Crematogaster stollii Forel

Longino, J. T., 2003, The Crematogaster (Hymenoptera, Formicidae, Myrmicinae) of Costa Rica., Zootaxa 151, pp. 1-150 : 109-114

Crematogaster stollii Forel 1885

Plate 3, 4

Crematogaster stollii Forel , 1885:373. Syntype workers: Guatemala, Retaluleu (Stoll) [ MHNG] (examined). Forel, 1899:83: description of queen. Santschi, 1918:182: combination in C. (Neocrema) . Santschi, 1922:244: combination in C. (Oxygyne) . Emery, 1922:136; Kempf, 1972:89: combination in C. (Orthocrema) .

Crematogaster stollii var. amazonensis Forel , 1904b: 682. Syntype workers: Brazil, Amazonas, Cachveira Jurua , in Cecropia No.5587 (Ule); and Pará ( Göldi) [ MHNG] (examined). Emery, 1922:136: combination in C. (Orthocrema) . NEW SYNONYMY

Crematogaster (Crematogaster) stolli subsp. autruni Mann , 1916: 441. Syntype workers: Brazil, near Manaos (Mann) . Emery, 1922: 136: combination in C. (Orthocrema) . NEW SYNONYMY

Crematogaster stolli var. guianensis Crawley , 1916:369. Syntype workers: Guyana (G. E. Bodkin) . Emery, 1922:136: combination in C. (Orthocrema) . NEW SYNONYMY

Tranopeltoides parvispina Wheeler, W. M. 1922a:12, fig. 3. Holotype queen: Guyana Kaieteur , 8 Aug 1911 (F. E. Lutz) . Kempf, 1960:174: junior synonym of stollii .

Crematogaster hyperphyes Kusnezov , 1953:215. Syntype worker, queen, male: Bolivia, Chapare . Kempf, 1968:384: junior synonym of stollii .

Range

Guatemala to Amazonian Brazil and Bolivia.

Description of worker

Color red with darker gaster, legs usually contrastingly lighter yellow brown; workers with pronounced size polymorphism.

In face view head subquadrate, wider than long in larger workers, with emarginate posterior margin; mandibles coarsely striate, shining; clypeus relatively flat, often with concave anteromedian impression, mostly smooth and shining with variable degree of weak etching laterally, frontoclypeal suture broad, deeply impressed, smooth and shining, such that posterior clypeus appears slightly raised relative to face; malar spaces longitudinally striate, striae forming concentric rings around antennal insertions, grading to fine longitudinal striae posteriorly, then to smooth and shining on posterior face, extent of striation variable; face with 10-20 medium length erect setae, amber colored, slightly stiff, underlain by sparse appressed pubescence; scapes with subdecumbent to suberect pubescence, largest workers often with 1-3 long erect setae; scapes short, in face view not attaining posterior margin of head when laid back; terminal four segments of antenna gradually lengthening and broadening, becoming increasingly densely pubescent, terminal two segments may be relatively more differentiated than proximal segments, thus club may appear two-segmented or indistinctly 3-4 segmented.

Promesonotum forming a single convexity in small workers; in large workers mesonotum elevated, differentiated from pronotum, flat to slightly concave, forming a slightly elevated boss at promesonotal suture, sloping posteriorly to broadly V- or U-shaped propodeal suture; posterior wall of propodeal suture steep, rising to long sloping declivity formed by posterior face of propodeum, posterior face usually sculpturally differentiated into anterior and posterior halves, with the anterior portion sculptured and flat, the posterior portion smooth, shiny, and concave; the anterior portion sometimes slightly elevated, meeting posterior half at broadly obtuse angle, thus forming weakly differentiated dorsal and posterior faces; propodeal spine short, broad-based, upturned, shorter in Costa Rica than in South America; propodeal spiracle very large, extending from base of propodeal spine to upper edge of metapleural gland bulla, with pronounced swollen ring around orifice; sculptural details highly variable within populations and across range, in Costa Rica pronotal dorsum irregularly punctatorugose and sublucid to subopaque, in parts of South America becoming finely striate and subopaque, striae longitudinal or variously oblique or whorled; in Costa Rica side of pronotum with faint microsculpture, sublucid, in parts of range in South America subopaque; in Costa Rica katepisternum and side of propodeum with irregular longitudinal rugae, interspaces smooth and shiny or weakly punctate, in South America rugae become less prominent, punctation more so, some specimens evenly punctate, opaque, with no rugae; area around propodeal suture coarsely rugose, lateral carinulae lacking in Costa Rica, becoming more developed in South America, but not obscuring propodeal suture in side view; in Costa Rica anterior half of posterior face of propodeum weakly punctatorugose, sublucid, becoming more punctate and subopaque in South America; promesonotal dorsum and dorsal propodeum anterior to spines covered with abundant, medium length amber setae, setae stiff but not flattened; femora and tibiae with long appressed to subdecumbent pubescence, small workers lacking erect setae, large workers with 0-5 differentiated erect setae on front tibiae, 0-2 on middle tibiae.

Petiole in lateral view subtriangular, faintly sculptured if at all, sublucid, ventral margin concave, dorsal margin convex; anteroventral petiolar tooth large, acute, with blunt tip; dorsal face subquadrate, about as long as wide, sides flat to slightly convex, face smooth and shining to faintly microareolate; two or more erect setae projecting from sides, about two posteriorly directed setae on each posterolateral tubercle; postpetiole lacking ventral tooth (a few South American collections with small ventral tooth), in dorsal view wider than long, subquadrate but with strongly convex sides, posterior margin emarginate, variably developed posteromedian sulcus, dorsum faintly microareolate, with five or more erect setae; fourth abdominal tergite smooth and shining or with faint microareolate sculpture, with sparse appressed short pubescence and abundant medium length amber erect setae distributed evenly over surface, erect setae stiff, somewhat flattened, flatter than more filiform setae of mesosoma.

Measurements

HL 1.574, 0.779, 1.555; HW 1.880, 0.860, 1.826; HC 1.782, 0.828, 1.753; SL 0.933, 0.526, 0.878; EL 0.357, 0.183, 0.399; A11L 0.385; A11W 0.209; A10L 0.190; A10W 0.168; A09L 0.121; A09W 0.135; A08L 0.086; A08W 0.116; WL 1.886, 0.858, 1.929; SPL 0.224, 0.081, 0.156; PTH 0.413, 0.180, 0.395; PTL 0.487, 0.229, 0.494; PTW 0.472, 0.245, 0.500; PPL 0.392, 0.177, 0.345; PPW 0.514, 0.269, 0.529; CI 119, 110, 117; OI 23, 23, 26; SI 59, 68, 56; PTHI 85, 79, 80; PTWI 97, 107, 101; PPI 131, 152, 153; SPI 12, 9, 8; ACI 0.32.

Queen

A normal queen (dorsal face of propodeum drops steeply from postscutellum and much of propodeum appears ventral to scutellum and postscutellum, Fig. 1) with general shape, sculpture, and pilosity characters of the worker; size characters as in Figures 4 and 5.

Biology

Crematogaster stollii occurs in lowland moist to wet forest habitats throughout the wet Neotropics. It may prefer seasonally dry habitats over very wet forest. For example, in Costa Rica it is relatively common on the Pacific slope but I have never seen it on the much wetter Atlantic slope. Colonies occur in both mature forest and in large trees along fencerows. It can be relatively common in agricultural landscapes where large trees occur along roadsides, pasture margins, and ravines.

Workers vary greatly in size, and the largest workers are the giants among the New World species. Crematogaster stollii has a highly distinctive nesting behavior, unlike any other Crematogaster species known to me. Workers construct carton tunnels that extend from the ground up the trunks of large live trees. The tunnels radiate out in the crown and extend out to the tips of branches. Forel (1885) reported Stoll's observations of the carton tunnels that this species constructs, and the fact that the workers exude a drop of white fluid from the gaster when disturbed. Stoll was never able to find the nest center, but supposed it to be inside the branches. I have made additional observations of stollii , and can expand on Stoll's observations.

The carton material is composed of short, coarse plant fibers. The tunnels are 1-2cm broad and flattened, less than 1cm high. At first glance the tunnels are very similar to the ubiquitous nasutiform termite tunnels that cover tree trunks in the habitats where stollii flourishes. However, close inspection reveals a coarser material with visible plant fragments, rather than the fine grained, mud-like material produced by termites. Occasionally stollii incorporates head capsules of its own dead workers in the carton. Construction is always in the form of tunnels, with no larger globular structures or external nests of any kind.

Colonies are very large, and may cover several contiguous trees. Tunnels always extend down to the ground. Short excavations reveal that the tunnels continue down the trunk and onto large roots well below ground level. Tree species used by stollii are highly varied and include Ficus , Brosimum (Moraceae), Erythrina , Inga , Pithecellobium saman (Fabaceae), Pouteria (Sapotaceae), Apeiba tibourbou , Luehea seemanii (Tiliaceae), Enalagma (Bignoniaceae), and Bravaisia integerrima (Acanthaceae).

I have never seen workers exposed on the surface. Beneath the carton tunnels workers can be found scattered along runways and in clusters in small pits. The pits are excavated a few millimeters into the trunk. The runway floors and the walls of the pits are black and damp, and abundant white entomobryid Collembola are always present. Low densities of coccoid Homoptera occur in the pits and along the runways. Where the tunnels extend out to branch tips, they cover entrance holes into the branch interiors. Branch tips are hollowed out and packed with coccoid Homoptera and workers. Thus a tree that appears clean from the outside, with little surface activity, can actually host an enormous population of Homoptera on the inside of most of its branch tips. Presumably the ants are using the Homoptera as their main food source, but this needs investigation. Like Stoll, I have never found a nest center. In one case I examined a Bravaisia integerrima tree that had been pushed over by a bulldozer. It was covered with the remains of a C. stollii colony. A large knot high up on the trunk revealed that the trunk was hollow, and galleries entered the hollow trunk. Males were present in the knot, and in many small pits beneath the tunnels. I dug into the root ball of the tree and turned up the occasional worker in the soil, but I failed to find any concentration of workers or brood.

When the carton is opened and the workers disturbed, they raise their gasters and exude large, frothy droplets of white liquid. The smell is very acrid and is similar to the odor of other highly chemically defended Crematogaster .

Nuptial flights are presumably nocturnal because queens and males are routinely taken at blacklights. Queens and males have large ocelli, a feature of nocturnal fliers. I am unaware of any collection of queens associated with workers, and the identification of the queens is based on large size and similar morphology to workers.

A striking convergence of nesting behavior is found in the widespread Neotropical ant species Azteca forelii ( Dolichoderinae ). This species makes carton tunnels on tree trunks and branches, and the branch tips are hollowed out and filled with workers, brood, and coccoid Homoptera. Azteca forelii is broadly sympatric with stollii , and usually occurs in the same habitats. The carton tunnels are so similar that I cannot determine which species inhabits a nest without opening the carton and finding workers. Like stollii , the behavior of A. forelii is unique among the diverse genus Azteca .

Mercado (1962) reported an association between C. stollii and the stingless bee Trigona compressa (Latr.) . The study is short and anecdotal, stating "A colony of these bees contains about 10,000 to 15,000 workers. Completely surrounding a natural hive of T. compressa one finds a nest of Crematogaster stolli ants. The anthill has a population twice or thrice that of the bees. Both bee and ant colonies have a single entrance divided by a fragile partition of wax and resins. The entrance for the bees is 4 cm, more conspicuous and 3 cm of diameter. The ants penetrate to the inferior part of the entrance and go down through a net of tunnels, in the tree trunk until penetrating the humiferous crest next to the tree base. [an illustration makes it clear that the net of tunnels is on the tree, not in it]... every colony of Trigona compressa found by myself so far was associated with a colony of Crematogaster stolli . When someone wants to collect the honey of these bees, at the first impact of the ax the ants become cross and practically cover the trunk with an ant sheet, furiously biting the invader. When the natural hive is opened, the ants never touch the spilled honey, nor the brood, neither the bees. If one allows time for them to reorganize they will start all over again, in harmonic association."

Comments

Crematogaster stollii has the habitus of crinosa complex species, with worker polymorphism, a broad head with emarginate posterior border, short scapes, and differentiated appressed pubescence and erect setae on the face. The setae on the mesosoma are longer and more filiform than most crinosa complex species. The giant spiracle is distinctive and alone is sufficient to differentiate workers from any other New World species. The large spiracle may be associated with the closed, high-humidity environment in which stollii workers live. Since workers never forage outside, water loss may never be a problem and thus a small water-conserving spiracle may be unnecessary.

The various subspecies and varieties of stollii have been based on details of color, degree of striation and punctation on the face and promesonotum, and size of propodeal spines. I see no evidence of sharply differentiated sympatric, allopatric or parapatric forms, and I interpret the variation as geographic variation within one polytypic species.