Glandular Sources and Specificity of Some Chemical Releasers of Social Behavior in Dolichoderine Ants

Edward O. Wilson, Mario Pavan
1959 Psyche: A Journal of Entomology  
Introduction Species of the ant subfamily Dolichoderinae are excellent subjects or the experimental analysis of chcmical communication. The group is advanced evolutionarily, and large, complex societies are the. rule. During foraging, workers of many species orm long, conspicuous columns tightly bound to persistent odor trails. When disturbed they void volatile secretions that seem likely to function, at least in part, as olfactory releasers of alarm behavior. Furthermore, the dolichodefines
more » ... e dolichodefines have recently been the object of extensive biochemical research by Pavan, Cavill and their associates (cf. Pavan, I95O; Pavan and Ronchetti, I955; Pavan and Trave, I958; Cavill, Ford, and Locksley, I956; Cavill and Locksley, I957). Several terpenoids including the previously unknown iridomyrmecin, iso-iridomyrmecin, and iridodial, have been identified as components of the anal gland secretion of various species. The availability of these substances in purified forfi has made possible for the first time the precise behavioral assay ot natural secretory products in ants. Trail S.ubstances Applying the artitcial trail technique used earlier by one of us (Wilson, 959) in studies of the myrmicine Solenopsis saevissima, an attempt vas made to determine the glandular source, of the trail substance in workers of Iriclomyrmex humilis Mayr. Groups of workers from laboratory colonies (grown from stocks originating from Baton Rouge, Louisiana) were attracted to drops of M sucrose solution, about 6 mm in diameter, on glass feeding tables. Selected body parts of freshly killed workers vere made into. artificial trails directed at the feeding workers. The number of workers following the trails beyond the trail midpoint, i.e., I5 cm. from the sucrose bait, were recorded. When positive responses were obtained, the duration of biological activity was timed to the nearest half-minute. Preliminary experiments showed that the trail substance is limited to the abdomen, [/Filson and Pavan Chemical Releasers in Ants 71 smears from other body parts proving consistently inactive. This result was not unexpected, since humilis workers can be clearly seen to drag the posterior portion of the abdomen over the ground while in the act of laying trails. Five organs in the abdomen are capable of emptying glandular secretions to the outside: (I) the hind-gut; (2) the paired "true" poison glands, with a large reservoir, the poison vesicle; (3) Dufour's gland (also called the accessory gland) (4) the anal glands, opening just above the anus; (5) the glandular "ventral organ" recently discovered by Pavan (1955) . In a series of experiments, the five organs were dissected out of freshly killed workers, washed in insect Ringer's solution, and bioassayed with artificial trail tests. The results, summarized in table I, show that the bulk, and perhaps the entirety, of ,he trail substance resides in the ventral organ. The occasional weak TABLE 1. lridomyrmex humilis workers. Organ Number Number of of tests positive responses* hind gut true poison glands, plus reservoir Dufour's gland anal gland ventral Results of artificial trail assay of selected abdominal organs of organ * Positive res Number of workers responding: range (with mean ) 0-2 (M<I) 0-2 (M<I) 0-30 (M=8) 0-15 (M--6) 65-339 (M=134) onses are defined, arbitrarily, as the crossing Duration of responses in minutes: range (with mean) 0-1/2 (M< 1/2) point by 10 or more workers. 0-1/2 (M< 1/2) 0-4 (M=I) 0-1 (M-l) 2-9 (M=5) of the trail mid- responses to the contents of the Dufour's and anal glands may not be biologically significant, i.e., may not indicate the presence of substances normally secreted into the trail, since responses of comparable magnitude have been obtained with pure methylheptenone, a substance not known to occur in humilis workers. In a second series of experiments, the ventral organ and anal glands of Iridomyrmex pruinosus (Roger) 4 were assayed. As in humilis, the ventral organ smears caused strong trail following while re-4Collected at Baton Rouge, Louisiana. 72 Psyche [December sponses to anal glands were negligible. Furthermore, artificial trails made with 2-heptenone, the anal gland secretion oi lruinosus (M. S. Blum, in litt. ), caused negligible responses. In a third series o experiments, the hind-gut, poison apparatus, Duour's gland, and ventral organ were assayed in a laboratory colony of the primitive dolichoderine Mo.n.acis bispin.osa (Olivier)2 The unusually restless and aggressive nature o the Monacis workers required the following modification in procedure. Straight trails 75 cm. in length were drawn down the middle o a clean glass plate. Three workers were then released, one at a time, and allowed to wander over the glass plate, crossing and re-crossing the trail. A positive response was recorded when the worker persistently followed the trail to its end. Of 8 workers thus. exposed to trails made successively from 6 ventral organs, 7 gave positive responses. The responses o workers in triplicated control experiments using the other abdominal organs were all negative. After the chief source of the odor trails in the three dolichoderine species had been located, bioassays were made across species, in order to determine the extent o: species-specificity o the trail substances. As shown in table 2, each species appears to have a different substance. M. bistinosa L. occidentale I. hu rnilis I. pruinosus T. sessile TARLF. 2. Intraand interspecific trail-substance tests employed in the present study. Only intraspecific tests produced positive responses (q-). The results of all interspeeific tests employed were negative (--). Interspecifie tests not attempted are indicated by a question mark. Source Recipient Monacis Liometol)um Iridomyrmex Iridomyrmex Tapinoma bispinosa occidentale humilis pruinosus sessile + ? / Further, the Dufour's glands and ventral organs of Monacis bispinosa were assayed with laboratory colonies of the myrmicine ants Crematogaster lineolata (Say), Solenopsis geminata (Fabr.), and S. saevissima (Ft. Smith). The ventral organ induced no detectable response in these species. The Dufour's gland caused no response in the Crematoyaster but, quite unexpectedly, caused strong trail-ollowing in 8olenopsis saevissima In act, single 3gonacis glands consistently drew out approximately the same number oi 8. saevissima workers as single saevissima glands. It was also determined that a single Monacis Collected at Palmar, southern Costa Rica. 1959] Wilson and Pavan Chemical Releasers in Ants 73 Dufour's gland caused a weaker response in S. yeminta, well below the intensity caused by a yeminata gland and approximately the same as that caused by a saevissima gland (see Wilson, 96o). These results are interpreted as indicating the chemical identity, or nearidentity, of the 3lonacis bispinosa and Solenopsis saevissima Dufour's secretions. Curiously, the Solenopsis use the secretion as a trail substance, but the Monacis do not. The function of the secretion in Monacis remains unknown. Some further comment is needed concerning the organ we are here referring to as the ventral organ. The gland was discovered in Iridomyrmex humilis by Pavan (955) and described in greater detail by Pavan and Ronchetti (955) . It is a paddle-shaped organ resting on the ventral body wall just over abdominal sternite V/ (sternite IV of gaster). It opens posteriorly via a neck-like constriction between the VI and VII abdominal sternites. The opening is fitted with an unusual beak-shaped deformation of the posterior border of sternite VI, a structure seemingly designed to guide the effluence of the gland contents. Until the present experimental work, the function of the gland remained unknown, although Pavan and his associates suggested
doi:10.1155/1959/45675 fatcat:t4fq4nwo35cinoxezwsp2qt32e