Crematogaster crinosa complex

Longino (2003) constructed this complex for a group of Neotropical Crematogaster species. They are united by shared morphology, ecology and life history attributes.

Species

 * Crematogaster ampla
 * Crematogaster crinosa
 * Crematogaster rochai
 * Crematogaster torosa
 * Crematogaster vicina

Diagnosis
Throughout the tropical and subtropical Americas, highly insolated habitats are often dominated by Crematogaster with the following characters:


 * Colonies large and polydomous.
 * Workers with pronounced size polymorphism.
 * Head of large workers emarginate posteriorly.
 * Scapes short, not attaining posterior margin of head when laid back (SI < 75).
 * Dorsal face of petiole short, nearly as broad as or broader than long, subquadrate or more often with strongly convex sides, usually widest at midlength.
 * Postpetiole lacking ventral tooth; in dorsal view globular to subquadrate, as wide as or wider than long, and sometimes with faint longitudinal medianimpression.
 * Face with short appressed pubescence and variable number of short erect setae, never with abundant long flexuous setae.
 * Setae of mesosomal and gastral dorsum always short, flattened, and stiff, never long and flexuous.

I refer to these as the crinosa complex, because Mayr's Crematogaster crinosa is the oldest described taxon that exhibits these characters. Habitats where these ants are dominant include mangroves, seasonal dry forests, savanna, the upper canopy of wet forests, scrub and thorn forests, and anthropogenic habitats such as pasture edges, roadsides, and urban areas. The habitats where they are absent or show greatly reduced abundance are wet forest interiors and montane areas. In many ways this group is intermediate between other Neotropical Crematogaster and the Crematogaster s.s. that dominate the North American temperate zone. Wet forest Crematogaster tend to have the petiole elongate, widest posteriorly, and tapering anteriorly. The crinosa complex shows a shortening and broadening of the petiole, and a shift of the widest portion to midlength or even more anteriorly. Crematogaster s.s. continues this trend, such that the dorsal face of the petiole has broad anterolateral lobes, is definitely widest anteriorly, and tapers posteriorly. Many rainforest Crematogaster have the postpetiole globular. The postpetiole of the crinosa complex may be globular or slightly bilobed. The postpetiole of Crematogaster s.s. is strongly bilobed with a pronounced median sulcus. Many rainforest Crematogaster have abundant long flexuous setae on the face and fourth abdominal tergite, heads that are more rounded posteriorly, and long scapes that extend to or beyond the vertex margin. The crinosa complex and Crematogaster s.s. are essentially identical in the short pubescence on the face, the generally sparse erect setae, and the short scapes. Given the prevalence of elongate petiole with posterior node among the most generalized myrmicines, the shortening and broadening of the petiole within Crematogaster is probably apomorphic. Paralleling these morphological transitions is a habitat transition from moderate to increasingly harsh environmental conditions. The pattern may reflect a phylogenetic history in which the crinosa complex was derived from a rainforest ancestor and moved out into open areas, dry areas, and the subtropics. Subsequently, under this scenario, Crematogaster s.s. was derived from a crinosa complex ancestor and moved across the frost line and into the temperate zone. The pattern could also be generated by parallel responses to natural selection, with the similarities between the crinosa complex and Crematogaster s.s. being due to similar adaptive responses to dry conditions. Also, the widespread occurrence of Crematogaster s.s. in the Old World has not been incorporated into this scenario, and begs investigation.

Other Neotropical species that share the general habitus of the crinosa complex but have long flexuous hairs are Crematogaster erecta, Crematogaster moelleri, Crematogaster stollii, and Crematogaster crucis. These are all more associated with rain forest and montane forest, compared to crinosa-group species.

Morphology
Taxonomy within the crinosa complex is difficult. The following is a synopsis of character systems that vary within the crinosa complex:

The face and clypeus may be largely smooth and shining, or with varying degrees of granular to finely longitudinally striate sculpture that extends from the anterior portion of the head. Often the clypeus, malar spaces, and space between eyes and antennal insertions is finely striate. These striae may extend up the sides of the head and medially, to the point where only a thin median strip remains smooth and shiny. In some cases the entire face is uniformly striate, causing a sericeous luster or mat surface.

The position and strength of the promesonotal suture varies. In some cases the promesonotum forms a uniformly curved convex profile, with the promesonotal suture not visible in profile or dorsally. In other instances the mesonotum is enlarged and the pronotal dorsum shortened, such that the remnants of the promesonotal suture can be seen as oblique impressions laterally, but the suture is effaced in dorsal view. In very large workers the suture may be entire but strongly arching anteriorly, such that the pronotal dorsum is very short (approaching the condition in the queen). Finally, the mesonotum may be less enlarged relative to the pronotum, and the promesonotal suture visible as an impression that more evenly divides the promesonotum. In lateral view this results in a promesonotal profile that is more flat-topped.

The dorsal face of the propodeum, which is always very short relative to the posterior face, may drop steeply to the propodeal suture, such that the juncture of the dorsal and posterior faces is strongly produced as an anteroposteriorly compressed ridge. In some cases the dorsal and posterior faces meet at an obtuse angle, such that in lateral view there is a distinct v-shaped propodeal suture, a subhorizontal dorsal face, then a sloping posterior face. In some cases the dorsal and posterior faces are in the same plane, such that the propodeal suture appears very shallow with no posterior wall, and the propodeum forms a single declivity from the propodeal suture to the petiolar insertion. The transition is like a wrinkle in a rug being gradually smoothed out.

The anteroventral petiolar tooth may be long, narrowly acute, and sharp. If long and sharp, it may project anteriorly, with the entire ventral margin of the petiole, including the tooth, in the same plane, or the ventral margin of the tooth may curve resulting in an upwardly concave ventral margin. Alternatively, the tooth may be reduced to a nearly right angle.

The fourth abdominal tergite may have an even vestiture of erect setae over the entire surface. In other cases the setae are clustered anterolaterally, often leaving a medial strip bare. The tergite may be completely bare, or with only 1-3 setae anterolaterally.

The above character variation is continuous rather than discrete, and all combinations of characters seem to occur. However, variation is not random, and clusters of correlated characters occur that suggest distinct sympatric species. Within Costa Rica, for which abundant material is available, I have come to the conclusion that there are three sympatric species: crinosa, Crematogaster torosa, and Crematogaster rochai. However, occasional specimens exhibit combinations of characters that blur the distinctions. I have not discovered characters that are uniformly diagnostic. These three "character clusters" seem to occur widely in the Neotropics. For example, abundant collections from southeast Texas in the USA almost all match torosa, but one collection exhibits all the characters of crinosa, suggesting that two species are sympatric there. However, among the other collections I have been able to examine from scattered localities in the Neotropics, many show combinations of characters that do not match one of these three. As in other complex taxa, there is probably a combination of widespread lineages and narrow endemics. Additional collections and character analysis will be necessary to gain a clearer picture.

Additional forms
Given the ecological prevalence and the large character variation in the complex, many taxonomic names have been generated. Besides the more clearly delineated species listed above, there are numerous other names that are associated with this complex. I summarize below the available names I associate with the crinosa complex (or remove from the complex). All require further taxonomic study.

Forms with completely striate face:
 * Crematogaster atra
 * Crematogaster sericea
 * Crematogaster uruguayensis
 * Crematogaster peristerica

It is unclear whether the variation in face sculpture from smooth and shining to completely striate represents intraspecific or interspecific variation. It may be that selection gradients cause parallel changes in face sculpture among multiple sympatric lineages, such that the above forms may be synonyms of crinosa and torosa. For example, I have examined five collections from Colombia that have completely striate faces. Other than the striate face, one collection has the characters of crinosa, three have the characters of torosa, and one has the characters of rochai. Alternatively, further research may reveal that the forms with completely striate faces are found in sympatry with shiny faced forms, or show discontinuous character variation, such that they are best interpreted as distinct species. If so, it must then be determined whether these forms are distinct from each other or should be interpreted as one widespread species.

Forms with deep propodeal suture:
 * Crematogaster goeldii
 * Crematogaster stigmatica
 * Crematogaster heathi
 * Crematogaster chodati
 * Crematogaster unciata

The distinctive feature of these forms is a deeply invaginated propodeal suture and an elevated and compressed dorsal face of propodeum. From my cursory examination, goeldii, stigmatica, and unciata were very similar. They were bicolored, with red brown head and mesosoma and contrasting black gaster. The anteroventral petiolar tooth was moderately well developed and the pilosity was much like torosa. The types of chodati were uniform red brown, and the setae on the mesosoma and fourth abdominal tergite were longer and less flattened. The types of heathi had the coloration of goeldii and the pilosity of chodati. The invaginated propodeal suture is itself a variable character, grading into the condition seen in crinosa and torosa, and so the discussion regarding the forms with striate face applies here as well. Crematogaster goeldii and related forms may be synonyms of other species, multiple distinct species, or one variable species.

Crematogaster rochai-like forms:
 * 4 Crematogaster brevispinosa quadrinomials (unavailable names)
 * Crematogaster russata
 * Crematogaster subtonsa

Other unresolved crinosa complex names:
 * Crematogaster brevidentata
 * Crematogaster thalia
 * a Crematogaster brevispinosa quadrinomial (unavailable name)
 * Crematogaster montana