The genus Linepithema was erected by Mayr (1866) for his male-based species L. fuscum. In this study a new species is described also based on male morphology, L. paulistana sp. nov., collected in the city of São Paulo, Brazil, which is attributed to the fuscum group (Formicidae: Dolichoderinae). Linepithema paulistana sp. nov. is the only species of fuscum group present in the eastern part of South America. It is easily distinguishable from the other species of the group because of the presence of a triangular volsellar tooth, which is distally situated between the digitus and the basivolsellar process. By using SEM and optical microscopy, the external genitalia of L. paulistana sp. nov. were analyzed and illustrated and some characters and previous interpretations have been re-evaluated in the Linepithema fuscum group. The male external genitalia are also comparatively analyzed in three species representative of the three Linepithema species groups, those of fuscum, humile, and neotropicum. The present work confirms that the morphological characters of male ants, especially those of male external genitalia, are effective for the identification of genera or species. Given the discrete morphological differences between the external genitalia of the fuscum group and the other species of this genus, a re-evaluation of the generic status of Linepithema is suggested.

External genitalia, Linepithema paulistana, male ants, taxonomic status, winged ants

The Neotropical ant genus Linepithema Mayr, 1866 (Formicidae: Dolichoderinae) includes 20 species that are widely distributed in Central and South America and the Caribbean (Wild 2007; Bolton 2022). Workers of this genus are monomorphic, and the males and queens are winged. The best known and most studied species is the invasive Argentine ant L. humile Mayr, 1868, which shows a cosmopolitan distribution due to its ability to adapt to different climatic conditions and to the plasticity of its reproductive strategies (Ward 1987; Aron et al. 1994, 2001; Aron 2001; Roura-Pascual et al. 2004; Menke and Holway 2006). The genus was erected based on males collected in Lima, Peru, which was described as Linepithema fuscum Mayr, 1866. Shattuck (1992a) subdivided the dolichoderine genus Iridomyrmex Mayr, 1862, transferring the Neotropical species to Linepithema, including L. humile, based on the examination of workers associated with males. In his generic review of Dolichoderinae, Shattuck (1992b) provided a diagnosis of Linepithema for the worker, queen, male, and the larval stage, and divided the genus into two male-based species groups, those of fuscum and humile.

Linepithema was later revised by Wild (2007), who followed Shattuck’s species-group classification and suggested that the fuscum group may be monophyletic, and the humile group may be paraphyletic, as it lacks distinct synapomorphies. Subsequently, Wild (2009), confirmed the monophyly of the fuscum group via molecular phylogenetics, and subdivided the humile group into three species groups: humile group, iniquum group, and neotropicum group. Unfortunately, L. fuscum could not be included in this phylogenetic analysis. According to Wild (2007, 2009), the fuscum group includes seven species, two of which, L. cryptobioticum Wild, 2007 and L. flavescens Wheeler & Mann, 1914, have unknown males. The highly distinctive characters of the males of the Linepithema fuscum group, compared to the other species groups, have been highlighted by both Shattuck (1992b) and Wild (2007), and mainly refer to: i) the submarginal cells of the forewings; ii) the ventral petiolar process; iii) the pygostyle; and iv) the paramere and volsella, which are parts of the external genitalia. Notably, the female castes of L. fuscum remain unknown, with the intersex association based solely on geography (Wild 2007), and no COI DNA barcode analysis was performed (Wild 2009).

In the present study we describe a new species of Linepithema, collected in the city of São Paulo (Brazil), based on males. Due to its genitalic form, we attribute the species to the fuscum group. The morphology of the new species is described by using optical and scanning electron (SEM) microscopy; these are the first SEM illustrations of a fuscum group male and include figures of the external genitalia. The male genitalia are compared to three species, representing three of the four species groups (fuscum, humile, and neotropicum). The morphological interpretation and terminology used for external genitalia in previous Linepithema descriptions is revised. Finally, an updated key to males of the species Linepithema in the fuscum group is presented, and a new species distribution map of the Linepithema fuscum-group is supplied.

The genus Linepithema was designated by Mayr (1866) for males that he described as L. fuscum, for which workers and queens remain unknown. Regardless, males of the fuscum group are easily distinguished from other Linepithema species due to the synapomorphic features indicated by Shattuck (1992b) and Wild (2007). We report these features in Table 1 together with new or updated characters, while in Table 2 we report the male morphometrics for all species of the L. fuscum group with the addition of our proposed new species. Currently, males of the L. fuscum group are known for five out of seven species and are recognizable by the peculiar morphology of the volsella (digitus and basivolsellar process, Fig. 6), which was used as the main species-specific feature by Wild (2007) in his dichotomous key to males. Linepithema paulistana sp. nov. shows all diagnostic features of fuscum group and is distinguished from the other species by the presence of a triangular tooth situated on the volsella distally, between digitus and basivolsellar process (Figs 1G, 4A–F). For this reason, in the present description of L. paulistana sp. nov. we emphasized the morphology of the external genitalia (in particular of the volsella).

The first description of the external genitalia in L. fuscum was made by Mayr (1866), later revised by Emery (1912), who supplied a new description of all the three paired valves, figured by a schematic drawing. Subsequently, the external genitalia of the genus Linepithema fuscum group have been described by Shattuck (1992b) and Wild (2007). According to careful studies of several authors, the volsella is composed of two separate sclerites: the lateroventral parossiculus and the medial digitus (Snodgrass 1941); both sclerites are united longitudinally by a narrow, sclerotized suture (Schulmeister 2001); the parossiculus is a unique sclerite (Schulmeister 2001), consisting in a basal basivolsella, a ventral basivolsellar process and a lateral cuspis (Snodgrass 1941). The distal apex of the cuspis and the ventral and apical basivolsellar process are usually recognizable by the presence of setae (Boudinot 2013). By using SEM and optical microscopy we analyzed the external genitalia of L. paulistana sp. nov. and re-evaluated some characters and previous interpretations in the Linepithema fuscum group.

In L. paulistana sp. nov. we found that: i) the cuspis is very reduced but clearly recognizable as a ridge by the presence of long setae on apical part (Fig. 4C–E), probably due to a lateromedial flattening of the volsella, and not absent as claimed by Shattuck (1992b) and Wild (2007) in the diagnosis of the fuscum group. In our opinion, the presence of the cuspis in males of other species of the fuscum group should be re-considered. ii) The “ventrodistal process” of the volsella described by Wild (2007) is the basivolsellar process, which is reduced and with apical and ventral setae (Fig. 4A–F). The L. fuscum group is, to our knowledge, an exception in the subfamily Dolichoderinae: the paramere shows a telomere very elongated, narrow, ribbon-like and digitiform apically, extending anterodorsally up the proctiger (Fig. 4A–C). Otherwise, the external genitalia of the subfamily Dolichoderinae, based on the diagnosis by Boudinot (2015), is characterized by a paramere, with basimere dorsally and distally developed, and a telomere strongly reduced and lobed, representing a synapomorphy of the subfamily (Yoshimura and Fisher 2011; Boudinot 2015). As for the paramere (basimere+telomere), in the males of the fuscum group, including L. paulistana sp. nov., the term telomere should be preferred to gonostylus, utilized in the previous descriptions (Shattuck 1992b), consistently with the same lobate structures of other dolichoderine males, as suggested by Boudinot (2013).

In order to show the morphological diversity of the male external genitalia within the Linepithema species-groups, we also analyzed and illustrated by SEM the species L. humile (representative of humile group) and L. neotropicum (representative of the neotropicum group), and we compared these species with L. paulistana sp. nov. (representative of the fuscum group). We highlighted, in particular, that: i) in L. humile and L. neotropicum the basivolsellar process is well developed, located ventrally, with lateral cuspis and medial digitus (Fig. 5C–F). In males of the fuscum group, due to elongation and flattening lateromedially of the volsella, the cuspis is reduced laterally to the digitus and the basivolsellar process is reduced ventrally (Fig. 4A–F). ii) The morphology of the dentate valviceps lamina in L. humile and L. neotropicum has a rounded profile dentate ventral edge strongly convex and rounded, while in L. paulistana sp. nov. has a dentate ventral edge straight (Figs 1J, 4D, 5G, H). iii) The proctiger in L. paulistana sp. nov. is well developed (Figs 1G, I, 4A, C), probably representing a morphological character of the fuscum group, given that in L. tsachila and L. fuscum is present (AntWeb 2022). The proctiger is very reduced in the males of the other Linepithema species-groups (Fig. 5C, D). Relevant morphological characters that distinguish the species of the Linepithema fuscum group from the other species of the same genus are reported in Table 1.

Concerning the morphometric characters reported in Table 2, L. paulistana sp. nov. shows several similarities with L. fuscum, but differs in having smaller head (CI), much smaller wings (WL), and longer profemora (FL). However, the possibility that L. paulistana sp. nov. could represent an allopatric population of L. fuscum, with some morphological differences, seems highly unlikely due to: i) different morphology and dimensions of the volsella, an acknowledged species-specific character; ii) the great geographic distance between the species, L. fuscum from the Pacific side and L. paulistana sp. nov. from of the Atlantic side, separated by the Andes Mountains; iii) the different climatic and ecological conditions where the two species live.

As the fuscum group is male-based, the assignment of the species to this group should be verified by using male characters. However, based on worker morphology (Wild 2007), two species of Linepithema with unknown males, L. flavescens and L. cryptobioticum, were assigned to the fuscum group (Wild 2007). For L. cryptobioticum this assignment was supported by a molecular study (Wild 2009). Since the workers of the new species here described are still unknown, we cannot exclude a priori the possibility that males of L. paulistana sp. nov. could be part of the aforementioned species. However, this is unlikely based on the following arguments: 1) The species L. cryptobioticum, L. flavescens and L. paulistana sp. nov. have allopatric distribution: L. flavescens is very far from the other species and shows very localized distribution in the Northern hemisphere, being only present in the Hispaniola Island; L. cryptobioticum is only present in Paraguay; L. paulistana sp. nov. is only present in São Paulo city. 2) The species L. cryptobioticum was included in the molecular phylogeny of Wild (2009) and was found to be a daughter species of L. angulatum. For this reason, we hypothesize that male of L. cryptobioticum is morphologically very similar to that of L. angulatum. Comparing the males of L. angulatum and L. paulistana sp. nov. we find the following distinctive differences that widely reduce the possibility of a close relationship between the two species: the volsella is very different in shape (Fig. 7A, B) and the morphometric dimensions are very different (see for example HL, HW, FL, MML, OI in Table 2). However, we cannot exclude that L. paulistana sp. nov. is a derived offshoot within the “angulatum” cluster. For these reasons, in order to have more clarity about the relationships in this intricate species group, we will have to wait until the discovery of males of L. cryptobioticum, as well as L. angulatum males from other representative localities within the wide range of distribution, spanning from Meso-America to the Pantanal.

Finally, it is remarkable that L. paulistana sp. nov. is the only species of the fuscum group present in the eastern part of South America, extending the former distribution of this species group (Fig. 6), widely distributed i western South America. This finding represents an important range expansion of the genus, which opens the possibility to find additional species of L. fuscum group also in other parts of eastern South America.

The present work confirms that the information on morphological characters of male, especially of male external genitalia, is effective for the identification of ant genera and species, as in most insects (Bolton 2003; Yoshimura and Fisher 2007; Yoshimura and Fisher 2011; Boudinot 2013, 2015; Cantone 2017). We are strongly convinced that a greater knowledge of the male caste would greatly help the identification of most ant species, genera, or species groups, like Linepithema, where workers are very difficult to identify at the species level. Unfortunately, in the family Formicidae the knowledge on the male caste is very limited and the taxonomy is traditionally mainly based on workers. Linepithema is one of the very rare genera described on male caste and allows to describe a new species, like L. paulistana sp. nov., based on male specimens. In our opinion the great morphological differences between the external genitalia of species of the fuscum group and all the other species formerly assigned to the genus Iridomyrmex, like L. humile, and now included in three different Linepithema species groups, suggest that a re-evaluation of the taxonomic status of this genus is urgently needed.

We thank Alexander L. Wild, author of the great revision of the genus Linepithema, Jeffrey Sosa-Calvo, and an anonymous referee who carefully revised the manuscript and greatly improved this paper with their helpful suggestions.