In Central and northern South America, a complex of species occurs that have (1) the frontal carinae separate from the torulus, passing above it and onto the clypeus; (2) the face sculpture varying from completely striate to strongly clathrate (composed of high, sharp, well-separated rugae, which form irregular polygons over face surface); and (3) AT4 varying from striate to completely smooth and shining. The complex occurs as a series of allopatric populations restricted to montane forests. There is abundant material from Costa Rica, which reveals the presence of two sympatric species. In contrast, material from South America is sparse. Character variation is high even within Costa Rica, and discordant character variation occurs across the material from northern South America. I have not been able to discover evidence for multiple sympatric lineages over broad geographic areas in South America, although there is weak evidence for two to three sympatric species in parts of Colombia. It appears that each local mountain range hosts a unique community, perhaps shaped by a combination of dispersal history, local selection, and occasional hybridization.
A discrete character that separates sympatric forms in Costa Rica, and largely maintains its integrity in the material from northern South America, is the presence or absence of a torulus trough. The torulus trough is an effect of the shape of the lateral portion of the clypeus. In specimens lacking a trough, the lateral portion of the clypeus falls perpendicularly to the torulus, such that in anterior view (looking at the head with mandibles forward) the torulus is relatively exposed. In specimens with a trough, the lateral clypeus is somewhat produced, such that it falls to the torulus at an acute angle, and in anterior view the torulus is relatively obscured by the lateral "wings" of the clypeus (figure).
The variation in rudis group material is described below, beginning with material lacking a torulus trough and ending with material having a very pronounced and deep torulus trough.
Costa Rican mayri: HW 1.3-1.7mm (n=4); mandibles coarsely to weakly striate; torulus trough lacking; face sculpture composed of high, sharp, well-spaced rugae; spaces between rugae smooth and shining; rugae often anastomosing to form polygons, with little longitudinal orientation (clathrate sculpture); this condition grades into increasing degrees of longitudinal orientation of rugae, especially anteriorly; rarely, rugae may be almost entirely longitudinal and subparallel; posterior face of forefemur smooth; AT4 smooth and shining, or occasionally with faint rugae anteriorly near petiolar insertion; AT4 with sparse, whitish pubescence under the erect setae, or lacking underlying pubescence; legs black or occasionally red.
Lectotype of mayri, "Colombia": differing from Costa Rican material primarily in larger size, HW 1.95mm; also, face more uniformly striate, less clathrate; legs red. See under mayri.
Holotype of reichenspergeri, "Brazil": nearly identical to mayri type, but legs black; HW 1.89mm. See under mayri.
VENEZUELA: Aragua: Rancho Grande [10¡21'N 67¡41'W], 1100m, 11 Aug 1967 (R. W. Poole) LACM ENT 139866 [MCZC]; same data, 23-27 Jun 1971 (W. L. & D. E. Brown) LACM ENT 139865 [MCZC]. These two specimens are nearly identical to mayri type except face relatively more clathrate; HW's 1.95mm and 2.12mm. Also, the eyes are relatively smaller than any other rudis group specimens (EL/HW 0.17).
COLOMBIA, Valle: Medio Calima and Rio Bravo [3¡55'N 76¡35'W], 3 Sep 1989 (W. Mackay) LACM ENT 139779, LACM ENT 139780, LACM ENT 139781 [JTLC, LACM, Mackay collection]. These three specimens are similar to mayri type except face more clathrate; legs black; HW 1.72mm (n=1).
PERU: Junin: Hacienda San Juan, Colonia Perene [10¡53'S, 75¡13'W], June 1920 (Cornell Expedition) LACM ENT 139867 [LACM]. This specimen similar to mayri type except face more clathrate; HW 1.62mm.
ECUADOR: Pichincha: Maquipucuna, 5km ESE Nanegal, 0¡07'N 78¡38'W, 1350m, 17 Aug 1991 (P. S. Ward #11498) LACM ENT 139844, LACM ENT 139845, LACM ENT 139846, LACM ENT 139847, LACM ENT 139848, LACM ENT 139849, LACM ENT 139850, LACM ENT 139851, LACM ENT 139852, LACM ENT 139853, LACM ENT 139854 [JTLC, LACM]. This nest series lacks torulus trough; clypeus somewhat produced anteriorly in lateral view; size relatively small compared to Costa Rican and Colombian mayri, HW 1.35mm (n=1); AT4 with relatively coarse longitudinal rugae near petiolar insertion; legs black.
On most mayri from Costa Rica, and the series from Valle Dept., Colombia, the portion of the frontal carina that curves mesad above the torulus is well separated from the torulus, and remains relatively elevated (and higher than wide) to the point where it joins or parallels the lateral clypeal carina. On a few Costa Rican specimens, the types of mayri and reichenspergeri, the specimens from Rancho Grande, Venezuela, and the specimen from Peru, the frontal carina tapers as it curves mesad, and becomes little more than a raised line, no higher than wide, that crosses the dorsum of the torulus and approaches the lateral clypeal carina. The series from Ecuador is intermediate in the degree of tapering of the frontal carina.
This figure shows variation in the dorsal view of the mesosoma for the material near mayri.
Holotype and paratype of xanthopilosus, Ecuador. The two workers of this species have unique pubescence on AT4: the erect setae are underlain with a dense, yellowish, decumbent pubescence, which gives a distinctive woolly appearance to gaster. Other characters include lack of a torulus trough; mandibles coarsely striate; clypeus not produced anteriorly in lateral view; face clathrate; HW 1.36mm (n=1); forefemur coarsely striate; AT4 with closely-spaced, subparallel, longitudinal striae near postpetiolar insertion, gradually breaking up and fading posteriorly; disc irregularly and faintly rugulose punctate, intervals shiny, rugulae strengthening somewhat, becoming more longitudinal at posterior border.
Holotype of virgatus, Ecuador: mandibles smooth, not striate; torulus trough weak; clypeus produced anteriorly in lateral view; face finely longitudinally striate, with few cross linkages, much finer than most rudis group material; HW 1.33mm; forefemur smooth; AT4 largely smooth and shining, with barest traces of longitudinal rugulae near petiolar insertion, and weak rugulae at posterior margin, converging to a point immediately anterior to posterior margin; legs red.
ECUADOR: Pastaza: Mera [1¡28'S 78¡8'W] 26 Jan 1923 (F. X. Williams) LACM ENT 141499 [BPBM]. This specimen is from very near the type locality of virgatus. It is similar to the virgatus type but clypeus not produced anteriorly; more extensive clathrate sculpture on the face; HW 1.40mm; AT4 completely smooth, with no rugulae.
ECUADOR: Tungurahua: Banos [1¡24'S 78¡25'W], 1829m, 26 Oct 1922 (F. X. Williams) LACM ENT 141498 [BPBM]. This specimen is from very near the type locality of virgatus. It is similar to the virgatus type but mandibles coarsely striate; clypeus not produced anteriorly; HW 1.33mm; AT4 completely smooth, with no rugulae.
COLOMBIA: Risaralda: nr. Pueblo Rico [5¡12'N 76¡6'W], Mar 1992 (F. Fernandez) LACM ENT 139777, LACM ENT 139778 [JTLC, LACM]. These two specimens are similar to the virgatus type but have weakly striate mandibles; clypeus not produced anteriorly; HW 1.34mm (n=1); AT4 completely smooth, with no rugulae.
Lectotype and four additional workers from the original syntype series of rudis, from Bogota, Colombia: lectotype with torulus trough present; clypeus produced anteriorly in lateral view; face clathrate; HW 1.31mm; posterior face of forefemur smooth; propodeal spines strongly diverging; AT4 coarsely striate over most of surface (ASW 0.035mm), leaving a narrow smooth band posteriorly; AS4 with coarse longitudinal striae anterolaterally; legs dark red black. Two of the other workers are similar to the lectotype. One worker has subparallel spines and coarsely striate forefemur. Another worker has light red-orange legs.
Five collections from the Cauca Valley in Colombia's Valle and Risaralda departments: I consider these conspecific with rudis; they are very similar to the lectotype in size, shape, and sculpture. The torulus trough is slightly weaker; the forefemur is coarsely striate on all the specimens; AT4 is coarsely striate (ASW 0.037mm) over entire surface (4 collections) or fading half way back (1 collection); and the legs vary from orange to dark red to almost black, but are uniformly colored within collections.
batesi from the central ranges of Costa Rica and western Panama: torulus trough well developed; clypeus not produced anteriorly; face clathrate; mean HW 1.29mm (range 1.11-1.50, n=4); propodeal spines subparallel or less often somewhat divergent; forefemur smooth; AT4 smooth or with traces of short striae radiating from petiolar insertion; if anterior striae present on AT4, never as regular as on carbonarius (figure); legs always black. Costa Rican material is abundant and the limits of variation well understood. The type of laeviventris (from Volcan Chiriqui, Panama, just across the border from Costa Rica) falls within this group. laeviventris was named as a subspecies of carbonarius and later synonymized with it by Kempf (1951). As construed here, laeviventris is a synonym of batesi, but corresponds to populations whose known range extends from Cerro Campana near the Panama Canal to Cerro Guanacaste in northern Costa Rica, probably in all areas of montane wet forest above about 600m elevation.
Four specimens from the Osa Peninsula of Costa Rica, a somewhat isolated patch of cloud forest in a region known for its biogeographical distinctness: similar to other batesi specimens but rugae on face and mesosoma more rounded, less sharp than all other rudis group material, giving specimens a slightly "melted" look; AT4 with irregular coarse striae anteriorly and posteriorly, with disk smooth. These specimens are identified as batesi.
Holotype of batesi from "Colombia": HW 1.35mm; propodeal spines divergent; forefemur smooth; AT4 completely smooth and shining. This specimen is within the range of variation of laeviventris, but the size and propodeal spine orientation are also similar to rudis. Further characterization of variation in Colombia is needed (if possible; are there still cloud forest habitats around Bogota, where various early Procryptocerus types were collected?).
ECUADOR: Napo: 25km WSW Loreto, 0¡44'S, 77¡31'W, 1050m, 8 Aug 1991 (P. S. Ward) LACM ENT 139756 [LACM]; VENEZUELA: Miranda: Origin Caracas [10¡30'N 66¡55'W], in quarantine, San Francisco, 21 Jul 1947, LACM ENT 139752, LACM ENT 139753, LACM ENT 139754, LACM ENT 139755 [JTLC, LACM]. These two collections are similar to batesi but clypeus anteriorly produced in lateral view (especially in Ecuador series, figure); clypeus clathrate or foveate rather than striate, especially on bend; HW 1.27mm (n=1); forefemur striate; AT4 smooth with traces of coarse striae near petiolar insertion; AS4 with a group of about 6 subparallel, oblique striae on each side anteriorly. The Venezuelan workers are all light orange brown, and appear to be callows. The Ecuadorian workers are all black. These two collections are remarkably similar given the great distance between their collection localities. It is tempting to speculate that these and batesi represent an ancestral type, replaced by other forms (carbonarius and rudis) in the center of the former range.
Holotype of carbonarius from Bogota, Colombia: torulus trough deeply impressed; HW 1.04mm; propodeal spines subparallel; forefemur smooth; AT4 finely striate on anterior threefourths (ASW 0.021mm), striae more regular than any on batesi (figure); legs black. It is very distinct from the type of rudis, which is also from Bogota. Presumably the species are sympatric there.
Two collections from the Cauca Valley near Cali: similar to carbonarius in shape and size (HW 1.09-1.10mm (n=2)), but forefemur partially striate; striae on AT4 somewhat coarser, ASW 0.028mm; apices of femora and all of tibiae red. These collections are quite close to the carbonarius type and are identified as such. Paralleling the situation near Bogota, these collections of carbonarius appear to be sympatric with a local form of mayri and rudis.
Page author:
John T. Longino, The Evergreen State College, Olympia WA 98505 USA. longinoj@evergreen.edu
Date of this version: 6 December 2002.
Go back to top
