Lasius Fabricius, 1804

Boudinot, Brendon E., Borowiec, Marek L. & Prebus, Matthew M., 2022, Phylogeny, evolution, and classification of the ant genus Lasius, the tribe Lasiini and the subfamily Formicinae (Hymenoptera: Formicidae), Systematic Entomology 47, pp. 113-151 : 135-137

Genus Lasius Fabricius, 1804

= Donisthorpea Morice and Durrant, 1915.

= Acanthomyops Mayr, 1862 syn.rev.

= Austrolasius Faber, 1967 syn.nov.

= Cautolasius Wilson, 1955 syn.nov.

= Chthonolasius Ruzsky, 1912 syn.nov.

= Dendrolasius Ruzsky, 1912 syn.nov.

Type species. Formica nigra Linnaeus, 1758 (= L. niger ).

Subgeneric classification remarks. The modern body of taxonomic work on Lasius was initiated by Wilson’s revision of the genus ( Wilson, 1955), which was classified into four subgenera at the time: Cautolasius , Chthonolasius , Dendrolasius , and Lasius s. str. In this work, Wilson provided a phylogeny of Lasius ( Fig. 1A ), but this treatment was an intuitive account of what he considered trends in the evolution of the morphology and biogeography of the genus. Subsequently, the subgenus Austrolasius was erected for a few socially parasitic species ( Faber, 1967), andthe former genus Acanthomyops was included in Lasius as the sixth subgenus ( Ward, 2005).

The first phylogenetic inference based on molecular data was presented by Hasegawa (1998), who used COI to investigate relationships of four Lasius species (phylogeny not figured here). Since then, two major attempts at resolving the phylogeny were presented by Janda et al. (2004) and Maruyama et al. (2008) ( Fig. 1B, C ). Both studies used a combination of morphological characters and molecular data, including mitochondrial markers (COI, COII, tRNA-Leu, 16S). Amore recent effort focused on the phylogeny of European species related to L. niger ( Lasius s. str.) and included nuclear genes LW Rh and wg in addition to 16S and COI ( Talavera et al., 2015).

To date, the monophyly of the subgenera has not been questioned except for the nominotypical subgenus ( Janda etal., 2004; Maruyama etal., 2008). Recently, Seifert (2020, p. 21) stated that there is clear justification for elevating the subgenera to generic status. Such an action would considerably complicate the classification of the Lasius genus group because of the robustly supported paraphyly of the subgenera that we have uncovered here ( Figs 1D , 2 , S 1 , Table 4 ). Specifically, in order to retain monophyly at the genus rank, four new genera would need to be erected for the species groups of brunneus , nearcticus , atopus , and pallitarsis . An additional issue would be the placement of the species which have not been sequenced, particularly those of the niger species group and, for example, the recently described L. brevipalpus Seifert , which is incertae sedis in the niger clade. The strongest morphological reorganization at the generic or subgeneric levels would be to recognize the reciprocally monophyletic niger and flavus clades as Lasius and Acanthomyops , respectively, but we refrain from doing so here (for our rationale, see ‘Species group classification of Lasius ’ below).

Comments on extant species. We determine that one species, Lasius escamole Reza should be excluded from Lasius and considered a junior synonym of the dolichoderine Liometopum apiculatum Mayr syn.nov. Reza (1925) described L. escamole in the context of a cultural study on the eponymous traditional Mexican dish known tobe made from the larvae of L. apiculatum ( Hoey-Chamberlain et al., 2013) . Although Reza’s description and illustrations are extremely vague, it is possible to see details that point to a dolichoderine identity. In the figures of the original description, the mandibles have long masticatory margin and small, sharp, even denticles, the ventral metasoma is shown as a slit-like anal opening rather than a formicine-like acidopore, and various figures display the fine, dense, appressed pilosity characteristic of Liometopum Mayr , but no erect setae as expected for Lasius .

Many species have been added to Lasius since Wilson’s revision, mostly in Europe and Mediterranean, while North American taxa have largely remained untreated except for a thorough revision of Acanthomyops ( Wing, 1968a, 1968b). Careful research has revealed multiple Palaearctic Lasius species that show only subtle morphological differentiation from close relatives ( Seifert, 1983, 1990, 1991, 1992, 2020; Schlick-Steiner et al., 2003). There is no reason to believe that North America does not harbour a diverse fauna of such ‘cryptic species’. For example, the question of the putative Holarctically distributed Lasius species was resolved by Schär et al. (2018) who elevated to species rank the American representatives of L. alienus , L. flavus , and L. umbratus Nylander , recognizing the following revived taxa, in order: L. americanus Emery , L. brevicornis Emery , and L. aphidicola (Walsh) . Renewed focus on the Nearctic fauna is necessary, as is expanded sequencing at the global scale.

Comments on extinct species. Without having scored the Baltic Lasius fossils other than † L. schiefferdeckeri , that is, † L. punctulatus Mayr and † L. nemorivagus Wheeler , we are unable to quantitatively address their placement. Historically, † L. punctulatus and † L. schiefferdeckeri were considered to be members of Lasius s. str. ( Wilson, 1955; Dlussky, 2011), while the queen-based † L. nemorivagus was placed in Chthonolasius ( Wilson, 1955) later to be implicitly considered incertae sedis in the genus ( Dlussky, 2011).

Our combined-evidence dating analyses recover † L. schiefferdeckeri as sister to or within the Lasius genus group ( Figs 4 , S 7–S 9 ). As the specific relationship of the fossil to the extant species of the Lasius genus group is uncertain, we conservatively consider the fossil incertae sedis in Lasius . There remains the possibility that † L. schiefferdeckeri is ancestral to the extant niger clade and is indicative of low rates of phenotypic transformation, as suggested by Mayr (1868), Wheeler (1915), and Wilson (1955). The placement of † L. schiefferdeckeri may be refined in future study by scoring characters which are explicitly derived from comparison of the brunneus and niger groups within the niger clade.

The two differentiating traits proposed for the brunneus and niger groups are, on average, <8 mandibular teeth in the brunneus group ( Seifert, 1992; some niger group species with <8), and presence of a subapical cleft in the mandibles of brunneus group males ( Wilson, 1955). While † L. schiefferdeckeri demonstrates both a tooth count of <8, and presence of a subapical cleft, the latter character is probably plesiomorphic of the Lasius genus group, and the polarity of the former is uncertain. Notably, Wilson (1955) observed that the male mandibles of † L. schiefferdeckeri are observed to vary from the ‘ brunneus form’ to the derived ‘ niger form’. With these three traits in mind, it does seem reasonable that † L. schiefferdeckeri is stem to or directly ancestral to the niger clade.

Note on biology. Despite the interest in this genus, however, basic natural history remains unknown for many species, including the morphologically aberrant L. atopus and the species we sequence here, ‘ L. nr. atopus ’. Only a handful of recently published studies have addressed the behaviour of some of the more rarely encountered species (e.g., Raczkowski & Luque, 2011), indicating that more effort is needed to elucidate the biology of Lasius .

Species group classification of Lasius

To avoid following the undesirable trend where taxonomy is divorced from molecular phylogenetics ( Steiner et al., 2009; Ward, 2011), we propose an informal species-group classification of Lasius , which reflects the phylogenetic structure within the genus. As the ICZN rules do not apply to informal species groups, this arrangement is also more adaptable to future refinements and new phylogenetic findings, thus contributing to taxonomic stability. Two main clades are recognized, those of L. niger (70 spp. + 1 ssp.), and L. flavus (54 spp.), with three and seven species groups, respectively. In addition, all 21 valid extinct species are formally considered incertae sedis in the genus, while 14 extant species are considered unidentifiable and incertae sedis in the genus.

Should the resurrection of subgenera be deemed advisable and desirable, we recommend dividing the genus into Lasius s. str. and Acanthomyops along the boundaries of the niger and flavus clade split. Resurrection of Acanthomyops may be justified, as it is readily recognizable by the autapomorphic condition of its metapleural gland (see couplet 5 of key to lasiine genera above, Fig. 7E, F , also Appendix S1), and the relatively old split between the flavus and niger clades ( Figs 4 , 5 ). However, we have elected not to do so here because: (i) our molecular sampling is still fractional at the species level (9/70 niger clade, 12/54 flavus clade); (ii) transfer of Lasius subgenera or species groups to Acanthomyops may be disruptive to the practicing systematic community; (iii) Sanger data are being rapidly superseded by UCE data for ant systematics; (iv) morphological investigation is necessary to diagnose the newly recovered groups, and (v) the ‘ Lasius ’ gestalt is easy to recognize during field collecting and generally facilitates communication. In either case, the exceptional morphometric data being generated for Lasius (e.g., Seifert, 1982, 1992, 2020; Seifert & Galkowski, 2016) will be valuable for model-based phylogenetic analysis combining phenotypic and genotypic data (e.g., Prebus, 2017; Barden et al., 2017a), particularly in the light of the greater information content of continuous trait data relative to discretized characters ( Parins-Fukuchi, 2018). Ultimately, our goal is to encourage the consilience of morphological and molecular data, and we view both as critical for a complete understanding of evolution and systematics.

The following inter-group transfers are made: (i) Members of nonmonophyletic former Lasius s. str. Are dispersed among L. brunneus , L. niger , and L. pallitarsis groups; (ii) L. atopus and L. humilis Wheeler are removed from the L. umbratus group (former Chthonolasius ) and placed in the atopus and flavus groups, respectively and; (iii) L. nearcticus Wheeler and L. sonobei Yamauchi are removed from the flavus group (former Cautolasius ) and placed in the nearcticus group. For diagnosis of the niger and flavus clades, see the key to extant genera of Lasiini above. The task of morphologically distinguishing two pairs of groups, those of brunneus and niger and of flavus and nearcticus , is difficult. Future phylogenetic studies are expected to refine the boundaries of these species-groups. In addition, keys to the species within these groups, and especially that of umbratus , are necessary. Diagnostic remarks for each species group are provided to guide such future work.

Note that, in the list below, [e] indicates that specimen(s) of the taxon has/have been examined while empty brackets, [], indicate that specimens were unavailable and placed based on study of the literature. Species that were included in our molecular sampling are bolded. Genus group type species are marked with asterisks (*). Overall, with the exclusion of L. myrmidon and L. escamole , 118 of the 125 valid, extant species (124) and subspecies (1) were examined either directly or indirectly (via AntWeb). For authorities and taxonomic synopses, refer to Bolton (1995) or AntCat ( Bolton, 2021). Finally, it should be noted that some groups contain additional, undescribed species.

Figures

Fig. 1. Comparison of phylogenetic hypotheses for Lasius; red indicates paraphyly of Lasius sensu stricto, orange of Lasius (Cautolasius) and violet of Lasius (Chthonolasius). (A) Intuition-based topology from Wilson (1955). (B, C) combined morphological and mitochondrial topologies of Janda et al. (2004), and Maruyama et al. (2008). (D) Consensus topology from the current study; see Table 4 for Bayes factor tests of our constraint analyses.

Fig. 2. Molecular phylogeny of the Lasiini inferred from Bayesian analysis of the Lasiini_55t data matrix.Node support values are reported as Bayesian posterior probabilities (BI) versus maximum likelihood bootstrap (ML). Filled circles at nodes indicate full support from both analyses. An asterisk (*) indicates that the topology of the ML analysis diverged from the BI analysis at the given node.

Fig. 4. Results from phylogenetic analysis of the Lasiini_w_outgroups_w_fossils_154t combined morphological and molecular dataset analyzed with MrBayes using the fossilized birth-death branch length prior and the IGR clock variation prior. Fossil taxa are indicated by a dagger preceding the name. Node support values are in Bayesian posterior probability. Filled circles indicate full support. Horizontal blue bars at nodes are 95% highest posterior density (HPD) intervals.

Fig. 7. Some key features of the major clades of the Lasiini. (A) third abdominal tergite modified with dorsoventrally-oriented groove for reception of entire petiole (Pseudolasius breviceps); black line indicates contour of tergum and black ellipse the spiracle; (B) third abdominal tergite receiving only posterior base of petiole (Myrmecocystus mimicus); white line indicates contour of tergum, and white ellipse the spiracle; (C) frontal protuberance and effaced carinae low relative to face (P. breviceps); (D) frontal protuberance and effaced carinae raised dorsally from face (M. mimicus); (E) metapleural gland atrium grossly enlarged (Lasius subumbratus); black lines show larger size of the metapleural gland bulla (lower indicator) as compared to F, and relative to the propodeal spiracle (upper indicator) and space between the spiracle and bulla (middle portion); (F) metapleural gland atrium small (Lasius turcicus); black lines show smaller size of metapleural gland bulla as compared to E, and relative to the propodeal spiracle and space between spiracle and bulla.

Fig. 9. Gross phenotypic synopsis of the species groups of Lasius, plus two phenotypically similar fossil species; all images in profile view; scale bars = 1.0 mm; scale bar unavailable for K. (A) Lasius brunneus (CASENT0280440); (B) Lasius fuliginosus (CASENT0179898); (C) Lasius nr. niger (CASENT0106128); (D) Lasius nr. atopus (CASENT0234858); (E) Lasius carniolicus (CASENT0280471); (F) Lasius claviger (CASENT0103542); (G) Lasius brevicornis (CASENT0280456); (H) Lasius nearcticus (CASENT0104774); (I) Lasius pallitarsis (CASENT0005405); (J) Lasius aphidicola (CASENT0280468); (K) †Lasius schiefferdeckeri (HJF013); (L) †Kyromyrma neffi (AMNH-NJ1029). (Image credits: D = the authors; AntWeb: A = Will Ericson; B = Erin Prado; C = Michael Branstetter; E, G, J = Shannon Hartman; F, H = April Nobile; I = unknown; K = Vincent Perrichot; L = Dave Grimaldi & Vincent Perrichot.)

Fig. 5. Results of the BioGeoBEARS and life-history trait ancestral state estimation analyses on a chronogram pruned from the analysis of the Lasiini_w_outgroups_w_fossils_154t combined morphological and molecular dataset. Horizontal blue bars at nodes are 95% highest posterior density (HPD) intervals from the divergence dating. Coloured circles on nodes are inferred inherited biogeographic ranges, with the top half of each corresponding to the upper branch and bottom half with the lower branch; coloured circles at branch tips represent biogeographical coding for terminal; dashed lines at 60 and 30 Ma indicate the boundaries of the three BioGeoBEARS dispersal rate matrix time periods. Coloured boxes at branch tips represent presence or absence of two traits for Lasius: temporary social parasitism (black = absence, red = presence) and fungiculture (black = absence, green = presence). Stars on internal branches indicate the evolution of temporary social parasitism (red) and fungiculture (green) The figure background indicates generalized major climatological regimes during Earth’s history (Royer et al., 2004; Zachos et al., 2008; Hansen et al., 2013), while the inset globes visually represent palaeoclimate and landmass configuration modified from Scotese (1998).