Eighty male specimens of 15 of the 16 species in the genus Apodemia and five Emesis species were examined. The material was collected in 2015–2017 in Mexico and gathered from Mexican as well as international scientific collections (Suppl. material 1). These specimens were selected in order to cover the phylogenetic diversity and the geographic distribution of the genus, with an emphasis on type localities. Specimens from A. selvatica De la Maza & De la Maza were not reviewed; however, its characteristics are discussed based on the original description (De la Maza and De la Maza, 2017b). Specimens examined came from the following scientific collections: Colección Nacional de Insectos del Instituto de Biología, México (CNIN-IBUNAM), Colección Lepidopterológica del Museo de Zoología de la Facultad de Ciencias, UNAM, Mexico (MZFC); Colección Lepidopterológica del Museo de Zoología de la FES Zaragoza, UNAM, Mexico (MZFZ); McGuire Center for Lepidoptera and Biodiversity, Florida Museum of Natural History, University of Florida, USA (MGCL) and Colección de Curtis Callaghan, Colombia (CJC).

The length of the right anterior wing from the base to the upper apex was measured. Labial palpi and legs were dissected using an Olympus SZX9 stereoscopic microscope, with a planar objective 1.5. The structures and wings were diaphanized by soaking them in alcohol and 5.25% NaClO solution (bleach). Then they were digitized and sketched using an Olympus DP12 camera attached to the microscope. Morphological terminology follows Comstock and Needham (1918), Harvey and Clench (1980), DeVries (1997), Penz and DeVries (1999, 2006) and Hall (1999, 2005). The following abbreviations were used: forewing (FW), hind wing (HW), dorsal (D), ventral (V). Male genitalia were extracted using an enzymatic digestion technique modified from Knölke et al. (2005). If needed, genitalia were soaked a few seconds in a hot potassium hydroxide solution (KOH 10%) to complete the cleaning. All structures were preserved in micro vials with glycerin solution and acetic acid at 4%. Terminology of genitalia descriptions follows Klots (1956), Eliot (1973), Harvey (1987), Penz and DeVries (1999, 2006) and Hall (2008). Digital images of dorsal and ventral views of the genitalia were taken using focus stacking of light microscopy with a Leica Z16 APO-A stereoscopic microscope, a Leica DFC490 HD camera, and the Leica Application Suite program. Cornuti images were taken with a ZEISS microscope AXIO Zoom. V16, with an AxioCam MRc5 camera, and the Zeiss Efficient Navigation program. All images were taken at the Instituto de Biología-UNAM.

Twenty-six specimens of Apodemia including eleven of the 16 currently recognized species were collected. Species not included were Apodemia chisosensis H. Freeman, A. virgulti (Behr), A. castanea (Prittwitz), A. planeca De la Maza & De la Maza, and A. selvatica. Some species (Apodemia chisosensis, A. castanea, and A. virgulti) were excluded due to the lack of sequences from nuclear genes which are more informative in generic level. Two other species (A. planeca and A. selvatica) have been recently described (De la Maza and De la Maza 2017a, b) and we did not have fresh tissue suitable for DNA extraction (Suppl. material 2). To evaluate the monophyly of Apodemia, sequences from 31 other riodinid species were included, representing all tribes of the family according to the most recent phylogeny of Riodinidae (Espeland et al. 2015). With the exception of three sequences of Emesis, other sequences of outgroups were taken from GenBank (Suppl. material 2).

DNA was extracted from legs or abdomen using the DNeasy Blood & Tissue kit (Qiagen, Valencia, CA, USA). Partial sequences of 623 bp of the mitochondrial gene Cytochrome Oxidase I (COI) were obtained. Also, two nuclear loci, 495 bp of the gene Elongation factor 1 α (EF-1a), and 402 bp of wingless (wg) were sequenced. These loci were selected because it has been shown previously they are informative at different levels of divergence within riodinid butterflies (Campbell et al. 2000, Campbell and Pierce 2003, Espeland et al. 2015, Proshek et al. 2015). Primer sequences for COI were taken from Proshek et al. (2013), for EF-1a from Monteiro and Pierce (2001), and for wg from Brower and De Salle (1998). All gene regions were amplified via polymerase chain reaction (PCR) in a 25 µL reaction volume containing 0.5–1.0 µL deoxynucleoside triphosphates (dNTPs; 10 mM), 18–19.25 µL double-distilled water, 0.2–0.5 µL each primer (10 mM), 2.5 µL 1X PCR buffer, 1.2 mM MgCl₂ (Fisherbrand, Pittsburgh, PA, USA), 0.15 µL Taq DNA polymerase (Fisherbrand), and 1.0–1.5 µL template DNA. For COI, DNA was denatured at 94 °C for 2 min, followed by 38–40 cycles of 94 °C for 30 s, 48–50 °C for 45 s, and 72 °C for 45 s. A final extension phase of 72 °C for 7 min terminated the protocol. For EF-1a and wg, DNA was processed according to a touchdown protocol suggest by Espeland et al. (2015) with an initial denaturation for 3 min at 94 °C, 20 cycles of 94 °C for 50 s, annealing temperature starting at 49 °C and ramping down 0.5° for every cycle for 40 s, 72 °C for 1 min, another 20 cycles of 94 °C for 50 s, annealing temperature (48–52 °C) for 40 s, and 72 °C for 1 min, and a final extension of 72 °C for 5 min. Double-stranded PCR products were checked by electrophoresis on a 1% agarose gel. PCR products were purified with polyethylene glycol precipitation (Lis 1980). DNA templates were sequenced in both directions with the Big Dye Terminator version 3.1 cycle sequencing kit (Applied Biosystems, Inc.) and an ABI 3100 automated DNA sequencer (Applied Biosystems, Inc.) using the amplification primers. Sequences were assembled and edited in the STADEN PACKAGE version 1.6.0 (Whitwham and Bonfield 2005).

Sequences were aligned using the MUSCLE algorithm (Edgar 2004) included in the software MEGA version 7 (Kumar et al. 2016). Sequences were uploaded to GenBank and accession numbers are listed in Suppl. material 2. Prior to the concatenated analysis, independent ML analyses were conducted for each gene. The phylogeny of the concatenated data set (n = 56 individuals, including outgroups) was inferred using Bayesian inference and maximum likelihood (ML) phylogenetic methods. For both methods, partitioned analyses were used to improve phylogenetic accuracy. The best-fitting substitution models and partitioning schemes were selected simultaneously using the Bayesian Information Criterion in the software PARTITIONFINDER version 1.1.1 (Lanfear et al. 2012). Bayesian inference analyses were conducted using MRBAYES version 3.2.1 (Ronquist et al. 2012). Four runs were conducted using the ‘nruns = 4’ command, each with three heated and one cold Markov chains with sampling every 1000 generations for 50 million generations. Output parameters were visualized using TRACER version 1.4 (Rambaut and Drummond 2007) to identify stationarity and convergence. Convergence between runs was assessed using AWTY (Nylander et al. 2008). After discarding the first 12.5 million generations (25%) as burn-in, parameter values of the samples were summarized from the posterior distribution on the maximum clade credibility tree using TREEANNOTATOR version 1.4.8 (Drummond and Rambaut 2007) with the posterior probability limit set to 0.1 and mean node heights summarized. Maximum likelihood analyses were conducted using RAxML version 7.2.6 (Stamatakis 2006) under the GTRCAT model, with 1000 nonparametric bootstrap replicates to assess nodal support. Nodes were considered strongly supported if their Bayesian posterior probability was ≥ 0.95 and their bootstrap value was ≥ 80 % (Huelsenbeck and Rannala 2004).

Finally, to obtain an estimate of genetic distances, pairwise genetic distances for COI were computed between and within the major clades obtained in the phylogenetic analysis (see results). The corrected pairwise genetic distances were calculated using the K2P model with MEGA version 7 (Kimura 1980, Kumar et al. 2016).

The results showed some incongruence between loci, but the major differences were between poorly-supported clades (Suppl. material 3–5). The final concatenated data set consisted of 1520 aligned nucleotide positions. The partitions and models that best fit the data were GTR+G (COI second positions, wg first and second positions, EF-1a third positions), and GTR+I+G (COI first and third positions, wg third positions, and EF-1a first and second positions). ML and Bayesian inference analyses resulted in highly congruent phylogenetic trees. The recovered relationships between the genera of Riodinidae were in agreement with recently published phylogenies (e.g., Espeland et al. 2015), thus providing a solid platform for the evaluation of the monophyly of Apodemia.

In the phylogenetic analyses, four clades can be distinguished (Fig. 1). The monophyly of Apodemia was not supported, as Apodemia phyciodoides W. Barnes & Benjamin was more related to Emesis than to other Apodemia. The rest of the species of Apodemia were included into three strongly supported clades. The first clade (Apodemia clade), included the North American taxa of the A. mormo complex: A. mormo and A. mejicanus (Behr), also A. duryi (W. H. Edwards) and A. multiplaga Schaus. The second clade included A. nais; and the third clade (Mexico clade) was composed by the Apodemia taxa distributed mostly in Mexico and Central America: A. hepburni Godman & Salvin, A. hypoglauca Godman & Salvin, A. murphyi Austin, A. palmerii W. H. Edwards and A. walkeri Godman & Salvin.

Figure 1. 

Phylogenetic relationships based on partial sequences of the mtDNA (COI), and nuclear genes (EF-1a, wg). Bayesian posterior probabilities and bootstrap values of major nodes are indicated with black dots for well-supported nodes. Vertical lines correspond to the major clades found in this study.

The relationships between the four major clades were as follows: the Apodemia clade was the sister group of A. nais, and this clade Apodemia + A. nais was the sister group of the clade Emesis + A. phyciodoides (Emesis clade). These relationships were strongly supported in the Bayesian analysis, but less supported in the ML analysis. The Mexico clade was the sister group of the rest of the species of Apodemia and Emesis.

Genetic distances within major clades ranged from 0.9% in A. nais to 9.8% in Emesis + A. phyciodoides, whereas distances between genera ranged from 7.4% in Mexico clade versus A. nais to 10.5% in the Apodemia clade versus Emesis clade. The genetic distance between Apodemia clade and Mexico clade was 8.7% (Table 1).

Diagnosis of Plesioarida Trujano-Ortega & García-Vázquez gen. n. and Neoapodemia Trujano-Ortega gen. n. are presented, comparing both with Apodemia (Table 2). The morphology of both genera is described, including photographs and illustrations of the structures; distribution maps of these new genera are also given. The extent of genetic divergence between them and the phylogenetic position within the Riodinidae are discussed. The morphologic examination of the specimens revealed that the genital structures of the species A. phyciodoides, A. planeca, and A. castanea are distinct from the other species of Apodemia; suggesting these are not congeneric with the Apodemia species of Central and North America. The inclusion of these three species in Apodemia is being evaluated (Seraphim et al. submitted, Trujano-Ortega unpublished data); therefore, the morphology of Apodemia reported here excludes these species.

Plesioarida Trujano-Ortega & García-Vázquez, gen. n.

http://zoobank.org/627AB6DD-9174-4175-A5AB-B6B34E586540

Figs 2, 3, 4, 5, 6

Type species

Apodemia walkeri Godman & Salvin, 1886 by present designation.

Diagnosis

The species of this new genus can be distinguished from other Riodinidae by a combination of characters (Table 2). Labial palpi are long, slender, pointed apically and projected forward and upward, the second segment is a little more than twice the length of the third segment, the third segment is barely visible from dorsal view (Fig. 2). Radial veins originate near the end of the discal cell, costal vein runs parallel to Sc+R1; these veins get close but never fuse together. Vein R4 reaches wing margin at the apex (Fig. 3). Prothoracic legs of males are slender and trochanter inserts at the middle of the coxa; tibia is as wide as tarsus and smaller than the length of the femur plus the trochanter. Most of the species present two tarsomeres, the second tarsomere is oval-shaped and the apex pointed, except P. hypoglauca comb. n. which has three tarsomeres with the last one oval-shaped (Fig. 4). Tegumen of male genitalia is typically oval-shaped, narrow and slightly sclerotized, posterior half is a hyaline area that Hall (1999) named ‘windows’ through which the subescafium can be observed; the uncus is rounded and with setae in the posterior margin. The vinculum is a narrow band not covering the whole margin of the tegumen, is mostly straight and convex toward the saccus, a little hump-shaped in the mid region. Valvae are bifurcated, the dorsal process is conical, elongated, and with a sharp end projected forward and exceeding the posterior margin of the uncus, the ventral process is shorter and blunt with many setae. Aedeagus is long and sigmoid, wider in the anterior edge, slender and pointed on the posterior edge, where it opens dorsally. Cornuti are a series of wide, strongly sclerotized spines that originate from individual bulbs (Fig. 5).

Figure 2. 

Left male palpus of Apodemia, Plesioarida, and Neoapodemia.

Figure 3. 

Wing venation of Apodemia, Plesioarida, and Neoapodemia. Upper, forewing; lower, hind wing. Vein abbreviations (black lettering): Sc subcostal, R radial, M median, Cu cubital, A anal. Scale bars: 3 mm.

Figure 4. 

Prothoracic legs of males of Apodemia, Plesioarida, and Neoapodemia.

Figure 5. 

Male genitalia of Apodemia, Plesioarida, and Neoapodemia.A lateral view B ventral view C dorsal view D aedeagus E cornuti.

Description

Male. Anterior wing length: 10–15 mm. Head. Ringed antennae with 30 to 32 flagellomeres of the same width, with white scales at the base of each flagellomere. Widen abruptly in the apical 10 flagellomeres to form the antennal club, which is dark and iridescent dorsally. Sometimes white or brown scales are present at the sides, ending in a whitish or yellowish tip, with a nudum from flagellomere 20 to the apex. Labial palpi white with black or brown scales mainly in the third segment. Wings (Figs 3, 6) with four radial veins. Three distinct shapes of anterior wings, rounded toward the apex (P. palmerii comb. n.), elongated and triangular (P. walkeri comb. n.) and triangular with the external margin curved and the apex slightly sickle-like (P. hypoglauca comb. n.). Background color in both wings varies from brown to dark gray. Some species present a series of white spots outlined with black in the anterior margins and a series of submarginal black dots, sometimes with white scales and occasionally with reddish scales toward the base of the anterior and posterior wings. In grayish species spots are black. Legs. Prothoracic legs with dense long scales generally whitish, mid and hind legs with multiple short and dense spines in the interior margin of the tibia and tarsus. Abdomen. Dark in the dorsum with reddish or whitish scales outlining each segment. Ventrally with dense scales varying from whitish as in P. hypoglauca comb. n. to brown-orange as in P. palmerii comb. n. Genitalia. Genital capsule small, uncus rounded with a groove of variable depth which gives it a lobulated or straight appearance. Tegumen oval-shaped and sclerotized in the anterior region; with large ‘windows’ that reach the gnathi. Gnathi are slender, sclerotized, slightly twisted ending in an upward hook. Vinculum generally is straight or slightly curved near the tegumen, a little wider near the valve, this swelling is weakly sclerotized and hard to notice, curved before the saccus and anteriorly projected. Dorsal processes of the valve conic and membranous toward the transtilla but strongly sclerotized toward the apex, which is a small upward hook, with setae lengthwise. The ventral process is long and blunt, of variable lengths but always with multiple mostly long setae. Aedeagus is slender toward the distal portion with a pointed tip, widening toward the anterior portion, straight or sinuous. Cornuti are thick sclerotized spines apparently each surging from independent bulbs forming a line.

Figure 6. 

Wing color patterns of Apodemia, Plesioarida and Neoapodemia A Neoapodemia chisosensis comb. n. B Neoapodemia nais comb. n. C Plesioarida murphyi comb. n. D Plesioarida h. hepburni comb. n. E Plesioarida h. hypoglauca comb. n. F Plesioarida p. palmerii comb. n. G Plesioarida walkeri comb. n. H Apodemia m. mormo I Apodemia duryi J Apodemia mejicanus K Apodemia multiplaga L Apodemia virgulti. Scale bars: 5 mm. Additional data of the specimens in the photos are shown in Suppl. material 6.

Etymology

The name comes from the Greek plesios meaning near or close to and the Latin aridus meaning dry, in reference to the desert and semiarid habitats of most of the species.

Distribution and habitat

This genus is distributed below 1750 m in the Pacific slope from central Arizona and in the Atlantic slope from the south of Texas to the dry forests of Guanacaste in the northeast of Costa Rica (DeVries 1997) (Fig. 7). In the USA, it has been collected in arid regions, in xerophilic shrubland of Arizona, California, and New Mexico, with isolated records in the south of Texas in Río Grande Valley (Warren et al. 2017). In Mexico it can be found in deserts and semiarid regions of Baja California Sur and part of Baja California Norte in the Chihuahuan Desert and in the Mexican Plateau, regions were xerophilic shrubland is dominant. It is also distributed in the deciduous tropical forest of the west of Mexico in the Pacific coast and the Balsas Basin, as well as in the east in the coastal plain of the Gulf of Mexico and the Yucatan Peninsula. Its distribution in deciduous tropical forests extends through Central America to Costa Rica. Finally, P. selvatica comb. n. and some populations of P. walkeri comb. n. inhabit the tropical forests in the south of Mexico in Veracruz and Chiapas. De la Maza and De la Maza (2017b) mention that P. selvatica comb. n. probably is present in Guatemala and Belize.

Figure 7. 

Known distribution of Apodemia, Plesioarida, and Neoapodemia. Black lines represent countries’ limits.

Natural history

Larvae of the species of Plesioarida Trujano-Ortega & García-Vázquez gen. n. are associated with the family Fabaceae, particularly with species of the genera Prosopis spp. and Acacia spp. (Ferris 1985, Austin 1988, DeVries 1997).

Plesioarida Trujano-Ortega & García-Vázquez, gen. n. Type species: Apodemia walkeri Godman & Salvin, [1886], Biol. Centr. Amer., Lepid. Rhop. 1(45): 468, no. 6, by present designation.

hepburni comb. n. (Apodemia).

Apodemia hepburni Godman & Salvin, 1886. Type Locality: “Mexico, Pinos Altos in Chihuahua”.

h. hepburni Godman & Salvin, 1886, comb. n. (Apodemia).

h. remota Austin, 1991, comb. n. (Apodemia). Type Locality: “México: Baja California Sur; Arroyo San Bartolo”.

hypoglauca comb. n. (Apodemia).

Lemonias hypoglauca Godman & Salvin, 1878. Type Locality: “Mexico”.

h. hypoglauca Godman & Salvin, 1878, comb. n. (Apodemia).

h. wellingi Ferris, 1985, comb. n. (Apodemia). Type Locality: “México: Yucatán; Pisté”.

murphyi comb. n. (Apodemia).

Apodemia murphyi Austin, [1989]. Type Locality: “México: Baja California Sur; Arroyo San Bartolo”.

palmerii comb. n. (Apodemia).

Lemonias palmerii Edwards, 1870. Type Locality: “Utah”

p. palmerii W. H. Edwards, 1870

= p. marginalis Skinner, 1920

p. arizona Austin, [1989], comb. n. (Apodemia). Type Locality: “Arizona: Cochise County, Arizona State Route 90, 10.8 miles north of Arizona State Route 82”

p. australis Austin, [1989], comb. n. (Apodemia). Type Locality: “México: Durango: 1 mi. S Nombre de Dios”.

selvatica comb. n. (Apodemia).

Apodemia selvatica De la Maza & De la Maza, 2017. Type Locality: "Estación Chajul" [México, Chiapas].

walkeri comb. n. (Apodemia).

Apodemia walkeri Godman & Salvin, 1886. Type Locality: “Mexico, Acapulco” [Guerrero]

Neoapodemia Trujano-Ortega, gen. n. Type species: Chrysophanus nais W. H. Edwards, 1876, Trans. Am. Entomol. Soc. 5(3/4): 291–292, by present designation.

= Polystigma Godman & Salvin, [1886]. Biol. Centr. Amer., Lepid. Rhop. 1(45): 469. Type-species: Chrysophanus nais W. H. Edwards, 1876, Trans. Am. Entomol. Soc. 5(3/4): 291–292, by monotypy. Preoccupied by Polystigma Kraatz, 1880, Dtsche. Entomol. Z. 24(2): 191.

nais comb. n. (Apodemia).

Chrysophanus nais W. H. Edwards, 1876. Type Locality: “Southern California... Prescott, Arizona”.

chisosensis comb. n. (Apodemia).

Apodemia chisosensis Freeman, 1964. Type Locality: “Chisos Mountains, elevation 5400 ft., Texas”

The phylogenetic analysis based on molecular data shows three well-supported clades which are also distinguished by their morphology. The clade Neoapodemia is more related to the clade of Apodemia (sensu stricto excluding A. castanea and A. phyciodoides) than with Plesioarida. Besides the phylogenetic position, the proposed genera can be easily distinguished morphologically as well as with Apodemia and Emesis. Labial palpi of Apodemia, Plesioarida and Neoapodemia are slender, long and projected upward and forward, while Emesis present small labial palpi, close to the head and directed upward. The second segment of labial palpi in Neoapodemia is smaller in proportion with the third segment and only one third is visible in dorsal view, in contrast with Apodemia in which the second segment is long and half or more of it is visible in dorsal view. Plesioarida differs from both genera having only part of the third segment visible in dorsal view. Regarding wing veins patterns, radial veins of Neoapodemia and Plesioarida originate near the end of the discal cell, whereas in Apodemia these veins originate at the middle of the discal cell. Veins C and Sc+R1 are closer in Plesioarida and Apodemia than in Neoapodemia, which present a clear separation between these veins. The number and shape of the tarsomeres of prothoracic legs are useful characters for separating the genera. Emesis has only one tarsomere, Plesioarida has two (except P. hypoglauca comb. n. which has three) and an oval-shaped tip, Neoapodemia and Apodemia presents three tarsomeres; the last tarsomere in Apodemia is conical, while Neapodemia the last tarsomere is smaller than in Plesioarida, wide at the base, tapering toward the apex and ending in a blunt tip.

Apodemia and Neoapodemia present large genital capsules, more similar than with those of Plesioarida. However, Neoapodemia gnathi are slightly twisted and the posterior tip is hooked and strongly projected upward; while gnathi of Apodemia are straight. The genital capsule of Plesioarida is smaller and rounded, with tegumen with a marked hyaline area; gnathi are slim and twisted. Vinculum is straight in Apodemia; convex in Neoapodemia, which makes it appear larger; and straight near tegumen and convex toward the saccus in Plesioarida. Dorsal processes of Plesioarida differ from the other two genera because are long and exceed the posterior margin of the uncus. The ventral processes of the valves of Neopapodemia are shorter than those of Apodemia and Plesioarida, which are also slender. Aedeagus bends in a smooth angle in Neoapodemia, but in a marked angle in Plesioarida, in both genera this bending appears near the posterior tip of the aedeagus. In Apodemia the bending angle is marked (approximately 45°), nearly at half of aedeagus or closer to the anterior tip. The cornuti are one of the most important differences. The cornuti of Apodemia are simple, long, and sclerotized; Plesioarida presents a series of aligned spines that surge from individual bulbs; finally, Neoapodemia has long, strong, sclerotized spines, flattened laterally and joined at the base, forming a crest. Emesis genitalia are diverse and the cornuti are simple when present.

The phylogenetic analysis with molecular data suggests that A. phyciodoides is part of Emesis. However, we took a conservative approach and decided that A. phyciodoides should remain in Apodemia until morphology was reviewed. Considering that Emesisneed a taxonomical review and that the Emesis + A. phyciodoides clade was well supported only by Bayesian but not ML method.

Natural history characters also help to distinguish between these groups. For example, it is noteworthy that each genus has distinct host plants (Ferris 1985, Scott 1986, Austin 1988, DeVries 1997, Brock and Kaufman 2006, Warren et al. 2017), which is evidence of differences in their life histories. Although none of these genera is exclusive of an environment, some of their species are characteristic of specific habitats. For example, Apodemia inhabits desert areas and shrubland environments (although A. multiplaga is from deciduous tropical forests). Neoapodemia is recorded from coniferous forest, chaparral, and submontane shrublands. Plesioarida inhabits desert areas and deciduous tropical forests.

Our results show that Apodemia is paraphyletic with respect to Emesis; furthermore, the morphological evidence and the preliminary molecular analysis (Trujano-Ortega unpublished data) suggest polyphyly. This is because, historically, diverse unrelated lineages have been placed within Apodemia and reassigned posteriorly. Harvey and Clench (1980) created the genus Dianesia for Apodemia carteri and suggested that A. castanea is not congeneric with North and Central American Apodemia. More recently, Penz and DeVries (2006) reassigned Apodemia paucipuncta to the recently described genus Hallonympha.

In a previous analysis of the phylogenetic relations of Riodinidae at the tribe or subfamily level, Apodemia and Emesis appear always closely related based on morphological evidence (Stichel 1910–1911, Harvey 1987) and molecular evidence (Saunders 2010, Espeland et al. 2015, Proshek et al. 2015). These two genera were grouped in the same tribe Emesini of Stichel (1910–1911). However, the taxonomic scale of these studies and the lack of recent material in scientific collections limited the resolution of lower taxonomic scale and the inclusion of Mexican and Central American species. That is why the two genera proposed here remained unnoticed. The morphological characters shared by Apodemia, Emesis, Neoapodemia, and Plesioarida are evident, as suggested by Harvey (1987), who analyzed the immature stages (pupae). These genera are more related to each other than with the genera of any other tribe of Riodinidae. Therefore, these two new genera should be included in a new tribe as mentioned by Espeland et al. (2015) for several Incertae sedis groups, like Emesis-Apodemia. The biogeographic patterns of these new genera are an interesting topic for further study. Espeland et al. (2015) proposed two independent dispersal events for these genera, from the Neotropical to the Nearctic region. One event of the ancestor of Apodemia mainly to the southwestern USA and the other event of the ancestor of Calephelis, to the eastern and central USA.

We consider that each of these genera is well supported by the morphological and molecular evidence. In order to resolve these phylogenetic relations a more extensive sampling is required, joined with a detailed review of the morphology of most species, including Emesis species. Also, the species A. phyciodoides, A. castanea, and A. planeca must be assigned to the correct genera in order to stabilize Apodemia. This can only be achieved with the use of diverse character systems and sufficient sampling. This study contributes to the systematics and classification of Riodinidae. It adds two genera to the family: Neoapodemia which, as Apodemia, is exclusive to North America, and Plesioarida of North and Central America.