NOTES ON NECROPHORIC BEHAVIOR IN THE ARCHAIC ANT MYRMECIA YINDEX (FORMICIDAE MYRMECIINAE) *

Ants of the Australian and New Caledonian genus Myrmecia apparently include the most archaic living Formicidae. Brown (t954) suggested that the genus may represent a relatively late evolutionary offshoot, specialized but fundamentally conservative, from a line ot: extremely generalized archaic forms represented as fossils by the genus Prionomyrmex of the Baltic amber and by the living Australian Nothomyrmecia macrops, of which only two workers have ever been found. Species of Myrmecia, therefore, may well illustrate the earliest patterns of Formicid social organization, and embody the most archaic patterns of Formicid social behavior, that we are likely to be able to study in detail in the laboratory or the field. The bodily habitus of Prionomyrmex and Nothomyrmecianot to mention of the far more archaic Mesozoic fossil genus 8pkecomyrma first described by Wilson, Carpenter, and Brovn (t967) which may well represent, as Wilson (97) suggests, an antecedent of the Myrmecioid complex of ants--all suggest active, epigeically foraging insects" a characteristic virtually universal in contemporary species of Myrmecia. This, combined with the fairly large size of the communities of many species of Myrmeciaspecific counts of 586 workers and over 2000 in total colony personnel have been made from larger colonies of M. gulosa (Haskins and Haskins, 95o) --raise the interesting question of how far these archaic forms may have evolved any pheromone-mediated patterns of communityintegrating behavior so conspicuous in many higher ants. This is an interesting and complex area of inquiry, the answers to which are far from obvious, as the recent investigations of Robertson suggest. A prior question may be significant in this context. Do ants of the genus Myrmecia exhibit characteristic behavioral responses to particular chemical substances normally encountered in the external


Introduction
Ants of the Australian and New Caledonian genus Myrmecia apparently include the most archaic living Formicidae. Brown (t954) suggested that the genus may represent a relatively late evolutionary offshoot, specialized but fundamentally conservative, from a line ot: extremely generalized archaic forms represented as fossils by the genus Prionomyrmex of the Baltic amber and by the living Australian Nothomyrmecia macrops, of which only two workers have ever been found. Species of Myrmecia, therefore, may well illustrate the earliest patterns of Formicid social organization, and embody the most archaic patterns of Formicid social behavior, that we are likely to be able to study in detail in the laboratory or the field.
The bodily habitus of Prionomyrmex and Nothomyrmecianot to mention of the far more archaic Mesozoic fossil genus 8pkecomyrma first described by Wilson,Carpenter,and Brovn (t967) which may well represent, as Wilson (97) suggests, an antecedent of the Myrmecioid complex of ants--all suggest active, epigeically foraging insects" a characteristic virtually universal in contemporary species of Myrmecia. This, combined with the fairly large size of the communities of many species of Myrmeciaspecific counts of 586 workers and over 2000 in total colony personnel have been made from larger colonies of M. gulosa (Haskins and Haskins, 95o) --raise the interesting question of how far these archaic forms may have evolved any pheromone-mediated patterns of communityintegrating behavior so conspicuous in many higher ants. This is an interesting and complex area of inquiry, the answers to which are far from obvious, as the recent investigations of Robertson suggest.
A prior question may be significant in this context. Do ants of the genus Myrmecia exhibit characteristic behavioral responses to particular chemical substances normally encountered in the external W'e have recently described one such set ot reactions: the stimulus to attack behavior in Myrmecia gulosa elicited by ormic acid, a normal exocrine defense product o its common prey-genus Camlohorus, but not, so ar as can be determined, o M. gulosa itself (Haskins,Hewitt,and Haskins,973). The present investigation is concerned with the reactions of workers of a species o Myrmecia to substances which commonly stimulate necrophoric behavior in higher ants, such as oleic and related tatty acids. Wilson (958) and Wilson,Durlach,and Roth (958)  prefers open, partially xerophytic woodland habitats, but ranges much further, to South Australia and probably New South Wales and Victoria as well. It is an unusually hardy species and its colonies can be maintained in laboratory culture or many years. The primary reason t:or choosing it or this investigation, however, was that, in the populations from which nests were selected for both field and laboratory work, even mature colonies are typically quite small. This and other characteristics suggest that the species is one of the socially less-evolved members of the genus and thus lend particular interest both to the reactions of its workers to necrophoric substances and to pheromones of higher ants.

Laboratory Tests
Reactions oi: workers of Jllyrmecia vindex to Oleic Acid A portion of a colony of M. vindex consisting of approximately 3o workers, :zo cocoons, and 25 larvae, collected at Kings Park, Perth, West Australia, on December 28, 96, and long established in the laboratory in an earth-containing Lubbock type glass nest of dimensions 7" X :z" and a//4" in depth, was set up in a foraging arena consisting of a rectangular plastic box, 48" X 24" X 6.5" in inside dimensions, to which the ants had free access. The bottom of the arena was covered with clean brown paper.
After the colony had become thoroughly conditioned to the environment, piles of brood which had been impregnated with to 2 drops of various substances were placed on cardboard sheets, 6" )< 8", at some distance i:rom the nest entrance. There was early indi-Counts of four completely excavated apparently mature colonies from this population were made in January, 1964, with the following results: cation that larvae were only marginally suitable for this work. Treated larvae were sometimes returned to the nest, as though larval odor, even when overlaid by that of oleic acid, still dominated behavior. With cocoons, however, worker reactions were more consistent.
At 5:3o p.m. two piles of IO larvae and IO cocoons each were placed on top of the Lubbock nest. The larvae were quickly restored to the nest. By 6 :I5 p.m., 5 cocoons of the untreated group had been transported to the nest, together with 3 from the fraction treated with oleic acid. By 6:38 p.m. all cocoons from the untreated pile had been transported to the nest, while no further oleic-acid-treated cocoons had been removed. By 8:55 the following morning, however, 4 of the oleic-acid-treated cocoons had been deposited in refuse middens in the arena, 2 being deposited 45" from the nest entrance and 2, IO" away. Three were still in original position on the cardboard, but had been well covered with earth. Earth grains obtained from inside the nest had also been deposited on spots of oleic acid wh;ch had drained from the cocoons to the cardboard.
The experiment was then repeated more precisely in the pattern of the work reported with P. badius. Ten squares of paper, " X 'P, were soaked with ole:c acid and scattered at random over the surface of the nest, together with lO identical but untreated squares, appropriately identified. Two larvae, likewise treated with oleic acid, were included. The trial was begun at 9"Io a.m.
Workers coming into contact with the treated obiects immediately executed marked cleaning movements and a rubbing of the g:ular surface of the head against the substrate. At c):2o, larva was picked up by a worker after much hesitation and held in the mandibles for 4 minutes, when it was grasped by a second, and then deposited by both at the nest entrance, but not taken inside. The impression of a conflict of drives was strong. One minute later this larva was again picked up, carried about 3" from the nest entrance, and dropped. By 9:5., it had been brought into the nest, while the second larva was held within /'' 72 Of the nest entrance. Two minutes later it also was brought into the nest.
At Io:oo a.m. a worker was be;nning to deposit earth trains on the oleic-acid-treated papers, and by lO:53, papers carried earth grains. By :oo p.m., one of the treated larvae had again been brought out of the nest, and a worker was preparing to drop it on the midden located 45" from the entrance. By 5:00 p.m. both treated larvae were on this midden. Buryin of the oleic acid squares continued but they were not moved toward the middens.

Psyche
[June At 5"IO p.tn., 2 Tenebrionid larvae, treated with oleic acid and the other untreated, were placed outside the nest entrance. At 5 :3 p.m. the untreated larva was carried into the nest, and minute later the treated larva also was seized and carried in. At 5:3o p.m., however, the treated larva was brought out, carried to a distance o 34", and discarded. The untreated larva was retained within the nest and shortly consumed.
These trials gave strong evidence that patterns both of burying and of rejection of objects impregnated with oleic acid are exhibited by M. vindex. It then seemed of interest to compare the reactions to oleic acid with those o other substances tested by Wilson et al. in their work with P. badius.
Reactions of workers of Myrmecia vindex to Methyl benzylamine; Caproic acid; Oleic acid; Formic acid Using the same colony and with the same experimental arrangements as above, 4 cocoons, ranging from reshly spun to nearly ready to eclose, were impregnated as ollows" Methylbenzylamine (3 drops) All these cocoons were placed on cardboards, and mounted on top of the Lubbock nest as before, at 2 :o4 p.m. At 2:9 p.m. the 2 controls and the 3 cocoons treated with ormic acid had been permanently returned to the nest. At 5:4o p.m., cocoon from the caproic acid group had been removed 4o.5 ' from the nest and deposited on a refuse midden. All other cocoons remained untouched.
At 8:45 a.m. on March 3 r, no further cocoons had been removed rom the cardboards, but a ?ew grains of earth had been deposited on the cardboards of the caproic-and oleic-acid treated objects. During the night the caproic-treated cocoon deposited on the refuse midden had been retrieved into the nest, but at 8:55 a.m. it was again brought out and carried to the midden, 44" from the nest entrance. Almost immediaeely it was again retrieved and returned to the nest. At 9:IO a.m., it was once more removed and dropped about 9 from the nest entrance. By 2:o3 p.m. this cocoon had been carried to a midden 36 from the entrance, but minute later 1974] Haskins & Haskins Myrmecia vindex 263 it was returned to the nest once more, only to be finally removed by 2:33 p.m. and dropped about 32" from the nest entrance. _At this time the 3 cocoons treated with oleic acid, the 3 treated with methylbenzylamine, and the remaining 2 treated with caproic acid were still in place on the cardboards, and a ew grains of earth had been deposited about each. They were permanently ignored. Triethanolarnine proved a powerful attractant, and the cocoons treated with it aroused immediate and intense attention. Within 3 minutes, was dragged into the nest..At 2:5I a second triethanolamine-treated cocoon was dragged to the nest enerance and dropped. At 2:52 the third was carried to the oleic-acid-treated group outside the nest and dropped there. Two control cocoons were carried into the nest at the same time. A spot of triethanolamine stain on the containing card was repeatedly examined by workers.

Reactions of workers of ]Plyrmecia vindex to
The second triethanolamine-treated cocoon, which was deposited initially near the nest entrance, at 2:56 p.m. was carried to a midden 45.5" away; it was then picked up almost immediately and redeposited at a distance of 40". _At 3:02 the triethanolamine-treated cocoon which had first been brought into the nest was taken out, carried to a point about 9 r away, and dropped. Earth grains had been deposited about the n-valeric acid group. The 3 oleic-acidtreated cocoons had not been moved nor buried. By 4:00 p.m. all three triethanolamine-treated cocoons had been discarded in a group 39' from the nest entrance; control cocoon was still in place; and all 6 cocoons treated with n-valeric and oleic acid remained in their initial position. At 4:50 p.m. there had been no urther change and the run was terminated.

264
Psyche [June On April 2, 965, the same experiment was repeated, substituting n-butyric or n-valeric acid. Results were essentially the same. Fiftyfive minutes ater the start, 3 o the 4 triethanolamine-treated cocoons had been transported or distances o 4o", 2", and 25.5" from the nest entrance and dropped; although none of the n-butyric or oleic acid cocoons had been moved, there was a considerable deposit of earth particles about them. One control cocoon had been moved about 2", but none had been carried into the nest. Twenty minutes later, the triethanolamine group was still being moved at frequent intervals, and the oleic acid group was being heavily banked with earth grains. No attention was paid to the n-butyric acid groups. The strong impression left by these trials, and others like them, was that oleic acid, and to a lesser degree caproic acid stimulate in Myrmecia vindex both a burying reaction and sometimes the transport of objects so contaminated to refuse middens. No evidence of overt digging was seen. Formic acid seemed without effect. Triethanolamine, however, functioned as an efficient excitant and attractant in higher concentrations. In lower concentrations it stimulated necrophoric behavior. The "conflict" behavior exhibited in the handling of some larvae and cocoons treated with several of these substances, when immatures were repeatedly taken into the nest, brought out again and discarded on a midden, then returned to the nest, sometime through three or our cycles, was in several cases striking.

Field Tests
Reactions of workers of Myrmecia vindex to Oleic acid; Caproic acid; Formic tcid; Methylbenzylamine It was of interest to conduct essentially similar tests with wild colonies in the field. A large and active colony of M. vindex in Kings Park, Perth, West Australia (a part of the same population from which the colonies had been taken for laboratory tests), was selected. It included approximately 2oo workers, had a single large entrance-hole, and a well-defined crater. Since workers of this population show a predominantly crepuscular and nocturnal foraging pattern in warm weather, tests were begun in the evening.
At 7:25 p.m., as dusk was gathering, 5 white cards, of dimensions 3 pp )4 5 pp, were mounted on the nest crater, approximately equidistant from the entrance. Twenty-five cocoons, obtained from a neighboring colony, were impregnated with 3 drops of test substance, with 7 retained as untreated controls. The arrangement and treatments were as shown below: Control Immediately on exposure a worker on the crater seized one control and dragged it inside the nest. _A_ second worker, encountering an oleic-acid-treated cocoon, started away. At 7:32, caproic-acidtreated cocoon was dragged into the nest, followed at 7:35 by a second control cocoon. At 7:37 a cocoon from the ormic acid group was taken to the nest. Between 7:37 and 7:40, 3 additional ormicacid-treated cocoons were taken into the nest, and one-half minute later the last control cocoon was taken in. One of the remaining two ormic acid cocoons was inadvertently toppled into the nest by another worker, leaving on the card. The other groups were untouched. At 8:02 p.m., when darkness forced cessation ot: observation, all cocoons treated with oleic acid, methylbenzylamine, and five ot: those with caproic acid remained in place on the cards. One cocoon treated with caproic acid had been deposited with the oleicacid-treated group.
At 6:00 a.m. the :ollowing day, 3 of the caproic-acid-treated cocoons had been removed, but it was not possible to determine whether they had been taken into the nest or discarded. Some grains of earth had been scattered on the caproic-acid-treated card.
Observations were continued throughout the day, but no urther attention was paid to the remaining cocoons. However, when the cards were removed from the crater at 7 :oo a.m. on the third day, it was found that a new entrance-hole had been excavated under the card carrying the oleic-acid-treated cocoons. It was clear that oleic acid, caproic acid, and methylbenzylamine effectively inhibited return of the cocoons to the nest by the workers, in marked contrast to those treated with formic acid and the controls, which were brought in promptl. Digging behavior was observed in the vicinity of oleic acid. In these experiments, however, the oleic acid objects were not buried nor transported to kitchen midcLens-perhaps because such middens of M. vindex are often located near the nest entrance. Indeed. it is an interesting behavioral characteristic of M. vindex, at least in this population that, at the seasons of maximum brood production, empty cocoons and pupal exuviae are requently piled near the nest entrance. That this disposition may be more than accidental is suggested by the behavior of some colonies Psyche [June in the laboratory (not those with which the above-mentioned tests were conducted), where such exuviae, originally scattered about a foraging arena, may be gathered into dense craters about the entrances to Lubbock nests.

Summary and Conclusions
In a parallel series o experiments to those reported by Wilson (I958) and Wilson, Durlach, and Roth (I958) with Pogonomyrmex badius, the effects o.f oleic acid, caproic acid, methylbenzylamine, n-butyric acid, n-valeric acid, ormic acid, and triethanolamine as behavioral releasers in the ant Myrmecia vindex were investigated both in the laboratory and the field. The comparison was deemed particularly interesting because o the archaic character of the Myrmeciinae and their societies.
As with Pogonomyrmex and other higher Formicid genera, oleic acid was i:ound to act as a releaser o both necrophoric and digging behavior, suggesting the early establishment o this set of reaction patterns in Formicid social evolution a not unexpected situation in view of the highly adaptive character o this pattern in ridding the nests o dangerous animal decomposition products, perhaps in response to the bacterial production o oleic and related fatty acids.
In Pogonomyrmex. n-butyric acid, n-valeric acid, and n-caproic acid stimulated weak to moderate alarm behavior, passing over into digging behavior. In Myrmecia, caproic acid stimulated mildly necrophoric behavior, associated with some digging and burying behavior. Exposure to /lter papers of cocoons impregnated with n-butyric acid resulted in deposition of earth grains on the treated object. No reaction was observed to n-valeric acid.
Thus responses to these substances were o the same general quality as those reported for P. badius. There were, however, very marked differences in the responses of I. vindex to two presumed releasing substances: formic acid and triethanolamine. Formic acid neither stimulated necrophoric nor digging behavior nor prevented immediate transport into the nest of formic-acid-treated cocoons. This result was somewhat surprising to the authors in view o the highly positive and specific attack reactions stimulated by formic acid in Myrmecia gulosa mentioned earlier, which have been reported elsewhere (Haskins, Hewitt, and Haskins, 1973). The most marked contrast between behavior patterns of P. badius and M. vindex, however, occurred with those released by triethanolamine. With M. vindex, vapors of this compound consistently stimuated the most co.nspicuous necrophoric behavior of any substance tested, while also acting as a

1974]
Haskins & Haskins Myrmecia vindex 267 decided attractant. In P. badius on the contrary, the reaction to triethanolamine was reported to be virtually neutral.
In this connection it is interesting to note that, in the populations studied, M. vindex commonly piles discarded cocoons and pupal exuviae on the crater of the nest, near the entrance, where they characteristically accumulate in some numbers, rather than transporting them far afield. Similar behavior has been described for P. badius (Wilson,Furlach,and Roth,I958), though in the latter case, removal of the discards is quickly accomplished by other foraging ants. The situation with M. vindex, however, appears more specific.
Indeed established colonies in the artificial nest may collect discarded cocoons and pupal exuviae scattered about the foraging area to form a crater about the entrance. It may be that subtle behavioral balances are involved, depending on the nature and "mix" of the included decomposition products, as suggested by the reversal of behavior, apparently in response to vapor concentration, released by triethanolarnine.