Bert Hölldobler and Hiltrud Engel-Siegel.
On the metapleural gland of ants.
Psyche 91:201-224, 1984.

Full text (searchable PDF, 6436K)
Durable link: http://psyche.entclub.org/91/91-201.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.

ON THE METAPLEURAL GLAND OF ANTS
Department of Organismic and Evolutionary Biology, MCZ-Laboratories, Harvard University
Cambridge, Massachusetts
The metapleural gland (also called metasternal or metathoracic gland), a complex glandular structure located at the posterolateral corners of he alitrunk is peculiar to the ants. Although the gland was noted by Meinert (1860) and Lubbock (18771, it was Janet (1898) who conducted the first detailed anatomical study of this organ, as part of his classic work on Mvrmica rubra. Additional details have been added by Tulloch (1936) on Myrmica laevinodis; by Whelden ( 1957a, b, 1960, 1963) on Amblyopone (Stigmatomma) pallipes, Rh~tidoponera convexa, R. metallica, Eciton burchelli, E. ham- alum; by Tulloch et a1 (1962) on Myrmecia nigrocincta; and by Kiirschner* ( 1970) on Formica pratensis. It is generally assumed that the metapleural gland is a universal and phylogenetically old character of the Formicidae. Even the extinct species Sphecomyrma freyi of Cretaceous age appears to have possessed a metapleural gland (Wilson et a1 1967a, b) and the organ is well developed in the most primitive living ant species Nothonijwnecia macrops (Taylor 1978) (see Fig. 2). In the course of our current comparative study of the internal and external anatomy of exocrine glands in ants, we discovered that the metapleural gland is absent or significantly reduced in several ant genera where such reduction had not been previously suspected. In addition we observed a widespread absence of the metapleural gland in males among ant species.
Our survey is far from complete, even at the generic and tribal levels. We think, however, that the pattern revealed by our observations is important enough to warrant a short publication at this time.
*Kurschner apparently was not aware that the paired thorax gland near the petiole she described was the metapleural gland. Manu'nript received bj, the editor March 16, 1984.



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

Psyche [vo~. 91
The metapleural gland is a paired structure. Each side consists of a cluster of glandular cells, and each cell is drained via a duct into a membranous collecting sac that Tulloch et al. (1962) called secretory recess. The collecting sac leads directly into the storage chamber or atrium (receptacle, sensu Tulloch et al. 1962), a sclerotized cavity. Externally the metapleural gland is often marked by a pronounced vault (bulla), and a slit-shaped opening to the outside (Fig. 1). Although the metapleural gland is present in most ant species, it nevertheless varies greatly among them in size and shape (Figs. 2 & 3). Table 1 lists all of the ant species for which we obtained complete series of longitudinal sections through the mesosoma. The speci- mens were fixed in alcoholic Bouin or Carnoy, embedded in methylmethacrylate and sectioned 6 to 8p thick. The staining was Azan (Heidenhain). We attempted to obtain approximations of the number of glandular cells either by counting the cells with clearly visible nuclei or by counting the number of duct openings in successive sections.
The data reveal a considerable variation in the size of the metapleural gland among different species. Even more significantly, our study established that the gland is absent in Oecophylla longinoda and 0. smaragdina (Fig. 4), in all species of Can7ponotus and Pobrhachis sectioned, and in Dendromyrmex chartijex (Tab. 1 ).
We extended this list by an additional survey of the external features that indicate the presence of the metapleural gland, using light-stero-, and scanning electrone microscopy. Of 27 species of Camponotus investigated, only C. gigas showed a slit-shaped opening in the posterior metapleural region (Fig. 5c). In all other Camponotus species the metapleural gland is clearly absent (Fig. 5a, b; Tab. 2). This confirms the suggestion of Ayre and Blum (1971) based on external inspection of Camponotus pennsfianicus workers that this species might not possess a metapleural gland. In none of the species of Polyrhachis investigated did we detect any signs of a metapleural gland (Tab. 2). In addition our study revealed that in several species whose workers and queens have well- developed metapleural glands, the males do not possess this organ; whereas in other species the males have large metapleural glands (Tab. 1).




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

9WHolfdohler & Engef-Siege1 - Met apieural gland 203
Fig, 1.
The metapleural gland of Attu. illustrating the major anatomical feature if this organ. a. SEM micrograph of the rncsosoma of A. rt'phato~~, showing the arge pronounced vault (bulla) which covers the iitorage chamber. Arrow points to he slit-shaped opening (meatus). b. Longitudinal section through the me*>o%oma of 4. .wA-n.s, .showing the large .region of the metaplcural gland (MPG). c. Longi- udinal iiection through meia pleural gland: CS = collecting bat; OC = glandular cells; &4 = meatus; R = storage chamber or receptacle.



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

Psyche [vo~. 91
The metapleural glands have been considered characteristic of all ants with the very few exceptions given by Brown (1968). From a survey of external criteria Brown listed four categories where the metapleural gland appears to be atrophied: "1. Males of army ants, subfamily Dorylinae. 2. Males of a few other genera, mainly in subfamily Myrmicinae (e.g. Leptothorax duloticus, Tetramorium, Strongvlognathus, Rhoptrom.~*rmex, Huberia striata). 3. Workers of the specialized slave makers of genus Polyergus. 4. Queens of certain scattered ant species that are known (or assumed, on grounds of other morphological peculiarities) to be social parasites, i.e., those species which found their colonies in the nests of other ant species". From these findings Brown developed an intriguing hypothesis about the function of the metapleural gland: "the gland produces a substance that, when tasted or smelled, says to another ant colony, especially one of the same species, 'I am an enemy'." According to Brown's hypothesis "an individual either with the same odor-or-taste, or with none at all, would be treated by its host colony as neutral". This would explain why certain species whose individuals have to enter a foreign colony (social parasites; doryline males) often do not possess a metapleural gland. This hypothesis was challenged by Maschwitz (Maschwitz et a1 1970, Maschwitz 1974). He was unable to experimentally demon- strate an enemy identification effect in the metapleural gland secretions, but he could show that in a number of ant species the metapleural gland secretions serve as powerful antiseptic substances that protect the body surface and nest against microorganisms. For example, the active antibiotic component of Atta sexdens was found to be phenylacetic acid, of which one ant stores an average of 1.4 fig (Maschwitz et a1 1970). In Crematogaster (Physocrema) difformis the secretions of the enlarged metapleural gland serve as antiseptics, but when discharged in larger quantities they can also repel animal enemies. Finally, in Crematogaster (Pksocrema) inffata, which also possesses a hypertrophied metapleural gland, Maschwitz (1974) discovered that the sticky secretions function primarily as an alarm- defense substance. He hypothesized that in this case the antiseptic gland has evolved to become an alarm defense gland. Our discovery of the atrophy of the metapleural gland among more genera than previously suspected places this organ in a new



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

19841 Holldohler & En~el-Siege/ - Meiapteural gland 205 Fig. 2. Metapleural gland of a Noiho~rm~ murrup,v worker, a. SEM micrograph of meatus. b. Longitudinal section through the metapleural gland. CS = collecting sac; GC = glandular cells: R = storage chamber.



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

206 Psyche [vol. 91
Fig. 3. (Above and facing page) Longitudinal sections through metapleura glands of workers of a. Mynnvi in piiwuh: b. Crrapm hv^ !uriwri: c. Rhriitloponerc mdlira: d. P'.i'uficm~ I nmx pai/idu+; e. Niivome.v.wr a!hi\eit>.~i: (. M ITII;~VI~C~~.V~~; mnulax. CS = collecting sac; GC = glandular cells; R = storage chamber.



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

I 9~41 HoUdohier & En&-Sied - Mefapieiiral gland 207



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

208 Psyche [vo~. 91
light. Since the gland is absent in a number of ant species known to be extremely aggressive and discriminatory towards conspecific foreigners and interspecific competitors (for Oecoph.~~lla see Holl- dobler and Wilson 1978, Holldobler 1979, 1984; for Camponotus see Carlin and Holldobler 1983) it is obvious that at least in these species the metapleural gland secretions have no function in enemy identification. The absence of the metapleural gland in male ants is also much more widespread than previously assumed. In fact, it appears that species in which males lack this organ or possess it in a very reduced state outnumber those in which the gland is well developed. Most of these males never have to enter a foreign colony in order to mate. Thus Brown's argument concerning the absence of the metapleural gland in doryline males is further weakened. In our view Maschwitz's experimental evidence concerning the antiseptic effect of most metapleural gland secretions is very convincing. We have repeatedly heard the argument that the secretions of other pheromone glands, such as the mandibular gland or poison gland are also acidic and have the potential of suppressing the growth of Escherichia coli in test plates. Thus, it is argued that Maschwitz's tests, although demonstrating an antiseptic effect, do not necessarily prove a primarily antiseptic function of the meta- pleural gland. Maschwitz himself has pointed out that other exocrine glandular secretions frequently have antiseptic power. In fact, he hypothesized that most epidermal glands originally were antiseptic devices before they became more complex glandular structures that produce either repellent secretions against predators or alarm pheromones used in social communication (Maschwitz 1968, 1974; Maschwitz et a1 1970).
Metapleural gland secretions can freely flow out of the storage chamber. The meatus is sometimes densely covered with bristles (Fig. 6), and often there are hairs and dispenser bristles inside the atrium along which the secretion can easily flow to the outer surface (Fig. 6c). As Brown (1968) pointed out, "some ant species have been seen to draw the legs, especially the tibia and tarsi of the forelegs, repeatedly over the meatus of the gland and then rub these leg parts over the rest of the body". In this way the metapleural gland secretion is probably spread over the whole body. It might also be distributed among nestmates by mutual grooming. Thus, it appears



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

19841 Holldohler & Ensel-Siege1 - Metapleurd gland 209 Fig. 4. a. SEM micrograph of the mesosoma of a worker of Oerophytta lon~i- noda, b. Close-up of the posterolateral corners of the alitrunk. Arrows indicate the area where the opening of a metapleural gland should be located.



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

210 Psyche
Fig, 5.
(Above and facing page) SEM micrographs of the mesosoma of workers of several formicine species. Arrow indicates region where the opening of the metapleural gland shou Id be iacaied a. Cat~ipunofui penn.i!-/runic-us: b. Campo- nuius con.ioArinu.i. In both species there are clusters of hairs visible in the area of the metapleural gland opening. Both species, however, lack a slit-shaped opening. c. The iilit-shaped opening of the metapleural giand of Cawpunofus gigns. No opening can be detected in d. Ci>lohci/ni.'i ;rum: e. Dendrompnzex rhanifex;
f. Pukrhachis ( Crrt imyrmu)? dmtdi.




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

19841 Hdldohler & Engei-Siege1 - Meiapleural gland 2 1 I



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

212 Psyche [vo~. 91
that the central location and general structure of the metapleural gland makes it ideally suited for distribution of an antiseptic secretion.
Why then is the metapleural gland absent or strongly atrophied in Oecophvlla, Polvrhachis, Dendrornyrmex, most Camponotus, cer- tain social parasitic ants, and many male ants? Maschwitz et a1 (1970) offered the following explanations for the last two cases. Social parasitic ants, they argued, are usually highly attractive to the host ants, which groom them very frequently, so that the social parasites benefit from the social distribution of the antiseptic secretions of their host ants. This relieves the parasitic species of the burden of producing their own antiseptics and allows them to deploy the freed energy into other organs and functions. The absence of metapleural glands in male ants was given a different adaptive significance. Males live only a relatively short time inside the nest. They are also much less numerous than workers. Therefore, there exists no particular need for them to produce large amounts of antiseptic secretions.
The latter hypothesis, of course, raises the question why in some species the males do have relatively large metapleural glands (Fig. 7f, Tab. 1). The reason could be that in those cases the ratio of males to workers might be much higher and/or the males might reside inside the nest for longer periods and therefore would present a considerable "antiseptic burden" to the colony. This would favor the selection of males capable of producing their own antiseptic secretions. Furthermore, the metapleural gland of male ants could also have another, secondary function, which does not exclude the primary antiseptic function; that is, it could produce sex phero- mones and hence be an important character maintained by sexual selection. During mating females might thus favor males with well developed metapleural glands, and the capacity to produce larger quantities of pheromone.
It is interesting that in all weaver ant species studied the metapleural gland was atrophied. The species we checked included Oecoph~~lla, Polyrhachis (Cwomyrma)? doddi, Dendromymex, Camponotus xenex. It is reasonable to speculate that these arboreal ants are much less exposed to microorganisms than terrestrial ant species, and therefore an antiseptic metapleural gland became unnecessary.




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

Fig. 6. The metapleurai gland of a worker of Irittornvrmex purpureux a. SEM
micrograph of mesosoma. The arrow points to the opening of the metapieural gland. b. Longitudinal section through the melapleural gland c Close-up of a section through the glandular cells (GC) and dispenser bristles (6); D = glandular duct; R = collecting chamber.




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

Psyche
Fig. 7. (Above and facing page)
SEM micrographs oft he exterior structure', of the metapleural glands of several ant species, a. Alitrunk of Putl0111~~rn1a pukhra
worker, b. Close-up of bulla. slit-shaped opening and sensory hairs (?) of the metapleural gland of P. pub-hra worker. c. Alitrunk of Creniafuf;a.'sier sp 10 (ANIC) worker. d. Alitrunk of Ca!aiat us intrudm.\ worker. e. metapleural gland opening with sensory hairs (?) of C intrudem- worker. f, Aliltunk of Crenmogasier
sp 10 (ANIC) male. Arrows point to the opening of ihe metapleural gland.



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




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

Table I.
Evaluation of the histological preparations of the metapleural glands of ants. Species which could not be identified with certainty are listed either by the accession number of our histological collection, or by the species number assigned by the Australian National Insect Collection (ANIC). where voucher specimens have been deposited. metapleural gland present (+),
absent (-), and approximate
number of cells in
subfamily/ species collector and locality worker queen male MYRMEC11NAE
Mvrmecia pilosula B. Holldobler, NSW, Australia + (239) - PONERINA E
Amb&upone ansiratis
Ifrpoponera sp. 3 (ANIC)
B. Holldobler, Canberra.
Australia
J. Traniello, Massachusetts,
USA
B. B. Lowery, Corregidor lsl.
Philippine Rep.
B. Holldobler, Eungella,
Queensland, Australia
B. Holldobler, Eungella,
Queensland, Australia
B. Holldobler, Kuranda,
Queensland, Australia
B. HOlldobler. Shimba Hills,
Kenya
B. Holldobler. Shimba Hills.
Kenya




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

Onvchomyrmex doddi
Onychomyrmex sp. 2 (ANIC)
Onychomyrmex sp. I (AN IC)
Pachyconwa crassa
Pachyron&la obscuricornis
Rhytidoponera merallica
Sphinrtomyrmex sreinheili
DORYLINAE
Eciton hamatum
Neivamyrmex nigrescens
P. J. Greenslade, Solomon
Islands
R. W. Taylor, Mt. Santubang,
Borneo
R. W. Taylor, B. Holldobler.
L. Eacham, Qld, Australia
P. Ward, Lake Eacham, Qld.,
Australia
R. W. Taylor, B. Holldobler,
L. Eacham, Qld., Australia
R. W. Taylor, Mt. Lewis, Qld.,
Australia
B. Holldobler, Shimba Hills,
Kenya
J. Traniello, BC1, Panama
C. P. Haskins. Sydney, NSW,
Australia
B. Holldobler, Canberra,
Australia
R. Silberglied, BCI. Panama
B. Hdldobler, Portal,
Arizona, USA
worker:
+ (120)
soldier:
+ (180)
+ (201)




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

Table 1. (Continued) Evaluation of the histological preparations of the metapleural glands of ants. Species which could not be identified with certainty are listed either by the accession number of our histological collection, or by the species number assigned by the Australian National Insect Collection (ANIC), where voucher specimens have been deposited. metapleural gland present (+),
absent (-), and approximate
number of cells in
subfamily/ species collector and locality worker queen male PSEUDOMYRMECINAE
P.seudomj~rnie.v ferrusinea R. Silberglied, Cerro Gallera, + (65) Panama
P,seudomyrmex pallidus P. Ward, Mustang Is.. Texas. USA + (35) Leptothorax alardycei
M,~*rniira aniericana
J. Traniello, Concord, Mass.,
USA
N. Weber, Guyana
B. Holldobler, Shimba Hills,
Kenya
E. 0. Wilson, Fazenda Esteio,
Brazil,
B. Cole, Keys, Florida. USA
J. Traniello, Massachusetts.
USA
B. Holldobler, Portal, Arizona.
USA
B. Holldobler, Portal, Arizona,
USA
media:
+ (473)
+ ('9




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

Verome'ixor pwguncki
NOTHOMYRMECIINAE
Nothomvrmecia macrops
ANEURETINAE
Atieuretu.~ simoni
DOLlCHODERINAE
Irihivrrt~ie.~ purpurrus
Cononi,vrti~a hicolor
FORMICINAE
Camponofus consohrinu.~
B. Holldobler, Eungella, Qld.,
Australia
E. 0. Wilson. Fazenda Esteio.
Brazil
B. Holldobler, Tampa. Florida,
USA
B. Holldobler, Gerbrunn,
Bavaria, Germany
G. Alpert, Tucson, Arbona, USA
R. W. Taylor, Poochera,
S. Australia
A. Jayasuriya, Sri Lanka
B. Holldobler, Canberra.
Australia,
B. Holldobler, Arizona, USA
B. Holldobler, Canberra.
Australia
B. Holldobler, Ochsenfurt,
Bavaria, Germany
minor:
+ (87)
media:
+ (79)
major:
+ (89)
soldier:
+(31) +(65)