sign in sign up
<<
Home , Armaniidae, Ectatomminae, Formiciinae, Gerontoformica, Myrmeciinae, Myrmicinae

PERSPECTIVE

The rise of the : A phylogenetic and ecological explanation

Edward O. Wilson*† and Bert Ho¨ lldobler‡§ *Museum of Comparative , Harvard University, Cambridge, MA 02138-2902; ‡School of Life Sciences, Arizona State University, Tempe, AZ 85287-4501; and §Theodor-Boveri-Institut fu¨r Biowissenschaften (Biozentrum) der Universita¨t, Am Hubland, D-97074 Wu¨rzburg, Germany

Contributed by Edward O. Wilson, March 18, 2005

In the past two decades, studies of , behavior, and, most recently, DNA sequences have clarified the phylogeny of the ants at the subfamily and generic levels. In addition, a rich new harvest of and Paleogene fossils has helped to date the major evolutionary radiations. We collate this information and then add data from the natural history of the modern fauna to sketch a his- tory of major ecological adaptations at the subfamily level. The key events appear to have been, first, a mid-Cretaceous initial radia- tion in forest ground litter and soil coincident with the rise of the angiosperms (flowering plants), then a Paleogene advance to eco- logical dominance in concert with that of the angiosperms in tropical forests, and, finally, an expansion of some of the lineages, aided by changes in diet away from dependence on , upward into the canopy, and outward into more xeric environments.

͉ evolution ͉ phylogeny ͉

umanity lives in a world recently divided (4), comprising the myrmine and more derivative traits. The largely filled by prokaryotes, abundant and diverse and Burmese (10), containing spheco- fungi, flowering plants, nem- five other less prominent subfamilies. myrmines and a possible myrmeciine, atode worms, spiders, mites, The workers of , the may be of comparable or even some- Hand six ecological keystone best known sphecomyrmine , were what older provenance. groups: , hemipteran ‘‘true’’ a mosaic of and traits (5, 6). In still other very early deposits have bugs, phytophagan , cyclorraphan Specimens have been found in deposits been found additional products of the , glossatan moths, and hymenopter- of amber (fossilized initial ant radiation, which appear to be ans, the last comprising , — resin) in Asia, Siberia, and North Amer- members or precursors of the bulldog and ants (1). Ants are especially notable ica, hence across much of the northern ant subfamily , including among the for their ecological supercontinent of Laurasia (7, 8). the living ‘‘dawn ant’’ dominance as predators, , and Unlike the ants of later, Paleogene macrops. The subfamily is today limited indirect . Although the deposits, the sphecomyrmines are very to Australia, with one very rare Ϸ11,000 species of the ant (For- rare. Among the many thousands of of on New Caledonia. A pos- micidae) make up Ͻ2% of the known Cretaceous insect specimens examined sible but still uncertain myrmeciine or global insect fauna, they compose at from New Jersey, Canada, and Burma, myrmeciine precursor, Cariridris bipeti- least one-third of its biomass (2). In the only seven specimens have enough fea- olata, has been described from a single, Brazilian Amazon, as judged by one sur- tures preserved to be called definitively poorly preserved specimen in the San- vey (3), the biomass of ants is approxi- sphecomyrmine workers. Of the biology tana Formation of Brazil, Ϸ110 million mately four times greater than that of of these ancient ants it can be said only years old. What seem to us to be other all of the land vertebrates (, that they lived in mesic forests with rich myrmeciines are 10 rock fossils from the , reptiles, and amphibians) floras and insect faunas. The exact tim- Upper Cretaceous Orapa deposits of combined. ing and causes of their in the Botswana (Ϸ90 million years B.P.) (12). The combination of advanced colonial final 10 or 20 million years of that era Several later myrmeciines of Paleogene life and worldwide environmental influ- remain unknown (9, 10). age, Ameghinoia and Polanskiella of Ar- ence gives special significance to the Early in the history of the sphecom- gentina, Archimyrmex of the U.S. Green evolutionary history of the ants. Recent yrmines, a wider radiation beyond the River Formation, and two species of phylogenetic analyses, when combined stem genus Sphecomyrma occurred. In recorded from the Baltic with data from a rich new harvest of addition to clearly derivative spheco- amber, bear witness to the spread of the Cretaceous and Paleogene fossils and myrmine genera were a sphecomyrmine- subfamily around the world, followed by the natural history of the modern fauna, ponerine intermediate (8), an apparent its retreat to Australia. It is reasonable permit a defensible reconstruction of true poneromorph, a member of the ad- to suppose that the Myrmeciinae di- the broad-scale ecological features of vanced subfamily (9), and a verged from the sphecomyrmine stem ant evolution. possible member of the subfamily Aneu- and spread extensively by Late Creta- retinae, considered to be precursors of the ceous or times (13, 14). The Earliest Ants modern subfamily (10). By times, as evidenced in Ants evidently arose during the Creta- What appear to be the oldest certifi- 10 specimens of ants from the presumed ceous period at somewhat more than able ant fossils include Paleocene amber of Sakhalin, doli- 100 million years ago. Their earliest cretacica, from the Early to choderines and aneuretines (the latter known fossils fall into two groups. The Uppermost Abian (Lower Cretaceous) with one contemporary species, Aneure- first is the very primitive sub- amber of France, dated to Ϸ100 million family . The second years B.P. (11). Because of imperfect group consists of primitive members of preservation, it cannot be placed with †To whom correspondence should be addressed. E-mail: the extant subfamily Formicinae and the reference to fossil or existing subfamilies [email protected]. poneromorph group of subfamilies, as but evidently contains a mix of spheco- © 2005 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0502264102 PNAS ͉ May 24, 2005 ͉ vol. 102 ͉ no. 21 ͉ 7411–7414 Downloaded by guest on September 29, 2021 tus simoni, surviving in Sri Lanka) had made their appearance alongside of poneromorphs, myrmeciines, and for- micines (15). By the Early to Middle Eocene, as revealed by the recently described Fushun amber fauna from northeastern China that is Ϸ50 million years or some- what younger in age, the diversification of the major groups of ants was in full play (16). Among some 20 identifiable specimens of workers and queens is a variety of primitive ponerines and myr- micines as well as primitive members of the ‘‘formicoid’’ complex, comprising formicines and an assortment of what are either primitive dolichoderines, an- euretines, or both. Most also bear traits shared with their presumed spheco- myrmine or sphecomyrmine-like ances- tors, including short mandibles with small numbers of teeth, circular or ovoid head shapes, and relatively unmodified middle body segments. Several sedimen- tary compression fossils of Early Eocene provenance recently discovered in Brit- ish Columbia also represent relatively primitive poneromorphs (B. Archibald, personal communication). Emergence of the Modern Fauna The next glimpse into the history of the ants has come from a set of three speci- mens found in a middle Eocene amber Fig. 1. A schematic of the evolution of ants (Formicidae) at the subfamily level. Minor subfamilies are deposit at Malvern, Arkansas, that omitted. The four most diverse, abundant, and geographically widespread living subfamilies are bold- represents three of the dominant sub- faced. The ‘‘dorylomorphs’’ are driver and army ants () and an array of other, smaller subfamilies related to them. The schematic is based on our interpretation of the consensus of recent phylogenetic families in the present-day fauna: reconstructions (8, 9, 23–26). The were a Cretaceous of the Formicidae that were , Dolichoderinae, and believed not to be social; i.e., they had no anatomically distinct worker class (7). Formicinae (17). That the modern radiation had not only begun by Eocene times but was Origin and Phylogeny of the Mesozoic. The myrmeciines re- full-blown is disclosed by the Baltic am- The phylogeny of ants has been the sub- treated to Australia and New Caledonia, ber fossils, which are Middle Eocene in ject of several steadily improving recon- and the aneuretines retreated to Sri age. In examining 10,988 specimens as structions during the past half-century, Lanka. But the myrmicines, formicines, part of their pioneering ant fossil stud- starting with morphological data of and dolichoderines, together with the ies, G. Mayr (18) and W. M. Wheeler living faunas (9, 14, 21–23), then ad- ponerines (the dominant poneromorph (19) distinguished no fewer than 92 gen- vancing with additional morphological subfamily), not only flourished, but era. Ants were among the dominant in- information supplemented by fossil evi- spread worldwide as ecologically pre- sects, and the species as a whole had a dence, as reviewed here, and, finally, by eminent insect groups. The history of distinctly modern aspect, despite its the addition of DNA sequence data (24, the Ponerinae is of special interest in great age of Ϸ45 million years. 25). Fig. 1 presents a simplified version this story. Evidently monophyletic and Not surprisingly, the richly docu- of what we interpret to be the consensus comprising three tribes and 25 genera mented and younger amber of the among recent published reconstructions. (4), the ponerines are as variable overall Dominican Republic, which is Early The exact origin of the ants (family For- in anatomical characteristics and pat- in age (and thus roughly 20 micidae), involving an ancestral form terns of colony organization as any of million years old) has an even more presumably close to Sphecomyrma,as the other subfamilies. Yet, despite being modern composition. Of 38 genera and well as the timing of the origin, remain with the Formicinae and Myrmeciinae well defined subgenera identified (20), to be clarified. One estimate from DNA members of the oldest documented liv- 34 have survived somewhere in the New divergence extrapolation places the ori- ing phylogenetic assemblage and despite Ϸ World tropics to the present, although gin of ants at 140 million years B.P. having achieved such prolific diversifica- Ϯ all of the species studied thus far are (ma) (26); another puts it at 185 36 tion and geographic breadth, they re- extinct. Of the surviving genera and ma (95% confidence limits) (24). main oddly primitive in their general subgenera, at least 22 persist on His- social organization (27). paniola (Dominican Republic plus The Ponerine Paradox Haiti). Fifteen genera and subgenera What was the fate of the first evolution- Y Queens and workers of ponerine spe- have colonized the island since amber ary radiation of the ants? The Sphecom- cies are much closer in size than are times, restoring the number to 37. yrminae evidently vanished by the end the castes in the ‘‘higher’’ ant subfam-

7412 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0502264102 Wilson and Ho¨lldobler Downloaded by guest on September 29, 2021 ilies Myrmicinae, Dolichoderinae, and Possibly toward the end of Cretaceous poneromorphs (mostly ponerines) com- Formicinae and subfamilies. times, but more likely during Paleocene posed 22.2% of the species and 12.4% Y Most ponerine queens have relatively and Early Eocene times, after the end- of the specimens, compared with 10.6% low fertility, seldom producing more of-Mesozoic extinction spasm of many of the species and 12.9% of the speci- than five eggs each day. plant and groups, the ponerines mens for the formicines and a relatively Y Corresponding to the low degree of underwent most of the adaptive radia- paltry 1.1% of the species and 0.5% of queen fecundity, colony sizes of tion in taxonomic tribes and adaptive the specimens for the dolichoderines. ponerines are small. types that persist to the present time. But these three subfamilies of the ‘‘big Y Young ponerine queens of many, per- The fossil record of this interval, four’’ were dwarfed by the myrmicines, haps most, species start new colonies spanning some 30 million years, is still which made up 65.2% of the species and independently but not claustrally. poorly represented for ants, preventing 73.7% of the specimens. Clearly, the That is, they leave their natal nest, an exact placement of the ponerine radi- myrmicines, many of which (such as mate, and construct a nest of their ation in geological history. members of the tribes Dacetini and Ba- own or find a preformed cavity. But During this primary expansion, the sicerotini) are specialized predators like then, like some wasps and other pon- ponerines became entrenched worldwide the ponerines, are the rulers of the eromorph ants, they forage outside as predators, especially in world’s forest litter. the nest to obtain at least part of the warm-temperate and tropical mesic for- The soil and ground litter of the food with which they rear the first ests. They also evolved to favor ground world’s angiosperm forests, and espe- brood of workers. and leaf-litter sites. In effect, they pre- cially tropical forests, comprise the Y With a few, sometimes striking excep- empted this array of opportunities, par- habitat with the highest density and tions (28, 29), workers of ponerine tially blocking the sites from the later, species diversity of ants. Because all of species forage alone and do not use otherwise more successful radiations of the subfamilies of ants since their ori- odor trails or other signals the advanced subfamilies Dolichoderinae gins in the mid-Cretaceous have living to recruit nestmates to food sources and Formicinae. Of the ‘‘big four’’ in representatives, save the Sphecomyrmi- they encounter. present-day diversity and geographic nae and (the latter were Y True oral , the exchange of reach, only the Myrmicinae rivaled the giant ants, not to be confused with regurgitated food among adults and Ponerinae in invasion of the forest-floor Formicinae), and most of the genera between workers and larvae, is wide- predator niches. Either the myrmicine originating since the late Eocene also spread in the higher subfamilies radiation coincided with the peak of the have them, it is reasonable to suppose Myrmicinae, Dolichoderinae, and For- ponerine expansion or followed close be- that the tropical forest soil and ground micinae. It is evidently much less so in hind it. litter has always played the same role. the Ponerinae and, to date, has been The diets of by far the most known This habitat is reasonably interpreted observed only in the closely related ge- ponerines consist primarily or exclusively to be the headquarters of ant evolu- nus and (30), with of fresh insect and other arthropod prey, tion, from which major ant groups liquid otherwise transferred from drops supplemented by the scavenging of ar- have spread into other habitats or, in a held between the mandibles (31). thropods newly killed by other causes. great many cases, failed to do so. This, then, is the ponerine paradox: Many ponerine species are, moreover, The picture changes radically for the globally successful yet socially primitive. specialized predators (22, 27). ponerines and most other ponero- The puzzle might be partially resolved if Ants foraging as solitary huntresses morphs, for example, in habitats away the more advanced ant subfamilies can have small colonies, a necessity imposed from the tropical and warm-temperate be shown to be derived from a ponerine by the relative paucity of their food, and forests. They are notably scarce in cool- stock or, put in the more exact language the more so when they specialize on temperate forests, deserts, and arid of cladistics, the paradox is partly solu- particular kinds of prey. Predation as a grasslands. (In contrast, one other pon- ble if the ponerines prove to be way of life and small colony size, in eromorph subfamily, the , polyphyletic (of multiple origins). But turn, render other social traits simple has done well in these peripheral habi- even if such proves to be the case (con- and hence ‘‘primitive.’’ Foraging tends tats, especially in Australia.) trary to the opinion of systematists to be solitary, recruitment and alarm The picture changes in equal degree who consider the ponerines overall simple, and worker subcastes few to across a few vertical meters in the for- monophyletic), there likely remain di- none; workers are prepared to repro- ests and thence on up into the forest verse modern subgroups that are surely duce on their own in the absence of the canopy. In the Amazonian and Bornean monophyletic, such as the very large and queen, and brood care tends to be rela- forests, for example, subfamily domi- global tribe . So the paradox tively unorganized. nance is nearly flipped upside down. would not be truly resolved by taxo- The ponerines and other poneromorphs There, the formicines and dolichoder- nomic splitting. do well as predators and also as occupants ines have risen sharply in numbers The full and definitive solution to the of nest sites on forest floors in warm re- relative to the myrmicines, while the ponerine paradox and other mysteries of gions around the world (32). The favored ponerines have dropped to very low ant evolution requires more information sites are small spaces in the many dimen- levels (33, 34). from paleontology and ecology than sions of the litter, including the interiors The great majority of the canopy ant now exists. In the interim, and in the of rotting logs and stumps, tree limbs and species appear to be specialized for ar- hope of stimulating research in the twigs lying on the ground, clusters of dead boreal life, as opposed to being invaders critical areas, we offer the following leaves, masses of bryophytes, and the root from ground nests. Oddly, the arboreal combined phylogenetic and ecological systems of living trees, shrubs, and herba- ants, as a whole, are so abundant and ‘‘dynastic-succession’’ hypothesis (under ceous plants. possess such a large part of the animal the ‘‘Dynastic-Succession Hypothesis’’ The ponerines hold their own in this biomass as to be inviable if they lived heading below), consistent with existing complex and nutrient-rich environment. exclusively as predators and scavengers. knowledge (commentary added above In samples recently analyzed from for- Although tropical arboreal ants are, in and below heading). ested localities around the world (32), general, highly efficient in these roles

Wilson and Ho¨lldobler PNAS ͉ May 24, 2005 ͉ vol. 102 ͉ no. 21 ͉ 7413 Downloaded by guest on September 29, 2021 (35), there are not enough herbivorous gymnosperm flora worldwide, forest lit- empted there by these two groups, but insects available as immediate sources of ter became more complex. (In all re- more successful at creating homopteran protein to support them. This additional spects, especially structural but also symbioses. However, they were able to paradox now appears solved: It has been chemical and microclimatic, the litter of penetrate environments less available to learned that in the Amazon forest, at angiosperm forests is much better for predators, including cool-temperate cli- least a large fraction of the ant popula- ant colonies than that of gymnosperms.) mates and tropical forest canopies. Their tions of the canopy are not primarily Insects inhabiting the soil, ground litter, success is reflected in their high levels of predators but ‘‘cryptic herbivores,’’ sub- and vegetation of the forests and savan- abundance in amber (especially worker sisting on the liquid exudates of nas grew correspondingly diverse and specimens) and rock fossils (winged speci- hemipterous sap-feeding insects, includ- abundant. In Paleocene and Early Eo- mens), as would be expected from a pre- ing scale insects and treehoppers (36, cene times, the ponerines experienced ponderance in arboreal habitats. 37). They further tend many of these an adaptive radiation, and some of the In a phrase, the breakout of the doli- insects as a kind of cattle, protecting genera appearing back then have sur- choderines and formicines, and to some them from parasites and predators. Ex- vived to the present time. extent that of the myrmicines, was due trafloral nectaries are yet another signif- During, or perhaps more precisely, to a change in diet. This shift, in turn, icant plant-based source of nutrition. toward the end of the ponerine expan- was aided by the rising dominance of Around the world, in all major habitats, sion, and probably no later than the angiosperms over much of the land envi- a large fraction of the species of doli- Early Eocene, the myrmicines began ronment, an expansion that began in the choderines and formicines have invested their own radiation. They became formi- Cretaceous and culminated in the Paleo- their biology heavily in symbioses with dable competitors of the ponerines for cene and Eocene. It was furthered by the aphids, mealybugs, and other homopter- both prey and nest sites. In time, they expansion of the -producing ho- ous insects. Myrmicines have done so to equaled and then surpassed the poner- mopterans and lepidopterans, groups also lesser degree, and ponerines almost never. ines in biomass and diversity. Many also favored by the angiosperm dominance. Although quantitative data are lacking, added seeds and elaiosomes to their di- these disparities are commonly observed ets and, at least partly as a result of Future Research in natural history studies, and it seems these important new sources of oils and The ecological history of the ants through clear that the dolichoderines and for- carbohydrates, were able to expand geological time, culminating in the profu- micines have benefited in the growth of more effectively into deserts and dry sion of complexly social creatures around their biomass and diversity through coevo- grasslands. us today, must be accounted one of the lution with symbiotic homopterans. These Most importantly, some of the myr- great epics of evolution. Its unfolding, observations to the final part of the micines added symbioses with homopter- however, can still be glimpsed only in dynastic-succession hypothesis. ans to their repertory: largely, scale insects fragments. Large gaps remain in the fossil and treehoppers in tropical and warm- record, especially across the critical period The Dynastic-Succession Hypothesis temperate vegetation, aphids in cool- of major radiation extending from the late ͞ ͞ The K P (Cretaceous Paleogene, or temperate vegetation, and mealybugs Cretaceous into the Paleogene. Of equal ͞ ͞ end-of-Mesozoic), or K T (Cretaceous underground everywhere. Similar symbio- consequence, the life cycles and natural Tertiary, as commonly denoted) extinc- ses were contracted with the caterpillars history of the vast majority of living spe- tion event may have terminated the of honeydew-secreting butterflies. cies, which still bear the indelible stamp of sphecomyrmine ants, or these insects Dolichoderines and formicines also di- this history, remain unexplored. could have vanished before the event, versified, perhaps with the myrmicines but but anatomically primitive ponerines more likely later, in Early to Middle Eo- We thank B. Archibald, D. Grimaldi, M. soldiered on. During the reassembly and cene times. They were less successful than Kaspari, C. Saux, T. Schultz, and M. S. Wang continued expansion of the flowering the ponerines and myrmicines in the for- for their invaluable help during the prepara- plants, which replaced much of the old est litter environment, having been pre- tion of this article.

1. Grimaldi, D. (2000) Proc. XXI Int. Congr. Ento- 14. Ward, P. S. & Brady, S. G. (2003) Invert. Syst. 17, 27. Peeters, C. (1997) in Social Behavior in Insects and mol. 1, xix–xxvii. 361–386. Arachnids, eds. Choe, J. C. & Crespi, B. J. (Cam- 2. Wilson, E. O. (1990) Success and Dominance in 15. Dlussky, G. M. (1988) Paleontol. J. 22, 50–61. bridge Univ. Press, New York), pp. 372–391. Ecosystems: The Case of the Social Insects (Ecology 16. Hong, Y.-C. (2002) Amber Insects of China (Bei- 28. Baroni Urbani, C. (1993) Insectes Soc. 40, 233– Institute, Oldendorf͞Luhe, Germany). jing Scientific and Technological Publishing 250. 3. Fittkau, E. J. & Klinge, H. (1973) Biotropica 5, 2–14. House͞Henan Scientific and Technological Pub- 29. Witte, V. & Maschwitz, U. (2002) J. Insect Behav. 4. Bolton, B. (2003) Mem. Am. Entomol. Inst. 71, lishing House, Beijing), in 2 vols. 15, 195–217. 1–103. 17. Wilson, E. O. (1985) Psyche (Cambridge, MA) 92, 30. Hashimoto, Y., Yamauchi, K. & Hasegawa, E. 5. Wilson, E. O., Carpenter, F. M. & Brown, W. L. 205–216. (1995) Insect Soc. 42, 137–144. (1967) Science 157, 1038–1040. 18. Mayr, G. (1868) Beit. Natur. Preussens Ko¨nigliche 31. Ho¨lldobler, B. (1985) Isr. J. Entomol. 19, 89–99. Physikal.-O¨ konom. Gesells. Ko¨nigsberg 1, 1–102. 32. Ward, P. S. (2000) in Ants: Standard Methods for 6. Wilson, E. O. (1985) Paleobiology 13, 44–55. 19. Wheeler, W. M. (1914) Schrift. Physikal.-O¨ konom. Measuring and Monitoring , eds. Agosti, 7. Dlussky, G. M. (1983) Paleontol. J. 17, 65–78. Gesells. Ko¨nigsberg 55, 1–142. D., Majer, J. D., Alonso, L. E. & Schultz, T. R. 8. Grimaldi, D. A., Agosti, D. & Carpenter, J. M. 20. Wilson, E. O. (1985) Science 229, 265–267. (Smithsonian Institution, Washington, D.C.), pp. (1997) Am. Mus. Novitates 3208, 1–43. 21. Brown, W. L. (1954) Insectes Soc. 1, 21–31. 99–121. 9. Grimaldi, D. A. & Agosti, D. (2000) Proc. Natl. 22. Ho¨lldobler, B. & Wilson, E. O. (1990) The Ants 33. Bru¨hl, G. A., Gunsalam, G., Edward, K. & Lin- Acad. Sci. USA 97, 13678–13683. (Harvard Univ. Press, Cambridge, MA). senmair, K. E. (1998) J. Trop. Ecol. 14, 285–292. 10. Grimaldi, D. A., Engel, M. S. & Nascimbene, P. C. 23. Baroni Urbani, C., Bolton, B. & Ward, P. S. (1992) 34. Wilson, E. O. (1987) Biotropica 19, 245–251. (2002) Am. Mus. Novitates 3361, 1–71. Syst. Entomol. 17, 301–329. 35. Floren, A., Biun, A. & Linsenmair, E. (2002) 11. Nel, A., Perrault, G., Perrichot, V. & Ne´raudeau, 24. Crozier, R. H., Jermiin, L. S. & Chiotis, M. (1997) Oecologia 131, 137–144. D. (2004) Geol. Acta 2, 23–29. Naturwissenschaften 84, 22–23. 36. Tobin, J. E. (1995) in Forest Canopies, eds. Low- 12. Dlussky, G. M., Brothers, D. J. & Rasnitsyn, A. P. 25. Astruc, C., Julien, J. F., Errard, C. & Lenoir, A. man, M. D. & Nadkarni, N. M. (Academic, New (2004) Insect Syst. Evol. 35, 1–13. (2004) Mol. Phylogenet. Evol. 31, 880–893. York), pp. 129–147. 13. Baroni Urbani, C. (2000) Ecol. Geol. Helv. 93, 26. Brady, S. G. (2003) Proc. Natl. Acad. Sci. USA 100, 37. Davidson, D. W., Cook, S. C., Snelling, R. R. & 471–480. 6575–6579. Chua, T. H. (2003) Science 300, 969–972.

7414 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0502264102 Wilson and Ho¨lldobler Downloaded by guest on September 29, 2021