James F.A. Traniello.
Population Structure and Social Organization in the Primitive Ant, Amblyopone pallipes (Hymenoptera: Formicidae).
Psyche 89:65-80, 1982.

Full text (searchable PDF, 1268K)
Durable link: http://psyche.entclub.org/89/89-065.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.

POPULATION STRUCTURE AND SOCIAL ORGANIZATION IN THE PRIMITIVE ANT AMBLYOPONE PALLIPES (HYMENOPTERA: FORMICIDA E)
BY JAMES F. A. TRANIELLO'
Harvard University,
Museum of Comparative Zoology Laboratories Cambridge, Massachusetts 02138, U.S.A.
The genus Amblyopone contains the most morphologically and behaviorally primitive species in the poneroid complex of ants, and a detailed examination of their social structure could significantly contribute to the reconstruction of social evolution in the Formi- cidae. But because of their cryptic habits and distribution, the biol- ogy of the majority of species of Amblyopone and the related genera Mystrium, Myopopone, Prionopelta, and Onychomyrmex remains almost entirely unknown. Previous investigations have provided information on colony foundation (Haskins, 1928; Haskins and Enzmann, 1938; Haskins and Haskins, 195 l), ecology, behavior, and taxonomy (Wheeler, 1900; Brown, 1960; Gotwald and Levieux, 1972; Baroni Urbani, 1978), and physiology (Whelden, 1958). Still, many of the details of social organization in Amblvopone are lack- ing. I present in this paper the results of a two-year study on the behavior and ecology of Amblyopone pallipes. Study areas and nest collection
Thirty-one colonies of A. pallipes were collected under stones in a damp, white pine woodland in Westford, Massachusetts. A single colony was taken under the bark of a rotting log. Nests generally consisted of one or two shallow (6-lOmm) depressions in the soil immediately beneath the stone, from which a single gallery opened to subterranean chambers. Gentle excavation usually revealed one or two additional loosely structured chambers. Workers, queens, 'Present address: Department of Biology, Boston University, Boston, Massachusetts 02215
Manuscript received by the editor February 18, 1982. Pu&e W:65-81 (1982). hup ttpsychu einclub orgtS9tS9-065 html



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

66 Psyche [vol. 89
sexuals, and brood were found at all levels of the nest and were quickly aspirated.
Distribution and natural history
A. pallipes has been found in the eastern United States and in the St. Lawrence Valley in Canada in cool, moist, forested areas
(Brown, 1960). General references on the natural history of this species are given by Wheeler (1 900) and Haskins (1 928). Laboratory arrangements
Colonies were housed in artificial nests composed of a thick, moist filter paper bottom with cotton sides approximately 6mm high covered with a glass plate. The nests were placed in 15 X 22cm plastic boxes in which the humidity was kept high. The tot a1 nest area was roughly 10 cm2. A second chamber, similar in structure but somewhat larger, was connected to the nest as a foraging arena, where live prey were offered. Colonies were fed on whole, live geo- philomorph and lithobiid centipedes; in addition, elatyrid, bupres- tid, and tenebrionid beetle larvae were acceptable to the ants. This method of culture proved successful and greatly facilitated studies of social interaction since the activity of an entire colony could be monitored on the stage of a dissecting microscope. Ethogram data were compiled in this manner, and were analyzed using the methods of Fagen and Goldman (1977).
1. Life cycle and population structure.
Nest distribution and colony size. The spatial distribution of col- onies at the principal study site in Westford is presented in Fig. 1. An interesting feature of this population, in addition to its clumped distribution pattern is that three colonies were collected under stones in 1978 precisely where colonies were found the year before. This suggests that the colonies that were collected represented sub- units of a large, subterranean population. Each unit is small (modal size class = 9-16 workers). Complete collection data are presented in Fig. 2. Although distributional data are not given, a population of seemingly comparable density was discovered by Wheeler ( 1 900), who uncovered 30 nests in a three hour period. Also, the colony size data correspond closely to the data of Francoeur (1965, 1979, and personal communication).




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

9821 Traniello - Amblyopone pallipes 67
Fig. 1. - The spatial distribution of nests at the principal study site. Queen number. The frequency distribution of the number of queens in a colony is given in Fig. 3. Of 19 queenright colonies, 10 (52%) contained more than one dealate female. Observations of multiple queened colonies in the laboratory revealed that in at least some of these colonies each queen was functionally reproductive. However, many queens in apparently polygynous colonies did not lay eggs, and engaged primarily in worker tasks. Life cycle, colony reproduction, and population structure. Be- cause colonies were collected and censused throughout the spring and summer of 1977 and 1978, it is possible to outline the life cycle of A. pallipes (Fig, 4). Eggs are laid in late April or early May and larvae hatch and develop throughout June and July. Mature larvae pupate in mid-July and early August, and adults eclose approxi- mately two to three weeks later. Although small numbers of eggs and larvae are present in most colonies throughout the spring and summer, it appears that only one brood matures per year. The large number of eggs found in colonies collected in August hatch before September and overwinter as larvae (Talbot, 1957). It is possible that the winter chilling results in the determination of these larvae as sexuals. In late August and early September workers and sexuals simultaneously eclose unassisted from their pupal cases. The adults which eclose at this time are predominantly workers. In four colo-



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

Psyche
[Vol. 89
NO. WORKERS PER COLONY
Fig. 2. - Frequency distribution of colony sizes for 35 nests. nies collected in late August which reared brood in the laboratory, the ratio of 3:Q: tj
was as follows: 2:1:36; 0:5:13; 3:0:7; and 0:4:19. In all cases the worker population of a colony was at a maximum at this time. If this is considered in conjunction with the available information on nuptial flights in A. pallipes, then it is possible to speculate on colony reproduction and population structure.



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

Traniello - Amblyopone pallipes
2 4 6
NO. OF QUEENS PER COLONY
Fig. 3. - Frequency distribution of the number of dealate females in 19 queenright colonies.
Although the complete sequence of colony reproduction has not been observed, the studies of Haskins (1 928, 1979) and Haskins and Enzmann (1938) provide some evidence of its organization. Early in September, winged females leave the nest and disperse over short distances, finally alighting on the ground or low vegetation. Then, with gaster arched and sting partially extruded, they "call" males with a chemical sexual attractant. Males quickly locate females, copulation ensues, and soon after insemination females shed their wings and re-enter the soil; perhaps they return to the parent nest.



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

70 Psyche C-
At this point in the life cycle, the workerlqueen ratio is high- described above, yet colonies collected in the late spring are - smaller in size (approximately 50%). Therefore, colony rep r - tion by budding may occur if one or more fecundated queens with a portion of the worker force. This hypothesis has p r e w c - been considered by Wheeler (1900) and Brown (1.960), and %- ported by my data on colony growth, nest distribution, an.. structure. An additional feature of the nest distribution patter- ports the hypothesis of limited dispersal. The most dense p o p u = - of colonies occurred on the south side of an early stone wall <^- high), although nest sites were abundant on both the nort south sides, and soil, vegetational, and exposure parameter = peared to be identical. Also, laboratory observations indicac- alate females may shed their wings before mating occurs. On s - occasions newly eclosed females left the nest, shed their win,= - returned to the nest. Because mating occurs on the ground behavior does not exclude the possibility that these indivs- could eventually become inseminated. These females m a y return to the parent nest or may be adopted by a nearby colo - several laboratory experiments queens were introduced into - queenright nests or orphaned colonies. In all cases they accepted by both workers and queens. Similarly, workers c o u ZI transferred from one colony to another without aggr.ession. T
fore, populations of A. pallipes appear to be unicolonial and s - darily polygynous. Ecologically, the patchy distribution of t =
correlates with this type of population structure. 2. Social organization
The social ethogram. Social ethogram data were gathered. five colonies which were observed for a total of 73 hours, d - which 6,500 individual acts were recorded. The behavioral c= - of a single colony of A. pallipes (2 queens, 18 workers, brood- was studied for 25.7 hours is given in Table I. The total n u n - - acts observed was 42 (95% confidence interval for catalog siz- 47]), and the sample coverage was 0.9992. Behaviors listed %r ethogram having a frequency of 0 were observed in other c o l- and are included as part of the species repertory. The maj o r 5 - activities are common to many ant species; those that are un- -
will be discussed briefly. Antenna1 tipping is a behavior prev-^- described in Zacryptocerus varians (Wilson, 1975) which ace-



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

Traniello - Amblyopone pallipes




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

Traniello - Arnblyopone pallipes




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

Psyche [Vol. 8'
Table 1. - The social ethogram. N = number of acts observed in each caste. Behavioral Act
Workers (16) Queens (2)
N=2525 N= 158
I. Self-groom
2. Allogroom queen
3. Allogroom worker
Brood care:
4. Lay egg
5. Carry egg or egg pile
6. Lick egg
7. Lick larva
8. Carry, drag, or role larva
9. Bank mature larva with soil
10. Carry pupa
1 1. Lick pupa
12. Place larvae on prey
13. Assist removal of meconium
14. Assist larval molt
15. Lick ecdysial skin
Aggressive display:
16. Undirected
17. To worker
18. To queen
Predatory behavior:
19. Forage
20. Sting prey
2 1. Drag prey to nest
22. Drag prey within nest
23. Lick prey
24. Handle prey within nest
Nest maintenance and defense:
25. Guard
26. Handle nest material
27. Repair nest wall
28. Lick nest material
29. Excavate nest
30. Bury noxious object
3 1. Carry or drag dead worker
32. Carry or drag live worker
33. Extrude sting
34. Remove empty pupal case
35. Jitter
36. Jolt body
37. Lick meconium
38. Tip antennae
39. Flick antennae




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

19821 Traniello - Amblyopone pallipes 73 Workers (16) Queens (2)
Behavioral Act N=2525 N= 158
40. Pinch larvae 0
4 1. Cannibalize larva 0.0 170
42. Discharge subpharyngeal
pellet 0
Totals: 1 .O 1 .O
infrequently in A. pallipes. During this behavior the body is raised, the gaster is curved forward, and with the mandibles agape the antennae are held forward with their terminal funicular segments slightly inclined toward each other. The significance of antenna1 tipping is unknown, but it appeared to be part of a grooming sequence. Vibrational displays were given by workers if the nest wall was breached or if an individual was mechanically disturbed. If the stimulus was intense enough, other workers would show the same vigorous jittering behavior, consisting of rapid vertical movements of the head and thorax. This behavior had the effect of producing a general arousal within the colony and resulted in an increase in the number of workers appearing at the source of stimulation. In the case of nest damage, building behavior eventually occurred but did not immediately follow. This signal appears to be a primitive form of mechanical communication, in which alarm is propagated di- rectly through body contact. A similar vibratory display has been documented in A. australis (Holldobler, 1977). Workers and queens of A. pallipes have retained a number of behaviors that appear to reflect their wasp ancestry. Queens were seen grasping larvae and squeezing them in the neck region with their mandibles, thus causing them to regurgitate a droplet of clear liquid which they then consumed. Workers were never observed to regurgitate with other workers, queens, or larvae, and all individuals fed directly on prey. Aggression was observed between workers and queens. An aggressive display typically consisted of opening the mandibles and rising up on the extended legs. This behavior was usually exhibited by queens in the area of the egg pile and seemed to produce avoidance in contacted workers. These observations raise the question of whether queens maintain their reproductive status through behavioral dominance or inhibitory pheromones. Polyethism. Studies on the division of labor within the worker caste have revealed that temporal castes are absent in the species. A



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

74 Psyche [Vol. 89
complete account of polyethism in relation to the life history of A. pallipes is given by Traniello (1978).
Predatory behavior. Prey were found in only three of the colonies collected. In two colonies larvae were found clustered around litho- biid centipedes (length = 1.5-2.0cm), and in the third colony a carabid beetle larva was taken. In the laboratory, colonies were offered a variety of live arthropods that workers might encounter in leaf litter, soil, or rotting wood. Wood lice (Oniscus), house centi- pedes (Scutigera), and various millipedes were consistently rejected while small elatyrid, tenebrionid, and buprestid beetle larvae were carried to the nest and fed upon. The diet of A. pallipes appears to be restricted to live, linear-shaped arthropods that can be captured by workers. A related species, A. pluto, is entirely specialized on geophilomorph centipedes (Gotwald and Lkvieux, 1972). When large, robust-bodied centipedes (Lithobius sp.) were offered to col- onies of A. pallipes, workers were unable to grasp the prey due to its escape movements and body diameter. It is difficult to imagine a condition under which large prey could be captured, even if they were "cornered" in a narrow gallery. When Lithobius of similar size were held with forceps, workers were still unable to subdue the centi- pede. Freshly killed centipedes were not accepted. It is therefore difficult to support the hypothesis of a nomadic life style for A. pallipes. Although this species of Amblyoponini does not appear to move its colonies to the location of large, previously captured prey, other species, such as Onychornyrrnex do provide evidence linking group predation and nomadism in this primitive group of ants (Wilson, 1958).
Prey capture and retrieval is very stereotyped, and solitary hun- tresses stalk prey in a highly methodical manner. As prey are approached, workers advance cautiously, apparently orienting to odors or air microcurrents produced by the prey. When within strik- ing range (2-3mm) the mandibles are opened and the head is oriented orthogonal to the long axis of the prey. Then in a single motion the mandibles close around the prey, the legs elevate the body, and the gaster is swung forward. The prey is then repeatedly stung and the venom soon shows its paralytic effects. Initially, only the area adjacent to the cuticle penetrated by the sting is immobi- lized, and stinging continues until escape movements stop. Subse- quently, the retrieval of the prey begins after a brief period of self-



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

19821 Traniello - A mblyopone pallipes 75 grooming. The retrieval process varies in duration depending upon prey size, but even long (4.0-5.Ocm, 2.0-2.5mm in diameter) geo- philomorph centipedes are easily dragged to the nest. A number of orientation trips made between the prey and the nest generally pre- ceded retrieval. During these orientation runs, which were made throughout the retrieval process, workers continually checked their position relative to the nest. The prey was then dragged several centimeters; the worker then stopped, released the prey, and con- tinued homeward until she contacted the nest entrance. She then returned to the prey and repeated this sequence, alternating prey movement with orientation trips. Once the prey was in the nest, other workers approached and began vigorously licking the areas of the prey's body opened during capture. Larvae were either carried to the prey or, if close enough, moved toward it and adjusted their position on its body on their own accord. At times workers assisted in positioning larvae. Additional details of feeding behavior are nearly identical to those described by Gotwald and Lkvieux (1972). Communication during foraging. At times, two or three ants attempted to jointly carry prey, but cooperative efforts were hap- hazard and inefficient. But cooperative retrieval seems unnecessary due to the physical capabilities of individual ants. The critical ele- ment in prey capture is probably not retrieving, but subduing rela- tively large arthropods. Often after a worker began stinging a prey item, a second or third worker approached and assisted in para- lyzing the prey. The fact that workers were attracted to the point of prey capture suggested that additional ants may be recruited over short distances by orienting to prey odor, air currents, or some signal produced by the forager. To test the hypothesis that phero- mones are involved in this process I stimulated foragers to grasp and attempt to sting the tip of a pair of forceps and then lowered the worker, still attacking the forceps tip, in front of the nest entrance. The response of workers in the nest was dramatic. In five replicates, 5.8 k 2.3 workers/ 0.5 min approached the nest entrance under the experimental conditions. Only 0.2 k 0.1 workers/0.5 min were attracted to the nest entrance in controls (agitated forceps alone). This difference is statistically significant (.00 1 < p < .0 1 ; t = 6.1, Student's t-test). Although the possibility that stridulatory signals might be involved could not be ruled out, the results of these experi- ments suggest that chemical cues are involved in the attraction



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

Psyche [vo~. 89
response. Subsequently, crushes of the head, thorax, and gaster were offered on applicator stick tips at the nest entrance. Also, crushes of dissected poison and pygidial glands (Holldobler and Engel, 1979) were offered. Only head crushes produced attraction. Whelden's (1958) studies, in addition to our own histological investi- gations, revealed a group of large glandular cells at the base of the mandible. The indirect evidence described above suggests that dur- ing prey capture the contents of these cells are released, attracting workers in the vicinity to assist in subduing prey. 3. Ecology and social evolution
The results of this study and previous investigations suggest that populations of A. pallipes are unicolonial. Workers from different subnesfs within a population show no aggression toward each other. Such worker compatibility has been demonstrated in Rhytidopon- era metallica (Haskins and Haskins, 1979), whose populations appear to be structurally similar to those of A. pallipes, but occupy larger area geographically. Workers taken from nests three miles apart were not mutually hostile. The lack of aggression was consist- ent within, but not between populations. Ambyloponepallipes col- onies appear to be similarly viscous, but do not occupy as extensive an area.
Observations in the laboratory are in accord with Brown's (1960) position which states that after mating, females "always or usually return to the parent nest". Secondary polygyny in this species, in addition to its patchy distribution, indicates that this species isin the terminology of Holldobler and Wilson (1977) a habitat specialist. The characteristic A. pallipes habitat is cool, damp, heavily shaded woodland. Nest site and prey abundance are also important fea- tures. Populations apparently grow slowly, and through reproduc- tion by budding, eventually saturate the habitat. Such a scheme does not rule out the occurrence of dispersal flights, which have been witnessed on occasion (Haskins, 1928). As colonies become more populous within a habitat, dispersal flights should occur more frequently in order to colonize additional areas. Once a founding queen locates a preferred habitat, colony reproduction again is accomplished through budding. The strategy may be similar to that of the mound building species Formica exsectoides. However, it must be noted that in laboratory situations, A. pallipes queens have never been observed to successfully found colonies (Haskins, per-



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

19821 TVctniello - Ambtyopone pallipes 77 sonal communication). But it is difficult to determine whether this S s
an abnormality which occurs only in the laboratory or represents a inability of A. pallipes queens to found a colony alone. New 1' inseminated queens of A. australis found colonies in the p a r t i a n y claustral mode {Haskins and Haskins, 1951). However, A. austral'x.- is monogynous.
Within a habitat, A. paliipes escapes competition with the m a r- advanced groups of ants by additional specializations on micro- habitat and diet. This is in contrast to other unicolonial specÌà.e which are broad generalists.
Behaviorally, A. pailipes exhibits both primitive and advanced social traits, and many of the primitive characters are more consew- - the than those of Myrmecia. Age polyethim is lacking, and cornmt u - nication between individuals is primarily mechanical, although rudimentary short-range recruitment system that is mediated by pheremones exists. Among the primitive trophic characteristics % s the use of the sting to paralyze prey, which are subsequently Fee directly to the larvae without prior dismemberment. Employing t- "å´-Ìö sting to paralyze prey appears to be widespread in the Ponerinse and recently Maschwitz et al. (1979) have demonstrated that t hi venom of the oriental ponerine species Harpegnathus sahaius an d Lepiogenys chinensis indeed has paralytic, and not toxic, effecs-t s - Prey paralyzation also occurs in Daceton armigerum and Paltot^x_y - reus mrsatus (Wilson, 1962; Holldabler, pers. camm.). This is CCB n - trasted to mymicine species which use the sting as a defens%v e weapon. The importance of paralyzing but not killing arthrop id prey in Amblyoponepatlipes is obviously related to the direct prow G - $ioning of larvae; prey must be kept from decomposing until t h e y are consumed. Also, immobilization is necessary for successfu X retrieval, and energetically it is more efficient for solitary foragers t- carry paralyzed prey. The absence of regurgitation which is charsic- teristic of the Ponerinae, also is a primitive trait. Although one S the more distinctive features of A. pallipes and other Amblyoponi- n prey specialization, appears to be a conservative formicid trait, it- also possible that specialization was a response to competition.