those groups in which queens (or queens and kings in termites) are irreversibly differentiated from the workforce, effectively equivalent to advanced eusocial insects .
An opposite approach – which appears to have gained somewhat greater acceptance – is to broaden usage to embrace a much greater array of species that characteristically live in groups during at least part of the life cycle [2, 3, 9, 10]. “Social insects,” then, can apply to any group of conspecifics organized in a cooperative manner. In this view, formal labels and especially the recognition of any hierarchy of social levels are sharply devalued, and even the concept of eusociality is not necessarily taken very seriously. Instead, the focus is on processes and especially communication as the key to sociality, with group members impacting each other’s fitness. In this broader view, the existence of a variety of insects and arachnids lacking at least one of the three criteria yet showing sophisticated communication and cooperation puts a strain on the acceptance of traditional eusociality as a phenomenon with any real unity.
- 2.Costa, J. T. (2006). The other insect societies. Cambridge: Harvard University Press. 767 pp.Google Scholar
- 3.Costa, J. T., & Fitzgerald, T. D. (2005). Social terminology revisited: Where are we ten years later? Annales Zoologici Fennici, 42, 559–564.Google Scholar
- 6.Hölldobler, B., & Wilson, E. O. (2009). The superorganism. New York: W.W. Norton. 576 pp.Google Scholar
- 7.Michener, C. D. (1985). From solitary to eusocial: Need there be a series of intervening species? In B. Hölldobler & M. Lindauer (Eds.), Experimental behavioral ecology and sociobiology (pp. 293–305). Sunderland: Sinauer.Google Scholar
- 10.Wcislo, W. T. (2000). Environmental hierarchy, behavioral contexts, and social evolution in insects. In R. P. Mertens, T. M. Lewinsohn, & M. S. Barbeitos (Eds.), Ecologia e Comportamento de Insetos (pp. 49–84). Rio de Janeiro: PPGE-UFRJ.Google Scholar
- 11.Wilson, E. O. (1971). The insect societies. Cambridge: Harvard University Press. 548 pp.Google Scholar