A Descriptive Morphology of the Ant Genus Procryptocerus (Hymenoptera: Formicidae)

Morphology is the most direct approach biologists have to recognize uniqueness of insect species as compared to close relatives. Ants of the genus Procryptocerus possess important morphologic characters yet have not been explored for use in a taxonomic revision. The genus is characterized by the protrusion of the clypeus forming a broad nasus and antennal scrobes over the eyes. The toruli are located right posterior to the flanks of the nasus opposite to each other. The vertex is deflexed posteriorly in most species. An in-group comparison of the external morphology is presented focusing on the workers. A general morphology for gynes and males is also presented. Previously mentioned characters as well as new ones are presented, and their character states in different species are clarified. For the metasoma a new system of ant metasomal somite nomenclature is presented that is applicable to Aculeata in general. Finally, a Glossary of morphological terms is offered for the genus (available online). Most of the terminology can be used in other members of the Formicidae and Aculeata.


Introduction
The genera Procryptocerus and Cephalotes comprise the tribe Cephalotini (de Andrade and Baroni-Urbani 1999. Emery (1922 demonstrated that the tribe possesses the synapomorphic anatomical trait of mushroomhead-shaped proventricular valves (Kempf 1951) (Figure 1). This observation is supported by the studies of de Andrade and Baroni Urbani (1999). Procryptocerus was created by Emery (1887) to include species of Neotropical ants that were considered similar to those of the Paleotropical genus Cataulacus of the tribe Catalaucini (Kempf 1951). Procryptocerus is a lineage composed of about 80 species inhabiting rainforests from the Isthmus of Tehuantepec in Mexico to northern Argentina. Due to their cryptic habits, living inside twigs, these ants are rarely collected (Mackay and Vinson 1989). At present, most species are known from Central America, Colombia and Brazil.
Procryptocerus has been the object of two revisionary studies. Kempf (1951) revised the entire genus and Longino and Snelling (2002) the Central American species. Kempf (1951) recognized 28 species, and 8 subspecies, while for Central America Longino and Snelling (2002) recognized 14 species, described four new species, synonymized two species, and elevated two subspecies to species level. Currently, 56 nominal taxa are included in the genus (Bolton et al. 2006).
Procryptocerus ants possess notoriously variable morphology. Different characters, such as propodeal spine length, form of the clypeus, type of sculpture, and other such characters vary remarkably, sometimes even within the same species. A diagnostic morphology of the genus is proposed to be used as a template for a revisionary study of the entire group following the recommendations suggested by Bolton (2007): "… [not] make an unwarranted assumption that previous authors have already investigated all the useful characters..."; "Initial dependence on previous publications has a strong tendency to restrict the scope of a new investigation…"; "… develop a personal insight into [the] morphology and variation that is not unduly influenced by what has previously been published." Knowledge of morphology and anatomy is incomplete for all species. New characters must be discovered, and old characters tested. Morphological descriptions are thus essential components of our understanding of species and their diversification (Wheeler 2008, Bert Hölldobler, Universität Würzburg, personal communication).
This approach is divided into two sections. In the first part, a diagnosis is presented that expands previous morphological diagnoses of the genus provided by Kempf (1951) and Longino and Snelling (2002). Additional observations, current morphological terminology, and figures are part of this new diagnosis. A unification of the terminology is proposed within ants and with other Aculeata regarding the specialized system used to name the metasomal somites (abdominal (Abd) somites II to pygidium) (see Materials and Methods, including definition of cinctus).
In the second part, a selected Glossary (available online) is offered containing terminology appropriate for Procryptocerus. The provided terminology might also be used for descriptions as well as identification keys of other taxa of ants. Working on the study of morphology and the associated terminology is a constant necessity in order to unify criteria for the basic descriptive work in taxonomy and comparative biology.
In the sense used here, ante means before (anterior to) the referred structure (e.g. antepropodeal refers to a structure anterior to the propodeum), and antero refers to the anterior portion of the actual structure. Anteropropodeum refers to the anterior region on propodeum. Nevertheless, very common literature uses of the prefix "pre" are not changed when, for example, making reference to presclerites as presternite, pretergite, etc.
The use of terminology for shapes is quite useful for describing different structures. A combination of technical and common (not universal) names describing shapes is present in the literature. For instance, terms such as crescentiform, fusiform, disciform, etc. are technical and therefore universal. Terms such as "neck" for a part of the antennae, "cheeks", "apron", etc. are not technical, not universal and therefore are avoided.
Propodeum (Figures 9,10,15,37,38) The propodeum is the first abdominal tergite fused to the thorax, which together comprise the mesosoma. The propodeum is differentiated into the anteropropodeum and the posteropropodeum. In turn, the anteropropodeum is divided into the dorsopropodeum and the lateropropodeum. The dorsopropodeum is the dorsal area of propodeum, anterior to propodeal spines and containing the anteropropodeal processes laterally. The lateropropodeum is the lateral area (laterotergite) of the anteropropodeum containing the propodeal spiracle. The posteropropodeum is located beneath the propodeal spines; it is the posterior vertical or declivitous area of the propodeum. (Figures 37, 38, Table 1). In referring to metasomal somites in ants, usually two different systems are superimposed (Bolton 1994). A general system regards homologous abdominal somites (Abd) throughout the Hexapoda. In the Formicidae, as is it in the entire Apocrita, Abd I is part of the second tagma or the mesosoma, which is formed from the thorax plus Abd I tergite. The remaining abdominal somites form the third tagma starting at Abd II (petiole). A second, specialized (functional) system divides the metasomal somites into a petiole, postpetiole and gastral segments (Bolton 1994(Bolton , 2003. Because different groups of ants contain forms with one-or two-petiolate metasoma, the current specialized system of metasoma nomenclature uses the name "gaster" inconsistently and incongruently with homologous somites in the non-formicid Apocritans. This situation shows that the development of a consistent system of naming specialized metasomata has passed behind the terminology for prosoma or mesosoma. Bolton (1990) introduced the term helcium. A second helcium is characteristic of twopetiolate metasomata; when it is not present, no specialized term is available for Abd III, and hence the term postpetiole is inconsistently used between different castes and subfamilies. Occasionally, authors have had to explain the need of petiole and postpetiole in males without helcial sclerites (see de Andrade and Baroni-Urbani 2003). A simple solution to the inconsistent use of the terminology regarding postpetiole and gastral somites in ants would be to abandon the ! specialized system of metasoma vocabulary. However, the use of a specialized system has shown interesting advantages in the comparative morphology of ants (see Bolton 2003), and is applied in recent classifications (Bolton 2003, Perrault 2004, Ward 2007. Therefore, improvements to the specialized metasomal terminology are desirable. This work proposes a proposal of reconciliation into a single morphological specialized system for what we believe are homologous metasomata within ants and other Aculeata (Table 1).

Metasoma
In Apocrita, the metasoma is composed of Abd II to the caudal segment, or periproct (Snodgrass 1935). Literally, "gaster" means stomach (Brown 1979). However, the word "gaster" has widely been used in the literature of Apocrita referring to the external third tagma minus the peduncle ("petiole" in no ants) Stange 1965, Naumann 1991). The gaster constitutes the abdomen without Abd I (propodeum) and the peduncle of Abd II .
The Abd II (petiole) is always specialized in ants (Bolton 1994). The petiole is also found in some groups of Tiphiidae, Scoliidae, Vespidae, Mutillidae and other Aculeata, where Abd II is a specialized somite as well. The petiole may be nodiform, squamiform or a much reduced subcylindrical segment (Bolton 1994). Pedunculate, sessile and subssesile petioles in ants are, in general, artifacts of the anteroposterior displacement of the petiolar nodus. In groups with a pedunculate petiole (e. g. Pheidole or Solenopsis), the node is posterior. In sessile petioles (such as in Procryptocerus), the tergum is augmented anteriorly (nodal) usually forming an anterior nodal truncation.
The metasoma comprises two general subdivisions: the anterior metasoma constituting the petiole (Abd II), and the posterior metasoma composed of metasomal (mtm) 2 (Abd III) to pygidium. In the posterior metasoma, Abd III-IV (mtm 4-5) constitutes the middle metasoma, Abd V to pygidium constitutes the caudal metasoma. The petiole (mtm 1) contains the cinctus 1 (cinctus: constriction between pre and postsclerites, see below), the mtm 2 contains the cinctus 2, and the mtm 3 may or may not contain the cinctus 3. The petiole (Abd II) and mtm 2 (Abd III) are joined by the first helcium (Bolton 1990) in all ants. In the middle metasoma, the mtm 2 and 3 (Abd III-IV) may or may not be joined by a second helcium. The mtm 2 is usually considered the postpetiole when separated from mtm 3 by the helcium-cintus 3 complex. This state is seen in females of Myrmicinae, Pseudomyrmecinae, and others. In other considerations of a postpetiole, the mtm 2 is separated from mtm 3 by the cinctus 3 only (e.g. see de Andrade and Baroni-Urbani 2003 for Proceratium). This state is known to occur in males of Myrmicinae, Pseudomyrmecinae, and in both sexes of Ecitoninae (including Cheliomyrmicini), Ectatomminae, Amblyoponinae, Cerapachyinae, Ponerinae, Paraponera clavata, Proceratium, and others.
For the posterior metasoma, we propose the use of the name gaster when a postpetiole is not present (e.g. Formicinae, Dolichoderinae), and the use of postpetiole and opisthogaster when at least a cinctus 3 is present. The posterior metasoma is a gaster when its metasomata 3 to pygidium (Abd IV to pygidium) are combined as a unit (e.g. Formicinae, Dolichoderinae); i.e. the gaster constitutes the metasoma minus the petiole when a postpetiole is absent. The opisthogaster constitutes the combined caudal metasomata 4 to pygidium (Abd V to pygidium); i.e. the opisthogaster is the posterior metasoma minus postpetiole. Even in few cases, when it is not clear to associate gaster and opisthogaster to general subfamilies, tribes or genera, the two terms would stay associated to the already explained metasomata.
In summary, we propose a consistent specialized (functional) and morphological system for naming homologous metasomata in ants. When the cinctus 3 is absent in the metasomite 3 (Abd IV), we propose using gaster for the combined metasomata 3 to pygidium. When a postpetiole is considered, due to the presence of at least the cinctus 3, the posterior metasoma is divided into postpetiole and opisthogaster.
Cinctus 1, 2, 3 (pl. cincti) (Figures 3,9,16). Cincti are the anteriad sulci, often in the shape of a belt or girdling constriction, located on metasomal 1 (Abd II = petiole), metasomal 2 (Abd III = postpetiole) and metasomal 3 (Abd IV) somites. The sulci separate the pre and post sclerites. Cinctus 1 is usually a dorsal, very slender sulcus concealed between the posteropropodeal lobes, anterior to the peduncle and sternopetiolar process in petioles possessing these structures. In some groups (e.g. Pogonomyrmex) there is no apparent cinctus 1. In such cases, cinctus 1 is interpreted as the anteriormost portion of the peduncle located between the posteropropodeal lobes. In general, in sessile metasomata (e.g. Procryptocerus) cinctus 1 is located directly anterior to the spiracle. The cinctus 2 is located in metasomal 2, anterior to the ventropostpetiolar ("subpostpetiolar") process in metasomal 2 that possesses such process. The cinctus 3 (the "girdling constriction"-Bolton 1994) is located between the pre and postsclerites of the metasomal 3, posterior to the second helcium (sensu Bolton 1990), in two-petiolate ants. When the cinctus 3 is present in some of the sensu stricto single petiolate ants, it is a very fine sulcus (e.g. different poneroids groups), or almost absent (e.g. Odontomachus). Few genera such as Leptanilliodes possess more than three cincti.

Color
Color usually varies from dark-orange or redbrown appendages, and black meson in some Andean species usually over 600 m of elevation, to completely black in most species found at low elevations. Minor color variations are as follows. Scape and pedicel yellow, orange, red, or brown; eye brown; palps yellow; mandible brown laterodistad; flagellum, tibiae and telotarsi orange-brown; remaining body black.

Torulus vs. Annulus (antennalis)
Torre-Bueno (1989) considers the annulus (antennalis) to be the ring sclerite of the head into which the basal segment of the antenna is inserted. Gordh and Headrick (2002) consider annulus to be the antennal sclerite forming a sclerotized ring on the head into which the basal segment (scape) of the antenna is inserted. We follow Bolton (1994).

Notopropodeal fusion
The mesosoma comprises the thorax plus the propodeum, the tergite of Abd I fused to the thorax. The pleurites and the sternite of Abd I are entirely reduced and the tergite remains. In workers of ants, the notal and propodeal sclerites are usually fused forming a tergal (notal) fusion between the notum and propodeum. This condition is a notopropodeal fusion. Externally, it is usually impossible to recognize the structures involved in the fusion. The line of fusion may be indistinct (notopropodeal fusion usually convex), obsolete or differently marked by a suture, groove, impression, depression, etc. The line of fusion has different names in the literature, such as "propodeal suture" (a suture on the propodeum), "metanotal suture" (a suture on the metanotum), "metanotal groove" (a groove on the metanotum), "metanotal impression" (an impression on the metanotum), "metapropodeal suture" (a suture on the posterior [meta] region of propodeum), antepropodeal suture, metanotal area, etc. Since these terms make reference to the line of fusion, we recommend using the adjective notopropodeal in reference to the line of fusion; for instance: notopropodeal suture, notopropodeal groove ( Figure  3), notopropodeal convexity, notopropodeal impression, notopropodeal excavation, or otherwise make reference to the notopropodeal fusion to describe specific characters, such as notopropodeal fusion flat, notopropodeal fusion convex, etc. In several groups, e.g. some Camponotus, the metanotum is clear and so are the mesometanotal suture and the metanotalpropodeal suture. In those cases, a notopropodeal fusion is not apparent. Occasionally, a mesometanotal fusion or a metanotal-propodeal fusion can also be identified in a very few cases.
Striate sculpturing refers to longitudinal lines on a non-impressed cuticle, running parallel between thin and low elevated costae or costulae; the costae (or costulae) and lines are narrow and about the same width. The sculpture should be named striate-costulate, but it is customary to call it striate. Striate is one of the most common forms of sculpturing in Procryptocerus; it is often present on the metasomal 3 (first opisthogastral) tergite. Costate sculpturing refers to costae (elevated ridges rounded at their edges, dim. costulae) in general running parallel or quasi parallel to each other, the interspaces are wider than costae. Rimose refers to nearly parallel excavations (rimae), often narrow, short or long, in the shape of wavy cracks, running into each other (Gordh and Headrick 2000); elevations between rimae are vermiculate, often wide and flat at their ridges, which are usually micropunctulate. Rimae refer to the longitudinal fissures, crack-shaped interspaces; ridges refer to the elongated elevations (costae). The costae are wavy. The combination between rimae and flat ridges, running in anastomosis, produces rimose (dim. rimulose) or rivose (dim. rivulose) sculpturing. Combinations of character states, shuch as rimose-vermiculate, or rivosevermilaculate could be more specific.
Punctures are slightly impressed points (pricks) on the cuticle that appear to be made by a needle (Gordh and Headrick 2000). Punctures constitute the smallest circularmacrosculpture. Derived adjectives describing this sculpture are punctate (with punctures), puncticulate (sparcely punctate), punctulate (closely punctate). When puncticulate and punctulate are present on the same surface, the difference between the sculptures is clear. When only one is present, the terms are interchangeable. When densely punctate, the cuticle has a farinose texture. Dots (dotted sculpturing) are non-impressed circular marks, they are superficial, rounded, and the same size as micropunctures. A costate integument emphasizes the costae and not the interspaces (striae or sulci); in these cases spaces between costae are in general wider than the costa edges and not impressed. The sculpture is porcate when a set of combinations of costae and impressions between costae are present, forming canalicular (sulcal) spaces. Anastomosed porcae are porcae that run into each other. Scrobiculate are surfaces where scrobiculae (parallel, short porcae) are uniformly organized in a contiguous, chainlike series. When there are septae within striae, the sculpturing is reticulate (quadri-or quasi quadriculate) [belti frons] or areolate (polygonal or quasi polygonal) [scabriusculus frons posteriad]. In a subsecuent stage there are septae within a porcate surface, and the sculpturing is alveolate or clathrate. Alveolae are lacunose, impressed spaces delimited by irregular rugae or "costae" with sharp rims at their edges. The alveolae are regular or irregular polygons, and the sculpturing is called alveolate. Alveolate cuticle is often present on the posterior frons and petiole. Surfaces having alveolate sculpturing in Procryptocerus form landscapes of lacunae between ridges (ranges of "hills") containing sharp or obtuse edges. Clathrate sculpturing refers to the most complex, irregular combination of irregular porcae and transversal septae (crossing costae), forming deep, alveolate, cancelled holes of different and irregular diameter. In clathrate sculpture, the costae run anteroposteriorly in irregular fashion prevailing over the short transverse costae forming the septae. High density of alveolae conform clathrate sculpture. This characteristic is the most important one to differentiate clathrate and reticulate sculpturing, which could be apparently similar when both are present on frons and when looking at them through a common microscope. Clathrate sculpture may be present on the frons and mesonotum [mayri, batesi, clathratus]. Reticulate sculpture is usually present on frons only [belti, convexus, hirsutus].
When a combination of sculptures is present, it is useful to hyphenate two, sometimes three, different words qualifying sculpture (e.g. rugo-costate, costate-foveate, foveatereticulate, micro-striolate-imbricate). In general, when differences between proportions occur, the first word should emphasize the most common sculpture, or emphasize the first sculpture when referring to an anteroposterior (or any directionality) sequence of the sculpture present on any surface. Discriminating thickness of raised sculpture (e.g. costate, carinate) and width of circular, impressed sculpture (e.g. punctate, foveate, foveolate) is often not clear when only one of these types of sculpture is present. In those cases, the closest terms might be used interchangeably (e.g. foveate or foveolate, costate or costulate, rimose or rimulose). Discrimination of those sculptures is easier when several types of sculpture are present in the same area of an ant.
Confusion occurs between sculpture texture (appearance) and sculpture structure (constitution). To recognize the constitution (nature) of the sculpture, textures (e.g. leathery, farinose, rugous, coriarious, corticinus, etc.) should be avoided. Appearance strongly depends upon the "momentary" criterion of the researcher or interpreter, and magnification, light or system (microscope, SEM images, photomontage) used to recognize it. Nonetheless, when using a common microscope or photomontage images, the appearance is sometimes quite distinct with some descriptive forms (e.g. politus, shiny, glossy, farinose). It is best to use SEM images of sculpture (For instance see http://www.evergreen.edu/ants/genera/Antsof CostaRica.html). On the other hand, drawings are the best way to convey information about boundaries of sclerites.
Sculpturing within Procryptocerus is a rich source of characters, which is helpful in stablizing and recognizing the identity of species and could permit the formalization of hypotheses of evolutionary trends.