Howard Topoff and Ellen Zimmerli.
Formica wheeleri: Darwin's predatory slave-making ant?
Psyche 98:309-318, 1991.

Full text (searchable PDF, 684K)
Durable link: http://psyche.entclub.org/98/98-309.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

FORMICA WHEELERI: DARWIN'S PREDATORY
SLAVE-MAKING ANT?
Hypotheses concerning the evolution of slavery in ants emphasize the role of excess predation (Darwin, 1859), territorial encounters (Alloway, 1980; Wilson, 1975), or brood transport among polydomous colonies (Buschinger, 1986). Although these views are not mutually exclusive, many investigators have noted the general absence of group predation in dulotic ants (Holldobler and Wilson, 1990). This has been thought to be true for facultative slave making ants belonging to the genus Formica, although most research has been conducted only on the European species F. san- guinea (Dobrzanski, 1965), and the related North American species, F. subintegra (Talbot and Kennedy, 1940). Relatively lit- tle is known about the numerous other species belonging to the sanguinea group. For example, although colonies of Formica wheeleri are common in western United States, we have but one account of their raiding and nesting activities (Wilson, 1955). In the mountains of southeastern Arizona, colonies of F. wheeleri are sympatric with Polyergus breviceps, an obligatory slave-making ant with which F. wheeleri competes for the same slave species. In this paper, we present the results of a field study revealing that F. wheeleri may be the only facultative slave-maker to exhibit the general predatory behavior predicted by Darwin's hypothesis for the evolution of slave-making ants.
MATERIALS AND METHODS
Seven colonies of Formica wheeleri were monitored from June 16 to July 16, 1991, during the time period 0730-1700 h (MST). All colonies were located in the Chiricahua Mountains, Cochise County, Arizona. The field site is located 14 km west of Portal, * Department of Psychology, Hunter College of CUNY, 695 Park Avenue, New York, N.Y. 10021, and Department of Entomology, The American Museum of Nat- ural History, New York, N.Y. 10024
Manuscript received 1 November 1991




================================================================================

Psyche
[Vol. 98
Arizona. At an elevation of 2750 m, the habitat is dominated by Ponderosa pine and Douglas fir. Although three species of Formica are abundant in this habitat (F. occulta, F. marcida, and F. neogugates), only F. occulta is parasitized in this study site. Raiding Activities: The raiding characteristics of five colonies of Formica wheeleri are summarized in Table 1. Prior to the onset of the summer monsoon rains (June 23-July 6), the median time for group raiding was 1405 h. During the rainy season, the median time for group raiding was 4.5 h earlier. It is likely that the earlier onset of raids after July 6 was due to the corresponding drop in daytime temperature when the rainy season began. The data presented in Table 1 are only for raids in which at least 20 workers simultaneously left the nest area and proceeded in an organized group toward a target nest. But before and after such raids, every gradation of foraging behavior, from individual explo- ration to group raiding, was observed. Indeed, for most of the day, the distinction between raiding and individual foraging was entirely moot. When individual foragers left the nest, they scav- enged for dead arthropods, as do their F. occulta slaves. Slave individuals did not participate in group raids on other colonies of F. occulta.
Predatory Behavior: During most group raids by Formica wheeleri, host workers fled from their nest with brood, fighting between parasite and host was short-lived, and few workers of Table 1. Summary of Formica wheeleri raiding activities Range Median Sample size
No. of F. wheeleri on raid 21-660 114 42 Time of raid onset (MST)
June 16-July 06
July 07-July 16
Distance to target nest (m) 2-3 1 13 42
No. of brood taken (larvae + pupae) 2-1 82 25 16



================================================================================

Topoff & Zimmerli
either species were killed. Nevertheless, we observed numerous instances when F. wheeleri killed the resident F. occulta workers and took the dead ants back to their nest. We also observed preda- tion on other species of ants. Our observations of group predation are summarized as follows:
June 26: During a foray conducted by only 13 raiders from colony 4, one worker of F. occulta was killed by three F. wheeleri, and dragged back to the nest.
June 27: During a raid by colony 3, two workers of Camponotus and one of Formica marcida were present near the nest of the tar- get F. occulta colony. All three ants were promptly killed and taken back to the F. wheeleri nest.
June 28: Workers from colony 3 killed eight F. occulta workers at the target nest. Again, all were hauled back to the parasite's nest. June 28: Raiders of F. wheeleri moved out in several bursts. At the target nest, the arriving ants extensively excavated the F. occulta nest entrance. As the resident workers came to the surface, they engaged in prolonged fights with the intruders. Fifty-five F. occulta workers were killed and immediately carried back to the F. wheeleri nest.
June 29: Approximately 1 hour prior to a raid by colony 4, a target nest of F. occulta was attacked by a colony of Polyergus breviceps, and approximately 45 F. occulta workers had been killed (but not removed) by the Polyergus. When the Formica wheeleri raiders arrived at the site, they promptly retrieved every dead worker.
July 12: A raiding party from colony 2 encountered a column of foragers belonging to the species Liometopum apiculatum. Fight- ing erupted immediately. Although this interaction aborted the slave raid, 28 Liometopum foragers were killed and taken back to the F. wheeleri nest.




================================================================================

Psyche [Vol. 98
July 12: A colony of Formica wheeleri raided a colony of Polyer- gus. Fighting was intense and workers of both species were killed. Although the Polyergus did not retrieve any dead Formica, the F. wheeleri raiders picked up at least 12 dead Polyergus workers, and transported them back to their own nest. Finally, throughout the month-long observation period, raiders of Formica wheeleri were observed attacking leaf hoppers, small caterpillars, and spiders that inadvertently wandered into the path of a slave-raid swarm. In all such instances, the arthropods were killed and carried back to the F. wheeleri nest. Brood retrieval: Our observations on predatory behavior suggested that Formica wheeleri might be less specialized than Polyergus with respect to species diversity of retrieved brood. To document this, 50 pupae of Camponotus ocreatus (a non-sympatric species) were placed 0.5 m in front of a F. wheeleri raid swarm. The raid halted and all pupae were removed in less than 5 min. When a similar batch of C. ocreatus pupae were placed in front of a Polyergus breviceps swarm, the raiders ran over the pupae with- out hesitating and none of the brood was captured. This procedure was repeated on three consecutive days, each with a different colony of Formica wheeleri and Polyergus. In all cases, every pupa was removed by F. wheeleri and none were removed by Poly- ergus.
A similar, and somewhat more natural, test was conducted on June 30. At 1330 h, a colony of Polyergus initiated a raid toward a colony of F. occulta located 45 m to the south. En route to the tar- get, the Polyergus raiders encountered a nest of Formica neoga- gates. Resident workers emerged from the nest, and intense fighting broke out as workers of both species grabbed legs and antennae. After approximately 5 min, panic alarm spread through the F. neogagates nest. The queen emerged, followed by hundreds of workers carrying brood. When most of the adult Formica work- ers had scurried away, the area immediately around the nest entrance was littered with more than one hundred undefended pupae. When fighting ceased, the Polyergus workers regrouped, and proceeded another 10 m to the target nest of F. occulta. Signif- icantly, not one of the F. neogagates pupae was retrieved. Eighty-



================================================================================

19911 Topoff & Zimmerii 313
five of these pupae were aspirated and placed in front of a raid swarm of Formica wheeleri. All the pupae were retrieved in less than 5 min.
Emigration: Two emigrations were observed during the field sea- son. The first colony moved 2 m to the east during the period June 23-June 27, and the second colony emigrated 4 m to the east from July 2 to July 6. Throughout the emigrations, both colonies main- tained their normal activity schedules. Thus, worker transport com- menced at approximately 1300 h, and ceased shortly after 1630 h. In the first colony, workers of Formica wheeleri transported all brood, callows (including callow alates of F. wheeleri), and adults of both species. Workers of Formica occulta were never observed carrying other individuals. Their activities were confined to nest maintenance at the new site. In colony 2, we did observe 1 individ- ual of F. occulta carrying an F. wheeleri adult. But again, all other brood and adult workers of both species were transferred to the new site by workers of F. wheeleri.
Facultative slave-making ants, like those in the sanguinea complex, represent an intermediate parasitic group, between freeliving species on the one hand, and obligatory dulotic species on the other. In recent laboratory tests by Mori and Le Moli (1988), slaves were removed from colonies of Formica san- guinea and Polyergus rufescens. Not surprisingly, the behavioral repertory of F. sanguinea expanded dramatically within 30 days of slave removal, with workers becoming self sufficient at feeding and brood care. Workers of Polyergus, by contrast, were unable to care for their brood, and experienced high mortality. The behavior of both species during emigrations and foraging (summarized in Table 2) also shows that Formica wheeleri is less specialized than Polyergus.
Most studies on the raiding behavior of species in the F. san- guinea complex confirm that slave raiders usually rout their oppo- nents, who typically flee in a state of panic alarm, and that aggressive encounters, when they occur, are brief and do not result in the death of adult individuals from either species (Talbot and Kennedy, 1940). Wheeler (1910) noted, however, that when large



================================================================================

314 Psyche [vo~. 98
Table 2. Comparison of behavioral characteristics between Formica wheeleri and Polyergus breviceps.
F. wheeleri P. breviceps
Emigrations F. wheeleri adults carry Only slaves transport adults both brood and adults of and brood of both species to both species to new nest new nest
Foraging Workers forage individually No individual foraging and on group raids
Brood retrieval Workers retrieve brood even Only brood from slave or closely
from different ant genera
related species retrieved
Adult ant capture Raiders kill and retrieve Raiders rarely kill, and never
adults even from different retrieve other adults ant genera
colonies of slave species offer resistance during raids, prolonged fighting is possible, and many workers of both species can be killed. Despite numerous instances of such fighting, Wheeler did not observe slave-raiders retrieving dead ants of either species. During the period 1955-1 960, Marikovsky (1963) transferred more than 400 colonies of Formica rufa into habitats occupied by F. sanguinea. The aggressive F. sanguinea virtually destroyed all the introduced F. rufa colonies. Marikovsky also reported that Formica sanguinea raided five other species of Formica, as well as Camponotus herculeanus, and Lasius flaws. Although adult work- ers were routinely killed during these forays, only larvae and pupae were transported back to the Formica sanguinea nest. More recently, Talbot (1985) studied raids by Formica gynocrates (another facultative parasite belonging to the Rapti- formica group), which pillages colonies of Formica vinculans. When attacked, resident workers pick up brood and evacuate their nest, and seek protection at the top of plant stems. Nevertheless, F. vinculans workers from large colonies can stage a counterattack, and may be killed in the melee. In one such vigorous raid observed by Talbot, 230 pupae and 89 larvae were captured. Near the end of the raid, when pupa transport dwindled, a few Formica gynocrates workers picked up dead Formica vinculans individuals and carried them home.




================================================================================

19911 Topoff & Zimmerli 315
Given the scant evidence for predation by facultative formicine slave-making ants, it is not surprising that researchers have consis- tently rejected Darwin's claim that dulosis evolved from predatory raiding. It is true that many F. wheeleri raids are similar to those reported for other sanguineu group species, in that brood is retrieved after driving resident adults from their nest, and that killing of adults typically only occurs when resident workers actively defend their nest. Nevertheless, it is clear from our obser- vations that Formica wheeleri is the most predatory sanguineu species known to date: (1) it takes adult ants as well as brood, and often just adults of other genera; and (2) this behavior is common. The discovery of significant predation by Formica wheeleri clearly does not call for a major shift in our ideas about the origin of dulotic behavior. First, it could still be argued that predatory behavior of F. wheeleri evolved from intraspecific, territorial raids. Although we observed no such interaction during this field season, we do have one past record of such a territorial encounter. Second, although an evolutionary hypothesis of dulosis based on predation can account for brood capture and the rearing of excess pupae, the important issue of non-independent colony foundation must also be addressed.
We have recently shown (Topoff et al., 1990) that newly mated queens of Formica wheeleri usurp colonies of F. occulta by driv- ing out the resident queen and workers, gathering the pupal brood into a pile, and rearing the pupae through eclosion. We have also argued that the ability to usurp the pupal brood of incipient nests belonging to closely related species would be of tremendous selec- tive advantage to a colony-founding queen (Topoff, 1990a, b). Through olfactory conditioning, rearing of raided brood in the nest of the queen's origin could specify the host species colony to be invaded, and could even provide the queen with some degree of host-species olfactory resemblance, similar to that hypothesized for Polyergus (Goodloe and Sanwald, 1985). Such host colony invasions would only be successful, however, if the invading queen usurped pupae from colonies of closely related species, thus ensuring similar ecological requirements. The role of predation in such an evolutionary process would be to facilitate the subsequent capture of additional pupae, with some being used as food, and others reared to become slaves. Although we observed raiders of



================================================================================

316 Psyche [Vol. 98
F. wheeleri capturing adult workers of Liometopum, Camponotus, and Formica marcida during slave raids, no nest of these species was invaded. To date, such raids on nests have only been directed toward other colonies of F. wheeleri, F. occulta, and Polyergus breviceps. If the species to be raided by F. wheeleri is influenced by olfactory experience from their own slaves, raids on these three species may be biased, because all contain workers of F. occulta. Since we have found no nests of F. wheeleri without slaves, we plan to create such homospecific colonies by providing a newly- mated queen with several hundred pupae. Our goal is to determine whether such homospecific colonies of F. wheeleri will exhibit even greater predation, by expanding the range of species whose nests are raided.
In southeastern Arizona, at an elevation of 2750 m, the facultative slave-making ant Formica wheeleri is sympatric with Polyergus breviceps. Although three additional species of Formica are common in this habitat, only colonies of F. occulta are raided for slaves. Like other species in the Formica sanguinea complex, every gradation of foraging behavior, from individual exploration to group raiding, was common. Unlike other dulotic species of Formica, however, slave raids often resulted in the killing and retrieval of adult ants (in addition to brood) from the invaded colony. We also observed a raid on a colony of Polyergus, an obli- gatory slave maker in direct competition with F. wheeleri, because it uses the same host species. In this raid, Polyergus adults were also killed and taken back to the F. wheeleri nest. Workers of Formica marcida, Camponotus, and Liometopum apiculatum who were foraging in the path of slave-raid swarms were similarly killed and retrieved. These field observations depict a slave-mak- ing ant far more predatory than other species in the Formica san- guinea complex.
The base of operations for this study was the Southwestern Research Station of the American Museum of Natural History. The study was supported by PSC-CUNY Grant 6-62219, and by



================================================================================

Topoff & Zimmerli
National Geographic Society Grant 458 1-9 1. We thank David Jas- par for his essential assistance in the field, and Stefan Cover for his typically astute comments on the manuscript. ALLOWAY, T. M.
1980. The origins of slavery in Leptothoracine ants. Amer. Nat. 115: 247-26 1.
BUSCHINGER, A.
1986. Evolution of social parasitism in ants. Tree 1: 155-160. DARWIN, C.
1859. On the origin of species. John Murray, London. DOBRZANSKI, J.
1965. Genesis of social parasitism among ants. Acta. Biol. Exp. 25: 59-7 1. GOODLOE, L., AND SANWALD, R.
1985. Host specificity in colony-founding by Polyergus lucidus queens (Hymenoptera: Formicidae). Psyche 92: 297-302. HOLLDOBLER, B., AND WILSON, E. 0.
1990. The ants. Harvard University Press, Cambridge, Massachusetts. MARIKOVSKY, P. I.
1963. The ants Formica sanguinea Latr. as pillagers of Formica rufa Lin. nests. Insectes Soc. 10: 119-128.
MORI, A., AND LE MOLI, F.
1988. Behavioural plasticity and domestic degeneration in facultative and obligatory slave-making ant species (Hymenoptera: Formicidae). Moni- tore Zool. 22: 271-285.
TALBOT, M.
1985. The slave-making ant Formica gynocrates (Hymenoptera: Formicidae). Great Lakes Entomol. 18: 103-1 12.
TALBOT, M., AND KENNEDY, C. H.
1940. The slave-making ant, Formica sanguinea subintegra Emery, its raids, nuptial flights and nest structure. Ann. Entomol. Soc. Amer. 33: 560-577.
TOPOFF, H.
1990a.The evolution of slave-making behavior in the parasitic ant genus Polyergus. Ethol. Ecol. Evol. 2: 284-287. l99Ob. Slave-making ants. Amer. Sci. 78: 520-528. TOPOFF, H., WEICKERT, T., AND ZIMMERLI, E. 1990. A comparative study of colony takeover between queens of facultative and obligatory slave-making ants. J. Insect Behav. 3: 813-817. WHEELER, W. M.
1910. Ants, their structure, development and behavior. Columbia University Press, New York.
WILSON, E. 0.
1955. Division of labor in a nest of the slave-making ant Formica wheeleri Creighton. Psyche 62: 130-133.
1975. Leptothorax duloticus and the beginnings of slavery in ants. Evol. 29: 108-1 19.




================================================================================




================================================================================