P

Poneroid , which included the current poneroid as well as , the formicoid Roberto A. Keller1 and Christian Peeters2 subfamilies and , 1Museu Nacional de História Natural e da Ciência and all the ▶ that do not display army- & cE3c-FCUL, Universidade de Lisboa, Lisbon, behavior (formerly included in the now obso- Portugal lete Cerapachyinae). Ponerinae sensu 2CNRS, Institute of Ecology and Environmental lato was defined by putative ancestral traits Sciences, Sorbonne Université, Paris, France exhibited by its members rather than the existence of shared derived characteristics. William Morton Wheeler remarked that “the Ponerinae comprises Poneroid ants are a composed of six well- unmistakably primitive and generalized forms and defined subfamilies that are morphologically het- therefore constitutes a group of twofold interest, erogeneous (, first, as the ancestral stock of the higher subfamilies, , Apomyrminae, Paraponerinae, and second, as the oldest existing expression of Ponerinae and ) (Fig. 1). Poneroids social life among the Formicidae” [13]. Wheeler’s include just 11% of extant ant , and molec- view persisted for a century, with subsequent ular data [4] recognize it as the of the authors explicitly treating Ponerinae sensu lato as formicoids [12], the clade with the bulk of species an all-encompassing from which the other diversity (Fig. 2). Not included in either of the subfamilies arose [5]. poneroids or formicoids are ▶ Leptanillinae and The twenty-first century has already brought Martialinae, two diminutive subfamilies com- major changes to the higher classification as sub- posed of highly specialized tiny subterranean families were redefined [2], morphological charac- workers. Their relationship to other extant ants ters were formalized in a phylogenetic framework remains inconclusive. The now obsolete term [5], and internal phylogeny was revealed by DNA “poneromorph” [2] should not be confused with data ([4] and included references). The current “poneroid.” The various “-morphs” were tentative scheme (Fig. 2) recognizes poneroids not as a groups of similar-looking subfamilies without any collection of primitive or ancestral ants, but as an claim of phylogenetic relationships or formal tax- evolutionary radiation that occurred independent onomic status. from and simultaneous to formicoids [4]. Prior to quantitative phylogenetic analyses, The poneroid radiation resulted in a set of there was a long taxonomic history of treating morphologically distinct subfamilies, several of ants that are superficially similar to ances- which have only one or two extant species, tors as a large and heterogeneous subfamily suggesting a high degree of past extinction events.

© Springer Nature Switzerland AG 2020 C. Starr (ed.), Encyclopedia of Social , https://doi.org/10.1007/978-3-319-90306-4_99-1 2 Poneroid Ants

Poneroid Ants, Fig. 1 Poneroid ants are almost all mixta (Proceratiinae) predate specifically on eggs ground predators (solitary or in a group; generalists or (yellowish, mixed with white ant eggs); (c) highly prey-specific), with colonies seldom reaching soror (Ponerinae) shows the archetypal limited size diver- more than hundreds of workers: (a) goblin gence between queens (dealate queen is bottom left) and (subfamily Amblyoponinae) workers are minute and blind workers (male is top left); (d) commutata cooperative hunters of linear ; (b) (Ponerinae) colonies have a few thousand workers that Poneroid Ants 3

Amblyoponinae is peculiar because most of its spe- between the two to understand the possible causes cies lack the diagnostic petiole of ants (Fig. 1a). This of their greater ecological success. alone led to the idea that amblyoponines represented In light of the current phylogeny, it is tempting the transition between and “higher” ants, to focus on Leptanillinae and Martialinae for clues strengthening the view that poneroids as a whole about early stages in ant evolution, as this clade were ancestral, but this atypical petiole is better lies outside the formicoids and poneroids (Fig. 2). considered a reversal. Apomyrminae (one species) However, we cannot assume that, relative to a was initially placed within Amblyoponinae, then sister clade that is morphologically diverse and Leptanillinae due to its similar underground habitus, highly speciose, a small clade (69 species) repre- and is now the sister clade to Amblyoponinae (Fig. sents the ancestral condition [6]. Rather, lone 2). Proceratiinae (Fig. 1b)and▶ Paraponerinae (Fig. branches are often evolutionary “experiments” in 1e) were historically associated with Ectatomminae adaptation for a particular niche that were not very as a within Ponerinae sensu lato and considered successful and an evolutionary dead end due to the link to subfamily . Indeed, their mor- morphological specialization. A better interpreta- phology includes a mixture of primitive traits shared tion is that the larger sister clade is a lineage that with other poneroids and derived traits common remained generalized, thus giving rise to greater among formicoids. The poorly-known diversity in the long run but also retaining species Agroecomyrmecinae has two extant monotypic gen- that more likely reflect ancestral conditions. Not era previously placed in Myrmicinae. Amid these only is it unlikely that ants radiated from a highly sits Ponerinae sensu stricto (Fig. 1c, d), specialized subterranean clade with reduced mor- containing the bulk of the poneroid species and phology [7], but the record is exhibiting a broad range of morphologies, particu- rich with unspecialized forms that lived above larly in their mandibles that reflect diverse feeding ground [1]. habits [11]: generalist to highly specialist predators, hunting exclusively underground or above ground. The impact from the shift of viewing poneroids How to Assess Social Complexity in as an ancestral stock to recognizing them as one of Ants? two main independent ant radiations cannot be overstated. Because ponerines-as- ants was Sociality is always a composite of interconnecting the dominant paradigm, authors turned to traits including age structure, genetic composi- poneroids for clues about putative ancestral states tion, group size, communication mechanisms, for everything from morphology to behavior, division of labor, and degree of task specificity. “ ” while formicoids were held as the higher ants In ants these interacting traits can be magnified by and their traits considered advanced. Poneroids as marked morphological differences among adult an independent parallel radiation to formicoids, nestmates (queens and workers) and often however, require reassessment of social complex- supplemented by enduring mutualisms with ity across different lineages. Traits uniquely pre- plants, bacteria, fungi, or sap-sucking insects. sent among poneroids might not represent The diversity in ant diets and ecology is unparal- examples of early stages in ant evolution, but leled among insects, but commonalities across rather advanced states. Likewise, the formicoids lineages are easier to grasp by focusing on the as a whole exhibit a mix of ancestral and advanced fundamental dichotomy in colony characteristics traits, and it is important to carefully discern (modified from Ref. [3]): (A) strategies,

i.e., how resources are acquired, and (B) how

ä ä

Poneroid Ants, Fig. 1 (continued) are column raiders on of the very few arboreal hunters (solitary generalists). ; (e) clavata (Paraponerinae) are one Body sizes range from 3 to 25 mm. (Photos © Alex Wild) 4 Poneroid Ants

Poneroid Ants, Fig. 2 Evolutionary relationships among extant subfamilies of ants. Area of triangles represents relative numbers of species (phylogeny after Ref. [4]; species count after AntCat.org [March 2020]) resources are allocated to produce new adults, i.e., hunting, mass raiding, scavenging dispersed car- body size of queens and workers, their degree of casses, collection of honeydew, or assembly-line dimorphism, and the number of workers (colony tasks like fungus-farming. Mature colony size is size) and gynes reared annually. Both worker determined by queen fecundity (queen number morphology (including body size) and colony also), the degree of caste dimorphism, and the size have likely coevolved with foraging strate- amount of resources extracted from the environ- gies across ants, while queen morphology (includ- ment. An increased dimorphism is mostly linked ing body size and ovariole numbers) has to a reduction in size of the workers, with the coevolved with the hazards of colony foundation consequence that more individuals can be and selection for optimal size and longevity of manufactured with finite colony resources colonies. Hence, social complexity in an organi- (¼ food intake) [9]. Increased caste dimorphism zational sense can be assessed by comparing also allows “claustral” ▶ colony foundation, i.e., degrees of dimorphism between workers and founding queens are able to raise the first workers queens. Unlike social wasps and bees that are using internal metabolic reserves only if the constrained by the need to fly for both castes, ant queen-worker size difference is sufficient. workers can evolve to be 10–20 times as small as Because claustrality requires gynes that are costly queens [9]. In addition, various ant species show relative to workers, only populous colonies have strong morphological specialization within the enough resources to produce them in sufficient worker caste (e.g., minors and majors) or a soldier numbers every year. caste with unique traits absent in workers. Hence the behavioral intricacies inherent to social life are fi ampli ed in the large proportion of ant species Social Complexity in Poneroid vs. showing pronounced morphological differences Formicoid Ants within colonies. Colony size is an essential attribute of ant The close phylogenetic relationship between ants species, because evolutionary increases in forag- and spheciform [12] suggests that the ing force allowed new lifestyles, i.e., group ancestor of ants was a solitary winged ground- Poneroid Ants 5 hunting parasitoid wasp. We can thus expect that The vast majority (90%) of formicoid species early ants expressed transition states including (i) belong to subfamilies , , slight divergence in the body size of queens and and Myrmicinae (Fig. 2) that often show a high wingless workers, (ii) “non-claustral” indepen- degree of queen-worker dimorphism. Claustral col- dent colony foundation (founding queens must ony foundation is strictly restricted to these three forage to feed the first brood), and (iii) small large subfamilies [9], even though shifts to depen- colony size. These expectations are supported by dent foundation evolved frequently, often accom- the biology of various extant ants, both poneroid panied by the evolution of ▶ non-flying queens and formicoid, e.g., sikorae, (ergatoid queens or short-winged queens). These and (all Ponerinae), formicoid subfamilies can also show huge () and Rhytidoponera increases in colony size, often resulting from the (Ectatomminae). In these, besides the presence/ evolution of minute workers [9]. Moreover, novel- absence of wings, queens and workers are highly ties in reproductive structure such as polyandry, similar in size and have the same ovariole num- biased sex allocation ratios, and thelytoky com- bers (Fig. 1c). Non-claustral ICF is the rule across bined with or instead of sexual reproduction are Amblyoponinae and Ponerinae, and also in the concentrated in these three clades. Similarly, formicoid subfamilies Ectatomminae, ▶ social-parasitic queens are unknown in poneroids, ▶ Myrmeciinae, and ▶ [8]. but ubiquitous within the large formicoid subfam- Similarly, many species in all these subfamilies ilies (and just a couple species of and have colonies with just a few dozens of workers or Myrmecia). Unlike Formicinae or Myrmicinae, the less [8, 11]. Therefore, many poneroids and a Dorylinae include taxa at both extremes of a gradi- minority of formicoids share these three ancestral ent in queen-worker divergence and social com- traits. Even more, sporadic shifts to “” plexity (e.g., monomorphic workers in Lioponera (mated egg-laying workers) reproduction (list of vs. highly polymorphic workers and huge queens in species at the end) are restricted to these subfam- ). ilies (except Pseudomyrmecinae). In formicoid lineages, larger colonies can be Importantly, the occurrence of trophallaxis dif- associated with striking variations in body size fers sharply between poneroids and formicoids. among workers, underlying an increased efficiency Trophallaxis underlies trophic exchanges and in division of labor. However, a few poneroids show storage within colonies, and anatomical adapta- the developmental ability to produce pronounced tions in the digestive track (e.g., crop, proventric- worker as well as strong queen- ulus) seem crucial for this. Trophallaxis worker dimorphism, e.g., australis, (excluding “pseudotrophallaxis”, where sweet and specialists such as liquid food is carried between the mandibles) is bequaerti, analis,andtheNeoponera absent in poneroids ( is an exception), and laevigata species-group [10, 11]. Others such as this may have constrained larger colony sizes. In lutea, castanea, contrast, trophallaxis is found throughout the tarsatus,andSimopelta pergandei formicoids, including the lineages that otherwise show strong queen-worker dimorphism only [10, show primitive traits. This unequivocal difference 11]. Intracolonial body size variability appears may reflect a fundamental shift in diets that charac- restricted to species with larger colonies, but both terize formicoids: almost all poneroids are strict pred- these traits never reach the magnitude characteristic ators, while many of the Ectatomminae, of many formicoids. Myrmeciinae, and Pseudomyrmecinae collect In summary, poneroid taxa radiated within the or honeydew in addition to live or dead arthropods. narrow confines of involving individ- This shift is associated with the evolution of arboreal uals or small groups (Fig. 1). In contrast, lifestyles that characterize many Dolichoderinae and formicoid subfamilies radiated in very distinct Formicinae. ways linked with a highly diversified diet together with trophallaxis: some kept their social traits 6 Poneroid Ants unchanged (and resemble poneroids), while others 5. Keller, R. A. (2011). A phylogenetic analysis of ant innovated with exaggerated queen-worker dimor- morphology (: Formicidae) with special reference to the Poneromorph subfamilies. Bulletin of phism and its numerous consequences on colony the American Museum of Natural History, 355,1–90. structure. Rather than a sequential series of radia- 6. Linksvayer, T. A., & Johnson, B. R. (2019). Re-think- tions, the four dominant ant subfamilies (Ponerinae, ing the social ladder approach for elucidating the evo- Myrmicinae, Formicinae, and Dolichoderinae) lution and molecular basis of societies. Current – fi “ Opinion in Insect Science, 34, 123 129. diversi ed together [11]. List of gamergate species 7. Lucky, A., Trautwein, M. D., Guénard, B. S., Weiser, https://antwiki.org/wiki/Category:Gamergate”“List M. D., & Dunn, R. R. (2013). Tracing the rise of ants – of species with ergatoid queens https://antwiki.org/ Out of the ground. PLoS One, 8(12), e84012. wiki/Category:Ergatoid_queen” 8. Peeters, C. (1997). Morphologically “primitive” ants: Comparative review of social characters, and the importance of queen-worker dimorphism. In B. Crespi & J. C. Choe (Eds.), The evolution of social behaviour References in insects and arachnids (pp. 372–391). Cambridge: Cambridge University Press. p. 552. 1. Barden, P., & Grimaldi, D. A. (2016). Adaptive radi- 9. Peeters, C., & Ito, F. (2015). Wingless and dwarf ation in socially advanced stem-group ants from the workers underlie the ecological success of ants cretaceous. Current Biology, 26, 515–521. https://doi. (Hymenoptera: Formicidae). Myrmecological News, – org/10.1016/j.cub.2015.12.060. 21,117 130. 2. Bolton, B. (2003). Synopsis and classification of 10. Peeters, C., & Molet, M. (2010). Evolution of Formicidae. Memoirs of the American Entomological advanced social traits in phylogenetically basal ants: Institute, 71,1–370. Striking worker polymorphism and large queens in – 3. Boomsma, J. J., & Gawne, R. (2018). Super- Amblyopone australis. Insectes Sociaux, 57, 177 183. organismality and caste differentiation as points of 11. Schmidt, C. (2013). Molecular of no return: How the major evolutionary transitions ponerine ants (Hymenoptera: Formicidae: Ponerinae). – were lost in translation. Biological Reviews, 93,28– Zootaxa, 3647, 201 250. 54. https://doi.org/10.1111/brv.12330. 12. Ward, P. S. (2014). The phylogeny and evolution of 4. Borowiec, M. L., Rabeling, C., Brady, S. G., Fisher, B. ants. Annual Review of Ecology, Evolution, and Sys- – L., Schultz, T. R., & Ward, P. S. (2019). Compositional tematics, 45,23 43. heterogeneity and outgroup choice influence the inter- 13. Wheeler, W. M. (1910). Ants. Their structure, devel- nal phylogeny of the ants. opment, and behavior. New York: Columbia Univer- and Evolution, 134, 111–121. sity Press.