Lectotype designation and redescription of four commonly collected Neotropical species of Strumigenys (Hymenoptera: Formicidae)

In 1887, Mayr described four species of the ant genus Strumigenys collected in the Brazilian state of Santa Catarina: Strumigenys unidentata, Strumigenys subedentata, Strumigenys denticulata, and Strumigenys crassicornis. All of them were described based on a series of one to several specimens, without designation of a holotype, as usual at that time. The same can be said about Strumigenys eggersi, described by Emery in 1890 based on specimens collected in Saint Thomas (U.S. Virgin Islands), without designation of a holotype. In 1961, Brown designated a lectotype for S. unidentata and synonymized it under Strumigenys louisianae. However, the specimens belonging to the type series of the other four species remain as syntypes. Considering that these are four of the most frequently collected species of Strumigenys in the Neotropical region, in this work we provide lectotype designations and complete redescriptions for S. crassicornis, S. denticulata, S. eggersi, and S. subedentata to ensure the taxonomic stability of these names.


Introduction
The hyperdiverse Strumigenys Smith, 1860 is a cosmopolitan genus of ants with more than 850 described species (Bolton 2021), particularly noted for its variation in mandibular morphology (Bolton 2000;Baroni Urbani & De Andrade 2007;Booher & Hoenle 2021;. Despite having many species occurring in temperate ecosystems, the genus is most diverse in the tropical rainforests around the world (Bolton 1998). In those regions, Strumigenys is often an abundant component of the litter community (Ward 2000).
In terms of regional diversity, the Neotropical region has more than 200 valid species, with possibly several more to be described (Silva & Feitosa 2019;Bolton 2021). Among them, some species belonging to the gundlachi group (Bolton 2000) are considered the most common ant taxa from the Neotropical Strumigenys fauna. During the work for a catalogue of the Neotropical fauna belonging to the genus, we found some concerning issues related to type designation in four extremely common Neotropical species from the gundlachi group: Strumigenys crassicornis Mayr, 1887, Strumigenys denticulata Mayr, 1887, Strumigenys eggersi Emery, 1890, and Strumigenys subedentata Mayr, 1887 Issues date back from their original description. Three of the four abovementioned species (i.e., S. crassicornis, S. denticulata, and S. subedentata), along with the now synonymized Strumigenys unidentata Mayr, 1887, were described by Mayr (1887) based on several specimens collected by Alfred Hetschko at an unknown locality in the southern state of Santa Catarina, Brazil. In his work, Mayr established the syntypes for the new species based on the specimens studied. Later, Brown (1961) designated a lectotype for S. unidentata and synonymized it under Strumigenys louisianae Roger, 1863. However, no other study has investigated the type status or designated lectotypes for the other three species. In the case of S. eggersi, Emery (1890) provided the description of this species, along with several other species and variants of Strumigenys collected in continental and insular Mesoamerican territories, especially from Costa Rica and Saint Thomas, and did not designate any type category for the specimens used in the description. This has serious implications for the establishment of accurate species boundaries for S. crassicornis, S. denticulata, S. eggersi, and S. subedentata, affecting subsequent revisionary works.
In this work, we provide lectotype designations for S. crassicornis, S. denticulata, S. eggersi, and S. subedentata along with redescriptions based on the designated lectotype for each species and commentaries on morphological variation and distribution of recently collected specimens from Brazil.

Material
The material upon which this study is based is located at the following institutions: Collection curators were contacted indicating that lectotypes and paralectotypes would be designated in their respective institutions. Unique identifi cation codes were generated for each designated specimen per the curatorial requirements of each depository institution; not all specimens received unique identifi ers.

Taxonomic treatment
Species redescriptions are presented in the format of 'Taxonomic treatments' -i.e., sections of a given publication documenting the features and/or distribution of a related group of organisms in ways adhering to highly formalized conventions (Catapano 2019). The taxonomic treatments are comprised of fi ve subsections (based on the types of semantic content by Schulz & Jansen 2013): (i) nomenclature (i.e., species name, species name history, and type designations); (ii) diagnosis (i.e., universal statements describing the defi ning properties shared by all instances of the kind the statement refers to); (iii) measurements and description (i.e., assertional statements that document empirical observations about particular entities); (iv) comments (i.e., contingent statements of sorts, that ascribe predicates to a class that may or may not be true for all its members); and (v) additional material examined.
Diagnoses and redescriptions of the species are elaborated following Oliveira & Feitosa (2019) and Ladino & Feitosa (2020); declarations are elaborated following an anterior-posterior axis of description, with sculpture and setae being described fi rst, followed by body structures. The redescriptions are additions and amendments to Bolton's (2000) abbreviated descriptions and are based on the designated lectotypes, with known variation (including those observed in paralectotypes) for each species being described in the commentaries. Morphological and qualitative (i.e., sculpture and setae shape) terms follow Bolton (2000). Digital color images were obtained from AntWeb.org and their corresponding credits are given in the legend of each fi gure in the following format: attribution, specimen code, photographer, and site address. The measurements and indices used in this study are based on those used by Bolton (2000), Lattke et al. (2018), Booher et al. (2019), Tang et al. (2019), and Brassard et al. (2020). When needed, measurement defi nitions were amended as to provide unambiguous procedural criteria for standardized measurement procedures. Whenever possible, when images were made available, paralectotypes were measured and their respective measurement values are presented after the lectotype measurements. Additional non-type specimens were also measured, and maximum-minimum values are given. Image measurements were taken using the measurement application ImageJ (Rasband 2018). Measurements are expressed in millimeters to three decimal places, while indices are expressed merely as non-metric units.

List of measurements
AB4L = Abdominal tergite IV length: the length of the fourth abdominal tergite in lateral view, measured from the posterior margin of the postpetiole (i.e., the rim of the postpetiolar foramen) to the posterior margin of the tergite DPW = Dorsal petiolar width: the width of the petiolar node measured in dorsal view EL = Eye length: the maximum diameter of the compound eye in lateral view. In cases where unpigmented ommatidia are present (generally located at the outer rim of the compound eye), the measure should be taken considering these structures. In cases where the eye is reduced to a single ommatidium, the maximum diameter of this structure alone should be taken HL = Head length: the length of the head capsule excluding the mandibles, measured in dorsal view in a straight line from the mid-point of the anterior clypeal margin to the mid-point of the posterior cephalic margin. In species where one or both of these margins is concave, the measurement is taken from the mid-point of a transverse line that spans the apices of the projecting portions HT = Head thickness: the thickness of head in lateral view, with maximum distance measured between two parallel lines, one tangent with the dorsal-most point of the head and the other tangent with the ventral-most point of the head. If ventral margin concave upward, then measured from the lower line tangent to the uppermost portion of the curve HW = Head width: the maximum width of the head in full-face view, excluding the eyes ML = Mandible length: the straight-line length of the mandible at full closure, measured in dorsal view from the mandibular apex to the anterior clypeal margin, or to the transverse line connecting the anteriormost points of the clypeus in taxa where the anterior clypeal margin is concave medially PH = Petiolar height: maximum distance measured between two parallel lines, one tangent with the node apex and the other tangent with the ventral-most point of the petiole in lateral view. If ventral margin concave upward, then measured from the lower line tangent to the uppermost portion of the curve. If present, spongiform processes ignored PL = Petiolar length: the length of the petiole in lateral view, measured from the propodeal foramen to the posterior margin of the petiole (i.e., the rim of the petiolar foramen).
HEAD. Masticatory margin of mandible with three to fi ve denticles between apicodorsal tooth and submedian tooth, with two to four denticles proximal of submedian tooth ( Fig. 2A). Apex of mandible with unknown number of intercalary denticles (but see Fig. 2A and Comments section for variation). Anterior clypeal margin, in dorsal view, slightly angular and projecting anteriorly. Eye, in lateral view, with four to fi ve ommatidia in longest row. Eye on anterior half of head. In dorsal view, scape narrows basally; anterior margin expanded and almost lobate near subbasal bend. Third fl agellomere smaller than fourth fl agellomere; length of former only a third of length of latter.
MESOSOMA. Humerus with small angular projection. Dorsum of mesonotum, in lateral view, slightly higher than dorsum of pronotum. Metanotal groove weakly impressed. Propodeal spine relatively long and triangular, linked to propodeal lobe by narrow lamella that extends throughout propodeal declivity. Femoral bulla ovate and located distally on dorsal margin of sclerite.
METASOMA. Petiolar node, in dorsal view, slightly wider than long; in lateral view, anterior margin slightly longer than dorsal margin. Postpetiole, in lateral view, swollen and globular. Anterior margin of postpetiole, in dorsal view, medially concave. Ventral spongiform process of petiole absent. Ventral spongiform lobe of postpetiole minute to absent (Fig. 2B). Lateral spongiform lobe of postpetiole reduced to a narrow lamella (Fig. 2B). Ventral basigastral spongiform pad (= specialized setae on fourth abdominal sternite) small.
According to Bolton (2000), different series of this species show slight variation in setae and sculpture, although maintaining the diagnostic traits for the species. He mentioned that some specimens possibly have short fi liform humeral setae, although this condition was not observed in the type specimen and a few other individuals observed in this study. Additionally, a pair of mesonotal erect simple setae, which was not mentioned by Mayr (1887) nor Bolton (2000) in their descriptions, was also observed in the lectotype and a few other specimens. Humeral and mesonotal setae are apparently lost during the lifetime of the ants, since many specimens, otherwise well preserved, did not have those setae and most of the ones which did have them appear to be young adults by the appearance of their cuticle. Also, some specimens had an extremely reduced lateral spongiform lobe in the postpetiole, appearing vestigial, agreeing with the description made by Bolton (2000). In Bolton's (2000) description, the author mentions that S. crassicornis have three to four minute intercalary teeth. Since we did not had access to the physical lectotype specimen, we could not confi rm this condition. However, while studying other non-type specimens, we observed that the intercalary dentition consists of up to six to seven teeth. Interestingly, these intercalary teeth count does not agree with the diagnosis proposed by Bolton (2000) for the crassicornis complex.
In specimens collected in Orleans and Tunas do Paraná (cf. list of examined materials), the katepisternum appears entirely reticulate-punctate, without smooth patches whatsoever. Different specimens collected in the same square meter (from Winkler leaf-litter samples) possess both reticulate-punctate katepisternum and various degrees of smoothness. One specimen from Viçosa (Minas Gerais State) and one specimen from the Reserva Biólogica Augusto Ruschi (Espírito Santo State) also have an entirely reticulate-punctate katepisternum. In Viçosa, all the other specimens examined matched the lectotype in having a smooth katepisternum. The morphological variability in this particular character, summed with the higher intercalary teeth count found in some non-type specimens observed, reinforces the need for a reevaluation of the boundaries of this species.
In the Amazonian region, non-type specimens identifi ed as S. crassicornis tend to depart further from the lectotype, differing in one or more traits, and do not entirely match the species' diagnosis. One specimen from Primavera (Pará State) has shallow reticulation on fourth abdominal tergite. A couple of specimens from Amazonas state (vicinities of Manaus; cf. additional material examined ZF-02) have an almost entirely reticulate-punctate katepisternum and shallow reticulation on fourth abdominal tergite. A specimen from Canarana (Mato Grosso State) has shallow reticulation on fourth abdominal tergite and humeral and mesonotal setae which are not simple, but slightly fl attened and subfl agellate. A couple of specimens, also from Primavera, have shallow reticulation on fourth abdominal tergite, basigastral costulae absent, spongiform process on postpetiole absent, and smaller and less abundant metasomal erect setae. Finally, one specimen from "Jaci Novo'' (Rondônia State) is much larger than all examined specimens, has both humeral and mesonotal setae fl attened and subfl agellate, and a comparatively larger postpetiole, with shallow reticulation on fourth abdominal tergite. Mayr, 1887 Figs 3-4

Diagnosis
Strumigenys denticulata can be distinguished from other local species by the combination of long mandibles, fl agellate humeral setae (Fig. 4B), reduced postpetiolar spongiform projections (Fig. 4B), and fourth abdominal tergite mostly smooth.

Description
SCULPTURE. Head entirely reticulate-punctate, including antennal scrobe. Mesosoma entirely reticulatepunctate, except for katepisternum and part of metapleura, which are smooth (Fig. 4B). Fourth abdominal tergite superfi cially reticulate-punctate near base. Length of basigastral costulae, in dorsal view, more or less equal to length of postpetiole.
HEAD. Masticatory margin of mandible with 5-10 preapical denticles (Fig. 4A). Apex of mandible with two minute intercalary denticles (Fig. 4A). Anterior clypeal margin, in dorsal view, convex medially. Eye, in lateral view, with three to four ommatidia along longest row. Eye on anterior half of head. In dorsal view, scape cylindrical. Third fl agellomere smaller than fourth fl agellomere; length of former only one-third of length of latter.
MESOSOMA. Humerus with small angular projection. Dorsum of mesonotum, in lateral view, convex, confl uent with dorsum of pronotum. Metanotal groove weakly impressed, almost absent in lateral view. Propodeal spine relatively long and triangular, linked to propodeal lobe by narrow carina that extends throughout propodeal declivity. Femoral bulla small, ovate and located distally on dorsal margin of sclerite.

Comments
According to Bolton (2000), specimens belonging to S. denticulata have a wide range of mandibular length variation, with individuals collected in a single leaf litter sample presenting MIs ranging from 72 to 85. It is important to notice that the MI of the lectotype falls near the maximum value established by Bolton (i.e.,MI 84.9). On the other hand, ML and HL measurements fall well within the range proposed by the same author as diagnosable for the species (i.e., ML 0.355 and HL 0.418). According to Bolton (2000), some specimens identifi ed as S. denticulata have the katepisternum entirely reticulate-punctate, while all specimens observed in this study (cf. examined material) have a smooth patch in the katepisternum, including the lectotype. All specimens observed had the fourth abdominal tergite mostly smooth, only with the base of the sclerite with reticulate-punctate sculpture.
One of the paralectotypes (BMNH(E)1013551) is missing the postpetiole and gaster, rendering it impossible to evaluate morphological variability of those body regions in this particular individual. Emery, 1890 Figs 5-6

Diagnosis
Strumigenys eggersi mostly resembles S. denticulata and can be distinguished from this species by the combination of shorter mandibles (ML 0.288 and MI 65.6 from the former opposed to the ML 0.355 and MI 84.9 from the latter), humeral setae fi liform (Fig. 6B), presence of a pair of erect setae in the anteromedial area of the pronotum (Fig. 6B), and fourth abdominal tergite mostly sculptured.
SETAE. Cephalic ground-setae remiform (Fig, 6A). Two pairs of remiform erect setae on cephalic dorsum; both pairs located in the posterior third of cephalic dorsum, one pair nearer to occipital margin than the other. Apicoscrobal setae fl agellate (Fig. 6A). Anterior margin of scape with one or more remiform setae curved towards antennal insertion. Humeral setae fi liform (Fig. 6B). Pair of erect setae located in the antero-medial area of pronotum fi liform (Fig. 6B). Setae on petiole, postpetiole and fourth abdominal tergite remiform to slightly clavate.
HEAD. Masticatory margin of mandible with six preapical denticles (Fig. 6A). Apex of mandible with two minute intercalary denticles (Fig. 6A). Anterior clypeal margin, in dorsal view, convex medially. Eye, in lateral view, with three ommatidia in the longest row. Eye located in the anterior half of head. In dorsal view, scape cylindrical. Third fl agellomere smaller than fourth fl agellomere; length of former almost a third of length of latter.
MESOSOMA. Humerus with a small angular projection. Dorsum of mesonotum, in lateral view, convex, confl uent with dorsum of pronotum. Metanotal groove relatively well impressed. Propodeal spine This species is widespread in the Neotropics, with its northernmost range in Florida, USA (Wetterer 2018) and southernmost range in Santa Fé, Argentina (Vittar & Cuezzo 2008). According to Wetterer (2018), this species can be commonly found in urban areas when occurring outside its original range, which was, according to Brown (1960), "probably south Brazil and Bolivia, though a lack of collections from central and northern Brazil prevents us from knowing how far north this species extends". Since Brown's (1960) work, the number of records of S. eggersi in the Neotropical region has greatly increased, especially due to recent sampling efforts conducted in ecosystems both within and adjacent to the Amazon basin. Nonetheless, there still remains a large record gap for the species in the center of the Amazon biome. Although the revision of the species was not the aim of the present work, it is important to consider that the continuous reduction of this 'distribution gap' is fundamental when addressing the specifi c boundaries for S. eggersi.
Among the specimens examined, dentition patterns varied greatly, both in size and number. In the lectotype, there are total of seven preapical teeth restricted to the distal third of the inner margin of the mandible, while in some other specimens observed there are fi ve or six preapical teeth. Bolton (2000) mentioned that specimens belonging to S. eggersi can have four to eight teeth in the inner margin of the mandible, indicating that teeth variation is expected in this species. However, Longino (2006) provided an important account on the usefulness of teeth variation in demarcating different species of Strumigenys belonging to the gundlachi group, indicating that dental variation (number and relative size of teeth) should be carefully considered when discriminating potential new species in this group. Mayr, 1887 Figs 7-8  (Fig. 8C). Length of basigastral costulae, in dorsal view, almost half length of postpetiole.
HEAD. Masticatory margin of mandible with fi ve preapical denticles ( Fig. 8A; see Comments). Apex of mandible with two minute intercalary denticles (Fig. 8A). Anterior clypeal margin, in dorsal view, angular medially. Eye, in lateral view, with fi ve to seven ommatidia along longest row. Eye located on anterior half of head. In dorsal view, scape dorsoventrally fl attened, expanded throughout its length. Third fl agellomere smaller than fourth fl agellomere; length of former almost half length of latter.
MESOSOMA. Humerus with small angular projection. Dorsum of mesonotum, in lateral view, convex, confl uent with dorsum of pronotum. Metanotal groove relatively well impressed. Propodeal spine long and triangular, translucent, and linked to propodeal lobe by narrow lamella that extends throughout propodeal declivity. Femoral bulla small, ovate and located distally on dorsal margin of sclerite.
METASOMA. Petiolar node, in dorsal view, as long as wide; in lateral view, anterior margin almost as long as dorsal margin. Anterior margin of postpetiole, in dorsal view, slightly concave, almost straight. Lateral and ventral spongiform processes of petiole absent. Ventral spongiform lobe of postpetiole small. Lateral spongiform lobe of postpetiole small. Ventral basigastral spongiform pad reduced to curved (U-shaped in anterior view) carina.

Comments
Bolton (2000) considered S. subedentata as a member of the gundlachi complex (see Comments in S. denticulata). The lectotype and the additional examined specimens did not show major morphological variation. Most specimens have an entirely reticulate-punctate katepisternum; however, some of them have a small but distinct smooth patch on the ventral-most area of the katepisternum. Although it is not possible to clearly determine if the katepisternum of the lectotype has a smooth patch based solely on images, direct observation of the physical specimen showed that this anatomical region is entirely sculptured (G. Broad pers. comm.).
In the lectotype specimen both mandibles possess fi ve preapical teeth each (G. Broad pers. comm.), while in almost every non-type specimen observed in this study this number varied, in a symmetrical manner, from four to seven in each mandible. However, in a few specimens, we observed asymmetric variation in dentition count, which, despite not being a prevalent condition, it is a noteworthy variation.
Despite the wide range of occurrence of the species, the observed morphological traits did not presented a high variability degree, agreeing with Bolton's (2000) own observations. Nonetheless, examination of specimens sampled in other ecosystems within the Neotropical region will provide a clearer understanding of morphological variation along the distribution range of S. subedentata, allowing for a better delimitation of the species' boundaries in future revisionary endeavors.

Discussion
With the overall increase in representation of these four Strumigenys species in myrmecological collections, the morphological variability described by Bolton (2000) is becoming frequently documented, urging for a reassessment of the limits for these species. Adjusting and normalizing nomenclatural acts, along with their associated criteria, is necessary to provide an objective assessment of the defi ning features of any given taxon and for providing support for validation of taxonomic identities. Future revisionary efforts should consider expanding the specimen distribution coverage as to better explore morphological variability in a broader geographical perspective, especially considering how several locations that were considered to be knowledge gaps are now better represented in collections. Nonetheless, with the increased deforestation rates occurring in localities that hold an important part of biological diversity, especially in the Brazilian Amazon Forest, it is possible that much of this knowledge is at risk of disappearing.
With the lectotype designations for S. crassicornis, S. denticulata¸ S. eggersi, and S. subedentata, we take a step back but two steps forward in establishing clearer boundaries for some of the most common Strumigenys species in the Neotropical region.