Cyphomyrmex transversus Emery, 1966

Kempf, W. W., 1966, A revision of the Neotropical fungus-growing ants of the genus Cyphomyrmex Mayr. Part II. Group of rimosus (Spinola) (Hym. Formicidae)., Studia Entomologica (N. S.) 8, pp. 161-200 : 193-197

14. Cyphomyrmex transversus Emery , n. stat.

(Figs. 12, 24, 30)

Cyphomyrmex rimosus transversus Emery , 1894: 226 (Worker, female; Brazil: Mato Grosso). - Emery, 1905: 161 (Brazil, Mato Grosso: Coxipo). - Wheeler, 1907: 723-4 (Worker, female, male; Syn.; Brazil: Ceara; Pernambuco: Olinda). - Bruch 1915: 529 (Argentina: Salta, Tucuman). - Wheeler, 1925: 45 (Worker; Brazil, Rio Grande do Norte: Natal). - Borgmeier, 1927: 126-7 (Brazil: Rio Grande do Sul). - Santschi, 1931: 280, fig. 3 (Worker). - Santschi, 1933: 118 (Argentina, Misiones: Loreto). - Weber, 1940: 412 (Worker; key). - Kusnezov, 1949: 436, 442, figs. 6-14 (Worker; key; Argentina, Tucuman: Rio Sali, Los Puestos; Salta: Rio Saladillo, Cafayate, Cerro San Isidro; Santiago del Estero; Termas de Rio, Hondo, Santo Domingo, Palo Errao, Siete Arboles). - Kusnezov, 1957: 260-1, (Discussion; Syn.). - Weber, 1958: 260-1 (Worker, female; except new syn.)

Cyphomyrmex dentatus olindanus Forel , 1901b: 337 (Worker; Brazil, Pernambuco: Olinda).

Cyphomyrmex rimosus var. pencosensis Forel , 1914: 281-282 (Worker; Argentina. San Luis: Alto Pencoso). - Bruch, 1916: 323-4, fig. 14, pl. 10, fig. 1 (Worker). - Gallardo, 1916: 324 (Argentina, Buenos Aires: Buenos Aires, Tandil; Cordoba: Alta Gracia). - Bruch, 1923: 201-2, fig. 13, pl. 2, figs. 1-3 (Worker; fungus garden). - NOV. SYN.

Cyphomyrmex rimosus transversus var. pencosensis : Santschi, 1931: 278-9, figs. 4, 8, 9 (Worker, female, male; Argentina, Cordoba: Alta Gracia, Tanti Viejo, La Paz; Santa Fe: Fives Lille).

Cyphomyrmex rimosus pencosensis : Weber, 1940: 411 (Worker, female, male; key). - Kusnezov, 1949: 436, 441 (Worker; key; Argentina, Salta: Cafayate, Aguas Blancas; Tucuman: Saladillo, Tucuman). - Kusnezov, 1957: 10 (Worker; key).

Types. - Workers and a female, collected by P. Germain at an unidentified locality in Mato Grosso, Brazil, presumably in the Emery collection; not seen. One syntype worker of olindanus Forel received on loan from the MCZ. Types of pencosensis presumably in the Forel collection; not seen.

Worker. - Total length 2.7-3.4 mm; head length 0.67- 0.83 mm; head width 0.64-0.80 mm; thorax length 0.88-1.09 mm; hind femur length 0.69-0.83 mm. Uniformly yellowish brown to more or less fuscous brown; especially cephalic dorsum and gaster are occasionally more distinctly infuscated. Integument finely and densely punctate-granulate, opaque.

Head as shown in Fig. 12. Mandibles reticulate-striolate and somewhat shining. Clypeus having the anterior border either straight or slightly concave, bearing on its corners a weak, blunt tooth. Frontal area impressed, without hairs. Frontal lobes semicircular, greatly expanded laterad; frontal carinae a bit sinuous and diverging caudad, attaining the slightly produced occipital corner. Midfrontal tumulus and transverse frontal groove extremely feeble; head disc nearly flat. Paired carinae on vertex blunt, low, extremely weak to vestigial. Preocular carina curving mesad above eye, not joining up with the feeble carina extending from the occipital lobe foreward to the postero-inferior border of eye. The latter with about 9-10 facets across its greatest diameter. Supraocular tubercle usually weak, contained in, and marked as a blunt angle of, the postocular carina. Inferior border of cheeks sharply marginate. Scape in repose surpassing the occipital corner by a distance subequal to its maximum width. Funicular segments II-IX not longer than broad; segment I a bit longer than II and III combined.

Thorax as shown in Fig. 24. Pronotum dorsally with four tubercles, the median pair smallest; antero-inferior corner with a prominent tooth; sides of dorsal disc feebly marginate in front of the blunt, lateral tubercles. Mesonotum shallowly impressed, flanked by two pairs of low, ridge or welt-like tubercles; both the anterior and the posterior pair often fused to each other forming transverse, semicircular ridges, somehow imitating the condition obtained in peltatus and dentatus . Mesoepinotal constriction usually rather shallow in profile, forming an extremely blunt angle. Basal face of epinotum subquadrate, laterally bluntly marginate, each side bituberculate, the anterior tubercle obtuse, the posterior usually more prominent and tooth-like, situated below the level of basal face on the upper third of the declivous face. Basal third of hind femora gradually incrassate on flexor face, then forming an obtuse angle; the distal two thirds attenuate; posterior border of flexor face sharply marginate or even carinulata especially on bent.

Pedicel as shown in Figs. 24 and 30. Petiolar node strikingly transverse, about thrice as broad as long, lacking a dorsally produced crest and teeth on posterior border; strongly constricted in front of postpetiolar insertion. Postpetiole likewise rather broad, with a usually deeply impressed midlongitudinal groove and a shorter and broader groove posteriorly on each side. Tergum I of gaster with an antero-median groove, at least as long as petiole and hairless; lateral borders of same tergum distinctly marginate.

Body hairs squamate and reclinate, unusually short, thick and conspicuous on head, thoracic dorsum and gaster; narrow, squamate and appressed hairs on scapes and legs.

Female. - Total length 3.5-4.2 mm; head length 0.80- 0.93 mm; head width 0.76-0.88 mm; thorax length 1.09-1.33 mm; hind femur length 0.80-1.04 mm. This caste resembles quite closely that of rimosus . The lateral ocelli, not prominent nor placed on raised ridges; the distinctly dentate antero-inferior corner of pronotum; the always well developed and salient epinotal spines; the striking width of the pedicelar segments, even better expressed in this caste than in the worker; the deep longitudinal furrow on the postpetiolar dorsum, distinguish transversus from rimosus . The squamate body hairs are of the same kind as in worker. Wings infuscated, venation as represented by Kusnezov (1949, pl. 1, fig. 15).

Male. - There is a scant diagnosis of this caste in Wheeler (1907: 724).

Distribution. - The present species is known to occur from northern Brazil to central Argentina. Being more xerophilous than the otherwise omnipresent rimosus , it even occurs in the dry northeastern Brazil as the only representative of the genus.

Specimens examined: 83 - 68 workers, 9 females and 6 males - as follows: Argentina, Chaco: R. Saenz Pena, October 19, 1933, A. A. Ogloblin leg. 1 female (MCZ); Santa Fe: Fives Lille, Weiser leg. 4 workers (CTB). - Brazil, Rio Grande do Sul: Porto Alegre, May-June 1926, Haertel and P. Buck leg. 4 workers (CTB); Sao Paulo: Agudos, 8 different collections between March 1952 and October 1957, W. W. Kempf and C. Gilbert leg. 29 workers (WWK); Rio de Janeiro: Cabo Frio, July 1926, T. Borgmeier leg. 4 workers, 2 females (CTB); Minas Gerais: Aracuai, Nov. 1926 and October 1929, P. Thiemann leg. 9 workers, 3 females, 2 males (CTB); Goias: Goiania-Campinas, April 1927, S. Schwarzmaier leg. 5 workers, 1 female (CTB), Leopoldo Bulhoes, May 1935, S. Schwarzmaier, leg. 1 female, 1 male (CTB); Bahia: Jacobina, December 1953, C. R. Goncalves leg. 2 workers, 2 males (CTB); Pernambuco: Olinda, without date and collector, 3 workers including a syntype of Forel's olindanus (MCZ, CTB), Tapera, without date, B. Pickel leg. 2 workers (CTB); Ceara: Ubajara, December 1948, C. R. Goncalves leg. 1 worker (CTB); Para: Obidos, January 1949, C. R. Goncalves leg. 1 worker (CTB), Vigia, November 1943, C. R. Goncalves leg. 2 workers, 1 female, 1 male (CTB); Amazonas: Manaus, October 1953, C. R. Goncalves leg. 2 workers (CTB).

Discussion. - Emery's original description of the present species, in spite of its brevity, contains the main distinguishing features. Further descriptive details and discussions are contained in the papers by Wheeler (1907, 1925), Santschi (1931, fig. 3, which represents a profile of the thorax of a syntype worker), "Weber (1940, 1958), Kusnezov (1949). Already in 1925 Wheeler suggested that transversus "might be properly regarded as a distinct species". However, the conservative view prevailed and transversus has constantly been considered a subspecies of rimosus up to the present time.

Although transversus is doubtless a close ally of rimosus , it exhibits nevertheless a set of good distinguishing characters, constant throughout its vast range. Furthermore, both forms occur side by side over a vast part of South America. Hence full specific rank is evidently indicated for transversus .

The chief separatory characters between transversus and rimosus s. I. have already been given for the female in the foregoing diagnosis. The worker differs from rimosus in the feeble and low pair of carinae on vertex; the distinctly dentate anteroinferior corner of pronotum; the low mesonotal ridges, as seen in profile, especially the posterior pair - both pairs encircling the slightly impressed disc much as in peltatus and dentatus ; the rather shallow mesoepinotal constriction, appearing as an obtuse angle in profile; the two pairs of tubercles on the posterior corner of the basal face of epinotum; the strikingly transverse pedicelar nodes, principally the petiole; the deeply impressed middorsal groove on postpetiole; the long and hairless antero-median groove on tergum I of gaster; the body hairs which are thickly squamate, especially on head, thorax and gaster. Although due to variation proper to this group some of the aforesaid characters may occasionally fail to reach their full expression - or rimosus in one or the other specimen may imitate one or very few of the characters of transversus - their ensemble will always be sufficient to separate transversus from rimosus .

Synonymy. - There is no doubt about the absolute identity between transversus and olindanus , first recognized by Emery (1905) and formally proposed by Wheeler (1907), who had received a. syntype of olindanus from Forel. Recently, Weber (1958) sunk his own rimosus venezuelensis as a synonym of transversus after a comparison made between the respective types. Having seen three syntype workers of venezuelensis , I cannot agree with this procedure. Indeed, venezuelensis has exceptionally low mesonotal tubercles and a rather broad petiolar node, but fails to reproduce all other distinguishing features of transversus . Hence I propose to reinstate venezuelensis and place it provisionally under rimosus . Its final status depends from a through revision of the latter.

A new synonym for transversus is herewith proposed, namely rimosus var. pencosensis Forel . This form has been amply quoted (perhaps some records in the literature don't even refer to this form), but only Santschi (1931) seems to have given any thought to its systematic relationship. He showed that the closest relative is not rimosus fuscus Emery , as propounded by Forel, but transversus , from which it differs merely and exclusively in its darker color.

Unfortunately, Santschi did not sink pencosensis right there, but satisfied himself by lowering it as a variety under transversus . Weber (1940), in his synopsis of the genus, raised pencosensis to subspecific rank, probably overlooking Santschi's previous results. Kusnezov, (1949, 1957), fully aware of Santschi's opinion, chose to follow Weber's classification, distinguishing artificially both forms by the development of the thoracic tubercles in the worker caste. This character, admitting intergradation, is simply useless, and I can see no other solution than placing pencosensis into synonymy of transversus .

Bionomics. - From my field experience in Agudos I have come to the conclusion that transversus nests in dryer situations (open fields, parkland) than rimosus which prefers the more humid environment of dense woodlands. The distribution of the former seems to confirm this rule.

Bruch (1923) has studied and pictured the fungus-garden and nest of " pencosensis " in the Argentine. In fact, this ant cultivates a yeastlike fungus on excrements of insects, principally acridid grasshoppers, much as the typical rimosus and its allies. Perhaps here lies the reason why Weber hesitated to separate the form from rimosus , being of the opinion that all Cyphomyrmex forms with this aberrant or - as he likes to put it - primitive type of fungus culture should be ranked under rimosus . 1 cannot follow this line of reasoning, inasmuch as species of Acromyrmex and Atta , cultivating the same basidiomycete fungus (Pholiota gongylophora) would have to be placed - a fortiori - into the same genus, if not to be declared conspecific!