Bert Hölldobler and Hiltrude Engel-Siegel.
Tergal and Sternal Glands in Male Ants (Hymenoptera: Formicidae).
Psyche 89:113-132, 1982.

Full text (searchable PDF, 4980K)
Durable link: http://psyche.entclub.org/89/89-113.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.

TERGAL AND STERNAL GLANDS IN MALE ANTS*
Department of Organismic and Evolutionary Biology, MCZ-Laboratories, Harvard University,
Cambridge, Massachusetts.
Introduction:
Several recent morphological investigations have uncovered a variety of hitherto unknown or neglected exocrine glandular struc- tures in ant workers (Holldobler and Haskins 1977; Holldobler and Engel 1978; Kugler 1978; Jessen et a1 1979; Holldobler et a1 1982; Holldobler 1982; Jessen and Maschwitz in press). The behavioral functions of several of these glands have already been determined (For review see Holldobler 1982).
These studies dealt almost exclusively with ant females and except for the results of Janet's (1902) classical histological investigations of the internal anatomy of males of Myrmica rubra, nothing is known about exocrine glandular structures in the gaster of ant males. Since we consider this information important not only for a further analysis of the behavior of ant males, but especially for our understanding of the evolution of pheromone glands and chemical communcation in ants, we have undertaken a histological study of exocrine glandular structures in ant males. In this paper we present a survey of the abdominal glands not directly associated with the gonads. The purpose of this paper is not to give detailed descriptions of each gland found, but rather to present a comparative account of abdominal glands detected in representative species in the different subfamilies.
Materials and Methods:
For histological investigations live specimens were fixed in alco- holic Bouin or Carnoy (Romeis 1948), embedded in methyl methac- rylate, and sectioned 8p thick with a D-profile steel knife on a Jung Tetrander I microtome (Rathmayer 1962). The staining was Azan (Heidenhain). Especially small objects were embedded in a water soluble plastic (JB-4 embedding kit, Polysciences, Inc., Pennsyl- *Manuscript received by the editor May 1, 1982



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

vania) and sectioned 4-6p thick with glass knives on a rotary - :Ìö tome. In this case the staining was Hematoxylin-Eosin x- strength). The SEM pictures were taken with an AMR lO a( Scanning Electron Microscope. In a few cases only specimens "'v available which had been preserved in 70% ethanol. Results:
The major results are summarized in table I. In the followin s will discuss some of the details of our findings. Penis and subgenital plate glands:
Janet ( 1902) described in males of the myrmicine species My -- ^-3 rubra two major glandular structures directly associated wit_ copulatory apparatus. (1) The first comprise the penis glands, a clusters of glandular cells located inside the penis valves (FÌö -. Each cell sends a duct through a membrane into the lumen fc- XI by the valves (sperm gutters). This gland was also detected in :~~^r'1 of Formica rufa (Clausen 1938) in Conom.~v-ma brunnei and 7, lius sp. (Marcus 1953; cit. in Forbes 1954), in Caniponotus perÌÔ's -^-a vanicus (Forbes 1954), in Neivam.~wnex harrisi (Forbes & Do - 7 Quy 1965) and we found it in representative species of all -: subfamilies of ants. The size of the paired penis gland el- s (which are also called aedeagal gland, Forbes 1954) varies g r e among different species. In some it is a major gland (Fig. R ^) others it is represented only by a few glandular cells, and some i we were unable to identify the opening of the glandular d u c ~ s. The other major gland. associated with the copulatory appara ^Ìà I locaied in the 9th sternite. which together with the coxop c comprise the subgenital plate (Weber 1954). We therefore n .^^a-i these paired clusters of glandular cells "subgenital plate gl- .^^^-I Each glandular cell sends a duct through the intersegnental - brane into the ventral part of the genital chamber (Fig. 1,2> - subgenital plate gland was found in representative species ^cz^j subfamilies studied.
Tergal glands:
In his study of the workers and males of M.rrmica rubra, 2 i (1898. 1902) discovered a pair of clusters of a few glandular under the 6th abdominal tergite. Each cell is drained by a d u c penetrates the intersegmental membrane between the 6th a m 4



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

19821 HMdo bier & Engel-Siege1 - Glands in Male Ants 1 1 5
I I I l l I
IV v V I VII VIII IX
Fig. I
A. Schematic drawing of a longitudinal section through the gaster of a Atorornfuur 8.
3. Longitudina! section through 61h. 7ih. 8th and 9th abdominal segments of a Novoniestor dhheros-UA 8. A=anus; P=parl of penis with pen is gland; PG=pygidial gland; PPG-post-pygidial gland; SPG=subpital plate gland .



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

116 Psyche [Vol, 89
Fig. 2 A. Longitudinal section through pygidial gland of Novomessor ulbiseiosus (3.
B. Longitudinal section through subgenital plate gland of N. albisetosus 3.
CS= cuticu!ar structure: GC=gtandular cells: D0=openings of glandular ducts.



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

19821
Holldobler & Engel-Siege1 - Glands in Male Ants 1 17 abdominal tergites. In recent investigations this gland was found in workers of representative species belonging to all subfamilies, except in the Formicinae. Although the structure and size of the gland varies greatly, its wide distribution led us to conjecture that this gland might be a primitive monophyletic trait in ants generally, perhaps reaching back to the typhioid (or mutilloid) wasp ancestors of ants. In fact, we have recently found first indications that this gland is also present in some living typhiid wasps. Since this gland is anatomically closely associated with the last exposed tergite in female ants (7th abdominal tergite = pygidium) Kugler ( 1978) suggested that it be called the p.\?gidial gland. Of the several tergal glands recently discovered, the pygidial gland appears to be the most frequent in occurrence. Moreover, in several species its secretions have been found to serve as pheromones (Holldobler et a1 1976; Holldobler and Haskins 1977; Maschwitz and Schonegge 1977; Kugler 1979; Holldobler and Traniello 1980 a,b; Traniello and Jayasuriya 198 1). The pygidial gland seems to be homologous with the "anal glands" of the dolichoderine ants described by Pavan and Ronchetti (1955). As we pointed out previously (Holldobler and Engel 1978) the term "anal gland" is misleading, because the gland does not exit from the anal or cloaca1 opening of the gaster, but between the 6th and 7th abdominal tergites. We therefore suggested to refer the dolichoderine structure to the pygidial gland. Recently Jessen and Maschwitz (in press) proposed to name the pygidial gland in honor of its discoverer Charles Janet. Thus we have now three names for this tergal gland: anal gland, pygidial gland and Janet's gland.
Because the anatomical designation of the organ in ant workers (a criterion we prefer) has been used in several recent publications, we will continue to call the tergal gland opening between the 6th and 7th abdominal tergites pygidial gland.
Table 1. (Following pages) List of species that were investigated histologically, and of the types of tergal and sternal glands found. When the histological series was incomplete and we could not make a definite statement, or when we could not clearly identify glandular ducts, the column is marked with a "?". r=with reservoir: c=with cuticular structure.




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

Collector and Locality
Subfamily /.species
B. Holldobler. Brindabella Ranges
Australia
B. Holldobler, Townsville. Qld.,
Australia
J. Traniello. BCI, Panama
J. Traniello, BCI. Panama
B. Holldobler. Kuranda, Qld., Aus -
J. Traniello, BCI, Panama
J. Traniello, BCI. Panama
B. Holldobler, Shimba Hills, Keny-
B. Holldobler, Brindabella Ranges.
Australia
DORYLINAE
Ecit on
A. Aiello, R. Silberglied, BCI. Pan -
A. Aiello, R. Silberglied, BCI, Pan-
PSEUDOMYRMECINAE
P.seudomvrme.x pallic1u.s
P. Ward, Texas. USA
B. Holldobler. Shimba Hills, Keny-
B. Cole Florida Keys, USA
B. Holldobler, Arizona, USA
B. Holldobler. Arizona, USA
B. Holldobler, Eungella, Q u e e n s l a i 4 Australia
B. Holldobler, Arizona. USA
NOTHOMYRMECIINAE
Nothomyrmecia macrops R. W. Taylor, Eyre Peninsula, Aus - ANEURETINAE
Aneuretus .sinmni
DOLICHODERINAE
Iridomvrme.~ purpureus
Anula Jayasuriya, Sri Lanka
B. Holldobler, Canberra, Australia
B. Holldobler, Arizona, USA
FORMICINAE
Formica perpilosa
M~vrinecoq~stus mendax
Oecophvlla longinoda
B.Holldobler, Arizona, USA
B. Holldobler, Arizona, USA
B. Holldobler, Shimba Hills, Keny;




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

Intersegmental tergal glands Intersegmental sternal glands 1x1 VIII! VII!
VII VI
IV/ 1x1 VIIl/
VII
VI/
VI IVI
V1I1 VII VI
v IV Ill VIII VII VI v IV III
r r r r
+ + + + +
r r r r
+ + + +
r r r
+ + + +
r r r




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

Other tergal glands
glandular cells in 7th and 8th segment
ducts open dorsally into genital chamber IXth tergite: ducts open into genital
chamber
postpetiole gland opens between H I rd
tergite and post petiole
postpetiole gland




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

Other sternal glands
VIIlth
Tergo-sternal
glands
Bet ween 4 15; 5 ! 6;
6/7 segments
Between 415; 516:
6' 7 segments
Between 4/5;5/6:
617 segments
glandular cells in petiole: ducts
open ventrally through cuticle
VIIlth >
VIIIth Y
Between 4/ 5: 5/ 6;
6/7: 7/8 segments
Illrd Vlllth IXth
lllrd VIllth IXth
genital
plate
gland
Penis Anus
gland gland




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

122 Psyche [vo~. 89
As mentioned before Janet found this gland not only in workers of M. rubra but also in males. Ant males differ from the workers in having one more exposed segment (8th segment); often even part of the 9th segment is visible. Thus in ant males the pygidial gland does not open at the last exposed tergite (Fig. 1). As can be seen from tab. 1 we found a pygidial gland in species of the subfamilies Myrmeciinae, Ponerinae, Dorylinae, Pseudomyr- mecinae, Myrmicinae, Nothomyrmeciinae and Dolichoderinae. In Aneuretus simoni (Aneuretinae) we detected a few glandular cells, but we could not clearly see glandular ducts. In the males, as in the workers, there exists a considerable variation in the morphology of the pygidial glands, even within a single subfamily. In some species large clusters of glandular cells are associated with a special cuticu- lar structure on the 7th tergite (Fig. 1, 2, 6c). Some species possess more or less developed reservoirs, composed of an invagination of the intersegmental membrane (Fig. 3). In other species there are only a few glandular cells that send dorsolaterally ducts through the intersegmental membrane.
In ant males a post pygidial gland is almost as common as the pygidial gland (Fig. 1). It also consists of paired clusters of glandu- lar cells that open through the intersegmental membrane, but between the 7th and 8th tergites (Tab. 1). This gland is especially well developed in Nothomyrmecia macrops males, where the inter- segmental membrane forms a large reservoir. There it closely resembles the pygidial gland found in workers. Interestingly, the males of this species have only a few glandular cells between the 6th and 7th tergites. Thus in comparison with Nothomyrmecia females (including queens) the major tergal gland in the males is shifted one segment posteriorly. In most other ant species examined, however, the gland between the 6th and 7th tergites (pygidial gland) is the major tergal gland in both workers and males. The doryline males are a remarkable exception. They, too, have large pygidial glands, consisting of paired complex glands and a large reservoir. But the same structure is present in the next 3 seg- ments anteriorly (6th-5th; 5th-4th; 4th-3rd) (Fig. 4,5). In addition, intersegmental glandular cells were found between the 7th and 8th tergites. In the 3rd tergite we also found paired groups of glandular cells, the ducts of which penetrate the sclerotized cuticle of the 3rd tergite dorsolaterally.




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

19821
HMdobler & Engel-Siege1 - Glands in Male Ants 1 23
Fig. 3
Longitudinal section through pygidial gland of Pogonomyrwx harbaru.~ 8. GC=glandular cells: R=reservoir.
Sternal glands:
More than in any other subfamily, the males of the doryline ants are also richly endowed with sternal glands (Tab. I; Fig. 4). We found major complex glands with intersegmental reservoirs between the 3rd and 4th. 4th and 51 h, 5th and 6t h sternites. Glandular clus- ters are also present between the 6th and 7th sternites-but without a pronounced reservoir. In addition, clusters of glandular cells whose ducts penetrate the sclerotized cuticle were found in the 3rd sternite, they are also strongly developed in the 8th and 9th sternites (Fig. 4). Similar sternal glands were found in the 8th sternite of males of Myrmecia pi/osuta, Pachycondyla apiacdis and P. obscuri- cornis. In both Parhycond~~la species the glandular ducts open in bundles into cuticular cups located in the 8th sternite (Fig. 7a). in the ponerine species Lepiogenys dimma the males possess a huge intersegmental sternal gland between the 7th and 8th sternites. This gland consists of large paired clusters of glandualr cells. Each cell sends a duct into wider collecting channels which lead into a paired large reservoir, consisting of ventro-lateral imaginations of the intersegmental membrane (Fig. 6). There is a second paired sternal gland between the 8th and 9th sternite: but this gland is



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

124 Psyche [VOI. 89
Fig. 4
A Schematic drawing of a longitudinal section through the gaster of a N~iwmriwx spec. 3. illu\trat mg ihe -iegmcntal glandular struct ures. 0. Longitudi-
nal -.ect ion through a inicrsegmemal complex-gland (between IVih and Vth tergites) ( - 1 1 . . . . " F
3.
Azanus: GC=glandular ceils: P=parl of penis with penis gland: R= reservoir.



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

19821
Hdldobfer & Engel-Siegel - Glands in Male Ants 1 Z2
Fig. 5
A. SEM micrograph of the tergite with the attached intersegmental m e brane of a Neiva~n\ rtwx &
B. Larger magnification showing clearly the gland u duct openings in the cut ide (GO) and the intersegmental membrane which cons i s tt a mat of hrisfle4ike srruciure>i. This mat-membrane forms the intersegmental gla mi lar reservoir. See also Fig. 4B.




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

1 26 Psyche [VOL 89
much smaller, and no pronounced reservoir could be detected - Lep- togen.jv.s workers have two well developed sternal glands between the 5th and 6th, and the 6th and 7th sternites respectively (Holldobler and Engel 1978; Jessen et a1 1979). Leptogenys males lack these structures but do possess sternal glands in the 7th and 8th sternites. In males of the ponerine ant Palthothyreus tarsatus we found large paired clusters of glandular cells in the 8th sternite. The glandular ducts open through the intersegmental membrane between the 8th and 9th sternites. In addition Paltothyreus males possess unpaired intersegmental sternal glands, similar to those found in Paltoth~ -reus workers (Holldobler and Engel 1978), but smaller, between the 5th and 6th. and 6th and 7th sternites.
Other abdominal glands:
As indicated in table 1 we found several other abdominal glands in males which cannot directly be assigned to the group of tergal or sternal glands. In a few species (Novomessor, Leptogen.17.~) we detected glandular cell clusters in the petiole. In Ecfatomma, Dia- camma, Paltoth.1~eu.s we found small tergo-sternal glands. The ducts of the glandular cells composing them open laterally through the pleural membrane. We found similar small glandular cell bunches in males of Pachycondyla and Rhytidoponera, but we could not clearly identify the glandular ducts. These tergo-sternal glands resemble closely similar structures described by Jessen et a1
(1979) in workers of several ponerine species. Finally we found small groups of glandular cells directly at the anus of males in Pachj*confiIa, Ectatornma, Nei~rmex, Eciton, M.\wnecoc\^stus and Liometopum (Fig. 7b). These anus glands vary considerably in size, and it is possible that they are present in more species than we were able to document. We first found them in workers of Dordus (Holldobler and Engel 1978). The anus glands should not be mis-
taken for the rectal gland, an invagination of a glandular epithelium of the rectum, recently discovered in Oecoph~~~lla workers by H oll- obler and Wilson (1978). It is interesting to note that males of Oecophylla also possess a small rectal gland. Discussion:
Except for the glands associated with the ovipositor and sting apparatus, which the males lack, ant males are as richly endowed with exocrine glands as the females. In many species of ants the males have well developed mandibular glands, pro- and post-




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

9821 ffolldobler & Engel-Siege1 - Glands in Male Ants 1 27
Fig. 6
A. Longitudinal seclion through gaster of a f.epsugetivi diminstta&. show- ingthe rescrvoir(R) of the large sternal gland between 7th and 8th sternites, Aeanus.
B, Section through the large cluster of glandular cells of the sterna! gland, opening into the reservoir between 7th and 8th sternite. S=secretion in reservoir. C. Longi-
tudinal section through the pygidial gland of Lrp!i)~en~i diminma 8. Note that the glandular cells (GC) are considerably iarger than the glandular cells of the sternal gland (Fig 6B). CS=cuticular structure




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

128 Psyche [VOL 89
pharyngeal glands, maxillary glands, salivary glands, and meta- pleural glands, although the size of these various structures can vary considerably between the female castes and males. In this paper we surveyed specifically the abdominal sternal and tergal glands in ant males.
In almost all species studied we encountered two major glandular structures that Janet (1907) had already described in Mvrrnica rubra, the penis glands and the subgenital plate glands. Also quite generally present in males (except in the Formicinae) are the pygi- dial glands. The males share these organs with the females, although less well developed in some species. An interesting case is Notho- m~l~rrnecia: here the males have a rudimentary pygidial gland but a well developed postpygidial gland (between the 7th and 8th tergites). The males of the doryline ants are unusually well endowed with abdominal glands, in which they differ markedly from the workers. Although doryline workers have well developed pygidial- and post- pygidial glands (Holldobler and Engel 1978), the males have mas- sive glandular structures in each segment. In this context the findings by Whelden (1963) are of considerable interest. Whelden described a series of exocrine glands in the gaster of Eciton queens as follows: "Each of the segments of the gaster, including those telescoped together in the posterior part, contains a pair of these glands which are smaller in the anterior segment than those in the following segments". We were not yet able to section a doryline queen and therefore cannot compare the queen organs with those we found in males. It appears, however, that the males possess a glandular equipment very similar to that of the queens. Presumably in doryline queens these massively developed exocrine glands play an important role in the queen's chemical control of the worker ants and in her high attractiveness to workers, (Watkins and Cole 1966). We hypothesize that the males imitate queen pheromones, which might enable them to penetrate a foreign colony in order to get access to the wingless virgin female reproductives (Franks and Holl- dobter unpublished). In fact, this might also be the function of the massively developed sternal gland in Leptogen.1s males. In this genus, as in the dorylines ergatoid reproductive females presumably mate in the nest, so that males flying in from other nests have to penetrate a foreign colony.