Hereditas 92: 321-325 (1980)

Chromosome numbers in the with special reference to the taxonomical position of Formica uralensis Ruzsk. and Formica truncorum Fabr.

MARITA ROSENGREN’, RAINER ROSENGRENz and VERONICA SODERLUNDI

Department of Genetics’ and Department of Zoologyz, University of Helsinki,

ROSENGREN, M.,ROSENGREN, R. and SODERLUND, V. 1980. Chromosome numbers in the genus Formica with special reference to the taxonomical position of Formica uralensis Ruzsk. and Formica truncorum Fabr. - Hereditas 92: 321-325. Lund, . ISSN 0018-0661. Received January 25, 1980 The chromosome numbers of the Formica aquilonia Yarr., Formica uralensis Ruzsk. and Formica pressilabris Nyl. were all found to be n=26. The chromosome number of F. uralensis thus does not accord with the view that this species belongs to be subgenus Serviformica. Chromosomal polymor- phism was found in Formica truncorum Fabr.. one population having the haploid number n=28, instead of n=26. The chromosome numbers obtained for South Finnish populations of Formica polyctena Foerst.. Formica pratensis Retz., Formica lugubris Zett., Formica rufibarbis Fabr. and Formica transkaucasica Nas. were the same as those reported from Central Europe. Marita Rosengren. Department of Genetics. University of Helsinki, P. Raurariekatu 13, SF-00100 Helsinki 10, Finland

The systematics of the genus Formica is still un- between n=3 and n=84, with a genus frequency clear, especially with respect to Formica sensu peak at 14 (CROZIER1975). A high degree of vari- strict0 (= the “rufa group” as defined by BETREM ability in chromosome number may occur within a 1960). Not only are the morphological criteria single genus (e.g. the genus Camponotus with ha- commonly used in delimiting species within the ploid numbers varying between 9 and 26), but subgenus Formica rather vague but the same spe- within the genus Formica the chromosome num- cies may be included in the subgenus or not, or bers appear to be very uniform (CROZIER1975). and arbitrarily given the status of a“true” member of the information given by it is, consequently, of the rufa group or not. A case in point is the limited value, unless complemented by studies of mound-building species Formica uralensis Ruzsk. the karyotype. Unfortunately, karyotype analyses This species is usually considered a member of the are very difficult in Formica due to the extreme subgenus Formica s.str. or the rufa group (BETREMsmallness of the chromosomes. Our intention was 1960; G~SSWALDet al. 1965; KUI-TER1977), but to use banding techniques, but the methods we DLUSSKY(1967, and personal communication) in- tried did not give satisfactory results. cludes it in the subgenus Serviformica (the “fusca The chromosome numbers of the species For- group”). Another unclear case is represented by mica aquilonia Yarr., Formica uralensis Ruzsk. Formica truncorum Fabr., which is included in the and Formica pressilabris Nyl. are reported here rufa group by BETREM(1960) and KUTTER(l977), together with that of a possible new Formica spe- but not by COLLINGWOOD(1979). Various problems cies described by COLLINGWOOD(1979, p. 152) of the rufa-group can be tackled suc- under the name Formica “nylanderi” . Chromo- cessfully by unconventional methods: gel electro- somal polymorphism occurs in many species phoresis of enzymes (PAMILOet al. 1979), analysis (CROZIER1975; IMAIet al. 1977), but has so far not of volatile compounds (BERGSTR~Mand UFQVISTbeen observed in the genus Formica. We have 1973), application of numerical taxonomy (DOUWES now found chromosomal polymorphism in Formi- 1979) and use of behavioural critera (ROSENGRENca truncorum Fabr. In other cases, however, and CHERIX1980). Chromosomal studies should (Formica polyctena Foerst., Formica pratensis naturally have a central position in this context. Retz., Formica lugubris Zett., and Formica trans- The haploid chromosome number of varies kaucasica Nas.) the chromosome counts made by 322 M. ROSENGREN ET AL. Hereditas 92 (1980)

us on Finnish populations are in ,-lose agreement Table I. Each row in the table corresponds to a nest. Column “No. nuclei’’ = the total number of haploid metaphase nuclei in with the chromosome numbers previously report- which chromosomes were counted (the number within paren- ed for European popu1ationsOf the Same thesis is the number of nuclei with the most common chromo- species by HAUSCHTECK-JUNGENand JUNGEN (1976). some number). Column “n” = the most common chromosome number

Species and nest Locality No. nuclei n

Material and methods F. (Formica) polyctena Grankulla 68(65) 26 All the chromosome numbers reported in this F. (Formica) aquilonia Esbo 11(11) 26 F. (Formica) aquilonia Vanda 37(37) 26 paper were counted on haploid cells of the testes, F. (Formica) lugubris Vanda 22(22) 26 although the brains of workers and males were F. (Formica) “nylanderi” Sibbo 28(28) 26 also studied in many of the species. Mitotic divi- F. (Formica) pratensis Grankulla 37(34) 26 sions in brain cells were, as a rule, found in pre- F. (Formica) truncoruni Ing% 1q7) 26 F. (Formica) truncorum Hitis 75(65) 28 pupae, but not in pupae, as was also observed by F. (Formica?) uralensis Vanda 102(86) 26 IMAI(1966) while meiotic figures in the testes were F. (Serviformica)rujibarbis Lappvik 1q5) 27 observed mainly in early pupae (the eyes unpig- F. (Serviformica)rranskaucasica Siikajarvi 14(13) 26 mented or only weakly pigmented). In addition to F. (Coproformica) pressilabris Sjundel 17(13) 26 cells with haploid numbers the testes contained some diploid and polyploid nuclei (e.g. triploid). The material was dissected in colchicine-hypo- haploid number n=26 (Table 1). The material ofF. tonic solution and in most cases incubated for rufibarbis was too small and the chromosome about 20 min in the same solution (see IMAIet al. number too variable (n=25-28) to permit any de- 1977 for procedure). We used exclusively the finite conclusion, but n=27 was the most common haploid number for this species and corresponds air-drying technique described by IMAIet al. (1977). The preparations were checked in the to the diploid number 2n=54 previously reported phase contrast microscope and stained in Giemsa for F. rujibarbis and five other European Servi- solution diluted 1:24 with Sorensen’s pH 6.8 buf- formica species by HAUSCHTECK-JUNGENand JUNG- fer. Mounts were prepared with cover slips and EN (1976). The number 27 seems to be the rule for the haploid set, having been reported for Nearctic neutral mounting medium (Gurr). According to and Asian Servifwmica as well (HUNGand IMAI, IMAIet al. (1977) this technique should give C-banding without subsequent treatment, but we cited in CROZIER1975; cf. FRANCOEUR1973). The were unable to obtain satisfactory C-banding, only known exception is F. transkaucasica (syno- though replacement of incubation in colchicine so- nyme: F. picea Nyl.), for which HAUSCHTECK- lution with incubation in pure water appeared to JUNGENand JUNGEN(1976) obtained 2n=52 in give slightly better results. Attempts to induce (brain tissue from workers). This ac- G-or C-banding with trypsin or barium hydroxide cords with the haploid number (n=26, testis) also failed. found by us for this species in Finland. Our material of Formica species was collected DLUSSKY(1967) considers F. uralensis a member along the coast of Southern Finland from Sibbo of the subgenus Serviformica, but BETREM(1960) east of Helsinki to the Hitis archipelago west of and KUTTER(1977) refer it to the subgenus Formi- ca. The view of DLUSSKY(1967, and personal Hango (for locality of each of the nests, see Table I.). The species were determined with the aid of communication) is based on the fact that the widely used identification guides (e.g. DLUSSKY mandibles of uralerisis males are of the same sup- and PISARSKI1971; KUTTER1977). In addition, posedly primitive dentate type as in some Servi- workers from each nest were investigated in the formica. Other features of uralensis diagnostic of Servifortnica are the complete lack of outstanding scanning electron microscope (ISM-U3) at 13 hairs on the eyes of all the castes, including the KW, after coating with gold. males, the dullness of the frontal triangle on the head and possibly the narrowness of the para- meres of the male genitalia (DLUSSKY,personal communication; see also KUTTER1977). In the Results and discussion Finnish DoDulations of F. uraletisis the male 1. All Formica species investigated, except F. rufi- mandibles are indeed of a markedly dentate type barbis and the Hitis nest of F. truncorum had the compared with those in Formica s.str. (in which Hereditas 92 (I980) CHROMOSOME NUMBERS IN THE GENUS FORMICA 323

Fig. I a-d Haploid chromosome sets of a: F. “nylanderi” n=26; b: F. rruncorum (Hitis) n=28; c. F. uralensis n=26; d: F. aquilonia n=26. Bar: IOpm. all the teeth on the mandibles, except the apical HAUSCHTECK-JUNGENand JUNGEN (1976) studied one, are either absent or reduced to one or two five species of the subgenus Formica (rufa, polyc- irregular knots) and the workers are differentiated tena, lugubris, truncorum and pratensis) and by the exceptionally strong sculpturing of the found 2n=52 in all of them (brain tissue of work- frontal triangle (Fig. 2). Arguments for not includ- ers). We have confirmed this result with testis ing F. uralensis in Serviformica are the facultative preparations of polyctena lugubris and pratensis social parasitism of uralensis females during nest (n=26, Table 1) and in addition demonstrated that founding (KUWER1977), the structure and shape of this number also holds for F. aquilonia. The same the nest (mounds of needles as in the subgenus chromosome number is found in the F. “nylande- Formica) and the general type of foraging behav- ri” population, which, although showing affinities iour (ROSENGREN1%9, 1971). The chromosome to F. lugubris in its worker morphology, has high- number n=26 found in the present study also dif- ly deviant queens and can possibly be separated as fers from the common pattern in Serviformica. a new species (COLLINGWOOD1979). There may thus be reason to consider that Formi- Some of the studied populations, e.g. F. “ny- ca uralensis represents a separate phylogenetic landeri” and F. aquilonia had a satellite chromo- lineage, although its enzymes provide evidence of some (see Fig. 1 .). some relationship with Serviformica (PAMILOet al. The number n=28 found for F. lruncorum 1979). (Table 1, Fig. I) is at variance both with the F. pressilabris Nyl. has the chromosome num- chromosome number of the other, more eastern F. ber n=26 (Table I). This accords with results for truncorum population investigated by us in Fin- F. execfa reported from Switzerland (brains of land and with the chromosome number (2n=52) workers, HAUSCHTECK-JUNGENand JUNGEN 1976). reported for a Swiss population of this species by The same chromosome number is found also in HAUSCHTECK-JUNGENand JUNGEN (1976). We ob- Nearctic Coptoformica (HUNG1969). tained the same result with male pupae (testes) 324 M. ROSENGREN ET AL. Hereditas 92 (1980)

Fig. 2 a4Frontal triangle and adjacent pans on the heads of workers (SEMI. a: F. lruncorum (n=26); b: F. lruncorurn (n=28); c: F uralmsis; d. F. aquilonia. Bar: 100pm.

from this nest in both 1978 and 1979 (results from subgenus Formica, except Formica yessetisis For. the two years pooled in Table I), but the material (specimens sent to us by Dr. Seigo Higashi from from 1978 contained one pupa with n=27 instead Sapporo, Japan). It is thus evident that F. trunco- of n=28. Morphologically, the ants of this diver- ruin has a very special position within the sub- gent population do not differ clearly from the genus Formica, a fact also indicated by other fea- normal (n=26) truncoruni population. This was tures (KUTTER1977; PAMILOet at. 1979). Our confirmed by our scanning electron micrographs. chromosomal findings can hardly be correlated In both the Finnish truncoruiii populations the with the taxonomic position of truricoruni, unless workers had a very characteristic chitin micro- n=28 is the rule in this species (which seems un- structure on the frontal part of the head (Fig. 2). likely at present). Chromosomal polymorphism According to our observations this type of micro- has been observed in several ant species, but has structure is not found in any other species of the not previously been reported for the genus Formi- Hereditas 92 (1980) CHROMOSOME NUMBERS IN THE GENUS FORMICA 325 ca, and the only chromosome numbers known DLUSSKY, G. M.and PISARSKI, B. 1971. Reqizja polskich ga- ealier for Formica species are n=26 and 27. tunk6w mfowek (: Formicidae) z rodzaju For- mica L. - Fragm. Faun. 16: 145-224 Due to the smallness of the chromosomes and DOUWES, P. 1979. Formica rufa-gruppens systematik. - Ento- our failure to obtain a banding pattern, we were mol. Tidskr. 100: 187-191 not able to determine what kind of chromosomal FRANCOEUR, A. 1973. Revision taxonomique des especes Ne- rearrangement had occurred to change the number arctiques du groups fusca, genre Formica (Formicidae, Hymenoptera). -MPmoirs Entomol. SOC.QuPbec 3: 1-316 from 26 to 28. Nor do we know whether only one GOSSWALD, K., KNEITZ, G. and SCHIRMER, G. 1965. Die or a few nests of F. truncorum are characterized geographische Verbreitung der hiigelbauenden Formica- Arten by n=28, or whether we have to do with a karyo- (Hym., Formicidae) in Europa. - Zool. Jahrb., Syst. 92: type “race” extending over a substantial part of 369-404 HAUSCHTECK-JUNGEN,E. and JUNGEN, H. 1976. Ant the archipelago of the Gulf of Finland. chromosomes I. The genus Formica. - Insecres Sociaux 23: For the future, it seems important to develop 5 13-524 banding methods suitable for the genus Formica, HUNG, A. C. F. 1%9. The chromosome numbers of six species so that the chromosomes can be used more effec- of Formicine ants. -Ann. Ent. SOC.Am. 62: 455-456 IMAI. H. T. 1966. The chromosome observation techniques of tively to investigate the taxonomic relations. ants and the chromosomes of and Myrmicinae. - Acta Hymenopterologica 2: 119-131 Acknowledgments.- We wish to thank Pekka Pamilo, M.Sc., IMAI, H. T., CROZIER, R. H. and TAYLOR, R. W. 1977. Karyo- for valuable informations and for providing us with some ant type evolution in Australian ants. - Chromosoma 59: 341-393 pupae, and Anna Damstriim, M.A., for checking the language. KUTTER, 1977. Hymenoptera Formicidae. - Insecta Helvetica (Fauna) 6: 1-298 PAMILO, P., VEPSALAINEN. K., ROSENGREN, R.. VAR- Literature cited VIO-AHO, S.-L. and PISARSKI, B. 1979. Population genetics of Formica ants II. Genic differentiation between species. - BERGSTR~M,G. and MFQVIST,I. 1973. Chemical congruence Ann. Entomol. Fenn. 45: 65-76 of the complex odoriferous secretions from Dufour’s gland in ROSENGREN, R. 1%9. Notes regarding the growth of a polycalic three species of ants of the genus Formica. -J. Physi- nest system in Formica uralensis Ruzsky. - Notulae Ento- 01. 19: 877-907 mol. 49: 21 1-230 BETREM. J. G. 1960. Ueber die Systematik der Formica ru- ROSENGREN, R. 1971. Route fidelity, visual memory and re- fa-Gruppc. - Tudschr. Entomol. 103: 51-81 cruitment behaviour in foraging wood ants of the genus For- COLLINGWOOD,C. A. 1979. The Formicidac (Hymenoptera) of mica (Hymenoptera, Formicidae). - Acta Zool. Fenn. 133: Fennoscandia and . - Fauna Entomol. Scand. 8: 1-106 1-174 ROSENGREN, R. and CHERIX. D. 1980. The pupa-carrying test CROZIER,R. H. 1975. Hymenoptera. - In cytogenetics as a taxonomic tool in the Formica rufa group. - Int. Symp. 3, Insecta 7, (Ed. B. JOHN), Borntraeger, Berlin, p. 2-95 Biosystematics ofsocial Insecrs, Paris, 1980 (summary) DLUSSKY, G.M. 1%7. Muravi roda Formica. -Moscow, 236 P.