Three new species in the Mesene monostigma group from the Transandean region (Lepidoptera: Riodinidae: Symmachiini)

75TROP. LEPID. RES., 35(2): 75-84, 2025HALL: Mesene monostigma group from Transandean region Three new species in the Mesene monostigma group from the Transandean region (Lepidoptera: Riodinidae: Symmachiini) Jason P. W. Hall Department of Entomology, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560-0127, USA Date of issue online: 19 December 2025 Zoobank Registered: urn:lsid:zoobank.org:pub:27834A97-F5A1-455B-8CFF-8003C6C0E0DA Electronic copies (ISSN 2575-9256) in PDF format at: https://journals.flvc.org/troplep; https://zenodo.org; archived by the Institutional Repository at the University of Florida (IR@UF), http://ufdc.ufl.edu/ufir; DOI: 10.5281/zenodo.17860789 © The author(s). This is an open access article distributed under the Creative Commons license CC BY-NC 4.0 (https://creativecommons.org/ licenses/by-nc/4.0/). Abstract: Three new riodinid species in the monostigma group of Mesene Doubleday, 1847 (Symmachiini) are described from the Transandean region. Two distinctive species, M. pirata Hall & Solis n. sp. and M. lojensis Hall n. sp., are described from moist forest habitats, the former from the low-lying Bay Islands of northern Honduras, and the latter from premontane elevations in southwestern Ecuador. The third, M. nemo Hall n. sp., is a cryptic species externally resembling M. monostigma (Erichson, [1849]) that is described from wet lowland forest habitats between eastern Panama and northwestern Ecuador. Mesene monostigma Auct. is shown here on the basis of wing pattern and particularly male genitalia differences to consist of three species, including the Transandean M. nemo and the west Amazonian M. discolor Stichel, 1929 (stat. rev.), with true M. monostigma ranging widely across much of wet tropical eastern South America. All five monostigma group species appear to be alloor parapatrically distributed throughout the Neotropics. The names celetes Bates, 1868, and rochesteri Weeks, 1906, are newly synonymized with M. monostigma (n. syns.). Key words: Andes, Central America, endemism, seasonal evergreen forest, species description, taxonomy. INTRODUCTION The island of Roatán, the largest of the Bay Islands, sits about 50 km off the northern Caribbean coast of Honduras. With only a tiny number of butterfly specimens present in historical collections from these moist forest islands, I was intrigued by the potential diversity of their butterfly fauna. Therefore, between January 2008 and August 2013, I made four visits to Roatán, the first two with M. A. Solis, to inventory the butterfly fauna of the island. During the course of approximately 20 collecting days, 111 butterfly species were recorded, including 45 Hesperiidae, 3 Papilionidae, 11 Pieridae, 30 Nymphalidae, 15 Lycaenidae, and 7 Riodinidae. The Riodinidae recorded were Euselasia sergia (Godman & Salvin, 1885), Pachythone strati (Kaye, 1925), Theope eupolis Schaus, 1890, Theope sp. cf. barea Godman & Salvin, 1878 (females only, preventing a definitive identification), Anteros carausius Westwood, 1851, Rhetus arcius (Linnaeus, 1763), and, most notably, an undescribed Mesene Doubleday, 1847 species. The most recent and comprehensive faunal works on the butterflies of Honduras (Miller et al., 2012; Gallardo & Diaz, 2022) contain minimal data on the Bay Islands, and T. eupolis seems to be the only riodinid species previously reported from Roatán (Hall, 1999). However, despite not previously being reported in the literature from Honduras, there is a female of P. strati from Roatán in London’s Natural History Museum (as well as another female from San Pedro Sula in Berlin’s Museum für Naturkunde). On two more recent solo collecting trips exploring moist forest fragments in southwestern Ecuador, as part of my ongoing 30+ year investigation of the butterflies of Ecuador with K. R. Willmott and others, I discovered another undescribed Mesene species. Morphological investigation has revealed that these two new Transandean Mesene species, neither of which is present in any of the institutional or private collections I have examined (e.g., see list in Hall (2018)), are closely related to each other and to the wet forest species Mesene monostigma (Erichson, [1849]). Furthermore, I have found that M. monostigma itself, which has been considered geographically widespread and variable, consists of three alloor parapatrically distributed species with distinct genital morphology, including one undescribed Transandean species. The widespread Neotropical genus Mesene contains approximately 33 small, aposematic, often sexually dimorphic, and typically orange, red, black, and sometimes white patterned described species in the tribe Symmachiini (Callaghan & Lamas, 2004; Zhang et al., 2021; Brévignon, 2025; Hall, unpubl. data). New species continue to be described in the genus (Brévignon, 1995; Callaghan & Salazar, 1999; Hall & Lamas, 2007; Dolibaina et al., 2016), including in the Transandean region (Hall & Willmott, 1995; Callaghan et al., 2011), where about a third of the species occur. The purpose of this paper is to revise the taxonomy of M. monostigma and describe the three aforementioned new Transandean Mesene species in the context of a newly characterized monostigma group. 76 TROP. LEPID. RES., 35(2): 75-84, 2025 HALL: Mesene monostigma group from Transandean region MATERIALS AND METHODS Morphology was studied using standard techniques, with the utilized genital dissection methods, including aedeagal vesica eversion (e.g., Dang, 1993; Sihvonen, 2001), following those detailed in Hall (2018). The terminology for male genital structures follows Klots (1956), Eliot (1973), and Harvey (1987), and the nomenclature for venation follows Comstock & Needham (1898), with cells named for the vein above. Mesene material was studied in the 27 personally visited institutional and private collections listed in Hall (2018), and the known type specimens or images were examined for all available names. TAXONOMY OF MESENE MONOSTIGMA The widely distributed and variable M. monostigma has had a confused taxonomic history, with a parade of ever evolving arrangements published over the past 150 years. The species is currently treated as having a nominotypical subspecies that ranges from the southern Transandean region to Argentina, with the names hya Westwood, 1851, and guttula Stichel, 1916, as synonyms, and the west Amazonian subspecies discolor Stichel, 1929 (Callaghan & Lamas, 2004). In the older literature, the species was often incorrectly referred to under the junior name hya, and the unrelated phareus group names fenestrella Bates, 1868, and paraena Bates, 1868, were often mistakenly included as subspecies along with hya and guttula (e.g., Stichel, 1910-11, 1930-31; Seitz, 1916-20). During the course of contextual morphological research for this paper, I dissected male specimens exhibiting the full known range of wing pattern variation and from throughout much of the geographic distribution of the species. I dissected three males from northwestern Ecuador, two from eastern Ecuador, and one each from southern Peru, French Guiana, Paraíba state, in coastal eastern Brazil, and Mato Grosso and Mato Grosso do Sul states, in southern Brazil. I found that there were three consistently quite distinct genital phenotypes, referable to three alloor parapatrically distributed species that exhibit only subtle wing pattern differences, namely M. monostigma, M. discolor, and a new Transandean species. Mesene monostigma, described from Guyana from one or more males (Erichson, [1849]), ranges from Venezuela, Trinidad, and the Guianas through central and eastern Amazonia to eastern and southern Brazil, Paraguay, and Argentina. Typical males, from northeastern South America, have the most reduced orange markings, with orange scaling on the dorsal forewing forming a variably small bar along the base of the anal margin (e.g., see figures in Barcant, 1970; d’Abrera, 1994; Gallard, 2017). Very similar males, described under the subspecies name guttula from Paraguay (Stichel, 1916), with a variably slightly larger triangle of orange scaling at the base of the dorsal forewing, predominate in southern South America (see Fig. 1A; also e.g., see figure in Canals, 2003). Males with a variably large triangle of orange scaling at the base of the dorsal forewing that can extend to the lower margin of the discal cell, as well as typically the narrowest area of black distal scaling on the hindwing, occur in the central Brazilian Amazon, coastal eastern Brazil, and pockets of southern South America (e.g., see figures in d’Abrera, 1994 [an apparently mislabeled male from Tobago purportedly collected in 1914 by G. H. Sworder, who also seems to have mislabeled two never subsequently seen ithomiine species from Tobago (Cock, 2017)]; Canals, 2003). Examined males of M. monostigma from João Pessoa, Paraíba, on the eastern coast of Brazil, very closely match the syntype male of hya, described from “Brazil” (Westwood, 1850-52), and this historically accessible region of Brazil seems like a plausible locality of origin for the hya syntype. These male phenotypes not only have complex recurring distributions, but all intergrade with each other, rendering the recognition of subspecies unjustified. The white subapical spot on the male forewing is locally variable in size, but is generally small, varying from vertically rectangular to square, and the other white submarginal markings, one in the tornus of the ventral forewing and two on the ventral hindwing, vary from small to absent. Typical females of M. monostigma (see Fig. 1B; also e.g., see figures in d’Abrera, 1994; Gallard, 2017), which occur throughout at least most of the range of the species, exhibit more limited local variation, with the orange and white patches varying slightly in size, and the white submarginal markings on the ventral hindwing rarely showing through on to the dorsal surface. However, I suggest that two additional names, celetes and its currently recognized synonym rochesteri, described from females from Pará, east Amazonian Brazil, and, Suapure, Amazonian Venezuela, by Bates (1868) and Weeks (1906), respectively, newly belong in synonymy with M. monostigma (n. syns.), expanding our understanding of female variation in the species. This female phenotype (e.g., see figures in d’Abrera, 1994; Canals, 2003) is known to occur more rarely throughout most of the range of M. monostigma, from Venezuela to Argentina, and is the only female phenotype I have seen from the east Brazilian states of Paraíba and Pernambuco. It has the same wing pattern as the typical female phenotype of M. monostigma, including the unusual white forewing fringe element in the forewing apex, but has the orange patch on the forewing expanded distally to exactly engulf and obscure the white subapical spot in cell M3. I have examined rare intermediate specimens with the white forewing subapical spot partially engulfed by the orange patch but still completely visible on both wing surfaces, and both Bates (1868) and Weeks (1906) mention some females among their type series in which this spot is faintly visible at the edge of the orange patch on the ventral forewing. Similar wing pattern variation, or dimorphism, was recently reported by Brévignon (2025) in both sexes of nominotypical M. nepticula Möschler, 1877, in the Guianas, suggesting a synonymy with M. simplex Bates, 1868. The name discolor was proposed by Stichel (1929) as a species, based on a female from Bolivia, and was downgraded to a subspecies of M. monostigma by Hall & Harvey (2002), in a paper on androconial morphology, based on the existence of apparent male wing pattern intergrades. However, a more thorough investigation, including examination of a longer series of specimens, has revealed that males of M. discolor possess several subtle wing pattern characters that, with varying degrees of reliability, distinguish them from M. monostigma males, in addition to having quite distinct 77 genitalia, necessitating the return of M. discolor to species status (rev. stat.). Mesene discolor is a west Amazonian species that ranges from Colombia to Bolivia, and into western Brazil, and occurs up to nearly 1000 m. A pair is figured here from Ecuador (Figs. 2A,B), and another pair was figured in d’Abrera (1994) from southern Peru and Bolivia. A male with the wing pattern and genitalia shown here to be characteristic of M. discolor was also figured by Dolibaina et al. (2016) from Acre state, western Brazil, although the name M. m. monostigma was used. These authors noted the existence of more than one male genital phenotype among South American M. monostigma Auct. males, but expressed uncertainty about the true identity of and variation within male M. discolor. Males of M. discolor differ from the most similar and geographically proximate males of M. monostigma (the hya phenotype) by nearly always having a variably larger and more horizontally rectangular white subapical spot in forewing cell M3, typically having a more prominent white submarginal streak in ventral forewing cell Cu2 that rarely shows through onto the dorsal surface, and always having an undulating instead of approximately straight distal margin to the hindwing orange patch, with variably small distal orange protrusions in dorsal hindwing cells Cu2 and M3 (sometimes barely visible in the latter cell) and variably larger white submarginal spots on the ventral hindwing extending distally from these orange peaks. The male genitalia of M. discolor (Fig. 7) (three dissections examined) consistently differ from those of M. monostigma (Fig. 6) (four examined) by having more horizontally elongate valvae, with a ventral margin to the lower posterior valve process that is prominently concave instead of approximately straight in lateral view and less prominently inwardly curled (creating a broadly square instead of narrowly rectangular gap between the base of the valvae in ventral view), a tip to the lower posterior valve process that is more elongate, posteriorly pointed, dorsally curved, and inwardly curled (sometimes curling horizontally), and an upper posterior valve process that is smaller, more narrowly pointed, and shorter than instead of similar in length to the lower posterior valve process tip. The everted aedeagal vesica of M. discolor also has a larger dorsobasal “hand”-like cornutus that consists of four to six instead of two to four anteriorly directed “fingers” (four were observed only in the Paraíba male of M. monostigma), and a posterior bundle of very long, thin, posteriorly directed spines that is longer by about a third. The distinctive female of M. discolor, which I have encountered flying in the vicinity of male hilltop perching leks in Ecuador, has the four white submarginal spots variably enlarged into rectangular streaks that extend nearly to the wing margins on both surfaces of both wings. In contrast, in the most similar females of M. monostigma, the distal margin of the orange patch on both wings is straighter, and the white submarginal marks are much smaller, with only the spot in forewing cell M3 prominently visible on the dorsal surface. The third M. monostigma-like species, M. nemo (Figs. 3A,B), is known to range from eastern Panama to northwestern Ecuador, and is described and diagnosed below along with the two newly discovered and similarly alloor parapatrically distributed Transandean species, M. pirata from Honduras (Fig. 4) and M. lojensis (Fig. 5) from southwestern Ecuador. All five species can be placed in the newly designated monostigma group. They all possess the small compact wings and unspotted orange, black, and sometimes white pattern common in Mesene, as well as the typical generic pattern of concealed androconial scales, which are present on male abdominal tergites four, five, and six (Hall & Harvey, 2002), but the genitalia are diagnostic. The male genitalia of M. monostigma have heretofore been considered unique in the genus, with the phylogenetic affinities of the species unclear (Hall & Harvey, 2002; Dolibaina et al., 2016). However, all five of the here-recognized monostigma group species possess a unique and unusually complex arrangement of cornuti on the everted aedeagal vesica that consists of a large, approximately triangular, dorsobasally positioned, outwardly and anteriorly directed “hand”- or “foot”-like cornutus with multiple “digit”- like spines, a variably large, narrowly triangular, variably sinuous, ventrobasally positioned, outwardly and anteriorly directed cornutus with fine striations and multiple typically small spines along mostly the outer margin, and a bundle of very long, thin, centrally positioned, posteriorly directed spines (see Figs. 6-10). The interspecific genital variation exhibited by the monostigma group is unusually significant compared to that within congeneric groups, most notably the large phareus group (sensu Hall & Harvey, 2002), whose members are collectively characterized by a distinctive “horn”-like posterior projection from the aedeagal pedicel but possess no or only minor genital differences between even the most obviously different species in terms of wing shape and pattern, ecology, and geography. SPECIES DESCRIPTIONS Mesene nemo Hall, new species (Figs. 3A,B; 8; 12) Description: MALE: Forewing length HT 11 mm, PTs 10.5-11 mm. Wings: see Fig. 3A; forewing fringe black with white elements in cell Cu2, cell M3, and across cells M1 to R4+5, hindwing fringe black. Head: Eyes brown and bare with dark brown marginal scaling dorsally and cream scaling ventrally; dorsal three quarters of frons dark brown, ventral third cream; labial palpi a mixture of brown and cream scales, with tip of segment three entirely cream; antennal length approximately 60% of forewing length, segments black with whitish scaling at base and narrow nudum section along inner margin, clubs black with an orange-brown tip. Body: Dorsal surface of thorax orange with orange tegula, ventral surface cream; dorsal surface of abdomen orange with a narrow band of black scales around tip (present in HT but not some PTs), ventral surface pale yellow with a narrow band of black scales around tip (present in HT but not some PTs) and dark orange scaling around genital opening; all legs cream, with femur of midand hindleg slightly darker; a narrow and continuous band of concealed androconial scales present dorsally across anterior margin of abdominal tergites four, five, and six; eighth abdominal tergite and sternite approximately rectangular. Genitalia: see Fig. 8; ventral posterior corner of lower posterior valve process prominently inwardly curled and posterior tip of lower posterior valve process slightly inwardly curled, with long sparse setae present along posterior margin of upper and lower posterior valve processes; lightly sclerotized transtilla band extends continuously over aedeagus from behind upper posterior valve processes; ductus ejaculatorius enters aedeagus anteriorly and slightly from left, vesica exits aedeagus posteriorly. FEMALE: Forewing length 11 mm. Wings: see Fig. 3B; forewing fringe dark brown with white elements in cell M3 and across cells M1 to R4+5, hindwing fringe dark brown. Head: Eyes brown and bare with orange marginal scaling dorsally and pale orange scaling ventrally; dorsal half of frons dark brown, ventral half pale orange; labial palpi pale orange with scattered brown scales; antennal length approximately 60% of forewing length, segments black with whitish scaling at base and narrow nudum section along inner margin, clubs black. Body: Dorsal TROP. LEPID. RES., 35(2): 75-84, 2025HALL: Mesene monostigma group from Transandean region 78 surface of thorax orange with orange tegula, ventral surface pale orange; dorsal surface of abdomen orange, ventral surface pale orange; all legs pale orange. Types: HOLOTYPE male: ECUADOR: Esmeraldas, ridge above San Francisco, km. 17 San Lorenzo-Lita rd., 200 m, 01˚06'26"N 78˚41'55"W, 26 Mar 2004 (J. P. W. Hall & I. Aldas) (National Museum of Natural History, Smithsonian Institution, Washington, DC, USA [USNM]). PARATYPES: PANAMA: Panamá, 1 female: Bayano, Piña, 12 Dec 1974 (G. B. Small) (USNM). ECUADOR: Esmeraldas, 1 male: La Punta, km. 44 Lita-San Lorenzo rd., 300 m, 01˚03'55"N 78˚39'00"W, 21 Jun 1994 (J. P. W. Hall & K. R. Willmott) (Jason P. W. Hall collection, Washington, DC, USA [JH]); El Durango, km. 40 Lita-San Lorenzo rd., 300 m, 01˚02'45"N 78˚38'06"W, 1 male: 29-31 Aug 1999 (K. R. Willmott) (JH); 2 males: 20 Nov 2012 (I. Aldas) (David H. Ahrenholz collection, Landrum, SC, USA [DA]); same locality data as HT, 2 males: 16 May 2004 (D. H. Ahrenholz) (DA); 1 male: 17-18 Mar 2018 (J. P. W. Hall) (Instituto Nacional de Biodiversidad, Quito, Ecuador [INABIO]); 1 male: 19 Aug 2023 (J. P. W. Hall) (INABIO); 1 male: Tundaloma Lodge, San Lorenzo-Lita rd., 100 m, 01˚10'40"N 78˚44'54"W, 3 Jul 2015 (K. R. Willmott) (JH); 1 female: hill W. of Ricaurte, San Lorenzo-Lita rd., 100 m, 01˚10'25"N 78˚44'45"W, 12 Apr 1996 (K. R. Willmott) (JH); 1 male: El Cerro, Maldonado-Selva Alegre rd., 250 m, 00˚58'22"N 78˚55'19"W, 25 Jul 2011 (J. P. W. Hall & K. R. Willmott) (JH); Pichincha, 2 males: km. 21 Pacto-Guayabillas rd., 900 m, 00˚11.2'N 78˚51.8'W, Nov 2006 (I. Aldas) (DA). Etymology: The name alludes to similarly colored species of clown fish, or anemone fish, epitomized in popular culture by the animated Disney character “nemo”. The name is considered a masculine noun in apposition. Diagnosis: The Transandean M. nemo n. sp. (Fig. 3) is externally very similar to the apparently allopatrically distributed east South American species M. monostigma (Fig. 1). Indeed, the two cryptic species do not seem to be consistently separable on the basis of wing pattern, given the significant phenotypic variation exhibited by M. monostigma, although M. nemo males typically have slightly paler orange patches on both wings, and a slightly larger forewing orange patch than most males of M. monostigma, especially those from northeastern and far southern South America. Males of M. nemo also consistently have a straight and vertical apical distal margin to the hindwing orange patch, whereas most males of M. monostigma have a proximally indented apical margin, a notable exception being an examined population from Paraíba state, eastern Brazil, that has a vertical apical margin. The females of M. nemo are externally indistinguishable from typical females of M. monostigma. However, the male genitalia of the two species (Figs. 6 and 8) are consistently quite distinct, exhibiting numerous differences in the valvae and everted vesica (see the Taxonomy of Mesene monostigma section for details of the dissections made). The valvae of M. nemo are more vertically elongate than those of M. monostigma, with a short and approximately evenly square instead of rectangular and posteriorly narrowed gap between the base of the valvae in ventral view, a shorter and less inwardly curved ventral posterior corner to the lower posterior valve process (this corner looks pointed in lateral view in Fig. 6, of M. monostigma, because of its more prominent inward extension and curvature, which relates to the previous character), a much narrower and more dorsally directed posterior tip to the lower posterior valve process, and an upper posterior valve process that is slightly longer than instead of similar in length to the lower posterior valve process tip. The aedeagus of M. nemo is slightly broader, with a shorter and broader everted vesica (multiple vesica eversions for both species confirmed that this is not an artifact of incomplete eversion), a posterior bundle of very long, thin, posteriorly directed spines that is longer, more anteriorly positioned, and basally overlaps the two basal cornuti, a larger dorsobasal “hand”-like cornutus that consists of seven to eight instead of two to four “fingers”, and a ventrobasal cornutus that is twice as large, more sinuous, and has a proportionately larger apical spine. The similar, but more distinctive, west Amazonian M. discolor is characterized in the Taxonomy of Mesene monostigma section. Wing pattern variation in the examined male specimens of M. nemo is minor, and limited to small variations in the size and Figs. 1-5. Mesene adults (dorsal surface on left, ventral surface on right) (2 x life size). 1A. M. monostigma (Erichson, [1849]) male, São Vicente, 90 km. E. of Cuiabá, Mato Grosso, S.W. Brazil (600 m) (USNM). 1B. M. monostigma female, Colegio Buriti, Chapada dos Guimarães, Mato Grosso, S.W. Brazil (700 m) (USNM). 2A. M. discolor Stichel, 1929, male, Finca San Carlo, Napo, E. Ecuador (550 m) (JH). 2B. M. discolor female, Lumbaquí, Sucumbíos, E. Ecuador (950 m) (JH). 3A. M. nemo Hall n. sp., HT male, San Francisco, Esmeraldas, N.W. Ecuador (200 m). 3B. M. nemo PT female, Ricaurte, Esmeraldas, N.W. Ecuador (100 m). 4. M. pirata Hall & Solis n. sp., HT male, Mud Hole Hill, W. Roatán, Bay Islands, N. Honduras (100 m). 5. M. lojensis Hall n. sp., HT male, km. 3 Chaguarpamba-Velacruz rd., Loja, S.W. Ecuador (1450 m). TROP. LEPID. RES., 35(2): 75-84, 2025 HALL: Mesene monostigma group from Transandean region 79 distal shape of the orange patch on both wings and the size of the white forewing subapical spot, and variation in the prominence of the other white submarginal markings on the ventral wings, which are usually less prominent than in the figured holotype and rarely completely absent on the ventral hindwing. A very narrow band of black scales around the tip of the abdomen is also variably present or absent. Wing pattern variation in females of M. nemo is similar to that observed in males. Compared to examined Ecuadorian females, the Panamanian female paratype has slightly smaller orange patches, with more black scaling along the forewing costal margin, a slightly larger and purer white forewing subapical spot, and slightly more prominent remaining white submarginal markings on both ventral wings. Biology: This species is known to inhabit wet lowland evergreen forest from near sea level to 900 m. Males were encountered perching as solitary individuals or in small groups in large hilltop lightgaps. They perched under leaves with their wings open 2 to 4 m above the ground from 1430 to 1700 hrs. A female was found flying about 1 m above the ground at the edge of a hilltop lightgap at 1400 hrs. Distribution: Mesene nemo is known to range from eastern Panama to northwestern Ecuador (Fig. 12). The species is represented in European museums from western Colombia, although my old notes do not include specific localities, and it may also occur in the central Colombian valleys. Mesene nemo is replaced to the east of the Andes by the similar M. monostigma (across eastern South America) and M. discolor (in western Amazonia) (see the Taxonomy of Mesene monostigma section). Mesene pirata Hall & Solis, new species (Figs. 4; 9; 11A-C; 12) Description: MALE: Forewing length HT 10 mm, PTs 10-10.5 mm. Wings: see Fig. 4; fringe on both wings entirely black. Head: Eyes brown and bare with dark brown marginal scaling; dorsal three quarters of frons dark brown, ventral quarter pale brown; labial palpi dark brown; antennal length approximately 60% of forewing length, segments black with a few whitish scales at base and Figs. 6-10. Mesene male genitalia in lateral view, with aedeagal vesica everted. 6. M. monostigma, Cayenne, French Guiana (USNM). 7. M. discolor, Lumbaquí, Sucumbíos, E. Ecuador (JH). 8. M. nemo n. sp., PT male, El Cerro, Maldonado-Selva Alegre rd., Esmeraldas, N.W. Ecuador. 9. M. pirata n. sp., PT male, Mud Hole Hill, W. Roatán, Bay Islands, N. Honduras. 10. M. lojensis n. sp., HT male, km. 3 ChaguarpambaVelacruz rd., Loja, S.W. Ecuador. Scale bar = 0.5 mm. TROP. LEPID. RES., 35(2): 75-84, 2025HALL: Mesene monostigma group from Transandean region 80 narrow nudum section along inner margin, clubs black with an orange-brown tip. Body: Dorsal surface of thorax orange with orange tegula, ventral surface pale brown; dorsal surface of abdomen orange with a narrow band of black scales around tip (present in HT but not some PTs), ventral surface pale orange with a narrow band of black scales around tip (present in HT but not some PTs) and orange scaling around genital opening; all legs pale brown; a narrow and continuous band of concealed androconial scales present dorsally across anterior margin of abdominal tergites four, five, and six; eighth abdominal tergite and sternite approximately rectangular. Genitalia: see Fig. 9; lower posterior valve process slightly inwardly curved, with long sparse setae present along posterior margin of upper and lower posterior valve processes; transtilla band that is especially lightly sclerotized dorsally extends continuously over aedeagus from behind upper posterior valve processes; ductus ejaculatorius enters aedeagus anteriorly, vesica exits aedeagus posteriorly. Types: HOLOTYPE male: HONDURAS: Islas de la Bahía, Mud Hole Hill, above Sandy Bay Point, Coxen Hole-Sandy Bay rd., W. Roatán, 16˚20'22"N 86˚32'11"W, 100 m, 3-4 Sept 2011 (J. P. W. Hall & M. A. Solis) (to be deposited in the USNM). PARATYPES: HONDURAS: Islas de la Bahía, 4 males: same data as HT (JH); 1 male: above Punta Blanca, N. of Oak Ridge, E. Roatán, 16˚24’39”N 86˚21’14”W, 100 m, 13 Sept 2011 (J. P. W. Hall) (JH). Etymology: The name is derived from the Latin noun “pirata”, meaning “pirate”, in reference to the fact that the Bay Islands of Honduras were a renowned hotbed of British, Dutch, and French pirate activity in the sixteenth and seventeenth centuries. Diagnosis: Mesene pirata n. sp. is most similar to the new species M. nemo, from wet evergreen forests of the southern Transandean region, but the male (Fig. 4) can rapidly be distinguished from that of M. nemo (Fig. 3A), as well as those of the South American M. monostigma (Fig. 1A) and M. discolor (Fig. 2A), by having no white subapical spot in cell M3 on either surface of the forewing, or any other white submarginal spots, and an entirely black forewing fringe without white elements Fig. 11. Habitats of Mesene pirata n. sp. on Roatán, Bay Islands, N. Honduras (A-C), and Mesene lojensis n. sp. in Loja province, S.W. Ecuador (D-F). A. Mud Hole Hill, Coxen Hole-Sandy Bay rd., W. Roatán (100 m) (type locality of M. pirata). B,C. Perching lightgaps on semi-open ridge extending northeast from antenna in (A). D. Large forest fragment straddling ridge on Chaguarpamba-El Rosario rd. (1100 m). E. Trail through forest fragment straddling ridge above Chaguarpamba (1450 m) (type locality of M. lojensis). F. A M. lojensis male perching in long grass beside small forest fragment along ridge near Jorupe on Chaguarpamba-Chiguango rd. (1200 m). in cell Cu2, cell M3, and across cells M1 to R4+5. The male of M. pirata is also slightly smaller on average and has different coloration on the body, with a small inconspicuous patch of pale brown scales in the ventral quarter of the frons instead of a large prominent patch of brighter cream scales in the ventral third of the frons, entirely brown palpi instead of predominantly cream palpi with scattered brown scales, entirely pale brown legs instead of cream legs that become slightly darker on the femur of the midand hindleg, and a pale orange instead of pale yellow ventral surface to the abdomen. Mesene pirata males additionally differ from those of M. nemo by having slightly darker orange patches, particularly on the ventral forewing, an orange patch on the dorsal forewing that extends solidly into the lower half of the discal cell instead of being absent in the discal cell or present as variably sparse streaks, and a slightly narrower orange patch on the dorsal hindwing, leaving a broader black distal margin. The male genitalia of M. pirata (Fig. 9) most conspicuously differ from those of M. nemo (Fig. 8) in the shape of the valvae, with M. pirata having a triangular instead of vertically rectangular lower posterior valve process, with an approximately straight instead of prominently angular ventral margin, a posterior tip to the lower posterior valve process that is slightly longer, narrower, much more pointed, and largely posteriorly instead of dorsally directed, and an upper posterior valve process that is slightly narrower and more pointed, and substantially shorter instead of fractionally longer than the lower posterior valve process tip. However, the everted vesica of M. pirata is also distinctive, with a dorsobasally positioned cornutus that is larger and shaped like a distally rounded fan in dorsal view, with bilaterally symmetrical spines in terms of length and inward curvature, instead of “foot” shaped, with TROP. LEPID. RES., 35(2): 75-84, 2025 HALL: Mesene monostigma group from Transandean region 81 spines becoming longer to the left and predominantly curving from right to left, and a ventrobasally positioned cornutus that is relatively small, squat, and “foot” shaped in ventral view, with approximately evenly sized spines along the distal margin, rather than very large, narrow, and tapered, with a large apical spine that is curved downward and to the left and increasingly smaller spines along the outer margin. Mesene pirata is distinguished from the similar new species M. lojensis, from southwestern Ecuador, in that species account. Wing pattern variation in the examined male specimens of M. pirata is minimal, and limited to the exact shape of the distal margin of the orange patch on both wings. A very narrow band of black scales around the tip of the abdomen is also variably present or absent. A photograph of an unidentified flowerfeeding Mesene in the Guide to the Butterflies of Honduras by Gallardo & Diaz (2022) appears to be a female of M. pirata. It differs from the male of M. pirata by having a much larger orange patch on the forewing that extends across the discal cell, then about half way to the distal margin, and has an inwardly rounded instead of straight and inwardly diagonal distal margin. It differs from the female of M. nemo (Fig. 3B) by having entirely black distal wings with no white submarginal spots (at least dorsally) or white forewing fringe elements. Biology: This species is known to inhabit lowland seasonal evergreen forest near sea level. Males were encountered at the type locality of Mud Hole Hill perching as solitary individuals or in pairs in large open areas at two high points along the undulating forested ridge extending northeast from the antenna (Figs. 11A-C). They perched from 1520 to 1630 hrs, sometimes under leaves with their wings open 1 to 3 m above the ground, but most commonly beneath the tips of long grass blades. At a hilltop site above Punta Blanca, a solitary male was encountered perching 8 m above the ground along the edge of a broad forested track at 1530 hrs. A female of M. pirata was photographed feeding on flowers on Guanaja (Gallardo & Diaz, 2022), an island about 25 km east of Roatán. The type series was collected in September 2011, in the early rainy season, and the species was not seen again on a repeat visit to both of the aforementioned Roatán localities in August 2013. Population growth and development on Roatán have been explosive during the last two or three decades (e.g., Tuholske et al., 2017), and, sadly, recent satellite imagery reveals significant deforestation at both localities, highlighting the fragile conservation status of this species in the Bay Islands. Distribution: Mesene pirata has only been collected at two localities along the central hilly spine of Roatán in the Caribbean Bay Islands of northern Honduras (Fig. 12), and photographed on the hilly eastern neighboring island of Guanaja. It seems less likely to occur on the western neighboring island of Utila, which is smaller, very flat and swampy. However, the species seems highly unlikely to have a relictual distribution that is confined to the coastal Bay Islands given the lack of otherwise known lepidopteran endemics there. The fact that M. monostigma group species all appear to be alloor parapatrically distributed, and the geographically closest species, M. nemo, is known to range only as far west as eastern Panama, suggests that M. pirata may have a more widespread distribution within similar moist forest habitats on the central mainland of Central America. The fact that it has thus far eluded detection there (e.g., DeVries, 1997; Maes, 1999; Meerman, 1999; Glassberg, 2017; Gallardo & Diaz, 2022) may be due to its rarity and the difficulty in finding its preferred open-summit, forested hilltop perching sites. Mesene lojensis Hall, new species (Figs. 5; 10; 11D-F; 12, 13) Description: MALE: Forewing length HT 12 mm, PTs 11.5-12 mm. Wings: see Fig. 5; fringe on both wings entirely black. Head: Eyes dark brown and bare with dark brown marginal scaling; frons dark brown with increasingly dense cream scaling in ventral third; labial palpi dark brown with sparse cream scaling overall, especially ventrally, and dense cream scaling at tip of segment three; antennal length approximately 60% of forewing length, segments black with a few whitish scales at base and narrow nudum section along inner margin, clubs black with an orange-brown tip. Body: Dorsal surface of thorax a mixture of black and dark orange scaling with dark orange tegula, ventral surface black with cream setae; dorsal surface of abdomen predominantly dark orange with some scattered black scaling and a black tip, ventral surface predominantly whitish with some scattered black scaling and dark orange scaling around genital opening; foreleg pale brown, midand hindlegs dark cream with a black femur; a narrow and continuous band of concealed androconial scales present dorsally across anterior margin of abdominal tergites four, five, and six; eighth abdominal tergite and sternite approximately rectangular. Genitalia: see Fig. 10; ventral posterior corner of lower posterior valve process slightly inwardly curled and posterior tip of lower posterior valve process prominently Fig. 12. Map showing the known distributions of the three new Transandean Mesene monostigma group species. TROP. LEPID. RES., 35(2): 75-84, 2025HALL: Mesene monostigma group from Transandean region 82 inwardly tilted, with long sparse setae present along posterior margin of upper and lower posterior valve processes; lightly sclerotized transtilla band extends continuously over aedeagus from behind upper posterior valve processes; ductus ejaculatorius enters aedeagus anteriorly, vesica exits aedeagus posteriorly. Types: HOLOTYPE male: ECUADOR: Loja, km. 3 Chaguarpamba-Velacruz rd., 3˚53'49"S 79˚38'07"W, 1450 m, 22 May 2022 (J. P. W. Hall) (to be deposited INABIO). PARATYPES: ECUADOR: Loja, 1 male: same data as HT (JH); 2 males: Jorupe, Chaguarpamba-Chiguango rd., 3˚51'33"S 79˚34'47"W, 1200 m, 17 May 2024 (J. P. W. Hall) (INABIO); 1 male: Cordillera de Ramos, ChaguarpambaEl Rosario rd., 3˚48'52"S 79˚38'08"W, 1100 m, 18 May 2024 (J. P. W. Hall) (INABIO); 2 males: El Alto de la Cruz, E. off Alamor-Puyango rd., 3˚53'46"S 79˚59'47"W, 1000 m, 12-13 May 2024 (J. P. W. Hall) (INABIO). Etymology: This species is named after the topographically and ecologically varied southwest Ecuadorian province of Loja, to which it may be largely endemic. Diagnosis: The male of M. lojensis n. sp. (Fig. 5) differs from that of the closely related new wet forest Transandean species M. nemo (Fig. 3A) by being slightly larger on average and having a slightly narrower and more vertically elongate wing shape, darker, duller, and more brownish orange patches on both wings, a much larger orange patch on the forewing that extends close to the distal margin and inward from the middle of cell Cu1 to a point at or just distal to the discal cell end, no white subapical spot in cell M3 on the dorsal forewing, only occasionally (in about a quarter of specimens) a dirty cream to pale orange subapical spot in cell M3 on the ventral forewing partially visible at the apex of the orange patch, no other white submarginal spots, more extensive dark brown scaling at the base of the ventral hindwing, and an entirely black forewing fringe without white elements in cell Cu2, cell M3, and across cells M1 to R4+5. The male of M. lojensis also has different coloration on the body. The abdomen is dorsally dark orange with some black scaling instead of pure bright orange, and more contrasted ventrally, being predominantly whitish with some black scaling rather than pure pale yellow, the frons and palpi have cream scaling that is less extensive and more whitish than yellowish, the forelegs are brown instead of cream, and the midand hindlegs are darker cream with a much darker femur. In the male genitalia of M. lojensis (Fig. 10), the valvae differ from those of M. nemo (Fig. 8) by having a lower posterior process with a more posteriorly triangular than vertically rectangular shape, a slightly straighter posterior margin, and a slightly longer, more asymmetrically shaped, prominently inwardly tilted, and more posteriorly than dorsally directed posterior tip, and an upper posterior process that is more asymmetrically shaped, slightly downwardly instead of upwardly posteriorly directed, and slightly shorter rather than slightly longer than the lower posterior process tip. The everted vesica of M. lojensis has a slightly smaller, narrower, and basally more finely pointed dorsobasally positioned cornutus in dorsal view, a shorter and basally more finely pointed ventrobasally positioned cornutus with smaller apical spines, and a bundle of long, thin, centrally positioned spines that lacks the basally anteriorly lengthened and bent ventral spine characteristic of the three monostigmalike species. The male of M. lojensis primarily differs from that of the new northern Transandean species M. pirata (Fig. 4) by being noticeably larger and having a narrower and more vertically elongate wing shape, larger orange patches on both wings that are a darker, duller, and more brownish orange, an abdomen that is dark orange and black instead of bright orange dorsally and whitish and black instead of pale orange ventrally, a larger patch of brighter cream instead of pale brown scales ventrally on the frons, brown and cream instead of entirely brown palpi, and cream rather than brown tibia to the midand hindlegs. The male genitalia of M. lojensis are quite distinct from those of M. pirata (Fig. 9), having a prominently protruding ventral posterior corner to the valvae, differently shaped valve tips, and smaller and more asymmetrically shaped basal cornuti. Wing pattern variation in the examined male specimens of M. lojensis includes the exact size and shape of the orange patch on the forewing, which is typically distally angular and ends at the discal cell end (as in Fig. 5), but sometimes has orange scaling extending slightly beyond the discal cell end and a wider block of orange scaling in cell M3 that corresponds to a variably weakly differentiated dirty cream to pale orange spot in the same space on the ventral surface, and minor differences in the width of the dark band along the distal margin of the hindwing and at the base of the ventral hindwing. Biology: This species is known to inhabit seasonal evergreen forest from 1000 to 1450 m (Figs. 11D,E). Its presence at multiple localities within this elevational range, but absence from suitable sites at higher elevations suggests that it is a lowland or premontane rather than montane species. Males were encountered perching in open sunny locations on hilltops and ridgetops, such as large lightgaps, broad tracks (Fig. 11E), and forest edges, from 1100 to 1500 hrs, with an earlier end to perching activity observed at higher elevations. At more heavily forested sites, males were encountered as solitary individuals perching beneath leaves with their wings open 4 to 8 m above the ground, but at more open sites, with lower vegetation, they could be found perching as low as 1 or 2 m. On a severely environmentally degraded ridge near Jorupe, a small group of males was found perching beneath the tips of long grass blades at the edge of a small forest fragment (Fig. 11F). The type series was collected in May, at the end of the rainy season. This species is not uncommon at its preferred perching sites, and the fact that it does not appear to have previously been collected is largely the result of its very restricted range (see below) and the severe fragmentation of its habitat. Southwestern Ecuador has been inhabited and farmed for centuries, and, exacerbated by extended dry seasons during which hillsides are frequently intentionally burned, the natural vegetation has become fragmented into ever smaller and increasingly isolated patches (e.g., Best & Kessler, 1993; Tapia-Armijos et al., 2015). Noh et al. (2020) identified the known habitat type for M. lojensis (Piedmont Seasonal Evergreen Forest of the Catamayo-Alamor zone [see below]) as one of Ecuador’s 10 Critically Endangered or Endangered forest habitats (out of 64), based on the IUCN’s (International Union for Conservation of Nature) Red List of Ecosystems criteria (Keith et al., 2013), highlighting the concerning conservation status for this species. Distribution: Mesene lojensis has only been found at a few localities along the western boundary of Loja province in TROP. LEPID. RES., 35(2): 75-84, 2025 HALL: Mesene monostigma group from Transandean region 83 southwestern Ecuador (Figs. 12, 13). The habitat zone within which the species has been recorded, and to which it may be endemic, was categorized as Piedmont Seasonal Evergreen Forest of the Catamayo-Alamor zone (400-1600 m) by Noh et al. (2020) (the brown-shaded area in Fig. 13). This habitat type is confined to western Loja and El Oro provinces, and the species may occur in ecologically similar areas of El Oro, although I have yet to find it there. However, as delimited by Noh et al. (2020), this habitat type encompasses both a broad range of elevations and a moderate range of rainfall zones, a substantial portion of which are probably not suitable for M. lojensis. For example, much of the seasonal evergreen forest between Marcabelí, Balsas, and Piñas (in El Oro), in addition to that categorized as Piedmont Evergreen Forest of the CatamayoAlamor zone by Noh et al. (2020) (the peach-shaded area in Fig. 13, typified by Jocotoco’s Reserva Buenaventura) may be too low, wet, and humid, and the low-elevation seasonal evergreen forest bordering coastal semideciduous forest zones and lining the length of the dry Río Puyango valley, which forms the entire southern border of El Oro, is almost certainly too dry. In addition, based on my significant experience exploring the back roads of this region, I would estimate that the forest extent depicted by Noh et al. (2020) is overly generous, even for the 2014 date that it was mapped (by Ecuador’s Ministerio del Ambiente). Mesene lojensis could conceivably also occur in Piedmont Seasonal Evergreen Forest of the Western Andes (the pink-shaded area in Fig. 13), which extends in a narrow, inaccessible, and very poorly sampled strip north from El Oro into the provinces of Azuay and Cañar, but semideciduous habitats extending south into Peru are almost certainly too dry for the species. The geographically and perhaps phylogenetically closest relative, M. nemo, is confined to very wet lowland evergreen forest, and known to range only as far south as the northern edge of northwestern Ecuador’s Pichincha province (see Fig. 12). Mesene lojensis genuinely seems to have a very small relictual range, and to be endemic to southwestern Ecuador. ACKNOWLEDGMENTS I thank M. A. Solis for her companionship in the field and forbearance in the pursuit of day-flying Lepidoptera during two trips to Roatán, Honduras; K. R. Willmott, my tireless collaborator on the Butterflies of Ecuador project, for a multitude of reasons, not least logistical and moral support during my solo explorations in Ecuador; all those museum curators in Europe and the Americas who gave access to the riodinid collections in their care (see list in Hall (2018)); the National Geographic Society (Research and Exploration Grant #5751-96) and the National Science Foundation (DEB #0103746 and #0639977) for some financial support of field and museum research; the Instituto Nacional de Biodiversidad and Ministerio del Ambiente, Agua y Transición Ecológica for arranging the necessary permits for research in Ecuador, most recently 2023-MAATE-DBI-CM-2023-0298; and Fernando Dias and Andre Freitas for helpful comments on the manuscript. Fig. 13. Map of the forest ecosystems of southern Ecuador (bounded to the south by Peru and to the west by the Pacific ocean) showing the known records for Mesene lojensis n. sp. The forest ecosystems classification system is based on 2014 forest-cover data presented for the entirety of Ecuador by Noh et al. (2020). A key to the 20 different western forest types of southern Azuay, El Oro, and Loja provinces is presented to the right of the map, ordered from lowest to highest elevation and, within each elevational zone, from driest to wettest and/or geographically from north to south. Note that adjacent southern Morona-Santiago and Zamora-Chinchipe provinces are included in the map for geographic context only, and their distinct eastern forest types, which extend slightly across the continental divide into the highest elevations of the eastern third of Azuay and part of the eastern border area of Loja, are shown in shades of gray. Notable cities and towns are pinpointed on the map with two-letter codes (keyed out below the map) to aid in visualizing the location of forest types in relation to the road network. TROP. LEPID. RES., 35(2): 75-84, 2025HALL: Mesene monostigma group from Transandean region