Azorsa, F., 2008:
(figs. 2, 4, 6, 8-16, 20-22)
Holotype worker. Measurements (mm): EL 0.66, GL 3.29, HL 3.21, HW 3.47, ML 2.34, PL 1.40, PPL 0.55, PSL 1.05, PW 3.13, SL 2.01, TL 13.9, WL 3.07. Indexes: CI 108, MI 73, PI 46, PSI 34, SI 58.
Polymorphic. Head heart-shaped; wider than long. Mandibles elongate and linear with apical fork consisting of two teeth of which the ventral one is the largest. Mandibles finely reticulate-punctate. Inner (masticatory) margin of mandibles with long hairs, lacking any short, thick hairs. Outer margin of mandibles with some decumbent hairs. Mandibles, in full-face view, somewhat short and stout [(MW/ML')* 100= 34-44] (fig. 23). Dorsum of clypeus with suberect to subdecumbent hairs. In some intermediate and major castes, dorsum of head with a small but conspicuous ocellus. Ocular carina absent. Ocular crest, in lateral view, with 1-3 erect , simple hairs. Antennal scapes not surpassing the posterior margin of head. Antennal scapes slightly thickening towards the apex, finely reticulate and shiny, and with most of their lengths covered with sparse decumbent to subdecumbent hairs. Base of mandibles, in lateral view, finely reticulate-punctate and ventrally rugose, rest of lateral margin of mandibles smooth and shiny. Sides of head lacking a broad gap between bases of mandibles and margins of head capsule when mandibles fully closed (except in two minors studied, in which case there is a narrow gap). Depressions, adjacent to and ventral to mandibular insertion, shallow (much deeper in D. armigerum ).
Pronotum with a pair of humeral tubercles that are more carina-like and a pair of lateral, single-tipped (rather than bifurcate) spines. Propleuron, in lateral view, strongly angulate. Posterior portion of promesonotum with a pair of low tubercles. Promesonotum with at least two pairs of standing simple hairs. Metanotal groove weakly impressed. Propodeal spines long and somewhat curved inwards (U-shaped) when seen in fronto-dorsal view. Propodeal spiracles appearing, in dorsal view, as lateral prominences of propodeum; opening of propodeal spiracle longer than wide (oval).
Peduncle of petiole long. Anterior-lateral margins smooth and shiny. Dorsum of petiole anteriorly with a pair of small spines that project latero-posteriorly. Disc of petiole finely reticulate-punctate, lacking a second pair of tubercles or spines. Postpetiole, in dorsal view, hexagonal. Disc of postpetiole finely reticulate-punctate and with appressed hairs. Posterior margin of postpetiole, in lateral view, angulate. Dorsum of first gastral segment mainly with subdecumbent to decumbent hairs in addition to some appressed hairs.
Paratype workers. Measurements (mm): EL 0.31-0.79, GL 1.69-3.82, HL 1.08-3.57, HW 1.17-4.13, ML 0.66-2.83, PL 0.67-1.58, PPL 0.24-0.60, PSL 0.28-1.20, PW 0.95-3.60, SL 0.74-2.11, TL 5.68-15.9, WL 1.34-3.55. Indexes: CI 106-116, MI 61-79, PI 41-52, PSI 21-40, SI 51-74, (29 measured).
Holotype worker, Peru: Loreto, Iquitos, ACTS Field Station, Canopy Walkway, 03°15'00'S72°55'12'W , 20-24.iii.2006 (F Azorsa). [Deposited in MUSM .]
Paratypes, 14 workers with same data as holotype; 15 workers, Brazil: Amazonas, Hwy ZF 2, Km 19, ca 60 Km N. Manaus, 02°30'S60°15'W , 16.viii.1979, Terra Firma (T.L. Erwin et al.). [Deposited in BMNH (1), IAvH (1), ICN (1), MCZC (2), MZSP (2), MUSM (6), USNM (16).]
Non-paratypic material examined. 6 workers, Brazil: Amazonas, Manaus, Reserva 41 WWF, iii. 1992 (F.P. Benton) [ CPDC ].
Gyne and male. Unknown.
Range. This species is known to occur in Iquitos, Peru, and Manaus, Brazil.
Etymology. It gives us great pleasure to name this ant in honor of Mr. Barry Bolton for his extensive contributions to the study of ant taxonomy and, especially, to the taxonomy of the tribe Dacetini . His worldwide revision of the tribe Dacetini is a monumental, well-documented work containing well-executed SEM micrographs and a very user-friendly taxonomical key that facilitates identification of the many miniscule, curious species of the tribe.
Habitat . This species seems to be an exclusively canopy-dwelling ant. The Peruvian specimens were collected on a Sloanea sp. (Elaeocarpaceae) tree.
Worker variation. Among the specimens studied, some morphological variation has been documented, including: (i) All castes with sides of head lacking a broad gap between bases of mandibles and margins of head capsule when mandibles are fully closed, with the exception of the two minor workers studied, in which case there is a narrow gap. (ii) Erect hairs on the ocular crest are present in all workers examined. However, the number of hairs varies among specimens. We suspect that these hairs are fragile and can be easily lost, which may account for the variation observed between specimens. This seems also to apply to the standing hairs on the median promesonotum and behind the posterior tubercles of the promesonotum. (iii) Humeral tubercles are strongly reduced, sometimes forming a carina or absent, especially in smaller workers. (iv) The propodeal spines of all of the Peruvian specimens examined converging at the tips (U-shaped, when seen in fronto-dorsal view), whereas in most of the specimens from Brazil the propodeal spines are diverging, more like the state in D. armigerum . (v) Petiolar spines short, almost absent in the smaller castes. The petiolar spines are more developed in the specimens from Brazil.
FIGURE 23. Relationship between ML' (mandible length in full-face view) and MW (mandible width) between Daceton armigerum and Daceton boltoni .
Key to the species of Daceton .
Pronotal humeral spines bifurcate, the anterior tip larger than the posterior one (fig. 3). First gastral tergite without standing hairs (sometimes with very short, appressed hairs) (figs. 3, 7, 17-19). Mesonotum and metanotum divided by a strongly impressed metanotal groove (fig. 3). Inner (masticatory) margin of mandibles with a row of short thick setae (fig. 5)............................................................. Daceton armigerum .
-. Pronotal humeral spines long and simple (figs. 4, 10, 14). First gastral tergite with suberect to subdecumbent hairs (figs. 4, 8, 11-12, 14-15, 20-22). Mesonotum and metanotum divided by a weakly impressed metanotal groove (figs. 4, 10, 14). Inner (masticatory) margin of mandibles lacking a row of short thick setae (fig. 6)....................................................................................................................... Daceton boltoni .
The worker caste of Daceton boltoni shares many important character states with that of its sister species D. armigerum , including the heart-shaped head, the large eyes located on a low cuticular prominence (Bolton 2000), the number of apical mandibular teeth, and general habitus (figs. 1-8). Daceton boltoni differs from D. armigerum by the absence of a specialized row of thick setae on the inner (masticatory) margin of the mandibles; by mandibles that are slightly shorter and more stout, which could indicate differences in prey preferences between the two species (B. Bolton, pers. comm.); by a broad gap, when seen in profile, between the bases of the fully-closed mandibles and the margins of the head capsule; by shallow depressions adjacent to and ventral to the mandibular insertions; by long and simple lateral pronotal spines; by a weakly impressed metanotal groove; and by subdecumbent to decumbent hairs on the tergite of abdominal segment IV.
Behaviorally, D. boltoni appears to be very similar to D. armigerum . However, drop tests conducted at the type locality indicate that D. boltoni individuals exhibit weak and inconsistent aerial gliding behavior relative to those of D. armigerum (S.P. Yanoviak, pers. comm.).
These characters strongly suggest that D. boltoni is a distinct species rather than a variety of D. armigerum . These character states are consistent across all 30 specimens examined from two distant localities in South America (Iquitos, Peru, and Manaus, Brazil) where both species co-occur. No intermediate forms were observed to suggest that the two forms are conspecific. Rather, D. boltoni is sympatric with D. armigerum . Although its known distribution is currently only two locations in the Amazonian forest, it is possible and indeed likely that D. boltoni shares a broadly overlapping distribution with D. armigerum .
The discovery of a new species in the heretofore monotypic genus Daceton , a widely distributed genus of large and conspicuous ants occurring in most South American rainforests, suggests the possibility of a similar pattern in two other ant genera currently regarded as monotypic and likewise widely distributed, Paraponera Smith and Gigantiops Roger. Individuals of these monotypic genera are, like those of Daceton , large and conspicuous. It is well worth considering that, as with Daceton , these features may have blinded myrmecologists to the possibility that these "monotypic" genera consist in reality of multiple cryptic species.