Pseudomyrmex triplarinus

P. triplarinus has been collected from Triplaris americana (=T. noli-tangere Weddell), and other Triplaris trees not identified to species. The habits of P. triplarinus appear to be similar to those of Pseudomyrmex dendroicus, i.e. workers readily attack and sting intruders and they clear vegetation from the base of the host tree. An ecological study by Oliveira et al. (1987) in Brazil demonstrated that P. triplarinus workers were more efficient than a co-foraging Crematogaster species at finding and capturing termite baits on host foliage. (Ward 1999)

Identification
Ward (1999) - P. triplarinus may be recognized by the following combination of traits: relatively large species (worker HW > 0.99, LHT > 0.81; queen HW 1.46–1.60, queen LHT 1.19–1.41); head broad (worker CI > 0.91, queen CI 0.90–0.95), covered with fine punctures separated by one to several diameters; head not contrastingly darker than the mesosoma; pilosity abundant on mesosoma dorsum and external face of metatibia (worker MSC > 40, worker HTC > 8) but absent from side of head (frontal view); legs relatively long (worker LHT/HL > 0.76); petiole longer than wide, especially in the queen (worker PWI <1.00, queen PWI 0.75–0.84), and with a subtriangular subpetiolar process which is directed ventrally. The relatively slender legs alone separate P. triplarinus from all other Triplaris inhabitants except Pseudomyrmex dendroicus, from which it differs by the more abundant pilosity, ventrally directed subpetiolar process, and several other features (see P. dendroicus). P. triplarinus shows a moderate amount of variation, especially in body colour. Light-and dark-coloured workers may occur in the same population and within single nest series.

Distribution
Southern Colombia through the Amazon basin to Bolivia and the pantanal of southwestern Brazil.

Distribution based on Regional Taxon Lists
Neotropical Region: Argentina, Bolivia, Brazil, Colombia, Ecuador, Peru.

Biology
Wheeler (1930), Columbia - When the narrow-leaved Triplaris is inhabited by Pseudomyrma triplarina, it ranks with the terrible nettle (Jatropha urens L.) that grows in much the same sort of place. Both plants offer tempting holds on the steep slopes along the brooks, and a few stings from either the nettle or the ants will cause the hands to swell painfully. Ants and their brood occupy cavities in the central stem and its lateral branches throughout the Triplaris, except near the base of the larger trees. The cavities, which are continuous and are apparently cut through the nodes by the ants, may be occupied up to the last or next to the last internode below the terminal bud. The entrances are perpendicular slits in the stem or branches and are about 120° around the stem from the base of the leaf next above. They are usually just below a node, but may be more than half way down to the node below. There is almost always at least one entrance to a node, and there may be several, one above the other.

In very young stems the medullary cavity is about half the diameter of the stem, nearly round in cross-section, except for• the depressions noted below, and somewhat enlarged between the nodes. The passage through each node is greatly constricted. In stems an inch and a half in diameter the entrance may remain open, but the cavity even in the oldest stems is scarcely or not at all larger than in young twigs. The walls of the cavities in the latter are brown; in old twigs, black.

The depressions mentioned above are vertical, slit-like grooves in the walls of the cavities and are of two kinds, those which open into entrances and those which extend only to the inner surface of the bark and are used to house Coccids. The position of the entrances has been described; the Coccid houses have a rather more irregular location, but often seem to be incompleted entrances, to which• in many cases they correspond in position. They are frequently placed just below leaf attachments, however, where there is often a smooth, slightly depressed area on the external surface of the bark. Whether these “Coccid houses” are natural or cut by the ants was not ascertained.

The Colombian Triplaris and its ants had been previously observed by Forel, and Dr. Salt. We quote the latter's notes of October 16, 1927: “Quebrada Rodriguez, Rio Frio. A small tree of Triplaris sp. (americana  ? according to Johansen) about 12-15 feet high and 2 inches at 5 feet above the ground, was felled and some of the ants collected. They sting viciously, the poison causing the fingers to swell and bringing out red flushes right up to the arm. The ant-brood and coccids were found in the branches and at the tip. In every case the coccids were found at the upper or distal end of the internode, frequently right up against the• partition, and the ant-brood at the lower or proximal end. In several places on the inside of the hollow trunk the wood was deeply pitted but not perforated. It was not determined whether the pits were made by the ants or were normal depressions. The exit holes of the ants are regular in position and undoubtedly are merely modifications of some previously existing structures (lenticels ?) of the plant.”

The following insects were found by the junior author associated with the narrow-leaved Triplaris:

Pseudomyrma triplarina Probably the commonest and certainly the most vicious ant in the Triplaris. It does not inhabit every tree, however, nor does it always take complete possession, for it was sometimes found sharing trees with Crematogaster.. Series of this ant were also taken by Dr. Salt in Triplaris, presumably of the narrow-leaved form, at Rio Frio and Aracataca. Apparently these three ants are all capable of maintaining their colonies without leaving the home tree.

Nomenclature

 *  triplarinus. Myrmica triplarina Weddell, 1850: 263 (footnote) (w.) BRAZIL. Wheeler, G.C. & Wheeler, J. 1990a: 464 (l.). Ward, 1999b: 505 (designation of neotype). Combination in Pseudomyrma: Wheeler, W.M. 1942: 186; in Pseudomyrmex: Wheeler, G.C. & Wheeler, J. 1956: 386. Senior synonym of arborissanctae: Wheeler, W.M. 1942: 186; of ecuadoriana: Brown, 1949a: 44; of cordobensis, rurrenabaquensis: Ward, 1999b: 505. See also: Kempf, 1961a: 408.
 * arborissanctae. Pseudomyrma arborissanctae Emery, 1894c: 147 (w.q.) BOLIVIA. Junior synonym of triplarinus: Wheeler, W.M. 1942: 186.
 * cordobensis. Pseudomyrma arborissanctae var. cordobensis Forel, 1914d: 265 (w.) ARGENTINA. Combination in Pseudomyrmex: Kempf, 1961a: 400. Subspecies of triplarinus: Wheeler, W.M. 1942: 188. Junior synonym of triplarinus: Ward, 1999b: 505.
 * rurrenabaquensis. Pseudomyrma triplarina var. rurrenabaquensis Wheeler, W.M. & Mann, in Wheeler, W.M. 1942: 188 (w.) BOLIVIA. Combination in Pseudomyrmex: Kempf, 1961a: 401. Junior synonym of triplarinus: Ward, 1999b: 505.
 * ecuadoriana. Pseudomyrma arborissanctae var. ecuadoriana Enzmann, E.V. 1944: 79 (w.) ECUADOR. Junior synonym of triplarinus: Brown, 1949a: 44.

Worker
Ward (1999) - Measurements (n=39). HL 1.03–1.46, HW 1.00–1.41, MFC 0.108–0.158, LHT 0.82–1.23, CI 0.92–1.00, REL 0.40–0.45, REL2 0.42–0.47, FCI 0.10–0.14, SI 0.43–0.48, FLI 1.45–2.09, FI 0.34–0.39, PLI 0.76–0.91, PWI 0.78–0.98, PPWI 1.37–1.76.

Relatively large species (HW > 1.00, LHT > 0.80). Basal margin of mandible with mesial tooth located notably closer to the apicobasal tooth than to the proximal tooth (MD4/MD5 approximately 0.75–0.80); masticatory margin of mandible with six or (less commonly) seven teeth. Palp formula 6,4. Median clypeal lobe anterolaterally rounded or bluntly angled (not sharply angulate), its anteromedial surface deflected ventrally. Frontal carinae well separated (MFC > 0.10), subparallel or converging slightly posteriorly (ASM ≈ PFC ≥ MFC). Median lobes of antennal sclerites not conspicuously exposed (PFC/ASD 0.67–0.83; FCI2 0.57–0.79). Scape of moderate length for the species group, expanded distally to twice its basal width; funiculus short, also expanded apically, the terminal segment about 1.5–1.6 times the width of the first segment; funicular segment 2 about as broad as long, funicular segment 3 as broad as long or slightly broader than long (FLI > 1.44); funicular segments 4–10 broader than long. Eye small for the genus, of moderate size for the species group (see REL and REL2 values), only moderately elongate (OI 0.58–0.65). Head rather broad (CI > 0.91), the sides convex and either subparallel or diverging slightly posteriorly; side of head rounding relatively abruptly into the posterior margin, which is flat or very slightly concave, in frontal view. Mesosoma dorsum more or less flattened, but mesonotum may be inclined slightly relative to the dorsal face of the propodeum. Metanotal groove deeply incised (MP 0.039–0.093, MPI 0.037–0.075). Dorsal face of propodeum varying from flat to rather convex, subequal to or somewhat longer than the declivitous face (PDI 0.92–1.28), and meeting the latter at a well rounded angle. Metapleural gland bulla well-developed and conspicuous. Legs relatively long, LHT/HL 0.78–0.93. Petiole slightly longer than high (PLI < 0.92), with a weakly differentiated anterior peduncle, followed by an increasingly convex anterodorsal face which rounds into the steep posterior face (lateral view). Summit of node behind the midpoint of the petiole (NI 0.52–0.64). Anteroventral petiolar process well developed, subtriangular in shape and pointing ventrally not posteroventrally. In dorsal view, the minimum (anterior) width of petiole generally less than or equal to half the maximum width (PW13 0.42–0.53), petiolar spiracles forming prominent lateral bulges, and the straight sides of the petiole notably diverging posteriorly. Postpetiole globular, notably broader than long (PPWI >1.35); ventral protrusion of postpetiolar sternite less prominent than in P. concolor; anteroventral process of postpetiole moderately developed. Mandible typically sublucid, with scattered punctulae and fine striolation; the striolae sufficiently dense in some workers to dull the lustre. Head dorsum with numerous fine punctulae (0.005–0.015mm diameter) on a smooth, shiny background. Punctures more or less evenly spaced and separated by about their diameters or less on anterior portion of head, becoming a little less dense on the upper half of head where they are separated by one to several diameters. Mesosoma dorsum sublucid, with numerous piligerous punctures (separated by one to several diameters), side of mesosoma with punctulae less well defined and with a tendency toward the development of corarious-imbricate and imbricate-costulate sculpture, with a corresponding dulling of the integument, especially on the metapleuron and lateral propodeum. Petiole, postpetiole and gaster sublucid, the reflectance dulled (especially on the gaster) by very fine punctulae and associated appressed pubescence. Standing pilosity common and conspicuous on most of body, including scape, mesosoma dorsum, and external faces of tibiae (MSC 43–84, HTC 9–21); erect setae absent from the side of the head, present or absent on the posterior margin. Appressed pubescence well developed on most of body; appressed hairs mostly short and inconspicuous on head, longer and moderately dense elsewhere, including the petiole, postpetiole and abdominal tergite IV where hairs are separated by much less than their lengths. Body varying in colour from uniformly light yellow-brown or orange-brown to dark-brown; head concolorous with or only slightly darker than the mesosoma; gaster may be constrastingly darker than rest of body. Scape, appendages, and frontoclypeal complex tending to be lighter in colour than rest of body.

Type Material
Ward (1999) - Syntype workers, Brazil, Bolivia, Peru (not located; see below). Neotype worker here designated: Estacion Biologica Beni, 42 km E San Borja, Depto. Beni, Bolivia, 210 m, 14°48’S, 66°23’W, P. S. Ward# 9075.

Pseudomyrma arborissanctae Emery, 1894:147. Syntype workers, queens, Bolivia (Balzan) [Examined].

Pseudomyrma triplarina var. rurrenabaquensis Wheeler & Mann, 1942b:188. Syntype workers, Rurrenabaque, Bolivia (W. M. Mann) (, MCZC) [Examined].

Pseudomyrma arborissanctae var. ecuadorianus Enzmann, 1944:79. Three syntype workers, Ecuador (MCZC).

Weddell (1850) introduced the name Myrmica triplarina in a footnote to a paper containing descriptions of new species of Triplaris and other plants. No type material is known to exist for P. triplarinus and Weddell’s herbarium specimens of Triplaris apparently do not contain ants (J. Brandbyge, pers. comm.). Hence I am designating a neotype which agrees with Weddell’s original (albeit scanty) description of the ant, and which corresponds to the species which has come to be known in the literature as P. triplarinus.