Biosystematic revision of Epimyrma kraussei, E. vandeli, and E. foreli (Hymenoptera: Formicidae).
Psyche 93(3-4):253-276, 1986.
This article at Hindawi Publishing: https://doi.org/10.1155/1986/43540
CEC's scan of this article: http://psyche.entclub.org/pdf/93/93-253.pdf, 2040K
This landing page: http://psyche.entclub.org/93/93-253.html
The following unprocessed text is extracted automatically from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.
BIOSYSTEMATIC REVISION OF EPZMYRMA KRAUSSEI. E. VANDELL AND E. FORELZ
The myrmicine genus Epimyrma Emery 1915 presently comprises 11 described species, inhabiting central and southern Europe and North Africa. They all are living as social parasites together with host species of the genus Leptothorax (subgenera Myrafant Smith 1950 and Temnothorax Mayr 1861), some as active slavemakers, e.g. E. ravouxi (Andre 1896) (Winter 1979), others as "degenerate slavemakers" (E. kraussei Emery 19 15 (Buschinger & Winter 1982)), and E. corsica (Emery 1895) as a workerless permanent parasite (Buschinger & Winter 1985).
The taxonomy of the genus is not yet completely consolidated. Thus, in the most recent revision, Kutter (1973) comes to the con- clusion that E. kraussei, E. vandeli Santschi 1927, and E. foreli Menozzi 1921, are so similar that a future comparison of larger series presumably would reveal their synonymy. It is the object of this paper to provide evidence for the accuracy of Kutter's predic- tion. E. kraussei was described by 2 29 and 1 $ (Emery 1915) from Sorgono, Sardegna. Menozzi (1921) established E. foreli on the basis of 4 colonies from the vicinity of Sambiase di Calabria, S'Italy, and E. vandeli was described after 6 colonies collected by A. Vandel near Miramont-de-Quercy and Touffailles, Dept. Tarn-et-Garonne, in S'France (Santschi 1927, Vandel 1927). The most distinctive characters of the 3 species were slightly different shapes of the peti- oli, different grades of coloration from light, yellow-brown in E. foreli to a nearly black in E. vandeli, and the lack of $9 in the latter as opposed to E. foreli and E. kraussei. FB Biologic, Institut fur Zoologie, der Technischen Hochschule, Schnittspahnstr. 3, D-6100 Darmstadt, FRG
2FB 2 (Biologic), Universitat Bremen, Postfach 330440, D-2800 Bremen, FRG Manuscript received by the editor December 23, 1985.
254 Psyche [VOI. 93
During the past years, we have collected E. kraussei from numer- ous localities in the mediterranean area, including the type localities of E. vandeli and E. foreli. We were studying their populations, 59 5 -production in the lab and in the field, their reproductive behav- ior, colony foundation behavior, and karotypes. Crossbreeding of several populations including E. vandeli and E. foreli was possible. All observations pointed towards a synonymy of the 3 species.
Finally, the types were examined, and morphological studies includ- ing the QS of the 3 species were carried out. This considerable body of evidence now clearly demonstrates that E. foreli and E. vandeli
represent but junior synonyms of E. kraussei. MATERIAL COLLECTED AND RANGE OF
EPIM YRMA KRA USSEI
A total of 337 colonies of E. kraussei (including E.v. and E.J) have been collected between 1975 and 1984 (table 1). Populations are numbered for an easier identification in the following text- Fig. 1 may provide a visual impression of the range of E. kraussei,- it also contains a few additional localities from the literature, main1 y those from North Africa (Cagniant 1968). Nests usually are found in cre- vices between flat stones, most easily in old dry walls of terraced vineyards and olive orchards, but also in rocky slopes underneath shrubs (Buschinger & Winter 1983). Colonies are small and can thus be aspirated almost completely. In the type locality of E. vandeli, we did not find the species in the exact sites of Vandel; however, we could collect a sample of 11 colonies near Lauzerte, only 5 k m W of the original site, in the limestone slopes of the Barguelonne valley (table 1, no 5). E. foreli had been found near Sambiase di Calabria, in moss covering the bark of olive trees (Menozzi 1921). We tried in vain to find Leptothoracini in such sites, presumably because the ants have been decimated there by pesticide treatment of the trees. However, in several localities around Sambiase (table 1, no 1 91, we found 22 colonies of a yellowish Epimyrma with Temnothorax hosts, again in terrace walls. We are convinced that they represent members of the same population as that studied by Menozzi. Unfor- tunately, the search for E. kraussei in its type locality, Sorgono in Sardegna, Italy, in April 1985, remained unsuccessful. Even the host species was quite rare in this area. From the map (Fig. 1) we may conclude that both the type localities of E. vandeli and E. ford are situated well within the area of E. kraussei.
19861 Buschinger et a1.- Revision of Epimyrma 255 Table 1.
Localities and numbers of colonies collected of Epimyrma kraussei Emery 1915 (no 5a: Type locality of E. vandeli Santschi 1927, no 19: Type locality of E. foreli Menozzi 1921).
no. locality n colonies
Calpe (Spain, E'coast)
Puig de Pani (NE'Spain)
Selva de Mar (NE'Spain)
Faro de Sarnella (NE'Spain)
Pont de Bar/Seo de Urge1
Tremp/ Tolva (Span. Pyr.)
Ainsa (Span. Pyr.)
Broto (Span. Pyr.)
Chapelle St. Pons (S Bouleterne,
Lauzertel Quercy (S'France)
Cabrespinel Aude (S'France)
La Couronnel Bouches-du-Rhhe
Nyons/ Drome (STrance)
Suze-la-Rousse/ Vaucluse (S'France)
Ste. Maxime, Puget Ville/ Alpes
Venacol Haute Corse (France)
Alassio. Albenga, Ranzo, Toirano
Ventimiglial Prov. Imperia and
Ossuccio/ Lago di Como (N'ltaly)
Biolo/ Valtelino (N'Italy)
Lovere (Lago d'lseo, N'Italy)
Tignale (Lago di Garda, N'Italy)
256 Psyche pol. 93
Table 1, continued.
no. locality n colonies
19821 101 12
15 1974/06/ 15 Salorno (Adige, N71taly) 1 16 198 1 / 091 23-26 Krk (Dalmatia, Yugoslavia) 4 17 1983/09/ 29 Pag (Dalmatia, Yugoslavia) 1 18 1978/ 081 22 Nacionaini park Paklenica (Dalmatia, 1 Yugoslavia)
19 19821 101 03- 10 Gizzeria, Rogliano, near Sambiase 22 (Calabria, SYItaly)
Comparison of the type material of E. kraussei, E. vandeli, and E. foreli with new material
The types of E.v. and E.f. are deposited in the Naturhistorisches Museum Basel, Switzerland. We could study 1Q E. foreli, and 19 E. vandeli, both from the type series. The Museo Civico di Storia Naturale "Giacomo Doria" in Geneva, Italy, has provided us with the types (19, 19) of E. kraussei.
With a close examination of these types we could only confirm the similarity of all 3 "species" as was already stated by Kutter (1973). We therefore refrain from a detailed presentation of mea- surements and structures compared. We also did not find any con- stant differences between the types and specimens from our newly collected material, with respect to size, shape of petioli, head and thorax, length of body hairs etc.; just the coloration was slightly variable between different populations. Thus, the population from Calabria (E. foreli), and one from Spain (pop. no. 3) exhibit a quite light, yellowish brown coloration of Q and 9. Other E.k. popula- tions appear brownish, whereas a dark brown or nearly black is typical for $ E. vundeli (pop. no. 5), for a colony from La Couronne (no. 6), and for population no. 9 from Corsica. Young Q $ are darker in coloration than old queens, and callow Q Q usually exhibit some darker spots in the thorax, and a yellow base of the gaster,
19861 Buschinger et a1.- Revision
of Epimyrma 257
Fig. 1. Distribution of Epimyrma kraussei Emery 1915. @: Our collecting sites listed in Table 1. No. 5: Type locality of E. vandeli Santschi 1927; no. 19: Type locality of E. foreli Menozzi 1921. A and C: Localities of E. vandeli in N'Africa cf. Cagniant (1968), B: Locality of E. kraussei cf. Cagniant (1968); D: Type locality of E. kraussei in Sardegna.
whereas the coloration in old queens is usually uniform. This age- dependent color variation is also typical for E. ravouxi (Andre 1896) (Buschinger 1982).
Male genitalia, wing venation, and shape of $$ petioli We studied wing venation and genitalia of E.k. S$, and the out- lines of the $ and $ petioli of specimens from Tignale and Biolo (Italian Alps), Calpe (Spanish Mediterranean coast), Calabria (S'Italy, E. foreli), and Lauzerte (S'France, E. vandeli). The same characters were investigated in E. ravouxi from several distant pop- ulations [Taubertal: Bavaria (D), Swiss Valley (CH), S'France, Cor- sica (F)], in order to compare their variation within and between the species. E.r. is clearly distinct from E.k. (Buschinger & Winter 1983, Winter and Buschinger 1983), and thus may serve as a reference species. Males preserved in alcohol were dissected, and permanent
Sagittae of Epimyrma m. a: Pop. no. 19 (E. forelf); b: Pop. no. 14 (E. kraussei); c: E. ravouxi from Corsica: d: Pop. no. 5a (E. vandeli). mounds were made of the subgenital plate, the sagittae, and volsel- lae with laciniae, as well as the forewings and antennae. The outlines of $ and $ petioli were drawn and superimposed following a slightly modified method of Wehner (1983). As far as possible we always studied 10 ($99 from each of the populations mentioned above.
Table 2 reveals that the numbers of sagittal teeth (Fig. 2) vary both within E.k. and E.r., but with higher mean values in E.k., including the populations of E. v. and E.J. The volsellae and laciniae (Fig. 3, table 3) exhibit a high confor- mity in E.k. and the two populations of E.v. and E.f, in that the cuspis (tip of lacinia) rarely reaches, and never overlaps the digitus (terminology following Bitsch 1979). In E. ravouxi, on the contrary,
19861 Vuschinger et al.- Revision of Epimyrma 259 Table 2.
Num-erS of sagittal teeth in 38 of Epimyrma kraussei Emery 1915 (= E. vandeli Sam -t=-schi 1927, = E. foreli Menozzi 1921), and of E. ravouxi Andre 1896) from different populations.
n teeth n sagittae
species/ populatior min E max checked
no 14 Tignale 11 13.9 16 19
no 12b Biolo 10 13.6 16 18
no 1 Calpe 10 14.3 19 20
no 19 Calabria (E- f- ) 12 14.4 18 19
no 5a Lauzerte (E. W ) 11 13.7 17 20
Bavaria (D) 8 11.5 15 19
Nyons (F) 10 12.6 15 18
Corsica (F) 7 10.3 13 18
Swiss Valley (CH) 10 12.4 18 2 1
the cuspis usuaHY overlaps or at least reaches the digitus, with very few exceptions -
The subgenit.^al plates did not differ between populations or species.
Male wing venazzon
Wing venation in Epimyrma $9 is quite variable (Andre 1896, cutter 1973). In 8 forewings the radial cell is short and open, the cubital cell long m d usually closed, the discoidal cell may be closed, open, or nearly lacking, and the recurrens can be complete, incom- plete, or absent - deductions of wing venation need not be symmetri- cal in the two forewings of a specimen. We compared mainly the shape of the discoidal cells, which exhibits sizable differences between the species, but varies also within E.k. and E.r. considera- bly (Fig. 4).
Thus, table 4 shows the numbers of wings with open or closed discoidal cell. This character apparently is not appropriate for a differentiation of species or populations. A slightly better distinc- tion is possible with the shape of the discoidal cell (table 4). In E. ravouxi this cell is near to quadratic, with a slightly shorter anterior border. This is also true for a good deal of the N'ltalian and the Spanish populations of E.k., but already in these populations, and more in the Calabrian (no 19, E.f.) and the Lauzerte (no 5, E.v.)
Volsellae and laciniae of Epimyrma m. a: Population no. 14 (E. kraus- sei); b: Pop. no. 19 (E. foreli); c: Pop. no. 5a (E. vandeli); d: E. ravouxi from Corsica. populations the anterior border becomes shorter until the discoidal cell is triangular.
Shape of the petioli in QQ and $9
In several publications (e.g. Menozzi 1931, Sadil 1953) the pro- files of $ and $ petioli and postpetioli were used as the most impor- tant characters for the determination of Epimyrma species. Kutter (1973), however, clearly demonstrated with $9 from a single E. ravouxi colony that these profiles may vary to such an extent that they are useless for species discrimination. Nevertheless, we again studied this character, using a slightly modified method of Wehner (1983). The outlines of the petioli of 10 Q$ and 10 $$ (exceptions: Population 12b: 5 $$, and population 5a: 39 9) per population were drawn with the aid of a Wild M5
19861 Buschinger et al.-Revision of Epimyrrna 26 1 Table 3.
Morphological comparison of the shape of volsella and lacinia in <å´?<- of Epimyrma kraussei Emery 1915 (= E. vandeli Santschi 1927, = E. foreli Menozzi 1921), and of E. ravouxi (Andre 1896) from different populations n volsellae and laciniae where
antrum reaches c. overlaps n a$
species/ population open digitus digitus checked E. kraussei
no 14 Tignale 16 2 - 10
no 12b Biolo 17 1 - 10
no 1 Calpe 19 - - 10
no 19 Calabria (E$) 20 - - 10
no 5a Lauzerte (E.v.) 2 1 1 - 11
Bavaria (D) - - 20 11
Nyons (F) 1 3 12 9
Corsica (F) 2 4 12 10
Swiss Valley (CH) 1 3 17 11
dissecting microscope and a drawing tube at about X 88. The draw- ings then were superimposed in such a way that they all were of the same size and overlapped to a maximal degree (Fig. 5). However, sizes and profiles of the petioli are varying within each population so much that a clear distinction of populations by this character is impossible. Even between E. kraussei and E. ravouxi we could not find any reliable differences in the petiolar outlines. The character, therefore, is useless for taxonomical purposes in the Epimyrma spe- cies investigated, and it can neither support nor contradict a syn- onymization of E.f. and E.v. with E. kraussei. Karyotypes were studied using the air-drying technique of Imai et al. (1977). Usually we made preparations from testes of <$ pupae, and a few from cerebral ganglia of prepupae. E. kraussei from sev- eral populations (pop. no. 1, 5b, 6, 7, 9, 12a) and E. vandeli (pop. no. 5a) were checked, whereas no preparations of E. foreli could be made.
A total of 215 metaphase cells of 16 E. kraussei-6 pupae from 8 colonies of 6 different localities showed 10 chromosomes each (Fig. 6). 6 cells had 9 chromosomes, and 5 cells had the diploid number of
262 Psyche [vol. 93
Fig. 4: Forewing? of &imyrvn~
Left: Variation of wing venation of K
r- from 3 populations; reduction of the subrectangular discoidal all. K Corsica, b: Swiss Valley, c: Taubertal, Bavaria.
Right; Variation of wingvenation -within one E, fcraiwei-population (no. 1, Wpc, Spain). d,e,f: Reduction of the sub- triangular discoidal cell.
2n = 20.9 cells of 2 additional, apparently Q, prepupae contained 20 chromosomes, 2 others had 17 and 15, respectively. Chromosome numbers of less than the haploid (n = 10) or diploid (2n = 20) number are probably due to loss of chromosomes during prepara- tion. Single diploid cells in haploid 8 8 were occasionally found in other species, too (e.g. Hauschteck 1962, 1965). In 2 E. vmdeii Q pupae from a colony from the type locality (pop. 5a), 26 and 15 cells, respectively, were checked. They all had 10 chromosomes each.
The karyotypes of E. kraussei and E vandeli with n = 10 chromo- somes are apparently identical. They consist of 6 small to medium- sized metacentrics, 3 medium-sized submetacentrics and 1 large subtelocentric. Ek. and E.v. share this karotype with all the species of this genus so far studied [E bernardi Espadaler 1982, E. corsica (Emery 1895), E. ravouxi (Andre 1896) and E. stumperi (Kutter
Buschinger et a[.- Revision of Epimyrma
264 Psyche [VOI. 93
1950)l7 and with Myrmoxenus gordiagini Ruszky 1902, a species very closely related to Epimyrma (Buschinger et al. 1983, Fischer unpubl.). No host species of Epimyrma and no other Leptothora- cine species having this particular karyotype could yet be found. Thus7 we may suppose that E. foreli as well has the karyotype of the genus, and no arguments for or against the synonymization of the 3 snecies in auestion can be derived from our karyological studies. Host specificity
The host species of E. kraussei in all populations investigated, including those ascribed to E. vandeli and E. foreli, is invariably Leptothorax ( Temnothorax) recedens (Nylander 1 856). All other Epimyrma species have different host species belonging to the sub- genus Myrafant (Kutter 1973, Espadaler 1982, Buschinger & Winter 19851, and no other Epimyrma species has ever been found with Ternnothorax hosts. In or close to the localities where we have collected E. kraussei (table 1) we usually found several other Lepto- thorax species, particularly often L. (Myrafant) un$asciatus (La- treille l798), which then was parasitized by the slavemaking ants, E. ravouxi or Chalepoxenus sp., but never by E. kraussei. Host speci- ficity, is thus apparently a good character for species discrimination in the genus Epimyrma, and the joint use of Temnothorax by E.v., Ef., and E.k- is an argument for their synonymization. Population Data
Reproductive biology and colony foundation Epimyrma species, as far as is known, may differ considerably with respect to their sex ratios. Thus, E. ravouxi has a sex ratio of about 1.5 (81 9); in E. kraussei from population no. 14 (Tignale) this ratio is about 0.3 in field colonies; and 0.2 in laboratory culture (Winter & Buschinger 19831, and in E. corsica it is 0.08 (Buschinger & Winter 1985). Sex ratios correspond well with the reproductive biology of the species concerned: E. ravouxi is characterized by extranidal mating, whereas E. kraussei (pop. no. 14 Tignale) and E. corsica mate inside the mother nests and thus continually inbreed. The inseminate& dealate !&? of E.k. and E.c. remain in the mother
19861 Buschinger et ale- Revision of Epimyrma 265 Fig. 5:
Shape of petiolus and postpetiolus in Epimyrma 9Qand 99. a,b,c: E. kraussei from populations no. 1 (a), 12b (b), 14 (c); d: no. 19 (E. foreli); e: no. 5a (E. vandeli). Usually the drawings of 10 specimens (b2: 5, eg : 3) were superimposed following the method of Wehner 1983.
Table 5. Production of sexuals and sex ratios in populations of Epimyrma kraussei (= E. foreli, = E. vandeli). Mean values in (1. (a, b): Sexuals found in field colonies in fall; (c, d, e): Sexuals produced in laboratory culture from colonies collected in spring. Colonies with relatively high numbers of host were selected from the field material for laboratory culture. production of sex ratio
population lab/ field n colonies host 99 88 $9 819 b
(a) no 14 Tignale (E.k.1 field 19 371 (19.5) 71 (3.7) 224 (I 1.8) 0.32 %
(b) no 19 Calabria (E.J) field 22 505 (22.9) 24 (1.1) 221 (10) 0.1 1 %
(c) no 14 Tignale (E.k.)
lab 23 1082 (47)
113 (4.9) 537 (23.3)
(d) no 19 Calabria (E$)
lab 12 298 (24.8)
16 (1.3) 140 (I 1.7) 0.1 I
(e) no 5a Lauzerte (E.v.)
86 (17.2) 8 (I .6) 54 (10.8) 0.148
19861 Buschinger et a1.-Revision of Epimyrrna 267 nests over winter, and colony foundation through invading of a host colony occurs in spring. E.r. young queens, on the other hand, begin with colony foundation immediately after swarming, in late summer.
In most of the populations of E.k., E.J and E.v., we found evidence of a reproductive biology identical to that of E.k. pop. no 14 (Tignale), where we first have observed this kind of behavior (Winter & Buschinger 1983). 3 of the 1 1 colonies of E. v., which were collected on 31 March and 1st April, contained young Epimyrma- QQ still engaged with throttling the host colony queens. The E.J- population, on the other hand, was studied in fall, October 3-10, and most of the colonies contained dealate young QQ, a few alate ones, and some 33. Reproductive behavior, thus, is identical in E.k., E.v. and E.J, with intranidal mating and colony foundation in spring. So far as it could be checked, also the production of sexuals and the sex ratios are quite similar (table 51, the sex ratios indicating a generally high Q-bias.
Slave-making ant species are characterized by the presence of a comparatively high number of gg in their nests, apart from incipient colonies. In the genus Epimyrma, we found a considerable variation of g numbers in different species, dependent upon their respective type of parasitism. Thus, E. ravouxi, an active slave- maker, has up to 77 E.-~Q (mean 24.9) in a nest, whereas the "degenerate slavemaker", E. kraussei, had an average of only 3.5 and a maximum of 10 E--9 $! (Buschinger & Winter 1983). E. corsica (Emery 1895) has lost the 9 -caste completely (Buschinger & Winter 1985). E, vandeli was originally said to be workerless, whereas $?g had been described of E. kraussei and E. foreli. We therefore censused the E.-9 9 in most of our field-collected colonies, and also the $!-production of a representative number of colonies in laboratory culture.
In table 6 we compare the Epimyrma 9-numbers of 4 larger populations including 2 ascribed to E. kraussei (no 14 and 2a), and the populations no 5a (E.v.) and no 19 (E-J), and of 5 local populations of E. kraussei from the Spanish Pyrenees with nests always found in close vicinity.
Fig. 6: Karyotypes of a: Epimyrma kraussei from pop. no. 5b (Aude, S'France) and b: Epimyrma ravouxi from pop. no. 12a (Lago di Como, N'ltaly). Most striking is the fact, that our 11 field colonies of E. vandeli did not contain any E.-$$. This corresponds to the original de- scription of 6 colonies without $9 (Vandel 1927). In laboratory culture, however, we obtained a few $$ from colonies of this population (see below).
Workerlessness is also found in a certain amount of colonies in most populations of E. kraussei. In part, this is due to the fact the newly founded colonies do not yet contain E.-9 $, and most of our collecting was done in spring during the time of colony foundation. Therefore, it is not surprising that the population of E. foreli is the only one where all colonies contained at least one E. $ : The sample was entirely collected in the fall. On the contrary, our material from the type locality of E. vandeli was collected in spring, and in 3 of the 11 colonies the E.-Q was still engaged in throttling the Temno- thorax queen. A few more colonies may as well have been incipient ones, where the host queen had already been eliminated. Further- more, 3 colonies in the laboratory produced unusually high amounts of E.-SQ, and when dissected, the queens proved to be poorly inseminated, having very few sperm cells in their receptacula. The lack of E.-$9 in our sample is thus at least in part explained by these facts.
The highly variable average and median values of $-numbers as well as the maximum values in other populations are also very remarkable. In some populations, like that of E. foreli, but also at
Numbers of ~~irn~rrna-w in field populations of E. kraussei Emery 19 15 (= E. vandeli, = E. foreli). %
n colonies n Epimyrma 5 $ /colony % colonies a-
population censused mean median range lacking E.-Q$ 2 no 14 Tignale (E.k.) 82 2.07 1.5 0-10 46.3 spring & fall 2 no 2a Banyuls (E. k.) 36 5.33 3.5 0-20 11.1 spring 5 I no 19 Calabria (E$)
22 6.64 5 1-26
0.0 fall 1
no 5a Lauzerte (E.v.)
0 0 100.0
no 3a Pont de Bar (E.k.) 7 1 .O 0 0- 7 85.7 spring S' no 2d Faro de Sarnella (E.k.) 16 2.3 1 1 0- 9 56.2 spring 5'
no 2c Selva de Mar (E.k.)
5 0-15 20.0 spring
no 3d Broto (E.k.)
4 0-16 20.0 spring
no 3b Tremp/ Tolva (E.k.) 6 10.5 10.5 0-24 16.6 spring ^
270 Psyche [Vol. 93
Banyuls (no 2a) and along the Spanish Pyrenees (no 2c, 3b, 3d), we found a few colonies with 15 to more than 20 E.-$$, which would be sufficient for an effective slave-raiding. As was suggested for E. kraussei from Tignale (Buschinger & Winter 1983), however, we believe that slave-raids do occur only exceptionally, if at all, in the other populations now studied: Most colonies comprise but very few E.-$9, and colonies with higher E.-$-numbers on average do not contain more host species workers than those with few or no E.-$ $. From table 6 we may conclude that Epimyrma kraussei has established numerous local populations in which the reduction of $-numbers has occurred to highly variable degrees. The population ascribed to E. vandeli then would be close to one end of the scale which is complete loss of the $-caste like in E. corsica (Buschinger & Winter 1985), and E. foreli is among the populations with highest E.-$ -numbers. It must be stated, however, that a geographical vari- ation of $-numbers, e.g., in the sense of a cline, is lacking: Popula- tions with low $-numbers have been found in S'France (no 5a, E. vandeli) and in N'Spain (no 3a), and high $-numbers occur close to the latter locality (no 3b) as well as in S'Italy (no 19, E. foreli), In laboratory culture the $-production of Epimyrma colonies roughly corresponds to the field data. Table 7 provides a compari- sou of $-production in colonies from 3 populations. Most impor- tant is the fact that E.-$$ appeared in 2 of the 5 laboratory-kept colonies from population no 5a (Lauzerte, E. vandeli). Worker numbers, thus, are not contradictory to a synonymiza- tion of E.v. and E.f. with E. kraussei.
Intranidal mating is an excellent condition for experimental crossbreeding of sexuals from different populations and even spe- cies. Colonies are kept in nearly natural annual cycles with a long hibernation of about 6 months at 10å¡ a "spring" and "fall" phase in daily temperature rhythms of 10å¡ (12h, dark) and 20å¡ (12h, light) for 2 weeks each, and a summer phase of 15OC (lOh, dark) and 25OC (14h, light) for 2 weeks, followed by 2 months of 17OC (lOh, dark) and 28' C (14h, light), and again 2 weeks of 15' C/ Zå¡ when pupation decreases. For details of formicaries, feeding etc. see Buschinger (1974). All $ pupae from colonies of 2 populations or species are exchanged. Further 8 pupae arising newly from the
19861 Buschinger et al. - Revision of Epimyrma 27 1 Table 7.
Worker-production in colonies of Epimyrma kraussei Emery 19 15 from 3 populations, in the first summer after collecting (including the populations of E. foreli Menozzi 1921 and E. vandeli Santschi 1927). - --
Epimyrma $ $2 produced
population n colonies total mean median range no 14 Tignale (E.k.) 23 12
0.52 0 0- 2
no 19 Calabria (E.f.)
2.0 1 0-10
no 5a Lauzerte (E.v.)
5 5 1 .O
0 0- 4
remaining brood are either removed or exchanged. Usually the foreign pupae are easily accepted, and also the sexuals hatching from them. After dealation of the young QQ a few of them are dissected for control of insemination. In the following spring the QQ leave the nest chambers and can be placed with host colonies, where they found their own colonies. The first sexual offspring usually develops from rapid brood in the year of colony foundation (Winter & Buschinger 1983).
It must be said, however, that the rate of successful colony foun- dations is generally low, both with cross-mated QQ and those having normally mated with brothers, Quite often this is due to insufficient insemination, and perhaps to not yet optimal laboratory conditions. We therefore present only a preliminary survey of successful cross- breedings (table 8) without giving data on numbers of replicates or numbers of offspring produced. These experiments are being continued.
Table 8 clearly reveals that crossbreeding between different E. k. populations, and also between E.k. and E.v. or E.$, is possible. This result, however, can only weakly support our supposition of the synonymy of the 3 species, since we also succeeded in crossbreeding E.k. with E. corsica, and with E. bernardi, both of which are morphologically and biologically distinct good and species. The meaning of the morphological and biological characters stud- ied in E. kraussei, E. vandeli and E. foreli, has been discussed with reference to the question of synonymy of the 3 species already in the respective sections. We found no morphological characters which would allow a clear distinction between them. The karyotype is apparently homologous in all Epimyrma species. The 3 species
272 Psyche [vo~. 93
investigated have a common host species, Leptothorax (T.) reced- ens, which is not parasitized by any other Epimyrma species. The numbers of ~pimyrma-Q $ are variable, but low in all the 3 species, which therefore should represent "degenerate slavemakers" as was already stated for E. kraussei (Buschinger & Winter 1983). Field data and laboratory breeding results indicate that the 3 species have a highly Q-biased sexual production, intranidal mating and in- breeding, that the young $$ overwinter in their mother nests and invade own host colonies in spring. Crossbreeding experiments reveal that a strict genetical isolation is lacking. The 3 original sam- ples, comprising only few specimens, were apparently described as separate species mainly because they were found in quite distant localities, and because the variability of their slight morphological differences could not be evaluated then. We therefore synonymize E. vandeli Santschi 1927 and E. foreli Menozzi 1921 with E. kraussei Emery 19 15. Population structure and reproductive biology in this species, however, are highly remarkable (Winter & Buschinger 1983). The inbreeding system with young queens spreading on foot, and thus over only short distances, must result in an extremely restricted gene flow, even if a rare mating of sexuals from neighboring colonies might occur. The populations from different continents (northern Africa, southern Europe) and islands (Sardegna, Corsica), but also from more neighboring localities (southern France, northern Spain), must have been isolated for a very long time. This isolation, in our opinion, is responsible for the differences in coloration, morphology of wings and genitalia, and worker numbers, which we observed in certain populations. The replacement of one of these characters by another one can only occur through interdemic selection, through supplantation of a local population by another one which is somewhat more effective. Since E. kraussei, however, does not inhabit large, continuous habitats, but instead forms numerous small, patchily distributed populations, this process must be slow and rare. The reduction of worker numbers in favor of a higher $ production should be highly adaptive in this species. Since, however, the genetical basis for this evolution cannot spread, e.g., through flying m, we may speculate that different demes just have reached different degrees of worker reduction. Crossbreeding ex- periments have been started in order to find out whether or not
Buschinger et al.- Revision of Epimyrma
274 Psyche pol. 93
worker number in E. kraussei populations is genetically determined. If so, we may predict that somewhere in the range of E. kraussei, populations will be found with high $-numbers, and still actively slave-raiding, and other perhaps truly workerless demes. The evolution from outbreeding and slave-raiding towards intranidal mating and reduction of worker numbers and slave-making behavior, is an apparently widespread trait in the genus Epimyrma. Intranidal mating has been found also in E. bernardi and in E. corsica, two species which are morphologically clearly separated from E. kraussei. Whereas E. bernardi "still" produces a consider- able amount of $$, E. corsica has lost this caste completely (Buschinger & Winter 1985). Future studies will be necessary to find out whether worker reduction in Epimyrma is developing in several species or species groups independently, in parallel evolution, or whether the species with different worker numbers form a series of descent. The present study of E. kraussei evidently favors the first alternative.
Epimyrma vandeli Santschi 1927 and E. foreli Menozzi 192 1 are junior synonyms of E. kraussei Emery 1915. A comparison was made of the type specimens and of newly collected material from the type localities of E.v. and E.J, and from numerous populations of E. k.. No reliable morphological differences could be found, despite a certain variation in $ genitalia, wing venation and body colora- tion of different populations. Karyotypes are homologous in all Epimyrma species and populations yet studied. The host species is Leptothorax ( Temnothorax) recedens (Nylander 1856) in all E. k. populations including E.v. and E.J, whereas all other Epimyrma species have different host species. Epimyrma $-numbers vary between populations, E.v. having a particularly low, and E.f. quite a high one, both, however, remaining within the range of the other E.k. populations. Sexual production is similar in all populations with a remarkably low (^-production. In all populations studied, sexuals mate within the mother nests, and inseminated, dealate young QQ remain there over winter until they leave for colony foun- dation in spring. E.v. and E.J could be successfully crossbred with E. k., and sexuals from several E. k. populations among each other. Differences between E. k. populations presumably are due to their
19861 Buschinger et a[.-Revision of Epimyrma 275 quasi-clonal structure with very restricted or lacking gene flow between colonies and demes.
This work was supported by grants of the Deutsche Forschungs- gemeinschaft.
ANDR~, E. 1896:
Description d'une nouvelle fourmi de France. - Bull. Soc. Ent France 2: 367-368.
BITSCH, J. 1979:
Morphologie abdominale des Insectes. In: Grasse, P.-P., Trait& de Zoologic, t. VIII, fasc. 11, 290-578; Masson, Paris. BUSCHINGER, A. 1974: Experimente und Beobachtungen zur Griindung und Ent- wicklung neuer Sozietaten der sklavenhaltenden Ameise Harpagoxenus sublae- vis (Nyl.). - Ins. soc. 21: 381-406.
BUSCHINGER, A. 1982: Epimyrma goesswaldi Menozzi 193 1 = Epimyrma ravouxi (Andre 1896) - Morphologischer und biologischer Nachweis der Synonymic (Hym., Formicidae). - Zool. Anz. 208: 352-358. BUSCHINGER, A. & WINTER, U. 1983:
Population studies of the dulotic ant, Epi- myrma ravouxi, and the degenerate slavemaker, E. kraussei (Hymenoptera: Formicidae). - Entomol. Gener. 8: 25 1-266. BUSCHINGER, A. & WINTER, U.
1985: Life history and male morphology of the workerless parasitic ant Epimyrma corsica (Hymenoptera: Formicidae). - Entomol. Gener. 10: 65-75.
BUSCHINGER, A., WINTER, U. & FABER, W.
1983: The biology of Myrmoxenus
gordiagini Ruzsky, a slave-making ant (Hymenoptera, Formicidae). - Psyche 90: 335-342.
CAGNIANT, H. 1968:
Du nouveau sur la repartition des Epimyrma d'Alg6rie (Hymenoptkres - Formicidae - Myrmicinae). - Bull. Soc. Hist. Nat. Toulouse 104: 427-429.
EMERY, C. 1915: Contribute alla conoscenza delle formiche delle isole italiane. -Ann. Mus. Civ. Stor. Nat. Geneva, Ser. 3a, 46: 244-270. ESPADALER, X. 1983:
Epimyrma bernardi nsp., a new parasitic ant. - Spixiana 5: 1-6.
HAUSCHTECK, E. 1962: Die Chromosomen einiger in der Schweiz vorkommender Ameisenarten. - Vierteljahresschr. Naturf. Ges. Zurich 107: 2 13-220. HAUSCHTECK, E. 1965: Halbe haploide Chromosomenzahl im Hoden von Myr- mica sulcinodis Nyl. (Formicidae). - Experientia 21: 323-325. IMAI, H. T., CROZIER, R. H. & TAYLOR, R. W. 1977: Karyotype evolution in Australian ants. - Chromosoma 59: 341-393. KUTTER, H. 1950. uber zwei neue Ameisen. - Mitt. Schweiz. Ent. Ges. 23: 337-346.
KUTTER, H. 1973: Beitrag zur Losung taxonomischer Probleme in der Gattung Epimyrma (Hymenoptera Formicidae). Mitt. Schweiz. Ent. Ges. 46: 28 1-289.
276 Psyche [VOI. 93
1921: Formiche dei dintorni di Sambiase di Calabria. - Boll. Lab. Zool. gener. Agr. Portici 15: 24-32.
MENOZZI, C. 1931: Revisione del genere Epimyrma Em. (Hymen. Formicidae) e descrizione di una specie inedita di questo genere. - Mem. Soc. Ent. Ital. 10: 36-53.
1953: Epimyrma zaleskyi nov. spec. (Hym., Formicoidea). - Casop. Cesk. Spol. ent. 50: 188-196.
SANTSCHI, F. 1927: Notes myrmecologiques. I. Sur quelques nouvelles fourmis de France. - Bull. Soc. Ent. France 1927: 126-127. VANDEL, A. 1927: Observations sur les moeurs d'une Fourmi parasite: Epimyrma vandeli Santschi. - Bull. Soc. Ent. France 1927: 289-295. WEHNER, R. 1983: Taxonomie, Funktionsmorphologie und Zoogeographie der saharischen Wiistenameise Cataglyphis fortis (Fore1 1902) stat. nov. (Insects: Hymenoptera: Formicidae). - Senckenbergiana biol. 64: 89-132. WINTER, U. 1979: Epimyrma goesswaldi Menozzi, eine sklavenhaltende Ameise. -Naturwissenschaften 66: 58 1.
WINTER, U. & BUSCHINGER, A.
1983: The reproductive biology of a slavemaker ant, Epimyrma ravouxi, and a degenerate slavemaker, E. kraussei (Hymenop- tera: Formicidae). - Entomol. Gener. 9: 1 - 15.
Volume 93 table of contents