Summary. We present field and laboratory data on the Neotropical formicine ant Gigantiops destructor (Fabricius), the sole species of its genus and tribe. Monogynous and poly- domous colonies of G. destructor from French Guiana are distributed along the rainforest edges or along streams with a nest density of about 300 nests/ha. The species presents rudi- mentary nesting habits, as most nests are found in pre-exist- ing cavities in the ground, but some can be found in the hol- lowed internodes of Cecropia trees fallen onto the ground. A worker, sometimes hidden in a separate “sentry box”, might guard the nest entrance. Colony size can reach several hun- dreds of workers. The foraging activity of the workers is strictly diurnal with a peak between 9 : 30 and 11: 30. This largest-eyed of all known ant species has remarkable leaping abilities even more impressive than in other jumping ants. Workers are generalist solitary foragers. They collect extra- floral nectar from different plant species and prey on various small live arthropods that they detect visually before tracking and jumping on them. The same individual can forage both on prey and sugary sources during a single foraging trip. For- agers can eat their prey on site and never recruit nestmates in the field or even after a starvation period in the laboratory. They feed larvae with chewed prey. The complete lack of cooperation between foraging workers that can also fight for a prey with a nestmate, combined with the absence of any recruitment for large food sources, constitute a cluster of individualist traits rather unusual for an eusocial insect. Key words: Gigantiops destructor, nest structure, nest distri- bution, foraging activity, food preferences. Introduction Firstly described as Formica destructor (Fabricius, 1804), the Formicinae Gigantiops destructor is noteworthy as it is the only species of its genus and of the Gigantiopini tribe (Ash- mead, 1905). This species is confined to the rainforests dis- tributed along a strip of South America east of the Andes, extending from about 10° north to 15° south of the equator (Wheeler, 1922; Kempf and Lenko, 1968; Tobin, 1989). Its life history is little known and limited to the following infor- mation. Workers forage solitarily on the ground, or some- times among the branches of trees, leaping from twig to twig. They run and jump away when pursued by human observers (Smith 1858; Emery, 1893; Mann, 1916; Wheeler, 1922; Tobin, 1989; Hölldobler and Wilson, 1990). As a result, it is rather difficult to follow workers returning to their nests, so that early attempts to localize them were unsuccessful. Wheeler (1922) found only two queenless nests in British Guyana, Tobin (1989) found two nests, one queenless and the other queenright, in eastern Venezuela and a total of six nests in the Manu Biosphere Reserve in Peru. Other attempts to capture entire colonies were unsuccessful and, consequently, males were described only in the second half of the twentieth century (Kempf and Lenko, 1968). For all these reasons, this formicine ant species still merits investigation. In addition, because this species presents impressive forward jumping abilities and has the largest and most prominent eyes of all known ant species to date, it has recently attracted the inter- est of neuroanatomists (Jaffe and Perez, 1989; Tautz et al., 1994; Gronenberg and Hölldobler, 1999). As foraging workers have been frequently observed in the Amazon basin, we hypothesized that G. destructor is not scarce, so that a systematic search would permit us to deduce the ecological conditions of nest site selection in this species and consequently to find nests for further study. We examine Insectes soc. 48 (2001) 347 – 351 0020-1812/01/040347-05 $ 1.50+0.20/0 © Birkhäuser Verlag, Basel, 2001 Insectes Sociaux Research article Colony structure and foraging behavior in the tropical formicine ant, Gigantiops destructor G. Beugnon 1 , P. Chagné 1 and A. Dejean 2 1 Laboratoire d’Éthologie et Cognition Animale, FRE-CNRS 2382, Université Paul-Sabatier, 118 route de Narbonne, 31062 Toulouse cedex 4, France, e-mail: beugnon@cict.fr 2 Laboratoire d’Écologie Terrestre, UMR-CNRS 5552, Université Paul-Sabatier, 118 route de Narbonne, 31062 Toulouse cedex 4, France, e-mail: dejean@cict.fr Received 21 November 2000; revised 19 March and 15 May 2001; accepted 21 May 2001.