(Figs. 9-14)
Atta (Mycocepurus) smithi Forel , 1893: 37(1-372 (Worker; St. Vincent Island: Bellisle, Brighton). - Wheeler, 1907: 717-718, 773-774, pl. 50, figs. 15,.10; pl. 52, fig. 44 (Worker; Cuba; Bion.). - Forel, 1912: 187 (Nest; Colombia: Dibulla).
Mycocepurus smithi : Wheeler & Mann, 1914: 42 (Haiti: Cape Haitien. Diquini). - Santschi, 1933: 123, fig. 9 (Worker). - Wheeler, 1936: 204 (Dominican Republic: San Lorenzo). - Weber, 1940: 417 (Panama Canal Zone: Gatun). - Weber, 1946: 128-129 (Bion., Distrib., British Guiana). - Kerr, 1961: 47 (Brazil, Sao Paulo: Rio Claro; Bion.).
Atta (Mycocepurus) smithi var. borinquenensis Wheeler . 1907: 718 (Worker; Puerto Rico: Vega Baja, Arecibo, Utuado, Monte Mandios). - Wheeler, 1908: 149 (Puerto Rico: Coama Springs). - Weber. 1934: 56 (Cuba: Soledad). - NOV. SYN.
Atta (Mycocepurus) smithi var. tolteca Wheeler , 1907: 718-719 (Worker; Mexico, Jalisco: Tuxpan). - NOV, SYN.
Mycocepurus smithi var. eucarnitae Forel . 1913: 235-236 (Worker; Santiago de Cuba). - NOV. SYN.
Trachymyrmex attaxenus Menozzi in Eidmann. 1936: 85-8b, fig. 4; pl. 1. - fig. X, 1-3 (Worker, female: Brazil, Rio de Janeiro: Mendes; Bion.). - NOV. SYN.
Mycocepurus reconditus Borgmeier , 1937: 246-248, figs. 34-36 (Worker, female; Brazil, Baia: Agua Preta = Uruguca). - Borgmeier. 1948: 204-205 (Peru: Valle Chanchamayo). - Borgmeier, 1948: 470 (Argentina, Formosa: Mojon de Fierro). - Kusnezov, 1953: 221 (Bolivia; Syn.). - Kusnezov, 1956: 49, fig. 69D (Worker). - NOV. SYN.
Mycocepurus smithi var. trinidadensis Weber , 1937: 378-379. fig. 1 (Worker, female; Trinidad). - NOV. SYN.
Mycocepurus bolivianus Weber , 1938: 155-156, fig. 8 (Worker; Bolivia: Rurrenabaque). - NOV. SYN.
Mycocepurus manni Weber , 1938: 156-157. figs. 1, 2 (Female; Bolivia: San Gregori). - NOV. SYN.
Worker . - Total length 2.8-3.0 mm; head length 0.67- 0.75 mm; head width 0.61-0.67 mm; scape length 0.54-0.61 mm; thorax length 0.83-0.93 mm; hind femur length 0.61-0.69 mm. Integument distinctly and finely reticulate-rugulose and punctate.
Head as shown in Fig. 9; somewhat elongate with less bulging cheeks. Occipital angles feebly obliquely truncate, the lateral angle of the truncation often quite indistinct. Frontal carinae usually obsolete on posterior half of head. Occiput in side-view obliquely curving forward and downward, not forming a conspicuous angle with gular' face. Mandibles rather narrow; chewing border with 5 teeth. Base of antennal scape not dilated in a ringlike fashion.
Thorax and pedicel as shown in Figs. 10 and 11. Premesonotal disc with- a circlet of only 4 well-developed pairs of teeth; the infero-lateral pair of pronotum and the pair in the center of the circlet lacking or- at best rudimentary; sometimes there is a convex transverse carinule between the antero-lateral pair of mesonotal teeth. Anterior pair of postero-mesonotal and epinotal spines very short and toothlike. Petiole rather slender, with a longer peduncle; body of node lacking a lateral horizontal carinule, the, anterior pair of teeth on top of node separated from the posterior pair of teeth by a very shallow excision; Postpetiole depressed, with a pronounced postero-median furrow on disc.
Erect hairs confined to dorsum of head, dorsum of scape; clypeus with just a few long hairs on anterior border. Hairs on dorsum of gaster appressed.
Female. - Quite similar to that of goeldii but strikingly smaller in size: Total length 4.1-4.4 mm; head length 0.80-0.85 mm; head width. 0.75-0.80 mm; scape length 0.64-0.67 mm; thorax length 1.20-1.31 mm. Specific characters as in workers, except for the thoracic spinulation. Pronotum usually, with only one well-developed scapular tooth, the antero-inferior tooth either rudimentary or completely wanting. Mandibles with 5 teeth on chewing border. Vermiculate-rugose sculpture of body finer, often quite weak, especially on dorsum of postpetiole and on sides of thorax. Wings infumated, venation as stated for the male in the generic diagnosis. Tibiae and dorsum of postpetiole completely lacking erect hairs.
Male. - Still undescribed. Three isolated males taken by Dr. W. E. Kerr at Rio Claro, Sao Paulo, Brazil, seem to represent this caste of smithi .
Total length 3.8 mm; head length 0.64 min; head width, compound eyes included, 0.69 mm; scape length 0.43 mm; thorax length 1.25 mm. Black; funiculus and tarsi brown. Integument densely reticulate-punctate and opaque. Differs from goeldii as follows: Much smaller in size (cf. measurements). Head (Fig. 12) more elongate, with less pronounced occipital angles; very little rugulose. Mandibles punctate and without distinct striae. Scape relatively shorter; funicular segments II-XI about three times as long as broad. Pronotum on each side with a single scapular tooth. Mesonotal scutum and scutellum rather faintly longitudinally rugulose. Rest of thorax practically without conspicuous rugae. Epinotal spines rectangular in side-view. Tergum I of gaster distinctly longer than broad. Genitalia quite distinctive (see Figs. 13 and 14). Wings infumated, venation as in goeldii . Pilosity of pedicel and gaster appressed.
Distribution. - M. smithi is widely dispersed, ranging from central Mexico and the greater and lesser Antilles through Central America to southeastern Brazil (Sao Paulo State) and northwestern Argentina (Formosa Province).
Specimens examined. - Over a hundred individuals, mostly workers, a few females and tentatively three males, from the following localities: Workers. - Argentina, Formosa: Mojon de Fierro (N. Kusnezov) (CTB). - Brazil, Sao Paulo: Guaratingueta(W. W. Kempf); Rio de Janeiro: Jardim Primavera (U. Kohnen), Mendes (H. Eidmann) syntypes of Trachymyrmex attaxenus Menozzi (CTB), Petropolis (C- Gilbert), Sao Bento (C. R. Goncalves) (CTB, DDSV); Guanabara: Rio-de Janeiro (W. W. Kempf, C. R. Goncalves) (WWK, DDSV); Minas Gerais: Teofilo Otoni (P. Thiemann, O.F.M.) (CTB), Tres Pocos(T. Borgmeier) (CTB); Espirito Santo: Vila Velha (0. Seifert, O.F.M.); Goias: Anapolis (W. W. Kempf); Mato Grosso: Dourados (R. Mueller), Jardim (R. Mueller); Baia: Agua Preta (= Uruguca) (G. Bondar) syntypes of M. reconditus Borgmeier (CTB); Para: Belem (C. R. Goncalves) (CTB, DDSV). - Bolivia: Espia, Rio Beni (W. M. Mann) (NAW), San Antonio (H. Marcus (CTB), Rurrenabaque (W. M. Mann) syntypes of M. bolivianus Weber (MCZ, NAW). - Peru: Valle Chanchamayo (W. Weyrauch) (CTB). - Surinam: Courantyne R. (N. A. Weber (NAW), Paramaribo (D. C. Geijskes) (CTB). - Trinidad: s. loc. (W. M. Wheeler, N. A. Weber) (MCZ, NAW), Diego Martins (Urich) (MCZ), Mayaro Bay (N. A. Weber) (NAW), Northern Range (N. A. Weber), syntype of M . smithi var. trinidadensis Weber (NAW). - Haiti: Diquini (W. M. Mann) (MCZ). - Dominican Republic: S. Lorenzo (s. coll.) (MCZ). - Puerto Rico: Utuado (W. M. Wheeler) syntypes of M. smithi var. borinquenensis Wheeler (MCZ). - Cuba: Aspiro Range), Pinar del Rio (A. Bierig) (CTB), Bolondron (W. M. Wheeler) (MCZ), Cayajabos (A. Bierig) (CTB). - Costa Rica: Bataan (N. A. Weber) (NAW). - Mexico, Jalisco: Tuxpan (McClendon) syntypes of M. smithi var. tolteca Wheeler (MCZ). - Females. - Brazil, Rio de Janeiro: Mendes (H. Eidmann) syntypes of Trachymyrmex attaxenus Menozzi (CTB); Bafa: Uruijuca (O. Bondar) syntype of M. reconditus Borgmeier (CTB); Mato Grosso: Jardim (R. Mueller). - Bolivia: S. Gregorio(W. M. Mann) holotype of M. manni Weber (NAW). - Surinam: Paramaribo (D. C. Geijskes) (CTB). - Males. - Brazil, Sao Paulo: Rio Claro (W. E. Kerr). (All specimens in WWK unless noted otherwise).
Synonymy. - All the forms herewith placed into synonymy of smithi are briefly discussed in the following. These comments will also show the range of infraspecific variation of the present species.
1. M. smithi var. borinquenensis Wheeler , 1907, worker. - Syntypes from Utuado, Puerto Rico, seen. The main distinguishing feature of this form is said to consist in the presence of a small tooth on each side of the occipital furrow at the postero-median border of the head. This character, which is here indeed well-developed, also occurs occasionally in specimens from other often distant localities and is not apt to circumscribe a taxonomically valid form.
2. M. smithi var. tolteca Wheeler , 1907, worker. - Syntypes from Tuxpan, Jalisco, Mexico, seen. They are of a slightly lighter, yellowish color, have the posterior epinotal spines more acute and upright, a feebler cephalic sculpture; the small denticles flanking the midoccipital furrow in the preceding variant are here substituted by low and pointed ridges. However, none of these characters is significant.
3. M. smithi var. eucarnitae Forel , 1913, worker. - Types from Santiago de Cuba not seen. According to the description they are of somewhat larger size, have longer promesonotal spines, the anterior pronotal ones being as long as those of the mesonotum. Teeth flanking the midoccipital furrow as in var. borinquenensis . Several Cuban specimens examined, although not visibly disagreeing with the afore mentioned diagnosis, do not vouch for the existance of a particular geographical race on that island. Hence eucarnitae is just a plain synonym of smithi .
4. Trachymyrmex attaxenus Menozzi i. litt., worker and female. - The paper by Menozzi supposed to contain the formal proposition of this species never appeared in print. The name was published by Eidmann (1936), who also figured both the worker and the female and gave an important account of the biology of this ant. Syntypes, received by Borgmeier from Eidmann, proved on examination that this is nothing but the common and widespread M. smithi .
5. M. reconditus Borgmeier , 1937, worker and female. - Syntypes examined. In the original diagnosis this species is differentiated from obsoletus according to the description and figures of the latter. The types , however, confirm that reconditus is conspecific with, and a junior synonym of, smithi .
6. M. smithi var. trinidadensis Weber , 1937, worker and female. - A syntype worker seen. According to the description "the workers of this variety differ chiefly in sculpture. The anteriorly directed convex and blunt ridge between the anterior mesothoracic spines is more reduced or practically absent. Between the sharply carinate sides of the first gastric segment the surface is longitudinally and finely rugulose". Since these characters vary at random and the examined syntype does not reveal a tangible difference, the present variety is best relegated to synonymy of smithi s. str.
7. M. bolivianus Weber , 1938, worker. - Syntypes examined. This species has been correctly synonymized by Kusnezov (1956) with reconditus Borgmeier , which in turn is a synonym of smithi .
8. M. manni Weber , 1938, female. - The holotype was examined. The specimen possesses somewhat heavier reticulate-rugose and vermiculate macrosculpture. The antero-inferior scapular spine is rudimentary. Otherwise, thise female is much like smithi from which it may not be separated specifically.
Bionomics. - The ensuing data have been compiled from papers by Forel (1893a): 371-372, 1912: 187), Wheeler (1907: 773-774), Wheeler & Mann (1914: 42), Eidmann (1937: 85-86), Borgmeier (1937: 248) and Weber (1946: 128-129). The contribution by Eidmann is by far the most complete.
The small and sluggish workers when foraging carry dry leaves and caterpillar droppings back to their nest. The nesting sites are either in open fields and woods or even in moist gullies. The nest proper is in the soil. On the surface it is marked by craters of earth crumbs, measuring not more than 8 cm in diameter. These superficial structures stand out by their color which is different from that of the top soil, indicating that the nest cavities are at some depth. According to Bondar (Borgmeier, 1937) nest chambers have been dug out at a depth varying from 80 to 100 cm. In Colombia, Forel (1912) found a rather shapeless fungus-garden of this species at very little profundity.
A fact reported by many observers and confirmed by my own field experience is that usually a small area contains many craters of the same species, whereas neighboring areas have none at all. H. H. Smith (Forel, 1893a) who first called attention to the phenomenon, suggested that the craters of a given area represent the entrances of just one common formicarium (as happens with goeldii during the mating season, according to Luederwaldt). This, however, has not as yet been established conclusively.
The nest cavity, measuring 4-5 cm in width to 2.5-3 cm in height, possesses a flat ceiling and an excavated bottom. From the ceiling without the support of a framework of plant rootlets hang narrow clusters or threads of fungus material. These threads, which are quite consistent, are made up of finely cut up leaf material connected by the mycelium. The fungus itself has not as yet been identified. Eidmann states that superficially it resembles that of Atta sexdens , whereas Forel (1912) glibly states that it is not Pholiota (Rizotes) gongylophora. Away from the nest chamber lead several fine and threadlike tunnels barely giving passage to the tiny workers. Eidmann (1936, fig. 4) gives a photograph of a nest chamber with the suspended fungus garden.
While collecting in Puerto Rico, Wheeler (1907: 774) made several attempts at excavation of the fungus garden of M. smithi but succeeded only once. In moist red clay under a stone he found a small irregular chamber with about 30 ants. The fungus garden, a small mass of approximately 2 cc in volume, consisted of caterpillar droppings studded with bromatia that scarcely differed from those of Cyphomyrmex rimosus and allies, the only Attine ants known to cultivate a yeast. Wheeler's discordant observation poses an interesting problem, but also needs further confirmation..
According to Eidmann, the colonies are polygynous. At any rate he found several dealated queens in a single nest chamber. The same author proclaims a lestobiotic relationship between M. smithi and Atta sexdens because he found a great many nest chambers of the former between the cavities made by the latter. However, if any such relationship exists, it is not obligatory since M. smithi also occurs in areas where no sign of an Atta sp. could be discovered. Perhaps this association, of which no details are known, dissolves itself in the loose relationship of facultative synoecetes.
In southeastern Brazil AT. smithi lives occasionally side by side with M. goeldii under the same ecological conditions. Kerr (1961) even found 3 males of the former in a swarm of 150 males of the latter species at Rio Claro, Sao Paulo State, Brazil. None of the smithi males attempted to mate with goeldii queens.
