Taxonomic and faunistic study of four almost unknown Brazilian Meloidae (Coleoptera)

ABSTRACT The Brazilian fauna of Meloidae is poorly studied, even though it includes more than 160 species. In this paper, we aimed at widening the knowledge on four species of blister beetles from this country. Specifically, we defined the uncertain range of Tetraonyx angulicollis, as extended in south-eastern Brazil rather than in Mexico, and implemented the description of the species with figures. We studied the taxonomy and distribution of three almost unknown species of Nemognatha from Brazil, São Paulo State, providing descriptions and figures of sexual characters and colour variability of N. beauregardi, to which is probably referable as a junior synonym of N. plaumanni, of N. rufoscutellaris and of N. cfr. gounellei. Moreover, we assigned these three species to the subgenus Pauronemognatha, recently recorded from South America.


Introduction
Meloidae is a widely distributed family of beetles including about 3000 species characterized by: (1) the production of Cantharidin; (2) a hypermetabolic larval development except in the basal subfamily Eleticinae and (3) the biology of the first instar larvae which can be predators of Hymenoptera, Apoidea or Orthoptera, Acrididae (Bologna, 1991).At the imaginal stage, blister beetles are quite heterogeneous but are characterized by an elongated body, a narrow "neck", due to a marked constriction of the postocciput, and a ventral blade on the claws (Bologna, 1991;Bologna et al., 2010).Among all the instar larvae, only the first one, the triungulin, is well known for most of the species (Bologna et al., 2010).Triungulins are well sclerotized, highly mobile, and campodeiform larvae, and represent the dispersal and food finding phase (Bologna et al., 2010).In some species, triungulins attain the food directly by crawling, while in others they show morphological adaptations to phoresy (Bologna, 1991;Bologna et al., 2010).The other phases, up to the pupal stage, are responsible for: feeding and body growth (First Grub), enduring periods of cold or dryness (Coarctate), and preparing the pupal chamber (Second Grub) (Bologna et al., 2010;Bologna & Di Giulio, 2011).
The blister beetle fauna of Brazil is still scarcely studied, even if it includes many species (more than 160: see Denier, 1935aDenier, , 1935bDenier, , 1940;; unpublished records from studied collections).
Among the Neotropical Meloidae, only the subfamily Eleticinae was revised at the genus level (Selander, 1966) together with some monotypic genera of Meloinae, or those including very few species.As for the more speciose genera of Meloinae, only some groups of species of Epicauta Redtenbacher, 1845 were partially analysed (Adams and Selander, 1979;Selander, 1981;Campos Soldini and Roig-Juñent, 2015;Campos Soldini et al., 2018).The genus Tetraonyx Latreille, 1805 shows its larger diversity in Brazil (at least 55 species over 102), but the only study of the species remains the late monograph of Haag-Rutenberg (1879).After that revision, only a catalogue of the New World (mostly Neotropical) tribe Tetraonycini was published (Selander, 1983), with additional contributions focused on the other three monotypic genera of this tribe (Selander, 1965(Selander, , 1983(Selander, , 1985;;Bologna and Pinto, 2007), one study on the Argentinian species of Tetraonyx (Selander and Martinez, 1984;Selander and Selander, 1992), and a faunistic study on the Brazilian State of Rio de Janeiro (Quintino and Monné, 2009).
A contribution to the monotypic genus Meloetyphlus Waterhouse, 1872 was recently published (Di Giulio et al., 2023).
The Nemognathinae subfamily, recently explored with a molecular phylogenetic approach considering their whole range (Riccieri et al., 2023), is insufficiently studied in the Neotropical Region, especially as concern the Brazilian fauna.The allocation of the species within the described genera still requires further investigations (Pinto and Bologna, 1999;Bologna and Pinto, 2002).Such an undefined taxonomic situation is so deep that Denier (1935a) in his catalogue of American blister beetles, erroneously considered the genus Nemognatha Illiger, 1807 as a synonym of Zonitis Fabricius, 1775.The Neotropical species of Nemognatha were never revised, even if Riccieri et al. (2023) pointed out that in South America are distributed some of the Nearctic subgenera studied by Enns (1956), together with other unknown intrageneric lineages.Moreover, the value of the genus Zonitolytta Pic, 1927 is questionable and it was considered by Selander (1991) as a subgenus of Nemognatha, but because he did not provide an explanation for this transfer, we continue to treat Zonitolytta at the generic level (Pinto and Bologna, 1999).Also, in the genus Zonitis, some species appear well distinct and possibly referable to new genera (Pinto and Bologna, 1999;Bologna, unpubl.), or a few others have been referred to the genus Pesudozonitis Dillon, 1952 (Pinto, 2019) or Rhyphonemognatha Enns, 1956 (Pinto andBologna, 2016).Recently, García-París et al. ( 2014) studied in detail one species of Nemognatha from Venezuela which was described from south-eastern Brazil (see below).A revisional work of the southern America Nemognathinae could permit to assess the generic status of the Neotropical species and to describe new species and synonymies.In particular, at least 10 species of Nemognatha were recorded from Brazil, and a few additional, as N. claveri Pic, 1914, probably belong to other genera.
In this nebulous context, this paper is aimed at taxonomically studying four species of almost unknown Meloidae from the São Paulo State, in Brazil, one belonging to the genus Tetraonyx and three to the genus Nemognatha.

Material and methods
The examen of numerous specimens from the Museu de Zoologia, Universidade de São Paulo, collected in a few localities of the São Paulo State, allowed to conduct this study.The examined material includes four specimens of Tetraonyx angulicollis Haag-Rutenberg, 1879 [one now in the Marco A. Bologna collection (MAB) at Roma Tre University], 61 of Nemognatha beauregardi Pic, 1910 (five now in MAB), 53 of N. rufoscutellaris Pic, 1910 (five now in MAB), and 7 of N. cfr.gounellei Pic, 1910 (one now in MAB).
Morphological study was carried out using an Olympus SZX 12 stereomicroscope.Photographs were taken with Visionary Digital LK Lab System (Visionary Digital, Palmyra, VA) equipped with a Canon EOS 6D mark II dSLR camera and an MP-E 65mm f/2.8 1-5× lens (Canon, Tokio, Japan).This device allowed the automatic capture of stacks of images on different focal planes, which were combined with the Helicon Focus 7 software.
Taxonomy.The description of this species made by Haag-Rutenberg ( 1879) is clear and permitted the identification of our specimens.The genus Tetraonyx includes two described subgenera, the nominate and the monotypic Paratetraonyx Kaszab, 1959 (Pinto and Bologna, 1999).The nominate subgenus was divided by Haag-Rutenberg into two lineages: one (A) characterized by the sides of pronotum concave in middle and with more marked angles; the second (B) with pronotum not incised medially and obtuse angles.According to the shape of pronotum (Figure 1c), T. angulicollis belongs to the lineage A and it is the only species with the dorsal body colour totally ochre-brown (Figure 1a).
Among the examined specimens, the abdomen could be entirely black or with the last three ventrites orange; meso-and metapleures totally or partially black; mesosternum orange or partially black; antennae black and antennomeres similar in both sexes.
As the other two following species, the head is very short and galeae are quite short, differently than in most Nemognatha, reaching the lateral-anterior angle of the eye.The male modified areas are present on ventrites II-V (see below; Fig. 2d) and are represented by scarcely depressed, sub-oval surface, with a distinct and denser punctuation and dense different setae; male genitalia as in Fig. 2e/f.Remarks.This species is recorded only from Brazil.Beauregard (1890) described it as N. atra using the name found on a specimen of the Brussels' Museum collections, but this name was preoccupied by N. atra Dugès, 1889 from Mexico (now Zonitis (Parazonitis) dugesi Selander, 1954).Consequently, the name was changed by Pic (1910b) as N. beauregardi.After Pic's paper this species has never been studied.Erroneously, N. atra and N. beauregardi were included by García-París et al. (2007) among the synonyms of N. chrysomeloides (Linnaeus, 1763), a species widely spread from Mexico to Argentina, which belongs to another lineage and is greatly distinct at least because of the ventrites not modified and galeae extremely elongate.
According to the classification of Nemognatha proposed by Enns (1956), very few South American species have been assigned to one of the four Nearctic subgenera.Riccieri et al. (2023) referred tentatively N. nigrotarsata (Fairmaire et Germain, 1861) from Chile and NW Argentina to the nominate subgenus, and N. mimula Borchmann, 1952 from Ecuador and Peru to the subgenus Pauronemognatha Enns, 1956, while they did not place in none of the described subgenera both N. chrysomeloides and N. nigronotata (Pic, 1914).We confirm the arrangement in Pauronemognatha of the first two species and refer to it also N. beauregardi, which shows modified areas on ventrites II-V (Figure 2d).
García-París et al. ( 2014) recorded for the first time N. plaumanni Borchmann, 1942 from Venezuela.This species was described from south-eastern Brazil, Santa Catarina, Nova Teutonia [erroneously considered by Borchmann (1942) as in the Rio Grande do Sul State], not far from the localities of N. beauregardi we studied in the São Paulo State.As noted above, García-París et al. ( 2014), who examined topotypic specimens of N. plaumanni in the Budapest' Museum, erroneously considered N. beauregardi as a variation of N. chrysomeloides and did not compare N. plaumanni with N. beauregardi.According to the figures of the topotypic N. plaumanni published by these authors, this species seems extremely similar to N. beauregardi and we suggest their possible synonymy.We cannot confirm this co-specificity without the examination of the male structures (ventrites and genitalia) of N. plaumanni, and because García-París et al. ( 2014) stated that this species cannot be referred to any described subgenera, while in N. beauregardi the modified ventral areas and the hind spurs typical of Pauronemognatha are clearly detectable.
Léo Correia da Rocha Filho communicated to us (pers.comm.II.2023) that specimens of N. beauregardi have been collected in the nests of Tetrapedia diversipes Friese, 1899 (Hymenoptera, Apidae), and those of N. rufoscutellaris and N. cfr.gounellei (see below) in the nest of Tetrapedia spp.Parasitization of T. diversipes was ascertained also in N. rufoscutellaris and in the species which probably corresponds to N. gounellei (Rocha-Filho and Garofalo, 2016).
Taxonomy.This species is black with elytra dark black-violet metallic, head, pronotum, procoxae and abdomen usually orange (Fig. 3a/b/c).The colour variation of the examined specimens is large and could support the hypothesis of co-specificity with other Brazilian taxa.First, the head colour could be totally orange (Fig. 3e) or with a black frontal area, present in almost 14% of the individuals, usually rectangular, extended from the suture to the middle of interocular space or totally covering the frons until the posterior margin of eye (Fig. 3d).The antennomeres are black, but in two specimens antennomeres I-II are orange as the base of III.Elytra in a few individuals are blue-violet or green-violet.The abdomen is orange, especially in males, but in about 38% of individuals is black or rarely dark reddish.
As in both other Nemognatha object of this study, the head is very short and galeae are quite short, reaching at most the anterior ventral angle of the eye.The modified ventral areas, which allow to refer for the first time this species to the subgenus Pauronemognatha, are depicted in Fig. 3f, and male genitalia in Fig. 3g/h.
Remarks.Both N. rufoscutellaris and N. subparallela Pic, 1927 have been described generically from Brazil and afterwards recorded also from Bolivia and Argentina.According to the description, they look very similar to each other, except for the head colour, black in the first and orange/red in the second species.We agree with Denier (1935b) who considered N. subparallela as a chromatic variation of N. rufoscutellaris with the head orange/red, rather than black.For instance, Martinez (1992) considered these two taxa as distinct, based on the unpublished opinion of R. B. Selander.Another species, N. pallidicollis Beauregard, 1890, which was compared by Pic (1910a) to N. rufoscutellaris, could be referred to the intraspecific variation of this complex, but we did not examine it.If the synonymy between the two species will be confirmed, the name N. pallidicollis should have the priority.
For information about the bee host see above.Taxonomy.This species (Fig. 4a/c) is black with the elytral colouration brown-hazelnut, variable at apex, as it could be metallic violet on the apical third-fourth (50% of the specimens; three localities) or apically only infuscate without distinct violet colouration, but this might be related to the scarce sclerotization being the individuals extracted from the nests of their hosts (37%; one of the previous localities).In one specimen elytra are completely broken posteriorly.Both head and pronotum are red (Fig. 4a/b), but in three specimens the head is black, and in one of them also the pronotum is dark.The abdomen is usually orange.
As in the previous two species, the head is very short and galeae are quite short, reaching at most the anterior ventral angle of the eye.The modified ventral areas (Fig. 4d), permit to assign for the first time this species to the subgenus Pauronemognatha.Male genitalia are represented in Fig. 4e/  Remarks.Until now, this species was recorded only from Brazil.We refer tentatively the examined specimens to N. gounellei because of the correspondence of most of the description (Pic, 1910a).Doubts are due to the lack of a small violet spot at the base of the elytra, but since the great variability in colouration of the Brazilian Nemognatha, numerous specimens should be examined to define their variation.
For information about the bee host see above.