Two new species of Euptychia Hübner, 1818 from the upper Amazon basin (Lepidoptera, Nymphalidae, Satyrinae)

Abstract Two new species of Euptychia Hübner, 1818 are described from the upper Amazon basin: Euptychia attenboroughi Neild, Nakahara, Fratello & Le Crom, sp. n. (type locality: Amazonas, Venezuela), and Euptychia sophiae Zacca, Nakahara, Dolibaina & Dias, sp. n. (type locality: Acre, Brazil). Their unusual facies prompted molecular and phylogenetic analyses of one of the species resulting in support for their classification in monophyletic Euptychia. Diagnostic characters for the two species are presented based on wing morphology, wing pattern, presence of androconial patches on the hindwing, and genitalia. Our results indicate that the projection of the tegumen above the uncus, previously considered a synapomorphy for Euptychia, is not shared by all species in the genus. The adults and their genitalia are documented, and distribution data and a map are provided.


Introduction
The nymphalid subtribe Euptychiina (Satyrinae: Satyrini) is one of the most poorly known butterfly groups. More than 400 predominantly Neotropical species in some 50 genera are recognised within the subtribe and many taxa remain undescribed (Lamas 2004;pers. obs.). However, generic classification of species within the subtribe is confused mainly because of a lack of clear morphological characters and morphological homogeneity (Peña and Lamas 2005). In addition, their drab coloration has probably contributed to this group being ignored by many lepidopterists in the field. Forster (1964) described 33 euptychiine genera which are mostly accepted today, but provided few reliable diagnostic characters for these taxa. As a result, placement of species in these genera is often tentative: by way of example, genera such as Magneuptychia Forster, 1964, Cissia Doubleday, 1848, and Splendeuptychia Forster, 1964 are recovered as polyphyletic or paraphyletic in recent phylogenetic analyses (e.g. Peña et al. 2010), indicating the confused generic-level taxonomy. On the other hand, the genus Euptychia Hübner, 1818 is relatively well known compared to other euptychiine genera, and is morphologically defined by the posterior projection of the tegumen above the uncus in the male genitalia. This character is considered to be a good synapomorphy (Freitas et al. 2012) and has been used to classify recently described Euptychia species (Neild et al. 2014. Many other described and undescribed Euptychia species also possess this character (Nakahara, unpub.), and its presence even contributed to the inclusion of Caenoptychia boulleti Le Cerf, 1919 in Euptychia (Freitas et al. 2012). However, we here describe two new species of Euptychia which lack this posterior projection of the tegumen. The placement of these two species in Euptychia is supported by molecular data, which are presented in this paper, and by many alternative possible diagnostic characters for the genus, which are discussed.

Morphology
Comparison of the morphology of the two new species was made with other Euptychia specimens in the collections listed below (museum acronyms are from Heppner and Lamas 1982): Type specimens of Euptychia species in the BMNH were also checked, as well as photographs of additional taxa on the website "Butterflies of America" (available online in Warren et al. 2015). In addition, recent works on French Guianan Euptychia by Brévignon (2005Brévignon ( , 2008 and Brévignon and Benmesbah (2012) were consulted, and our taxa were discussed with the authors of those papers. Wings were diaphanised using standard techniques for Lepidoptera. The abdomens of the Venezuelan pair of E. attenboroughi sp. n. and one of the Brazilian males of E. sophiae sp. n. were dissected to observe genital structures. Abdomens were dissected using standard techniques, with adult abdomens being soaked in hot 10% KOH solution for 3-10 minutes, dissected and subsequently stored in microvials in glycerine. External morphology and dissections were studied using stereomicroscopes and photographed using digital cameras. The terminology for genital and abdominal structures follows Scoble (1992), with additional detail derived from Klots (1956). Our use of the term "vinculum" agrees with Austin and Mielke (2008). Nomenclature for wing venation corresponds to that of Miller (1970: 46), and wing areas to Neild (2008: fig. 1.2).
The following abbreviations are used:

Molecular and phylogenetic analysis
A leg from the Brazilian male paratype (DZ 29.578) of E. sophiae sp. n. was used to obtain a DNA sequence for cytochrome c oxidase subunit I (COI) and elongation factor-1 alpha (ef1a). We extracted DNA using Qiagen's DNeasy Blood & Tissue Kit, following the protocol and using a final elution volume of 50 µl. Primers LCO_nym (forward, 5'TTTCTACAAATCATAAAGATATTGG 3') and HCO_nym (reverse, 5'TAAACTTCAGGGGTGACCAAAAAATCA 3') were used to amplify COI. Elongation factor-1 alpha was amplified by using primer pairs Ef44 (forward, 5'GCYGARCGYGARCGTGGTATYAC 3'), and EfrcM4 (reverse, 5'ACAGCVACK-GTYTGYCTCATRTC 3'); however, since this primer pair failed to amplify ef1a, the following primer pair was used to amplify short fragments of ef1a: Ef44 (forward, 5'GCYGARCGYGARCGTGGTATYAC 3') and Monica (reverse, 5'CATRTT-GTCKCCGTGCCARCC 3') (Monteiro and Pierce 2001). All PCR reactions were conducted in a 25 µl volume comprising 1 µl of template DNA, 9.5 µl ddH2O, 1 µl of each primer (10 µM), and 12.5 µl Omega 2x Taq Table 1). Sequences were aligned using MAFFT v. 7.107 (Katoh et al. 2002). Best-fitting models and partitioning schemes were jointly selected using PartitionFinder v. 1.1.1. (Lanfear et al. 2012). The program was run twice, once with the models available in MrBayes 3.2.3 (Ronquist et al. 2012) and once with those in RAxML v. 8.1.11 (Stamatakis 2006). The selected partitioning schemes and models can be found in Table 2. Phylogenies were inferred using maximum likelihood (ML) and Bayesian inference on the concatenated data set of 1041bp. ML analyses were run using RAxML v 8.1.11 (Stamatakis 2006, Stamatakis et al. 2008) with the GTR model for all partitions (as selected by PartitionFinder) and 1000 rapid bootstrap replicates followed by 200 thorough maximum likelihood searches with joint branch length optimization. Bayesian analyses were run in MrBayes 3.2.3. Analyses were run twice for 10 million generations with trees sampled every 1000 generations with the temperature parameter set to 0.15. Model parameters were unlinked across data partitions, and the prior for rate variation among partitions was set to variable. The first 25% of the samples were discarded as burn-in. Convergence was checked in Tracer 1.6 (Rambaut and Drummond 2007). RAxML and MrBayes were run on the CIPRES cluster (Miller et al. 2010).

Results
According to ML and Bayesian inference based on barcoding region and EF1a, E. sophiae sp. n. is sister to the clade consisting of E. picea Butler, 1867 and two undescribed species. Unlike Freitas et al. (2012), Euptychia boulleti (Le Cerf, 1919) is here placed as sister to the clade with E. enyo Butler, 1867, E. westwoodi Butler, 1867, and an undescribed species in the ML tree, although with relatively low support. However, in the Bayesian tree ( Fig. 1), this species appears as sister to the rest of Euptychia. Thus in this tree the ancestor of Euptychia (minus E. boulleti) has no bootstrap support since this node does not exist in the ML tree. Placement of E. sophiae sp. n. in monophyletic Euptychia is supported by both analyses. The results of this molecular analysis are supported by the ML analysis in Nakahara et al. (2015), which included over 150 euptychiine and other satyrine taxa and recovered the genus Euptychia as monophyletic and including E. sophiae (as Euptychia_n_sp).

Taxonomy
Diagnostic morphological characters seen in these two new species which appear to be unique among the genus Euptychia include: (i) presence of a single signum in female E. attenboroughi sp. n. (no female is known for E. sophiae sp. n.); (ii) lamella antevaginalis unsclerotised around the ostium bursae of female E. attenboroughi sp. n. (sclerotised in all other Euptychia examined), except on the ventral surface around the neck of the ostium bursae; (iii) reduced third segment of labial palpus of female E. attenboroughi sp. n. (Fig. 5); (iv) presence of a prominent ventral hindwing androconial patch in male E. sophiae sp. n. None of these characters are known in any other species of Euptychia we have examined and are considered to be diagnostic for E. attenboroughi sp. n. (states i, ii and iii) and E. sophiae sp. n. (state iv). For diagnostic characters for E. attenboroughi sp. n. and E. sophiae sp. n., consult the diagnosis for each species. Diagnosis. Differs from males of its congener E. sophiae sp. n. as follows (no females of E. sophiae sp. n. are known, but we expect characters indicated with an asterisk (*) will serve to differentiate this sex): (1) FW more produced apically, with outer margin straighter or more concave, and outer margin at an angle away from the central line of the body (nearly parallel in E. sophiae sp. n.); (2) dorsally with prominent orange scaling on the posterior DHW on and near the tornus* (E. sophiae sp. n. lacks this orange scaling); (3) VHW submarginal band does not increase in width at the tornus (increases substantially for E. sophiae sp. n.); (4) small ocellus at the anal margin of the VHW median band larger*; (5) androconial patch on VHW pale and barely visible, whereas prominent on E. sophiae sp. n.; (6) gnathos projecting nearly parallel to the uncus (projecting vertically in E. sophiae sp. n.); (7) distal one-fourth of the valva broad in lateral view (narrow in E. sophiae sp. n.); (8) aedeagus straighter, less curved.
The female resembles several species in the genus Euptychia, especially E. marceli Brévignon, 2005, but is distinguishable on the VHW through the unique combination of a very large and ovoid (not round) ocellus in Cu 1 -Cu 2 bordered on its tornal side by orange scaling, and by the presence of a tiny ocellus on the anal margin at the posterior end of the median brown band (this last found rarely on some specimens of E. marceli Brévignon, 2005). It can also be differentiated from all other known Euptychia females by the three characters (i, ii, and iii) elucidated at the beginning of the Taxonomy section. The male somewhat resembles a few taxa currently in Chloreuptychia Forster, 1964(see Warren et al. 2015, with its wing shape and greyish translucence, but differs by the absence of a bluish sheen on either wing surface, presence of orange HW tornal scaling on both wing surfaces, presence of a small VHW ocellus at the anal margin, absence of orange-and silver-lined ovoid VHW markings in M 2 -M 3 and M 3 -Cu 1 , and by the single silvery-white pupils in the VHW ocelli (double in most Chloreuptychia species).
Eyes. Dark brown, sparsely hairy; creamy-grey scales dorsally and laterally along posterior edge of eyes.
Palpi. Covered by long creamy-grey hair-like modified scales dorso-laterally, ventrally with long fine hair-like modified scales projecting like a Mohican, mostly black along outer margin, but interior wall of modified scales creamy-grey. Mohican highest in centre, gradually reducing anteriorly and posteriorly, and anteriorly reduced to a pointed tuft. First segment covered with black scales dorsally, black and white hair-like modified scales ventrally, second segment covered with short white hair-like modified scales and white scales laterally, black scales distal one-third of dorsal surface, ventrally adorned with long black and white hair-like modified scales 3-4 times as long as segment width, second segment slightly longer than eye diameter, third segment covered with black scales dorsally and ventrally, creamy-white hair-like scales laterally, about one-seventh of second segment in length.
Thorax. Covered in long light grey hair-like modified scales. Legs. Greyish. Foreleg tarsus about 2/5 of tibia in length, femur about 2/3 of tibia in length; tibial spurs absent on midleg and hindleg.
Abdomen. Eighth tergite and sternite well developed, apparently as equally sclerotized as other tergites and sternites, but weakly sclerotized towards posterior end.
Androconia patches. Modified wing scales, presumed to be androconia, present on either side of vein 2A on the HW at the base of the dark median band on the dorsal and ventral surface; visible on the dorsal surface as two tiny ovoid pale greyish-brown patches approximately 0.5 mm long, and on the ventral surface as a strip approximately 2.0 mm long in space 2A-3A; these patches, best viewed using backlighting, are homologous with the dorsal and ventral androconia patches of E. sophiae sp. n.
Wing venation. FW recurrent vein present, approximately 1.75 mm long; FW vein Cu not swollen at base; HW with humeral vein barely visible, very short (approximately 0.6 mm), curved anteriorly towards the costal lobe. Wing shape. FW costa gently convex to apex, apex relatively pointed, outer margin straight, or slightly concave between M 3 and Cu 2 , and angling about 20 degrees away from the central line of the body, inner margin almost straight; HW costa lobed in basal area then gently concave to apex at Sc+R 1 , apex rounded, outer margin scalloped, anal margin concave near tornus, basally convex.
DFW. Both wings slightly and variably translucent with greyish-brown to chestnut brown ground colour; fringes greyish-brown to brown; four diffuse dark brown to chestnut bands crossing from the costal to the anal margin, the first basal and barely visible, mostly ghosting through the slightly translucent wing from the ventral surface, centrally wide but tapering to a point anteriorly and posteriorly, the second submedian, wider, nearly straight and better defined, crosses the mid discal cell from mid costa to four tenths along the inner margin, the third median slightly wavy strongly defined and the widest, crosses from the costa near the discocellular veins, which it traverses, reaching to seven tenths distance along the inner margin, the fourth begins near the apex where it is narrow and very sinuous, runs parallel to the outer margin and curved in each interspace down to M 3 and then angles without curves in towards the body and widens reaching the submarginal area of Cu 1 , then again running parallel to the outer margin down to the inner margin near the tornus; the margin with a very fine dark brown line running parallel to the outer margin, beginning at the apex, incurved in each interspace to M 3 , then straight to the tornus; a white-pupilled black subapical ocellus in the centre of M 1 -M 2 , touching M 2 but not quite reaching M 1 .
DHW. Four diffuse dark brown to chestnut bands crossing from the costal to the anal margin, the first basal, mostly ghosting through the slightly translucent wing from the ventral surface, the second submedian also ghosting through, nearly straight, crosses the mid discal cell from mid costa to half distance along the anal margin, the third median and slightly wavy, in some specimens (two of the three males) curved inwards in the distal discal cell, better defined with dorsal scaling (less ghosting) and the widest, crosses from the costa two-thirds towards the apex, almost reaching 2A two-thirds along its length, and not passing to the anal margin, the fourth submarginal begins near the apex where it is narrow, runs parallel to the outer margin (curved in each interspace), gradually thickening, widest in M 2 -M 3 (where the basal edge points inward) and M 3 -Cu 1 , and then thinning gradually to just reach 2A near the tornus; the margin with a very fine clearly defined dark brown line running parallel to the outer margin, beginning at Sc+R 1 , incurved in each interspace to the tornus and entering the anal lobe to 3A; three ventral ocelli visible through the slightly translucent wings, the dark circular areas showing through from the ventral surface in cells Cu 1 -Cu 2 and M 1 -M 2 covered with small spheroid areas of diffuse very dark brown dorsal scaling, the latter very small, the former roughly half the diameter of the black ventral "iris"; orange scaling on the distal side of the large ocellus in space Cu 1 -Cu 2 continues to the outer margin and in the same areas of Cu 2 -2A.
VFW. Both wings slightly translucent with pale greyish-brown ground colour; fringes greyish-brown; one very thin well-defined submarginal dark brown to chestnut band and four more diffuse dark brown to chestnut basal bands, submedian, median and postmedian crossing from the costal to the anal margin, as described for the dorsal surface; a silvery-white-pupilled black subapical ocellus in the centre of space M 1 -M 2 , touching M 2 but not quite reaching M 1 , circled by a gold ring with a thin grey-brown outer edge which enters R 5 -M 1 ; in M 2 -M 3 the grey-brown edge of the ocellar ring breaks open and the yellow area spills posteriorly into the centre of the interspace; in one specimen (MGCL) there is an additional tiny ocellus in Cu 1 -Cu 2 with a reddishbrown "iris" (of the same colour as the transverse bands), a tiny pupil (or merely missing scales?), and an outer ring of the same colour as the background pale-grey-brown, surrounded by an indistinct scattering of brownish scales that define the edge of the outer pale ring. VHW. One very thin well-defined marginal dark brown band and four more diffuse dark brown to chestnut basal bands, submedian, median, and marginal crossing from the costal to the anal margin, as described for the dorsal surface, but the submarginal and marginal bands continue from the tornus along the anal margin to the base of the chestnut median band, while the marginal band (only) continues to the base of the submedian band; a very small ocellus on the anal margin at the base of the median brown band composed of a large black subovoid "iris", a narrow golden outer ring, and a narrow dark brown border; three postmedian ocelli, composed of a single small silvery-white pupil, a large black spheroid or ovoid "iris", and a narrow golden outer ring enclosed in a narrow grey-brown border; the smallest of these ocelli is spheroid, entirely within Rs-M 1 , not touching either vein, the second, nearly twice as large, spheroid and occupying the full width of M 1 -M 2 , with the grey-brown outer border just touching M 1 , and the black "iris" just spilling over into M 2 -M 3 , and in the same way as on the VFW, with the outer posterior border broken open in M 2 -M 3 and the golden outer ring protruding into the centre of the interspace, where a poorly defined wide suffused band of greyish brown scales links this and the third largest ocellus, ovoid, over three times the diameter of the first at its widest point (parallel to the outer margin), with the black "iris" fully occupying Cu 1 -Cu 2 , spilling into M 3 -Cu 1 but just touching Cu 2 , the golden outer ring double the width of the other two ocelli, reaching the centre of M 3 -Cu 1 , and entering Cu 2 -2A, surrounded by a grey brown ring with indistinct edges; yellow-orange scaling on the distal side of the large ocellus in space Cu 1 -Cu 2 continues slightly into M 3 -Cu 1 and fully to the submarginal band of Cu 1 -Cu 2 with more extensive orange scaling in the distal quarter of Cu 2 -2A just spilling over 2A onto the anal lobe.
Male genitalia (Fig. 4). Tegumen dorsally flattened, ventral edges concave, posterior margin projecting to form a short gnathos fused to the tegumen (approximately one-fifth length of uncus) almost parallel to uncus, but slightly ventrad, somewhat trapezoid in dorsal view; uncus anteriorly hairy, rather narrow and long, posterior tapered and slightly hooked in lateral view, evenly wide in dorsal view; ventral arms of tegumen fused to anterior margin of tegumen; appendices angulares absent; saccus slightly longer than uncus, dorsal arms of saccus combined with ventral arms from tegumen; valva sparsely hairy, basal three-fourths vaguely trapezoidal, distal one-fourth rounded, distal half of valva in dorso-ventral view resembles propodus of a lobster, but without the dactylus; aedeagus tubular, in lateral view rather straight, slightly broadening anterior portion which opens anterodorsally, posterior one third of aedeagus narrow, slightly bent upwards, cornuti absent. FEMALE (Fig. 3): Forewing length 16.0-17.0 mm (n = 3). Similar to male except as follows: FW shape not elongate, subtriangular, with distinctly convex outer margin; HW shape similar to male, but slightly less elongate; dorsal surface with all the same dark semitranslucent bands but the basal, submedian, and median bands with more scales present dorsally; the DHW dark circular areas showing through from the ventral surface in cells Cu 1 -Cu 2 and M 1 -M 2 almost covered with large spheroid areas of diffuse very dark brown, nearly black, dorsal scaling; the large subtornal ocellus entirely encircled   by diffuse orange scaling, and distally by a half-circle of golden scales homologous with the ventral outer ring of this colour; ventrally there are no obvious differences.
Etymology. We name this butterfly to honour the great English naturalist, author, and TV presenter, Sir David Attenborough, in gratitude for opening the eyes and hearts of millions to the natural world through his inspiring and edifying work. To prevent any future ambiguity, the name attenboroughi is considered to be a Latinised male noun in the genitive case.
Distribution, behaviour and habitat. The six specimens known to date were all collected within 500 kms of each other in the north-west of the upper Amazon basin, representing a very restricted distribution. It is impossible with such a small sample size to draw any concrete conclusions, but we hypothesise that this species is restricted to suitable habitat to the north of the Amazon river, and that its sibling species occurs only to the south, although a limited area of sympatry may exist. One of the senior authors first collected specimens of this new species in 1994 while conducting field work for the Butterflies of Venezuela book series (Neild 1996, Neild 2008 in south-western Venezuela. The pair that he collected were settled on low vegetation along a path inside tropical evergreen forest and were netted immediately, before any further observations could be made. One of the Colombian specimens was collected in similar habitat at 10:30 a.m., while the female was captured at 2:45pm outside the forest in scrubby secondary vegetation ("rastrojo"). The type series were all found at low elevations from about 100 m to almost 300 m above sea level. The two Venezuelan specimens were collected in the first half of March, during an especially strong dry season which drastically reduced butterfly numbers and species diversity. The Colombian and Brazilian specimens were taken in August and September, months that are also typically among the least wet of the year. The species is evidently very rarely collected, but this may not reflect reality in the field; rather its perceived scarcity may simply be the result of its apparently highly restricted distribution in an area of the Amazon basin that has been, and still is, very little explored.
Host plant. Unknown. Palpi. Covered by long creamy-grey hair-like modified scales dorso-laterally, ventrally with long fine hair-like modified scales projecting like a Mohican, mostly black along outer margin, but interior wall of modified scales creamy-grey. Mohican highest in centre, gradually reducing anteriorly and posteriorly, and anteriorly reduced to a pointed tuft. First segment covered with brown scales dorsally, black and creamy hair-like modified scales ventrally, second segment covered with short creamy hair-like modified scales and light brown scales laterally, brown scales distal one-third of dorsal surface, ventrally adorned with long black and creamy hair-like modified scales 3-4 times as long as segment width, second segment slightly longer than eye diameter, third segment covered with brown scales dorsally and black scales ventrally, creamy hair-like scales laterally, about one-seventh of second segment in length. Thorax. Uniformly covered by dark grey-brown hair-like scales. Legs. Greyish. Foreleg tarsus more than half-length of tibia, femur about 2/3 of tibia in length; tibial spurs absent on midleg and hindleg.

Euptychia sophiae
Abdomen. Eighth tergite and sternite well developed, apparently as equally sclerotized as other tergites and sternites, but weakly sclerotized towards posterior end.
Androconial patches. Two small (< 1.0 mm) pale grey androconial patches on DHW, barely separated by 2A, located at its distal one third; patch in cell 2A-3A prominent; patch in cell Cu 1 -Cu 2 restricted to width of median band and located at juncture of this band with 2A. A black and short (approximately 1.5-2.0 mm) androconial patch at the distal third of 2A on the VHW.
DFW. Greyish brown, darker along the costal and outer margins with a narrow ochre area on the first fourth of the costal margin length. Four dark brown to rufous bands, the former basal, dark brown (approximately 0.1 mm width) following the radial vein on its distal edge, the second submedian, dark rufous brown, from near the origin of R 1 to 2A, crossing the discal cell near the middle, the third median, rufous and slightly concave, crossing the cell end from the origin of R 3 across the base of Cu 1 to 2A, curving distally near the inner margin, and the final band submarginal, dark brown, narrower and slightly crenulated from R 4 to M 3 , rufous and posteriorly enlarged from M 3 to 2A. Ocellus of the VFW observable through transparency.
DHW. Greyish brown. Four dark brown to rufous straight bands, the first basal, dark brown, short and tapered, the second submedian, dark rufous brown, from costal margin to anal margin, crossing the discal cell near the centre, the third median, rufous brown, from costal margin to anal margin, crossing the discal cell in its distal quarter, the final band submarginal, rufous and crenulated in each cell from Rs to anal margin, except Cu 1 -Cu 2 , following the contour of the outer margin, but widened and curved inwards in M 2 -M 3 and to a lesser degree in M 3 -Cu 1 , wide and reddish orange from Cu 1 to 2A, thinned from 2A to anal margin where it nearly reaches the median rufous band. Ocelli of the VHW observable through transparency.
VFW. Light greyish brown, bands similar to DFW. One developed black ocellus from M 1 to anterior sixth in M 2 -M 3 , with a white pupil at the centre and a broad and yellow external ring, surrounded by a greyish brown area that extends posteriorly to the posterior half of M 3 -Cu 1 . Submarginal line dark brown, from R 4 to anal margin, crenulated in R 4 -M 3 , remainder and straight, distally surrounded by a thin yellow line. Fringes dark brown.
VHW. Light greyish brown, bands similar to DHW. Three postmedian black ocelli, the anterior the smallest in Rs-M 1 , the second twice as wide as the first, from M 1 to anterior third of M 2 -M 3 , and the posterior bigger, about three times wider than the second, ovoid, from the posterior third in M 3 -Cu 1 to the edge of Cu 2 , all three with a white pupil at the centre and a broad and yellow external ring. A fourth minute black ocellus, with a yellow outer ring but with no pupil, located on the anal margin at the base of the postmedian brown band. Marginal dark brown, thin, and crenulated line from Sc+R 1 to 3A. Fringes dark brown.
Male genitalia (Fig. 8). Tegumen dorsally flattened, trapezoidal, lateral posterior margin projecting ventrad as a short gnathos fused to the tegumen, subtriangular, nearly at right angle to uncus; uncus basally hairy, almost 10 times longer than wide, laterally apex curved downward; ventral arms of tegumen fused to anterior margin of tegumen; appendices angulares absent; anterior projection of saccus almost same length as uncus, dorsal arms of saccus combined with ventral arms from tegumen; valva sparsely hairy, basal two-thirds shaped vaguely as an elongated semi-circle, distal one-third rather narrow then widening to form a spatulate apex, distal half of valva in dorso-ventral view resembles propodus of a lobster (Decapoda: Nephropidae), but without the dactylus; aedeagus strongly curved upwards in lateral view, almost same length as valva, posterior portion opens latero-ventrally; cornuti absent. FEMALE: Unknown. Etymology. The specific epithet honours T. Zacca's niece, Laura Sophia. To prevent any future ambiguity, the name sophiae is considered to be a Latinised modern female noun in the genitive case.
Distribution and habitat. This species is only known from the type locality in Serra do Divisor National Park (SDNP), Acre, in the extreme west of Brazil, and from   across the border in the neighbouring department of Loreto, in north-eastern Peru. The Brazilian specimens of E. sophiae were collected in forest characterized as submontane dense ombrophilous forest in a landscape of "terra firme" forest with patches of seasonally inundated areas with a predominance of palms in the genus Mauritia Linnaeus f. (Arecaceae) at about 200 m a.s.l. (see Figs 9, 10). A small north-south mountain range surrounds the low area in the western portion of the SDNP, with some hills as high as 600 m a.s.l. Only one hilltop was sampled (visible in Fig. 9) but no specimen of E. sophiae was collected there, although other species of Euptychia were observed. Despite subsequent expeditions in June 2013 and August 2014 to the same area of the SDNP, with five and seven collectors respectively, we were unable to find additional specimens of E. sophiae. Unfortunately, no behavioural notes were taken for this new species, but the three specimens known to us indicate a correlation with periods of average to below average rainfall, an observation similar to that made for E. attenboroughi. Only future sampling will indicate whether these two species show a distinct preference for avoiding the months of highest annual precipitation.

Discussion
Males of E. attenboroughi and E. sophiae, especially E. attenboroughi, exhibit external wing pattern elements and wing morphology that are atypical for the genus, superficially recalling certain species in the genus Chloreuptychia. It was therefore important to support our generic classification for these two new taxa using more objective genetic analyses. Both new species are described in this genus on the basis of ML and Bayesian analyses of E. sophiae, reinforced by the ML analysis performed in Nakahara et al. (2015): all analyses suggest that E. sophiae should be a member of the monophyletic Euptychia clade with high support. Due to many external morphological similarities, including wing pattern and absence of a posterior projection of the tegumen in male genitalia, it is reasonable to consider E. sophiae to be a sister species of E. attenboroughi, which should therefore also be placed in Euptychia. The results indicate that the posterior projection of the tegumen is not shared by all members of the Euptychia clade, and that the absence of this character in these two species could be the result of secondary loss. Morphological evidence to support these two species in the Euptychia clade are: 1) presence of the forewing recurrent vein in the discal cell; 2) absence of basal swelling of the forewing cubital vein; 3) a relatively reduced humeral vein; 4) a developed 8 th tergite and sternite in the male abdomen; 5) presence of the sclerotized region of the 8 th abdominal segment in the female, located at the very basal side of the 8 th abdominal segment; 6) absence of the lateral sclerotisation of the 8 th abdominal segment of the female; 7) origin of the ductus seminalis at the posterior end of the ductus bursae; 8) absence of tibial spurs on the midleg and hindleg. States 1-7 are shared by all Euptychia species examined so far, and are rarely seen in other euptychiine butterflies. State 8 is shared by most species of Euptychia and those species that possess tibial spurs require further investigation regarding their classification. It is also noteworthy that within Euptychiina those Euptychia species whose early stage biology is known have both unique hostplants and larval characters (DeVries 1987, Beccaloni et al. 2008: known hostplants are non-seed plants in Selaginellaceae (Lycopodiophyta) and Neckeraceae (Bryophyta). States 1-8, hostplants, and early stage characters are possible apomorphic characters for Euptychia. Further in-depth study of euptychiine butterflies will hopefully determine this.