The Neotropical antlion genus Ameromyia Banks, 1913 (Neuroptera: Myrmeleontidae), systematics and redefinition under a phylogenetic approach

A taxonomic revision and the first phylogenetic hypothesis of the Neotropical genus Ameromyia Banks (Myrmeleontidae: Brachynemurini) is herein presented. The phylogeny is based on 45 morphological characters and recovered the traditional Ameromyia as paraphyletic in respect to the monotypic genus Venezueleon Stange, which is here synonymized under Ameromyia. Three species are synonymized (A. hirsuta Navás and A. stevensi Navás under A. nigriventris (Walker), and A. pentheri Navás under A. strigosa (Banks)) and two new species are described (A. clepsydra sp. nov. and A. explicata sp. nov.). Ameromyia sensu novo is a valid genus with 12 species restricted to South America, and divided into two species groups. Taxonomic keys are also presented to adults and larvae, as well as a discussion on the genus biology.


Introduction
Myrmeleontidae Latreille is currently the most diverse family in the order Neuroptera, comprising nearly 2,150 extant species (Oswald 2021), and divided into four subfamilies and 17 tribes (Machado et al. 2019). Commonly known as antlions and owlflies (the latter only for part of the Ascalaphinae) (Machado et al. 2019). The adults possess characteristic slender bodies and wings, clavate or capitate antennae, and usually are aerial generalist predators, adapted to arid and semiarid regions (Mansell 1999). The larvae, which have two or more mandibular teeth, are sit-and-wait predators that hunt prey while perched on rocks and/or trees, or, more commonly, buried on particulate substrate (Miller and Stange 1985;New 1986;Stange et al. 2003;Stange 2004;Badano and Pantaleoni 2014).
The tribe Brachynemurini (Myrmeleontinae) currently comprises 117 valid species in 28 genera, which are restricted to the Americas (Stange 1994;Machado et al. 2019). This tribe is highly variable in regard to morphological features such as leg chaetotaxy and wing venation (Stange 1970(Stange , 1994, but its most impressive feature is the male genitalia in which the parameres have folds (hinges) and articulate with itself (Addams 1956;Stange 1970). Banks (1927) erected Brachynemurini to accommodate the New World antlions which didn't exactly fit into his previous taxonomic classifications for the Old World antlions, which mostly relied on the number of hind wing presectoral crossveins. Afterwards, the tribe was divided into two subtribes, Brachynemurina and Austroleontina, based on the number of presectoral and MP2 crossveins in the hind wing (Banks 1943). The first large revision of the tribe focused on the North American species (Stange 1970), and disregarded Banks previous internal classification, as well as erected new diagnostic characters for the tribe: i) curvature of forewing 2A in relation to 3A; ii) forewing with a short CuP; iii) one to five presectoral crossveins on hind wing; iv) tibial spurs absent or smaller than the procoxae; v) female anterior gonapophysis platelike; and vi) hinged parameres of male genitalia. Posteriorly the tribe was reduced to a subtribe of Myrmecaleurini (now restricted to the Old World) based on larval characters (Stange and Miller 1990).
However, the first cladistic analysis of Brachynemurini reinstated its tribal status, but reclassifying its species in three different tribes: Brachynemurini (s.str.), Gnopholeontini and Lemolemini, mainly based on female genitalia and larval characters (Stange 1994). Posterior phylogenetic analysis recovered Brachynemurini (s.str.) sister to Dendroleontini (Michel et al. 2016), both sister to the remaining Myrmeleontinae (Badano et al. 2016), or sister to Gnopholeontini (Winterton et al. 2018). In the first phylogenomic analysis focused on Myrmeleontidae, Brachynemurini was also recovered sister to the remaining Myrmeleontinae sensu Machado et al. (2019). However, the tribe was recovered as paraphyletic in respect to Gnopholeontini, which was deeply placed within Brachynemurini and thus synonymized under the latter, returning to the initial larger concept of Brachynemurini defined by Stange (1970).
Ameromyia Banks, 1913 is a Neotropical genus of Brachynemurini which today comprises 12 valid species (Stange 2004). The genus is diagnosed as having setae on the frons, large pretarsal claws and clavate setae on male genitalia (Stange 1994). The adults live on grasslands and perch on grass stems which reflect their distinctive coloration (Fig. 1), and the known larvae live on sand dunes (Stange 1994(Stange , 2004. Ameromyia was first described circumscribing the Brachynemurini species that present a long CuP in the forewing and a long CuA in the hind wing, as well as the presence of both banksian lines on both wings (Banks 1913). In the same study, Amazoleon Banks, 1913 was described to include species with the same features, but with slender wings and longer abdomen. Amazoleon was posteriorly synonymized under Moza Navás 1912 (Navás 1914), however, this synonymization was not adopted by none of the following studies. Posteriorly, Foya Navás 1914was synonymized under Amazoleon (Esben-Petersen 1920 and Nemotolus Banks, 1943 was described based on an enlarged pillula axillaris, although similar to Amazoleon. Posteriorly Amazoleon and Nemotolus were synonymized under Ameromyia (Stange 1967). Nevertheless, Ameromyia was only redescribed afterwards, receiving its current diagnosis together with the description of the male and female genitalia, and the larvae of two species by Stange (1994). In the same study, Ameromyia was divided into two species groups (modesta and nigriventris groups, classified and named after the species with known larvae), placed deep within Brachynemurini, and having Venezueleon Stange, 1994 as the sister genus. Although, most recent phylogenetic analyses of Myrmeleontidae recovered Ameromyia sister to all remaining Brachynemurini (Machado et al. 2019).
Ameromyia male genitalia was described as bearing the same structure as its sister genus Venezueleon. However, both genitalia descriptions were conflicting in regard to the homology of both gonarcus and parameres (Stange 1994). Moreover, almost no photographs or illustrations are available in the literature, and the ones provided do not illustrate the whole insect nor diagnostic features, and even wing illustrations do not depict full wing venation or both wings. All these issues highlight the need for a taxonomic review of Ameromyia (Stange 2004), which is presented here, comprising a taxonomic revision with description of new species, including taxonomic key to adults and larvae, a phylogenetic hypothesis for Ameromyia, and a discussion about the genus biology.

Specimen collection and identification
A total of 607 Ameromyia specimens were herein analyzed. Type specimens analyzed are discussed under "Remarks" for each species. For species with syntypes, lectotypes were designated from the syntypes series according to the better-preserved specimen. Specimens were provenient from several collecting events throughout 2017-2019, and by loans or photographs from institutions. Collected data for the specimens studied is included in the section "Examined Material" for each species. New distributional records at country level are highlighted in bold in the section "Distribution" for each species. A map showing the distribution of each species, per species group and by collection sites was constructed in QGis 3.4.13 Essen using projection unit GCS, datum WGS 84 (QGIS Development Team, 2015). Specimens whose label locations stated only country or state/district/ province were plotted on the capital of the stated location. Specimens collected were preserved in 80% ethyl

Live specimen rearing
Immature specimens selected for rearing were placed in individual plastic containers. Plaster or styrofoam plates were fit into each container, and then coated with sand, soil and sometimes small pebbles from the habitat where the larvae were collected. Larvae were fed two to three times per week with either Drosophila melanogaster Meigen maggots, first to third instar Tenebrio molitor Linnaeus larvae, and/or Nasutitermes Dudley nymphs, depending on prey availability. After pupation, cocoons were transferred to a glass rearing cage with sand and vertical wooden sticks. Rearing boxes were kept in air-conditioned rooms with the temperature between 23-25°C. After adult emergence, the pupal exuvia and the larval sclerites were preserved in 80% ethyl alcohol, allowing for direct association between the different life stages of the same species. Cocoons were stored in vials with naphtalene and posteriorly associated with its respective preserved specimens. Adult specimens were placed in voil fabric enclosures of 60 × 60 cm during mornings, and of 2 × 2 m during afternoons and nights. Voil fabric enclosures were enriched with dry bushes and/or branches, and sand, dirt and rocks from the specimen's natural habitats. Adult specimens were fed adult Drosophila melanogaster Meigen, and handfed maggots, honey plus water, or gelatin with brown sugar and egg yolk. Hand feeding consists of grabbing adult specimens by their wings, and touching the food to their mouthparts, until they taste, grab the food and start chewing. When senescent, adult specimens were placed into a freezer for posterior pinning, or stored in 80% ethyl alcohol.

Phylogenetic analyses
Ingroup consisted of all valid Ameromyia species, including the new species. Outgroup was sampled upon specimen availability and taxonomic history according to previous systematic studies (Stange 1994;Machado et al. 2019), and consisted of two representatives of non-Myrmeleontinae genera (Dimares elegans (Perty, 1833) (Ascalaphinae) and Dimarella riparia (Navás, 1918) (Nemoleontinae)), and four Myrmeleontinae genera (an unidentified species of Myrmeleon Linnaeus as a non-Brachynemurini; and Argentoleon irrigatus (Gerstaecker, 1894), Austroleon immitis (Walker, 1853) and the historical sister group of Ameromyia, Venezueleon guaricus Stange, 1994, as Brachynemurini genera). The data matrix was produced in Microsoft Excel. The symbols '?' and '-' were used for missing data and non-applicability of characters respectively, and cases of polymorphy were coded between brackets. Characters and character states were described according to Sereno (2007), treated as discrete and unordered. The cladistic analysis was performed using TNT v.1.5 (Goloboff et al. 2008) under parcimony by traditional search, using equal and implied weighting, and all trees were rooted using Dimares elegans as the outgroup root. Implied weighting procedure followed Mirande (2009), with 15 k-values calculated for an average fit ranging from 50 to 90% of a perfectly hierarchical character. Higher sums of similitud index and Subtree Pruning Regrafting (SPR) indicates more stable k-values, and thus distortion coefficients and SPR distances of the consensus trees were used as criterions for selecting weighting usages. Also, different (consecutive higher) implied weighting k-values were tested until analysis recovered different topologies for comparison ends. Bremer support absolute and relative values (Goloboff and Farris 2001) were calculated with tree bisection-reconnection retaining 20,000 suboptimal trees with 10 suboptimal steps, and symmetric resampling frequency differences (GC) (Goloboff et al. 2003) with 33% removal probability and 10,000 replicates, were calculated for the trees recovered under equal weighting.

Characters
A total of 45 morphological characters were coded and included in the character matrix (Supplementary Material  Table S1). Some characters were adapted from Stange (1994), as indicated in the characters list below.

Cladistic analysis
Analyses under equal weighting resulted in four most parsimonious trees, with 82 steps. Two sets of trees were almost identical, differing only in regard to the outgroup (relationship between Dimarella riparia and Myrmeleon sp.), and therefore, only two trees with truly different topologies for Ameromyia internal relationships were produced (Fig. 11). Different (consecutive higher, k+1) k-values for implied weighting always recovered a single tree and with the same topology as one of the trees recovered under equal weighting (Fig. 11A), varying only in consecutive smaller fits, and thus Mirande's (2009) implied weighting procedure recovered the same single identical tree in all groups, with the same sums of SPR similarities and distortion coefficients (Supplementary  Material Table S2). Furthermore, the trees recovered under equal weighting differed only in regard to A. muralli position within the genus, and thus the tree under equal weighting which had the same topology as the trees recovered under implied weighting was chosen as the current proposed classification for Ameromyia (Fig. 11A), but both topologies will be discussed. species in this group (clade A, which is supported by one synapomorphy (10:0 Profemur with black decumbent setae only). Within clade A, A. pubiventris (Walker, 1860) was recovered sister to A. modesta (Banks, 1943) + V. guaricus (clade B), the latter supported by two synapomorphies (11:1 Femora with contrasting coloration between external and internal faces; and 12:1 Tarsomeres with contrasting coloration between external and internal faces).
The 'nigriventris' group ( Fig. 11) was recovered monophyletic in all analyses, but supported by different set of synapomorphies in both resulting trees. In our proposed classification for Ameromyia (Fig. 11A), this group is supported by a single synapomorphy (29:1 Hind wing posterior area cells at least as high as long), while in the alternative tree (Fig. 11B) this group is supported by two additional synapomorphies (16:1 Forewing mediocubital area with discontinuous infuscations and 17:1 Presectorial and sectorial areas with discontinuous infuscations). In both trees, Ameromyia clepsydra sp. nov. was recovered sister to all remaining species within this clade.

Diagnosis.
Fore and hind wings with both banksian lines well developed; hind wings vein CuA with at least four crossveins connecting with posterior branch of MP; male ectoproct postventral lobe shorter than ectoproct height; female gonapophyseal plate elongate; lateral gonapophysis of female genitalia fused; male genitalic sac with clavate setae; parameres lateral tooth basad to paramere plates; male gonarcus positioned anterior to parameres; and male medincus absent; larvae with thread-like setae.
Description. ADULT. Head: frons with short setae. Antenna of same length on both sexes, moderately short with about 30 flagellomeres. Antennal fossae separated by about two times pedicel width. Interocular distance less than eye width. Profemur without clavate setae. Vertex with many short decumbent setae. Ocular rim without setae. Clypeus with scattered setae, which are oriented ventrally. Labrum with two rows of short setae on ventral margin, all of which pointing ventrally. -Thorax: Pronotum broader than long when pretarsal claws are shorter than distal tarsomere, and longer than broad when pretarsal claws are otherwise. Pronotum with many long anteriorly oriented bristles on all margins, which are longer on lateral margins. Pronotum often with mirrored "P" shaped markings medially. Prescutum 2 and 3, mesoscutum and metascutum with many short, posteriorly-oriented white setae on posterior margins, and meso and metascutella with longer white setae on posterior margins. Thorax with many decumbent setae laterally, and mesepisternum frequently with a sulcus mesally. Lateral side of thorax sometimes with a longitudinal pale band, which covers episternum and epimeron of meso and metathorax, and contrasts with ventral thorax which is darker. Mesonotum without blade-like setae. -Legs: Legs with many short decumbent setae. Femoral sense hair present or absent. If present, profemoral sense hair as long as femur and much longer than that of mesoleg, which is shorter than mesofemur. Femur with or without black and/or white bristles. Tibiae with many black bristles, which are longer than femoral bristles, when present. Tibial bristles as long as tibial width, when femoral bristles are absent. Tibial spurs present, longer than basitarsus. Five tarsomeres, longer than broad. Distal tarsomere much longer than basitarsus. Pretarsal claws at least 2/3 of distal tarsomere length, more commonly longer than distal tarsomere. -Wings: Pilula axillaris well developed. Forewing as long as hind wing or slightly longer. Both anterior and posterior banksian line well developed and present in both wings. Wing membrane mostly hyaline, sometimes with a reddish or brownish tint. Forewing costal area simple at least until midwing length. Posterior area of forewing about as broad as that of hind wing. Forewing rhegmal area with brown infuscation on rhegmal crossveins. Forewing vein CuP runs along posterior fork of CuA, with at least four crossveins connecting them. Forewing vein 2A running towards posterior wing margin in a smooth curve. Forewing vein 2A connected to 3A by a crossvein, or touching 3A before running towards posterior wing margin. Hind wing CuA runs along posterior branch of MP fork with at least four crossveins connecting them. -Abdomen: Male abdomen longer or much longer than wings, female abdomen shorter than wings, or almost 1.5× longer than wings. Male tergite IX with many long, thick setae ventrally. Male tergite IX subquadrangular, with posteroventral margin slightly swollen and with many posterior oriented setae. Male ectoproct with a developed short postventral lobe, which is subequal to half of ectoproct height, without median or secondary lobes. Parameres posteriorly plate-like, arched, striated on external face of paramere plates, with a basal lateral hollow tooth on each side, and high sclerotized posterior folds. Parameres with a hinge mesally on internal face of each paramere plate, and with a row of short setae on each internal margin. Parameres anterior to hinge smooth, "spoon-like", positioned between parameres plates (in "folding" position), and fused dorsally with an anterior bifurcated projection. Gonarcus smooth, membranous, with long anteriorly projected arms which are flat and long. Gonarcus positioned anterior to parameres, arching barely dorsad to parameres  anterior bifurcation and dovetailing or almost dovetailing behind parameres plates lateral tooth. Genitalic sacs lateral to paramere tooth, with many clavate setae that reach beyond and/or above and posteriorly to paramere plates. Female terminalia with pregenital plate membranous with many setae. Gonapophyseal plate elongate, but shorter than posterior gonapophysis. Posterior gonapophysis digitiform, with many setae. Lateral gonapophysis fused, with well-developed digging setae. Ectoproct with well developed digging setae. -LARVA (third instar). Head: Larvae with three teeth. Mandible with distance between mandible base and basal tooth longer than between teeth. Mandible not enlarged at base. Some setae on external margin of mandibles base with length as long or almost as long as mandible base width. Internal margin of mandibles roughly straight until reaching distal teeth, and roughly about the same width. Distal tooth smaller than basal and medial teeth and oblique in relation to other teeth. Distal tooth closer to medial tooth than medial tooth is to basal tooth. Head capsule on dorsal view as broad as long or slightly broader than long. Head capsule on ventral view longer than broad or as long as broad. Head dorsal surface with thread-like setae near posterolateral margins. Head capsule dorsally with many short, thick, blunt dolichasters and very short, almost "spherelike" dolichasters, ventrally with short, thin dolichasters, and long and thick setae on lateral margins, which are straight on the anterior portion of head, and curved on posterior portion of head. Labrum with a row of thick, cylindrical dolichasters on posterior margin. Palpi with three segments, distal palpomere bigger than basal palpomere, which is bigger than mid palpomere. Distal palpomere enlarged at base with an acute end. Presence of rowed thick, cylindrical, white dolichasters near base of each palpi. -Thorax: Dorsal surface sometimes covered with thread-like setae. Dorsal surface prunescent, covered with short dolichasters. Thoraxic dorsal sclerites with short rowed dolichasters on all margins. Thorax ventral surface with long hair-like setae. Prothorax with many short, curved bristles on anterior margin. Mesothoracic spiracle borne on tubercle, tubercle length half of metathorax width at most. Tubercule covered with thread-like setae. Metathorax with two pedunculated setiferous processess on lateral margins, which bear many long bristles. Metathorax with a pair of eliptical dark spots submedially. -Abdomen: Abdominal tergites with thread-like setae, and sternites sometimes with thread-like setae. Abdominal spiracles not enlarged nor borne on tubercles. Setiferous processes on abdominal lateral margins bearing long white, black, or black and white bristles. Setiferous processes on abdominal lateroventral margins with hair-like setae. First abdominal tergite with two eliptical black spots submedially, almost aligned with metathoraxic dark spots. Odontoid process on sternite VIII longer than basal width. Ninth sternite covered with thread-like setae. Rastra with 4 thick setae, with innermost setae much shorter than remaining setae.

Remarks.
At first look, Ameromyia species can look similar to Argentoleon or to large Austroleon specimens, but can be readily differentiated from both genera by the short male postventral ectoproct lobes (which are much longer in Argentoleon and absent in Austroleon); the presence of both banksian lines in both wings; and the much longer hind wing CuA. Other Brachynemurini species bear a long hind wing CuA vein (such as Brachynemurus fuscus (Banks) and Brachynemurus nebulosus (Olivier)) or both banksian lines in both wings (B. nebulosus), but they are very different regarding distributional ranges and other Ameromyia diagnostic characters, such as the simple forewing costal area cells. Ameromyia male genitalia is also very conspicuous (Figs 9,10), as the paramere plates are enlarged and the paramere hooks are short and stout, and basad to paramere plates. The gonarcus is membranous and arches anterior to the parameres, with no mediuncus. The most similar-like genitalia to be found among Brachynemurini in previous studies is in Peruveleon Miller andStange (Stange 1970, 1994), but this genus is easily distinguished from Ameromyia by many morphological characters such as leg chaetotaxy, the length of the hind wing CuA vein, and also by molecular data (Machado et al. 2019).
Two species groups are determined based on taxonomical characters of wing veins, male genitalia, and cladistic results: the 'modesta group' and 'nigriventris group' (Fig. 11). These two species groups were previously named and identified by Stange (1994), but were classified only by larval characters. The modesta group can now be characterized by the abdomen length, which is longer than wings in females and almost double the length of wings in males, as well as the sickled club of the clavate setae on male genitalia (Fig. 7C), while the nigriventris group is characterized mostly by the hind wing posterior area cells, which are as high as long or higher than long.
Within groups, however, Ameromyia species are very morphologically similar, differing mostly in coloration. Coloration pattern on main wing veins such as Sc, R, Cu and A veins appears to be consistent across species, but markings pattern on wing membrane can be polymorphic within species. Thorax pronotal markings are also usually very similar within Ameromyia, but some species bear unique markings which are consistent within specimens, such as A. modesta (Figs 1C,2C,28A) and A. protensa (Fig. 39A). Leg characters are usually informative, and although badly conserved specimens can show misleading color patterns as they can faint or darken over time, leg chaetotaxy is very consistent within species. Genitalic structures are little informative. Ameromyia male specimens have hinged parameres (sensu Adams 1956) and a highly membranous gonarcus, and thus genitalic structure can be complex to visualize as the sclerites can bend and flex. At dorsal view, all known species bear almost identical male genitalia (Fig. 10A), and female genitalia presents close to no variation between species, even when across different species groups (Fig. 8). The digging setae on female terminalia can present some slight variation on thickness and/or length across specimens, but this is probably a regional adaptation for the granulometry of available sand pools across different habitats. Due to this great genitalic similarity between species, female termi-nalia and genitalia are only described in each description section when that species bear some morphological variation, and male genitalia is illustrated only in lateral view, in folding position, for each species section, as the lateral view provides more information as to the shape of the parameres (Adams 1956).
The larvae of most Ameromyia species are currently unknown, except for A. clepsydra sp. nov., A. guarica, A. modesta and A. nigriventris. Stange (2004) described the presence of thread-like setae on dorsal segments ( Fig.  14A-C) as a character for differentiating the two groups, as it is present in A. modesta but described as absent in A. nigriventris. However, it is present in all known Ameromyia larvae.
Biology. Very few data have been made available regarding Ameromyia biology and/or behaviour. Stange (1994) stated the known larvae lives on sand dunes, and later (2004) stated the adults rests on grass stems during the day. In this study, we have observed and/or reared in captivity adults from four species (A. clepsydra sp. nov.; A. explicata sp. nov.; A nigriventris and A. strigosa), and larvae from two species (A. clepsydra sp. nov. and A nigriventris). Adults from all species observed except for A. clepsydra sp. nov. were capable of flying high, at at least 2 m above ground, and when in captivity, they almost always perched at the top of the rearing enclosure. In their natural habitat, A. nigriventris females were seen flying close to the ground, presumably scouting for good ovipositioning spots, while all observed males which were near the ground were found slowly hovering in place, just above patches of loose earth which were apparently suitable for larvae. It is unknown if this is somehow a courting behaviour. Regardless of sex, when disturbed, specimens would quickly try to outmanouver their catcher, then rapidly ascend in the air and gain speed with the wind. Wild specimens of A. clepsydra sp. nov. were always observed flying very low, close to the bedrock, and no such ascending behaviour could be seen.
In this work, no feeding behaviour was observed in the wild, but in captivity, adult specimens of A. explicata sp. nov., A. nigriventris and A. strigosa accepted D. melanogaster as prey. Aphids, maggots, small catterpillars and green lacewings were offered, but none were succesfully preyed upon. Maggots were only eaten when handfed, and small catterpillars were only eaten when handfed, and if some hemolymph or gut content was exposed. All adult specimens accepted artificial diets such as honey + water, boiled egg white, and gelatin with sugar and egg yolk. However, all artificial foods were only eaten when -Abbreviations: an, antennae; bt, basal tooth; dgs, digging setae; ds, dolichasters; dt, distal tooth; mdt, medial tooth; met, metathorax; mt, mesothorax; oc, occelli; op, odontoid process; pt, prothorax; ra, rastra; sr, spiracle; st, sternite; ts, thread-like setae; tu, tubercle. handfed and individuals would starve to death otherwise, even if the food was left inside the enclosures. This suggests these Ameromyia species might rely solely on aerial hunting for feeding. Diets consisting of only sugar and water would only sustain individuals for a week, whether in every other diet (both natural and artificial) the specimens lived for approximately three weeks. Starving adults would die in two or three days.
The larvae of A. modesta and A. nigriventris were described as living in sand dunes, and A. guarica was described as living in shallow sand anchored to bedrock (Stange 1994). Of the larvae reared in the present work, that of A. clepsydra sp. nov. (Fig. 15B, C), lives in the same microhabitat as described for A. guarica: shallow, water-deposited sand on top of a river bedrock (Fig. 15A). The A. nigriventris larvae collected during this work were found on patches of loose, sandy wind-deposited earth on top of compacted clay soil (Fig. 16A). Despite the habitat differences, there is much similarity to the analyzed larvae microhabitat. Ameromia nigriventris larvae stays submerged in the loose, sandy soil, while using the compacted clay as a hard substrate which they anchor themselves to (Fig. 16B, C), and A. clespydra sp. nov. anchor themselves to the bedrock itself or the shallow compacted sand on top of the same bedrock (Fig. 15B, C). Almost al-ways, larvae were found on small "trenches", presumably dug out by the larvae themselves (Figs 15B,C,. All known Ameromyia larvae bear many long bristles on thoraxic and abdominal setiferous processes, which would help in escavating suitable holes in hardened substrate. When exposed, the larvae would slightly contract themselves into their holes, and remain still for a while. Then, they would try to dig backwards, or leave their spot and walk forwards, towards a new hiding place. When moving forwards, regardless of instar, A. nigriventris larvae are very quick and highly agile, while A. clepsydra sp. nov. are very slow and lethargic, sometimes refusing to move even if poked or disturbed. Coccoons (empty or with pupae still inside) in the wild were found in the same said "trenches" (Fig. 16D), or protected by bedrock (Fig.  15D). When in captivity, Ameromyia larvae were highly sensitive about the presence or absence of both the hard substrate to achor themselves to, and to said "trenches". Larvae from both reared species would easily startle and hardly feed in the absence of a similar hard substrate, and A. clepsydra sp. nov. larvae in the late third instar would not spin a coccoon nor pupate, and instead walk incessantly until it found an appropriate substrate or until death. To simulate the hard achorage of their natural habitat, plaster or styrofoam was "dug out" and then fit Figure 15. Collection site and habitat of Ameromyia clepsydra sp. nov., in Andaraí, Bahia, Brazil: A site surroudings and sandy substrate on bedrock. B Second-instar larvae exposed in its "trench". D Third-instar larvae exposed on its "trench". D Empty, open coccoon on bedrock crevice. -Abbreviations: cc, coccoon; sp, sand pools.
into the larvae plastic enclosures, which highly decreased larvae mortality. This highly suggests Ameromyia species are much more sensible to microhabitat selection and alterations, as other larvae collected from the same sites and in similar microhabitats (such as Austroleon immitis and Dimarella riparia) were able to feed, develop and pupate normally even in the absence of these specific mi-crohabitat conditions. Although larvae from other species were not found during this work, as the morphology and chaetotaxy of female terminalia is consistently similar throughout the genus, larvae habitat might also be similar, as the characteristics of female terminalia affects oviposition and the selection of larvae habitat (Stange and Miller 1990). First-instar larvae exposed in its "trench". C Third-instar larvae exposed on its "trench". D Coccoon, exposed on a dug-out "trench". -Abbreviations: sp, sand pools; tr, trenches.

Figs 17-19
Diagnosis. Black and white coloration. Thorax lateral side black with no apparent longitudinal pale band. Wing veins alternating black and white coloration. Wing membrane with many markings and infuscations, mostly on radial, rhegmal and substigmal areas. Sectorial and presectoral area crossveins sometimes with an infuscated biareolate "8" shaped pattern. Male genitalic clavate setae club subglobose.
Description. ADULT. Head: Antennae clubbed (4 mm) with 31 flagellomeres, light brown with yellow rings between the flagellomeres. Vertex white with grey patches, and dark brown spots on lateral margins; with a transversal and a longitudinal dark brown streak at anterior and posterior row, forming a "T" shape. Etymology. "Clepsydra" translates to "hourglass" in Latin. The characteristic pattern formed by transverse crossveins in forewing presectoral and radial sector area resembles an hourglass (Fig. 13B).
Biology. Larvae were found anchored to bedrock, buried in small pools of shallow water-deposited sand near a river (Fig. 15A). The bedrock was not protected from rain and/or sunlight, and the area is accessible by tourists and susceptible to trampling. able to pupate earlier due to the lack of a hard, suitable achorage. Dimarella riparia larvae collected at A. clepsydra sp. nov. type locality, and in the same microhabitat, were able to develop normally in captivity under the same conditions, which suggests the new species is much more sensible to microhabitat alterations. Other 17 larvae collected in vi.2021 were able to develop normally in captivity when provided with more appropriate microhabitats.
Remarks. This species is remarkably distinguishable from other Ameromyia species, mostly because of the contrasting black and white coloration (Fig. 1E) which is only found in A. explicata sp. nov. (Fig. 1F) and in A. strigosa (Fig. 1A) (black and yellow in the latter). The conspicuous "hourglass-shaped" venation patterns were most common to specimens collected over Bahia state. The postventral lobes of the male ectoproct in this species are also much shorter than in other species (Fig. 18A, B), and male genitalic sac bear much shorter, subglobose clavate setae (Figs 7D,18D), in contrast to other Ameromyia species. This species wing venation patterns and crossvein density on both wings seems to be very plastic. Specimens collected at Bahia state have a biareolate pattern which resembles an hourglass or an "8" shape on presectorial and radial areas of both wings (Fig. 17B), while the specimens from Rio Grande do Norte state lacks this characteristic (Fig. 17C). Stange (1970) suggests that crossvein density may have relation with habitat humidity, with in this case makes sense as the Bahia specimens were all collected near rivers, while Rio Grande do Norte specimens are from deep into a semiarid region on Caatinga biome (Fig. 12). Distribution. Argentina; Paraguay (Fig. 12).

Diagnosis.
Wing veins all pale. Cells in mediocubital area at least half brown suffused until at least 1/3 of wing span. Legs pale colored in contrast with dark abdomen and ventral thorax. Male ectoproct posterior margin with a weak dorsal notch.
Description. ADULT. Head: Antennae clubbed, 33 flagellomeres, dark brown with flagellum darker than sca-pus and pedicel. Vertex dark brown anteriorly, and light brown with dark brown patches posteriorly to anterior row. Vertex decumbent setae black, sometimes white on interantennal area. Interantennal area brown, usually with dark brown patches under scapus, in the middle of frons and just above clypeus. Frons pale brown, dark brown, or pale brown with dark brown patches. Frons setae black. Interocular distance less than eye width. Clypeus and labrum light brown. Mandibles dark brown. Palpi light brown, with external face of palpomeres slightly darker. -Thorax: Thorax dark brown, with light brown markings. Thorax on lateral view dorsally light brown and ventrally dark brown, with an evident longitudinal pale brown band under wings. Pronotum as broad as long, dark brown, with lateral margins pale brown, and a longitudinal medial pale brown streak. Pronotum marginal setae white. Mesonotum dark brown, with small light brown patches. Mesonotum with scattered white setae and curved white setae pointing posteriorly on mesoscutellum posterior margin. Mesoscutellum with white curved setae pointing posteriorly on posterior margin. Metanotum medially light brown and laterally dark brown, with white setae on posterior margins of metascutum and metascutellum. -Wings: Wing tip acute. Veins pale brown. Wing membrane hyaline, with dark brown infuscation on forewing mediocubital area, which is heavily suffused at forewing base and fading until apical third. Rhegmal area hyaline. Forewing vein CuP and hind wing vein CuA extending well before CuA fork and MP fork respectively, with at least nine crossveins connecting them and posterior branch of fork. Pterostigma light brown, opaque, encompassing six veins. Cells delimited by transversal veins in hind wing posterior margin higher than long. -Legs: Coxae pale brown. Femur and tibia with black bristles. Femur pale brown, with external face slightly darker than internal face. Meso and metafemur pale brown with internal face slightly darker than external face. Decumbent setae on basal half of profemur white, and black on distal half. Profemural sense hair longer than profemur and mesofemural sense hair much shorter than mesofemur. Tibia pale brown. Tibial spurs slightly shorter than pretarsal claws. Five tarsomeres, pale brown with distal joint slightly darker. Pretarsal claws at least Remarks. The type of A. dimidiata has lost both its hind wings (Fig. 20A). The type of A. baronei could not be analyzed during this work, but another specimen with this name identified by Navás, who described this species, was analyzed through photographs (from FML collection) and it agrees with A. dimidiata. The original description of A. baronei also agrees with A. dimidiata and therefore, we chose to maintain Stange's (1967) synonymy.
A single male specimen labeled from Colombia were identified from the CASC collection. As Colombia is way beyond A. dimidiata distributional range, and this particular batch of specimens (the ones labeled as bought from F. H. Walz.) also had many other label issues, in this sense, we interpret this Colombian record as dubious data. Stange (1967) registered this species for Uruguay (to Colonia and Rivera provinces), but it is unclear where the specimens analyzed in his work are deposited. These specimens are probably deposited in the institutions mentioned in his aknowledgements (Universidad de la República de Montevideo, Museu Argentino de Ciencias Natrales, the MLPA, and Observatorio de Física Cosmica), but as it is currently uncertain, we chose to refrain from register this species to Uruguay, although it is highly probable that A. dimidiata can indeed be found in Uruguay. Stange (2004) also previously registered this species for Brazil, but upon analysis of the registered specimen, it was in fact a misidentified specimen of A. tendinosa collected at the state of Mato Grosso.
This species is distinguishable from other Ameromyia species due to its veins on forewing which are all pale, and the absence of a rhegmal infuscation on forewing (Figs 20A,21A). A. pleuralis and A. tendinosa look very similar, but the first has a brown tint on forewing mem- brane and a dark colored forewing CuP, and the latter has a much darker overall coloration and also has a dark CuP vein. Distribution. Brazil (Fig. 12).
Other material. Etymology. The name "explicata", which means "unfolded". This refers to the fact that, during this work, the only males whose genitalia were in an "unfolded" state ( Fig. 10B, D) were of this species.

Biology.
A live female specimen was observed in the wild at Santa Rita de Cássia, BA, Brazil (Fig. 1F), and posteriorly reared in captivity. The location where the specimen was collected is located at the Cerrado biome, with dry vegetation and hard, compacted soil. The specimen observed was a high flyer, flying almost two meters above ground. It came towards the light trap, but apparently it wasn't strongly attracted to light. When reared in-lab,  this adult specimen accepted as food a mixture of honey and water and also live maggots when handfed. It fed and hunted by itself when offered D. melanogaster as prey.
Remarks. This species is very similar to A. nigriventris. The most evident distinction is the completely dark coloration of the thorax lateral sides (Fig. 3C). Even so, it can be hard to differentiate poorly conserved specimens, as specimens become paler over time. Male genitalia of both species are also similar, but A. explicata sp. nov. has shorter gonarcus membranous arms, and the shape of the paramere teeth is broader than in A. nigriventris (Fig. 10). Females can be promptly distinguished without need of genitalia dissection by its slightly convex pregenital plate margin (Fig. 23D), while males need to be dissected for a proper identification when devoid of coloration. This species is the only species whose males were found with "unfolded" parameres ( Fig. 10B, D). This species has a northern distribution in Brazil, from the Amazon Forest to the Northeastern region. Specimens collected in Miguel Calmon, Bahia, Brazil, cohabit this locality with A. nigriventris. However, all A. explicata sp. nov. specimens were collected at January, and during a field trip to Miguel Calmon in June, only A. nigriventris specimens were able to be found in the wild, which might suggest these species are separated by seasons. The specimens collected in the Amazon were found in "campinas", which are open areas with sandy soil distributed within the Amazon rainforest. Those specimens have a few differences from the northeastern brazilian specimens, but those are mostly slight variations in thorax and leg coloration and are hereby considered to be the same species despite the biome and geographical disparity. A few female specimens from Maranhão state bear a noticeable posterior projection on the pregenital plate, slender and longer digging setae on ectoproct and lateral gonapophysis, and a different pattern on wing coloration. It is possible those females are in fact a different species. However, until a male specimen is analyzed, we chose to identify these specimens as A. explicata sp. nov., and opted to not include them among the paratypes. (Stange, 1994)  Distribution. Venezuela (Fig. 13).

Ameromyia guarica
Diagnosis. Pronotum broader than long; femoral sense hairs absent; profemur without leg bristles; tibiae leg bristles equal or subequal to tibial width; pretarsal claws shorter than distal tarsomere; forewing with long rhegmal infuscation connecting both banksian lines. Remarks. This species is known only by the specimens presented in the original description. Previously a monotypic genus, Venezueleon is hereby synonymized under Ameromyia. Both genera had very similar diagnosis, differing only in regard of the size of the pretarsal claws (large in Ameromyia and small in Venezueleon) and profemoral sense hairs (long in Ameromyia and absent in Venezueleon) (Fig. 3B). Ameromyia guarica comb. nov.  also had some other differences such as the reduction of leg setae in general, but those characters are highly variable within Brachynemurini species, and in Myrmeleontidae in general as discussed by former authors (Stange 1970). Ameromyia guarica also has the male genitalia extremely similar to the Ameromyia species, and shares the clavate setae with a sickled club with other species in the modesta species group (Fig. 7C). Stange (1970) also suggests that, in Brachynemurini, a reduction of leg setae and tibial spurs size could be an adaptation to a diet of aphids and pollen (although no gut sampling was performed in this study). According to Stange (1994;, there are three undescribed species, all known from single specimens, which were collected near A. guarica type locality. Two of these specimens, collected at Falcon, Venezuela, have been analyzed but it is uncertain whether they are indeed new species. They differed mostly in coloration on pronotal and abdominal markings. Furthermore, only two larval specimens of Venezueleon sensu Stange were located and analyzed (Figs 26,27). Both specimens were not labeled as V. guaricus, but one as "Venezueleon sp. nov.", preserved in alcohol (Fig. 26), and the other was pinned and unlabeled (Fig. 27), but placed in the same tray as other A. guarica specimens, in which the tray itself had a "V. guaricus" label. Both larval specimens appear to be from different species, but it was unclear which one was actually the larvae of A. guarica. Although Venezueleon was previously described as having no thread-like setae, both specimens bear this character. White or pale thread-like setae are very difficult to be visualized in alcohol ( Fig.  26A-C), but very evident when the specimen is dry (Figs 26D,E,27). Additionally, the alcohol preserved specimen appears to be covered in a mold-like substance, which is tangled in the thread-like setae and further difficults the visualization of the latter structure (Fig. 27D, E). Furthermore, as the specimen labeled as "Venezueleon sp. nov." is the one that could fit A. guarica original description in this scenario (Fig. 26), we are assuming this is the one reared and associated by the author (Stange 1994), which is here included among A. guarica examined material, but more data is needed to further confirm the association.    Remarks. Ameromyia modesta is recorded from Brazil and Suriname for the first time. The type locality known as "Akuriman" or "Acurima", refers to a mountain near the border between Venezuela and Brazil (Roraima), on the municipality of Gran Sabana. Stange (1994) describes A. modesta larvae as bearing thread-like setae, but on the remarks for the Venezueleon larvae, states that in A. modesta, the thread-like setae are absent. Upon analysis of this larval specimen, it is evident that the larvae of this species indeed bear thread-like setae on head capsule, and thoraxic and abdominal dorsal surfaces (Fig. 30). This species is very similar in coloration to A. guarica, but it bears profemoral and mesofemural sense hairs, larger pretarsal claws and a longer than broad pronotum. Regarding other species in the same group, A. modesta differs from A. pubiventris in the pronotal markings (in which the pronotum is mostly light brown with a longitudinal stripe) (Figs 1C,2C,28A) and in leg coloration (Fig. 28B), while A. protensa is much larger in size, and is restricted to southern South America.  brown band under wings. Pronotum as broad as long. Pronotum dark brown, with a yellowish-brown longitudinal stripe medially and near lateral margins. Mesonotum dark brown, with yellowish-brown on notal junctures, and medially on mesoscutellum. Metanotum dark brown with yellowish-brown on notal junctures. -Wings: Forewing apex weakly falcate, and hind wing falcate. Veins with dashed dark and pale brown pattern. Wing membrane hyaline, with a dark infuscations on crossveins of forewing mediocubital area, along CuA vein, and on rhegmal area. Forewing mediocubital area with continuous or dotted dark infuscations around vertical crossveins at least until CuA fork. Forewing CuP and hind wing CuA veins extending before CuA and MP fork respectively, with nine or more crossveins connecting them. Pterostigma opaque, dark brown on basal half and white on apical half, encompassing six to seven crossveins. Cells delimited by transversal veins on hind wing posterior area higher than long.
-Legs: Coxae dark brown. Femur and tibiae with black bristles. Legs light brown, with many dark brown spots. Femur decumbent setae white and black. Profemural sense hair longer than profemur and mesofemural sense hair much shorter than mesofemur. Tarsomeres light brown, with distal fourth dark brown. Pretarsal claws at least 1.5 times longer than basitarsus. -Abdomen: Abdomen dark brown, with white setae on first sclerites, and brown setae on remaining sclerites. Male parameres oblong or subtriangular on lateral view. Gonarcus arch positioned not above paramere plates on lateral view. Male genitalia clavate setae with globose club. -LARVA. Larvae unknown.
Remarks. Ameromyia muralli is recorded from Bolivia for the first time, more precisely in Puerto Suárez, right on the border with Mato Grosso do Sul state in Brazil. The analysis of the holotype was based on old photo slides (Fig. 31C), but the original description greatly fits the analyzed specimens, as A. muralli has a conspicuous combination of a falcate hind wing and homogeneous coloration on profemora. The type's label only states "Urucum" (without other information such as date and collector), and Navás (1932) believed it actually refered to "Urucumacuã", a municipality located in the state of Rondônia (RO), Brazil. However, there is a locality named Urucum in the state of Mato Grosso do Sul (MS). This displacement may be explained because at that time both RO and MS were considered part of the same large state Mato Grosso (MT), which was posteriorly divided in the three current states. We believe that Urucum (MS) is the actual type location (instead of Urucumacuã as interpreted by Navás), as it is the literal label locality and other specimens herein analyzed, and today also housed in European collections, were collected at Urucum around the same time the holotype was probably collected.
This species can be readily recognized by the falcate hind wings (Fig. 31B), which are only shared by some specimens of A. strigosa (Fig. 44B)  ly) and A. clepsydra (much more strongly falcate at wing apex) (Fig. 17B). It also presents a dark longitudinal infuscation on forewing, which can appear as a continuous streak (Fig. 31C) or as an evident dotted line on forewing mediocubital area (Fig. 31B), which differentiates it from A. strigosa characteristic oblique wing markings. Another difference from A. strigosa is the presence of both white and dark decumbent setae on profemur.  zil. Eight adult specimens (three males and five females) and 14 larvae were captured and subsequently reared in captivity. Adults were paired in different sex couples in order to observe possible matings, and females were provided with earth and sand from their natural habitat in order to observe possible oviposition. A single captive female was observed in posture, similar to what as described by Stange (1970) and Miller (1990), with raised wings, and curved abdomen with the terminalia buried in the substrate, but no eggs were found. No mating behaviour was observed. All adult specimens were capable of flying high, and whenever disturbed while flying near the ground, would quickly ascend and get dragged by the wind, without losing control of their flight. Adults would start actively flying and hunting at dusk, significantly reducing their activity approximately one hour later, limiting their flight to only flying from place to place. Whenever perched or resting, specimens would fold their wings over the abdomen or keep their wings raised, with the hind wings twisted in a slightly different angle than those of the forewings. Wings rested in different positions were usually a sign of senescence or health issues. Larvae were found on patches of sandy, loose earth above compacted clay soil. Almost always, the larvae were inside a dug out "trench", achored to the hard substrate ( Fig. 16B-D). These "trenches" usually had a slope side, in which the larvae would stay anchored and submerged in the sandy soil, but with the mandibles open and exposed (Fig. 16B). All suitable patches of loose earth in which the larvae were found were isolated and considerably distant from vegetation or any kind of pro-tection, such as shade, rocks or overhangs (Fig. 16A). Patches of loose earth with debris, such as sticks, leaves or pebbles were usually devoid of A. nigriventris larvae. These debris, whenever moved by the wind, would "dig out" the sand and the hard soil, which probably disturbs A. nigriventris microhabitat and pushes the larvae away to a new location.
Remarks. All the type specimens were analyzed, except for A. stevensi striolata holotype. Ameromyia nigriventris previous syntypes consists of one male (paralectotype) (Fig. 32C) and one female (lectotype) (Fig. 32A) with the same label, and another female with a different label (paralectotype) (Fig. 32B). Although all three are stated to be from "Amazon", only the female paralectotype has the country of origin (Colombia). The lectotype was chosen in regard to the most common morphotype among the specimens analyzed. Each type series specimen has a different degree of wing infuscation, reflecting this species great plasticity regarding wing characters. The female paralectotype have broader wings, has a clear forewing membrane on presectorial area, and bear a dotted infuscation pattern on forewing mediocubital area (Fig. 33C), which resembles the same pattern found on A. pubiventris (Fig. 5A). The male paralectotype shows a black CuP vein and an almost completely suffused mediocubital area on the forewing (Fig. 34A).
In this work, we partially agree with Walker circumscription for A. nigriventris. The female paralectotype morphotype is most common in central Brazil, and this is the only specimen analyzed with this morphotype outside Brazil. A single male specimen with this morphotype was analyzed, and male genitalia agrees with A. nigriventris. However, the male paralectotype is enigmatic. It has a dark forewing CuP and hind wing CuA, as well as a suffused forewing mediocubital area, which agrees with A. tendinosa, but forewing RP and hind wing MA and MP have dashed patterns, in contrast to A. tendinosa respective wing veins that are completely dark. The completely suffused mediocubital area on forewing was present only in this specimen among all A. nigriventris analyzed during this work. Additionally, A. tendinosa has not been found outside central South America. In fact, no other species that bear a complete suffusion of forewing mediocubital area has been registered for Colombia. Nonetheless, despite the inconsistencies, we are still identifying the male paralectotype as an anomalous specimen of A. nigriventris. However, it is possible that this specimen in fact does represent a different species and further analyses could better elucidate the identity of this specimen. Unfortunately, we were not able to analyse its genitalia.
Ameromyia stevensi holotype (Fig. 35B), a female also from Amazon, agrees exactly with the lectotype of A. nigriventris and is hereby synonymized. Regarding A. stevensi striolata, albeit the type specimen was not seen, the original description agrees with A. nigriventris and therefore, in this work we chose to maintain this species as a synonym as proposed by Stange (1967).
We also agree with Stange's synonymy of A. longiventris under A. hirsuta (Stange 2010) as both type spec-imens are extremely similar, however, in addition, we are proposing the synonymy of A. hirsuta under A. nigriventris, a synonymy that was proposed before by Esben-Petersen (1920). Ameromyia hirsuta (Fig. 35A) does not show any exclusive consistent characteristic from A. nigriventris while also exhibiting extremely similar male genitalia. The brown tint on forewing membrane along the costal area, which is somewhat characteristic of specimens from Argentina and Paraguay, can also be found in many specimens of A. nigriventris, which were collected far beyond A. hirsuta previous distributional range, such as northeastern Brazil. In fact, many other species of Ameromyia can bear a tinted wing membrane as a polymorphic character. Female specimens from Argentina and southern Brazil sometimes bear an opaque streak on hind wing apex (Fig. 35A), but this is not a consistent character as it is not present in all females. The broad wing areas of the holotype is also not present in all specimens.
Ameromyia nigriventris larvae was previously described as not having thread-like setae (Stange 1994). However, all analyzed larvae indeed beared short, white thread-like setae on head, thorax and abdomen dorsal surface (Fig. 36D). White thread-like setae are very difficult to be visualized in alcohol, but very evident when the specimen was dried, which could explain the apparent absence of these setae on previous descriptions.
Ameromyia nigriventris is the species with the largest degree of variation regarding wing infuscation patterns within the genus. Wings can bear color pattern equal to that of the lectotype, equal to that of the female paralec- totype; a combination of both; devoid of infuscation or with an almost imperceptible infuscation in the aforementioned patterns. Ameromyia explicata sp. nov. and A. tendinosa are very morphologically similar to A. nigriventris. However, the first has a fully dark meso and metathorax on latheral view, while A. nigriventris has an evident lateral pale band, and the latter has a complete suffusion on forewing mediocubital area, as well as a dark colored forewing CuP vein, in contrast to A. nigriventris, which suffusion on forewing mediocubital area varies, and has a dashed pattern on forewing CuP.  Distribution. Brazil (Fig. 12).
Diagnosis. Forewing veins C, Sc, and RA all pale, in contrast to MA which has a dashed pale brown and dark brown pattern, and CuP which are dark brown.
Description. ADULT. Head: Antennae with 31 flagellomeres. Flagellomeres dark brown, with ventral face of the club light brown. Vertex light brown with dark brown spots on setae insertion, and dark brown spots and patches on anterior and posterior rows. Vertex decumbent setae black. Interantennal area dark brown. Frons light brown, usually with dark brown medially and under scape. Frons setae black. Interocular distance less than eye width. Clypeus light brown, slightly darker medially. Labrum light brown, slightly darker on ventral margin. Mandibles dark brown. Palpi light brown, with exterior face slightly darker. -Thorax: Thorax on lateral view darker ventrally, with an evident longitudinal pale brown band under wings. Pronotum as broad as long, light brown, with two longitudinal dark brown markings medially. Pronotum marginal setae white. Mesonotum dark, with longitudinal light brown streaks on each side of mesoscutum, and one medially on mesoscutellum. Metanotum light brown, with metascutum laterally dark. -Wings: Wing tip acute. Forewing veins Sc and RA vein pale; CuA, anterior branch of CuA fork and MP with a dashed dark and light brown pattern, and remaining veins are dark brown. Hind wing veins C, Sc and RA vein pale, and remaining veins are dark brown. Wing membrane hyaline, except on Costal and Radial areas, which have a brown tint. Forewing mediocubital area with dark infuscations around vertical crossveins, in a continuous pattern or apparently dotted but with connecting suffusions pattern, at least until apical third. Forewing CuP vein and hind wing CuA extending well before forewing CuA fork and hind wing MP fork respectively, with nine or more crossveins connecting them and posterior branch of fork. Pterostigma opaque, light brown, encompassing five to six crossveins. Cells delimited by transversal veins in hind wing posterior area higher than long. -Legs: Coxae light brown on posterior face and slightly darker on anterior face. Femur and tibiae with black bristles. Legs light brown, dark brown on exterior face. Femur decumbent setae white and black. Profemural sense hair longer than profemur, and mesofemural sense hair much shorter than mesofemur. Tibiae with dark brown spots on bristle insertion. Tibial spurs slightly shorter than pretarsal claws. Tarsomeres light brown, with posterior tip dark brown. Pretarsal claws at least 1.5 times longer than basitarsus. -Abdomen: Abdomen dark brown with many dark brown setae. -LARVA. Larvae unknown.

Remarks.
Ameromia pleuralis holotype is labelled as "syntypus" (Fig. 37A), but is stated as holotype in the original description, as it is the only specimen being described. A. pleuralis was by far the less sampled species by the current work. Specimens analyzed were restricted to central Brazil. All sampled male specimens have lost their terminalia. It is possible that the analyzed specimens of A. pleuralis are simply decoloured or badly preserved specimens of A. tendinosa, but without a larger sampling or without analysing the male genitalia, it is difficult to ascertain if A. pleuralis is in fact a valid species or if it's a synonym of A. tendinosa. Through photographs of the type specimen (Figs 37A,C), the forewing CuP is clearly dark, but, while the RA vein appears to be pale, distalmost sections appears to be dashed and thus it is difficult to discern. Analyzed specimens had fully pale RA veins (Fig. 37B). Nonetheless, based on the analysis of the type specimen photographs and the differences in the wing coloration, we decided to keep A. pleuralis as a valid species.
This species is very similar to A. tendinosa and A. dimi diata, but the latter has a southern distribution and has a pale forewing CuP vein, while A. tendinosa has a forewing R vein with a dashed coloration distally.  Distribution. Argentina, Brazil, Paraguay; Uruguay (Fig. 13).   (Fig. 39A), which are absent for A. pubiventris. Therefore, in this work we agree and maintain Stange's (1967) synonymy of F. trapezia under A. protensa. The type specimens of A. alterna and A. decarloi were not located, and are still missing as previously mentioned by Stange (2004;2010) and Oswald (2021). However, the original description for those species also agrees with A. protensa, in the same way as F. trapezia, and therefore we chose to maintain both species as synonyms.

Ameromyia protensa
Ameromyia protensa is the only species from the modesta group that occurs in Argentina, Paraguay and Uruguay (Fig. 13). This species is restricted to southern South America, and its northermost distribution record continues to be its type locality, in São Paulo state, Brazil.   Remarks. Walker (1860), in Myrmeleon pubiventris original description, did not designate a holotype, but he provided species measurement variations, "Length of the body 19-23 lines; of the wings 27 lines". This suggests he analyzed more than one specimen as interpreted by Stange (2004), who mentioned syntypes in his catalog. We agree with Stange's interpretation of syntypes, however, at the BMHN there is only one male specimen labeled as type (Fig. 40B), and thus we are hereby designating this specimen as the lectotype. Furthermore, Walker (1860) did not state any data label, but only mentioned "Amazon Region". The lectotype label states "Santarem" which is a municipality in Pará state, located at the Brazilian Amazon region. During this work, only specimens from Brazil were analyzed. However, Banks (1943) registered this species from Surinam (which was posteriorly mentioned by Stange (2004)). Stange (2004) also mentioned that the species occur in Venezuela, but without further details. Furthermore, we did not find this Venezuelan specimen in Stange's collection, but we found one specimen of A. modesta from Venezuela misidentified as A. pubiventris. Therefore, we are considering this distribution data as du-bious and we are not displaying it on our distributional map.
Despite the records for Surinam and the type locality in the Amazon Region, this species appears to be more abundant in open areas in the Cerrado and Caatinga biomes in Brazil.
Ameromyia pubiventris can be easily differentiated from A. modesta and A. guarica by leg coloration, and from A. protensa by the much smaller posterior area of forewing, and dashed Radial sector veins (Figs 5A,40A). Ameromyia pubiventris also has slender wings in comparison to A. protensa.
Distribution. Argentina; Brazil; Bolivia; Chile; Paraguay (Fig. 12 Forewing with dark brown infuscations in wing membrane alternating around crossveins on presectorial and radial areas, and along crossveins on sectorial, mediocubital and rhegmal areas, forming five evident oblique streaks around MA vein. Forewing mediocubital area with dark infuscations around vertical crossveins which become sparser along MA vein, culminating on rhegmal area. Forewing CuP and hind wing CuA veins extending well before forewing CuA and hind wing MP fork respectively, with nine or more crossveins connecting them and posterior branch of fork. Pterostigma opaque, dark brown on basal half and white on apical half, encompassing six to seven crossveins. Cells delimited by transversal veins on hind wing higher than long. -Legs: Coxae dark. Femur and tibiae with black bristles. Femur dark on external face and light yellow on internal face. Profemur decumbent setae white. Meso and metafemur decumbent setae white and black. Profemural sense hair longer than profemur and mesofemural sense hair much shorter than mesofemur. Tibia light yellow with dark spots on setae insertion. Tibial spurs slightly shorter than pretarsal claws.
Tarsomeres light yellow, with distal fourth dark brown. Pretarsal claws at least 1.5 times longer than basitarsus.
-Abdomen: Abdomen sclerites black, sometimes with yellow on posterior margins. Abdomen with white setae, brown near terminalia. Male ectoproct obliquous in relation to tergite IX, with postventral lobe slightly elevated. Male paramere plates subtriangular on lateral view. Parameres hinge anterior projections not entirely covered by gonarcus. Gonarcus arch positioned not above parameres plates. Male genitalia clavate setae with globose club. -LARVA. Larvae unknown.

Biology.
A live male specimen was observed in wild at Santa Rita de Cássia, Bahia state, Brazil. The location where the specimen was collected was at the Cerrado biome, with dry vegetation and hard, compacted soil. This specimen hovered in the air almost two meters above the ground.
Remarks. This species, previously restricted to Chile, Paraguay and Argentina, is now registered for the first time to Bolivia. It is also now recorded to Brazil, as Ameromyia pentheri is hereby proposed as a synonym of A. strigosa. The holotype of A. pentheri (Fig. 41A-C) seems to be a teneral specimen and is devoid of coloration, which is problematic for identifying Ameromyia specimens. It also has lost its terminalia. However, the holotype chaetotaxy agrees with A. strigosa, as it bears only white setae on frons, thorax and profemur. Ameromyia strigosa distribution, though, was previously restricted to southern South America, with no records from Brazil, and the only brazilian specimens identified previously to the synonymy were from Mato Grosso do Sul state, which borders Bolivia and therefore are very close to A. strigosa previous distributional range. The locality on A. pentheri holotype label only stated it was from "Santa Rita" in Brazil, which relates to more than a dozen municipalities with the same name. Nevertheless, upon analyzing the course of the expedition that collected the type specimen (dubbed as "Penther expedition" by Navás), as detailed in Bohme (1996), it became clear that "Santa Rita" referred to the municipality of Santa Rita de Cássia, Bahia state, Brazil. To ascertain the occurrence of A. strigosa in Bahia, a field trip was made to that same locality, which after three days resulted in a single collected male specimen that agrees with both A. pentheri and A. strigosa type specimens (Fig. 42A). A female specimen that agrees with both A. strigosa lectotype and the male specimen found at Santa Rita de Cassia, was located at the MZFS collection, also collected from Bahia state, Brazil (Fig. 42B).
The type specimen of A. fidelis was not analyzed in this current work, but the original description greatly fits that of A. strigosa, and therefore we chose to maintain the synonymy proposed by Stange (1967). Regarding the combination Moza strigosa by Navás (1917), although this combination was never officially undone, it seems that the author himself disregarded his nomenclatural changes as he kept referring to this species as A. strigosa in all of his following publications.
Ameromyia strigosa is very distinctive as it bears four to five oblique markings along MA vein in the forewing (Figs 1A,43C,44A). Also, frons (Figs 2A,41D) and leg short decumbent setae (Figs 2A,41B,C,43B) are almost all white, and body coloration is black and yellow or yellowish-brown (Figs 1A,42). The only other species that looks somewhat similar is A. muralli, which can also show some yellowish-brown coloration. Some A. strigosa specimens bear a falcate hind wing (Fig. 44B)   Distribution. Brazil (Fig. 12).

Diagnosis.
Wing veins CuP completely dark brown in contrast to other veins which have a dashed dark brown and pale brown pattern. Forewing cells in mediocubital area almost completely suffused until at least the middle of wing span.
Description. ADULT. Head: Antennae with 30-32 flagellomeres. Flagellomeres dark brown, except for the ones at the club, which are light brown dorsally or ventrally. Vertex light brown, with latitudinal dark brown streaks in anterior and middle row, or completely dark brown with faint light brown laterally. Vertex decumbent setae black. Interantennal area dark brown. Frons light brown, frequently with dark brown patches centrally, or entirely dark brown, light brown near clypeus. Frons setae black. Interocular distance less than eye width. Clypeus light brown brown with dark brown patches. Labrum light brown brown. Mandibles light brown on basal half, and dark brown on apical half. Palpi light brown, with external face of palpomeres slightly darker. -Thorax: Thorax dark brown with few light brown markings. Thorax dark brown on lateral view, with an evident longitudinal pale brown band under wings. Pronotum as broad as long, dark brown, with light brown anteriorly and laterally. Pronotum marginal setae white, or dark brown. -Wings: Wing tip acute. Veins with dashed dark and pale brown pattern, except for forewing and hind wing veins CuP which are dark brown. Wing membrane hyaline, except for forewing mediocubital area which is almost completely dark brown infuscated until forewing apical third, and the rhegmal area which is faintly dark brown infuscated. Forewing and hind wing sometimes with a faint dark brown infuscation alongside gradates. Forewing CuP and hind wing CuA veins extending before forewing CuA and hind wing MP fork respectively, with nine or more crossveins connecting them and posterior branch of fork. Pterostigma opaque, dark brown on basal half and white on distal half, encompassing six to seven crossveins. Cells delimited by transversal veins on hind wing posterior area higher than long. -Legs: Coxae dark brown. Legs internal face light brown and external face dark brown, with many dark brown spots on setal insertion, and on internal face in meso and metalegs. Femur and tibiae with black bristles. Femur decumbent setae white and black. Profemural sense hair longer than profemur and mesofemural sense hair much shorter than mesofemur. Tibial spurs slightly shorter than pretarsal claws. Tarsomeres light brown, with distal fourth dark brown. Pretarsal claws at least 1.5 times longer than basitarsus. -Abdomen: Abdomen sclerites dark brown, with light brown in posterior margin near the terminalia, with brown setae. Male paramere plates oblong on lateral view. Gonarcus arch positioned not above paramere plates on lateral view. Male genitalia clavate setae with globose club. -LARVA. Larvae unknown.
Remarks. The taxonomic background for this species is confusing. Decades after its original description by Gerstaecker, Esben-Petersen (1920) synonymizes this species (as Myrmeleon tendinosus) under Amazoleon pubiventris along with Foya trapezia. In his words, M. tendinosus "is the same species" according to the type specimen. However, on the following paragraph the author also seemingly synonymizes it again, but under another species, A.     This species is restricted to Brazil. The male paralectotype of A. nigriventris somewhat agrees with this species and can possibly be a specimen of A. tendinosa, which would expand this species distribution to the Amazon region including Colombia. Additionaly, three female specimens labeled as from Colombia were identified from the CASC collection. However, as A. nigriventris male paralectotype genitalia was not analyzed in the present work, and as this particular batch from where these Colombia female specimens come from (the ones labeled as "bought from F. H. Walz.") had some other label issues, we interpreted these Colombia records as dubious data.
Ameromyia tendinosa can be distinguished from other species in the nigriventris group by the dark forewing CuP and by the forewing mediocubital area which is completely suffused with dark brown infuscation from wing base at least to CuA fork (Figs 45,46A). A. pleuralis, A. dimidiata and A. muralli shows the same suffusion on mediocubital area, but the first two species have pale forewing Sc and RA veins, and the latter shows a dashed forewing CuP with alternating coloration, as well as a falcate hindwing.
The setose frons as well the enlarged pretarsal claws were found to be synapomorphies of the more inclusive group, the sampled Brachynemurini genera (Ameromyia, Argentoleon and Austroleon), and although Ameromyia specimens frequently have much more density of setae on the frons than Argentoleon and Austroleon, this appears to be a variable character and some specimens appear to have fewer setae than others. On the other hand, character state 39:1 "Clavate setae on male genitalia appears as a synapomorphy of Ameromyia by the present analyses.
The long CuA of the hind wing was discussed by Stange (1970) as an evolutionary trend in some species of Brachynemurini, which caused the broadening of the hind wing posterior area. In his observations, species with broader posterior area of the hind wings were more rapid fliers than those with narrower wings, and this character was associated as a possible adaptation to live in arid regions, with less vegetation and less air humidity, which would pose more difficult flight conditions. However, Ameromyia comprises species with both narrow (modesta group species) and broad (nigriventris group species) hind wing posterior areas, and species with either wing patterns are able to cohabit in many localities (Figs 12,13), despite the characteristics of vegetation and air humidity. Among studies of wing functional morphology, broader wing bases are usually indication of a grater capability for acceleration, while slender bases are more characteristic of slower, hovering flight (Ennos 1989;Wootton 1992Wootton , 2003. This agrees with field observations. Ameromyia explicata sp. nov., A. nigriventris and A. strigosa specimens were able to fly slowly and hover, but were relatively quick to gain speed whenever disturbed. No species from the modesta group were observed live in the wild during this work, but A. clepsydra sp. nov., which has a short hind wing CuA similar to specimens from modesta group, and Argentoleon irrigatus specimens, which also bear somewhat similar, slender posterior wings, were observed in flight. Ameromyia clepsydra sp. nov. specimens were always flying close to the bedrock, making short, slow flights, and Argentoleon irrigatus specimens observed had similar behaviour, but were flying a bit higher and maintaining hovering flight. Some A. irrigatus specimens tried to flee by flying upwards, but once they were caught by the wind, they quickly lost control of their flight and dropped to the ground, in contrast to Ameromyia specimens from the nigriventris group which flied high up normally, even when caught by the wind. It is possible that species from nigriventris group are more adapted to flying and hunting higher up in the air, while species from modesta group might specialize in flying closer to the ground and in the vegetation, and that the different size of the hind wing CuA, and subsequently of the posterior area, reflects these adaptations between these species' groups. Ameromyia male genitalia was previously described as having no "hinge" (sensu Adams 1956); having modified rod-like parameres; a reduced gonarcus and an enlarged mediuncus which covered the parameres (Stange 1994). Extensive examination of Ameromyia male genitalia, however, revealed that the male genitalia is indeed hinged like most Brachynemurini genera, including Argentoleon and Austroleon male genitalia, which were also previously described as unhinged (Stange 1994). The structure described as a "reduced gonarcus", as it happens, are the parameres anterior to the hinge (which has two parallel rows of setae), and are fused anterodorsally (Figs 9, 10); the "rod-like parameres" are in fact the gonarcus (Fig.  10); and the enlarged mediuncus is actually the enlarged paramere plates (Figs 9,10).
In Brachynemurini males, parameres have folds that, upon genitalia evertion, produces a movement in which the parameres fold upon themselves and greatly changes the genitalic complex apparent conformation (Addams 1956;Stange 1970). Ameromyia male genitalia is most commonly found in an "folded state" (Figs 9, 10A, C), in which the paramere plates are parallel to each other and posterior to the gonarcus. After genitalia eversion, the paramere plates unfold upon the hinge, in a way such that: the paramere plates dorsoposterior margins touch the genitalia ventral surface; the external striated faces of the paramere plates faces one another; and the paramere ventral hooks move posteriorly, in a "pinch" movement ( Fig. 10B, D). Of the many dissected male specimens, only two specimens of A. explicata sp. nov. were found with the genitalia in the "unfolded" state ( Fig. 10B, D), and it is unknown what provokes or could provoke the parameres movement upon or after specimen collection.
The cladistic analyses recovered two species groups as it were previously suggested by Stange (modesta and nigriventris groups) (Stange 1994), which can be readily separated by the height of the cells on hind wing posterior margin. The first species group, named by Stange (1994) as modesta group, is supported in resulted analyses by the abdomen much longer than wings (tergite 3+4 length equals to hind wing length) and the sickled club of the clavate setae on male genitalia. Ameromyia protensa was recovered sister to the remaining species in this group, and the latter group was supported by the presence of black-only decumbent setae on profemur. Ameromyia pubiventris was found sister to A. modesta + A. guarica in all analyses, with the latter group sharing the same state for characters 11 and 12 (contrasting coloration on external and internal face of profemora and tarsomeres, respectively). Although this result shows little branch support, it makes sense geographically considering that both A. modesta and A. guarica are restricted to the same distributional range (northernmost South America), while A. pubiventris ranges from the Amazon rainforest to southeastern Brazil, and A. protensa is restricted to southern South America (Fig. 13). Even with the sheer number of modifications and homoplasies, A. guarica, previously placed in Venezueleon, was recovered nested deep within the modesta group in all present analyses, which suggests that the congruency of main wing veins and genitalic characters that supports the genus outweights the leg and chaetotaxy adaptations which diagnoses Venezueleon, as suggested by previous studies when in regard to Brachynemurini characters (Stange 1970(Stange , 1994. Stange (1994) also stated that wing venation in Venezueleon is different from Ameromyia, as is the latter the CuP follows the posterior branch of CuA fork for a longer distance, but the venation pattern in Venezueleon is not different from any Ameromyia species from the modesta group. Furthermore, it is also stated that the absence of thread-like setae on abdominal tergites in Venezueleon larvae could be a convergence with A. modesta. This was probably a mis-take, since A. modesta larvae is the one described by the author as bearing thread-like setae on abdominal tergites. Additionally, all Venezueleon larvae analyzed in this work bear thread-like setae not only on the abdomen (tergites and sternites), but on the thorax as well (Figs 26,27).
The functionality of the thread-like setae is not fully understood, but it might be related to helping the larvae camouflage by holding debris in the same way as conical dolichasters and plumose setae (Acevedo-Ramos et al. 2021;New 1986), as many larvae analyzed had threadlike setae tangled together with many small sand grains. However, some species also bear thread-like setae on the abdomen ventral surface (Fig. 27B), which suggests this structure might serve another function not yet comprehended, such as thermoregulative or mechanosensorial functions.
The modesta group corresponds to Amazoleon, previously delimited by Banks (1913). However, the high consistency of genitalic structures and chaetotaxy strongly suggests that both groups are indeed very closely related and are comprised in the same genus, agreeing with Amazoleon synonymy under Ameromyia (Stange 1967). Other Amazoleon key wing character, such as a hind wing CuP following posterior branch of MP for a short distance (Figs 4A,B,5A), are found in the nigriventris group in Ameromyia clepsydra (Fig. 17B), further supporting the strong relationship between these groups.
The second species group, nigriventris group, is composed of species that bear a broad posterior area of hind wing, with cells that are as high as long, or higher than long (Figs 4C,D,5B,C). This group inner relationships were different between the two recovered trees, as they disagreed on steps and state transformations of three characters (11 Profemur color pattern; 16 Forewing mediocubital area infuscation pattern and 17 Forewing presectorial and radial area infuscation pattern), and ultimately, the internal placement of A. muralli.
The first species to branch off inside the nigriventris group, A. clepsydra sp. nov., features wing synapomorphies that are common to both groups while having conspicuous autapomorphies. Ameromyia clepsydra has the forewing CuP and hind wing CuA following posterior branch of CuA and MP fork respectivelly for a short distance, which agrees with the modesta group, but hind wing posterior area is much broader, which agrees with the nigriventris group. The course of forewing vein 2A also fits the pattern of the modesta group species (forewing vein 2A connected to 3A by a crossvein). However, a single analyzed specimen had both patterns, one in each wing (Fig. 17D, E). Although this is most probably a mutation, as this population showed a higher density and intraspecific heterogeneity of wing crossveins, it raises the question if this might be an acquired modification that can ontogenetically influence the behaviour of venation patterns, even characters historically considered as conserved such as how forewing vein 2A relates to 3A (Banks 1927;Stange 1967;New 1984). An autapomorphy of Ameromyia, the clavate setae on male genitalic sac, is apparently modified in A. clepsydra, as in this species the tip of the genitalic setae is not clubbed, but slight-ly swollen (Fig. 7D). All analyses recovered A. clepsydra as more related to the nigriventris group, stemming from the clade base.
The remaining species, as in clade C in our proposed classification (Fig. 11A), can be diagnosed by the long forewing vein CuP and hind wing vein CuA, and forewing 2A vein which touches 3A vein. Although this clade seems to be a well established group, with well defined synapomorphies and good resampling supports, inner relationships do not seem to be well solved. In this group, remaining basal relationships in both trees are recovered based on the state of wing marking patterns on presectorial, radial and mediocubital areas, which is a very variable character both for Ameromyia as for Brachynemurini in general. In fact, those characters have evolved independently in A. protensa and A. guarica, and are polymorphic for every other species that presents it, with the exception of A. dimidiata, A. pleuralis and A. tendinosa (clade F).
The placement of A. muralli is different between both trees. In our proposed classification, this species was recovered as sister to clade D, while in the alternative tree it was recovered as sister to clade F. This rogue placement is the result of the great plasticity and homoplasy of the aforementioned forewing characters, which in the case of A. muralli, could allocate this species as sister to either clade under equal weighting analyses. Ameromyia muralli bear a profemur without contrasting colors, a suffused mediocubital area on forewing, and is polymorphic in relation to the state of the presectorial area infuscation pattern. In the proposed classification, the cladistic analysis indicates the profemur with contrasting coloration evolved independently, in both modesta and nigriventris groups (in clades B and D), while the presence of infuscation on forewing presectorial and radial area crossveins, and the multistate character of forewing mediocubital suffusion changed states many times inside Ameromyia. The alternative tree (Fig. 11B), however, suggests that the contrasting leg coloration also evolved once in both groups, but was reversed in A. muralli, and that the forewing mediocubital area infuscation with a dotted pattern that becomes sparser along the wing span, as well as the presectorial and radial crossvein infuscations would be in fact plesiomorphic states within the nigriventris group, but the first was afterwards modified into a complete suffusion of forewing mediocubital area, and the latter afterwards reversed in clade I.
All weighted analyses recovered the same tree as our proposed classification for the genus, suggesting that, although both results under equal weighting show the same number of steps, the character of leg coloration (11) has more congruency than those of wing membrane coloration (16 and 17). Indeed, whilst not verified by many phylogenies, historically, Brachynemurini taxonomy is heavily biased towards leg morphology, and the pattern of wing markings and crossvein infuscation are very variable across the tribe, as it is within Ameromyia as well (Stange 1970(Stange , 1994. Some patterns, as the complete suffusion of forewing mediocubital area, seems to be consistent within species, and even some Brachynemurini species diagnoses hinges on such characteristic (as in Argentoleon longitudinalis (Navás) or Scotoleon niger (Currie)), but forewing presectorial and radial crossvein infuscations seems very widespread, polymorphic and not very informative in regard to relatonship between groups.
Clade E, although supported by a single synapomorphy (character state 17:1 Forewing presectorial and radial area crossveins with infuscations), is recovered as an unresolved polytomy of A. explicata sp. nov., A. nigriventris and A. strigosa, reflecting the large morphological similarity of the species in this group. Clade F on the other hand is supported by two forewing synapomorphies: a forewing CuP vein with homogeneous coloration and a clear forewing presectorial and radial areas, although the latter is in fact a reversion of the state transformation of this character for the nigriventris group. In clade F, A. tendinosa is sister to a clade comprising A. dimidiata and A. pleuralis (clade G). Although both A. tendinosa and A. pleuralis possess a dark forewing vein CuP, the analysis recovered the latter more related to A. dimidiata based on the pale forewing veins Sc and R.

Conclusion
In this work, we present Ameromyia as a valid Bra chynemurini genus, with 12 valid species in two well defined species groups. Venezueleon was recovered deep inside Ameromyia and thus synonymized under the latter, which in retrospect makes taxonomical sense as the two genera had an overlap of their previously diagnostic characters, with some exceptions such as the size of pretarsal claws, which is a known plastic character at species level for Brachynemurini in general. Adult specimens from different species (but from the same species groups) bear almost identical male and female genitalia and are morphologically very similar, differing mostly in leg and wing vein coloration and wing marking patterns, although the latter appears to be very homoplasic and polymorphic within some species. More extensive analysis which utilizes different datasets such as larval morphology and/or molecular data could hopefully better elucidate Ameromyia internal relationships. This study also enlightens the need of better understanding the South American Brachynemurini genera, as results of this work partly contradict some statements from previous works, such as the case of all analyzed genera bearing hinged genitalia.