Revision of the genus Callipia Guenée, 1858 (Lepidoptera, Geometridae), with the description of 15 new taxa

Abstract. The vividly coloured Neotropical genus Callipia Guenée (1858) (Lepidoptera Linnaeus, 1758, Geometridae (Leach, 1815), Larentiinae (Leach, 1815), Stamnodini Forbes, 1948) is revised and separated into four species groups, according to a provisional phylogeny based on Cytochrome Oxidase I (COI) gene data and morphology. Fourteen new species are described using COI data and morphology: a) in the balteata group: C. fi edleri sp. nov., C. jakobi sp. nov., C. lamasi sp. nov.; b) in the vicinaria group: C. hausmanni sp. nov., C. walterfriedlii sp. nov.; c) in the parrhasiata group: C. augustae sp. nov., C. jonai sp. nov., C. karsholti sp. nov., C. levequei sp. nov., C. milleri sp. nov., C. sihvoneni sp. nov., C. wojtusiaki sp. nov. and d) in the constantinaria group: C. hiltae sp. nov., C. rougeriei sp. nov. One new subspecies is described: C. wojtusiaki septentrionalis subsp. nov. Two species are revived from synonymy: C. intermedia Dognin, 1914 stat. rev. and C. occulta Warren, 1904 stat. rev. The taxon hamaria Sperry, 1951 is transferred from being a junior synonym of C. constantinaria Oberthür, 1881 to being a junior synonym of C. occulta stat. rev. The taxon admirabilis Warren, 1904 is confi rmed as being a junior synonym of C. paradisea Thierry-Mieg, 1904. The taxon languescens Warren, 1904 is confi rmed as being a junior synonym of C. rosetta, Thierry-Mieg, 1904 and the taxon confl uens Warren, 1905 is confi rmed as being a junior synonym of C. balteata Warren, 1905. The status of the remaining species is not changed: C. aurata Warren, 1904, C. brenemanae Sperry, 1951, C. parrhasiata Guenée, 1858, C. fl agrans Warren, 1904, C. fulvida Warren, 1907 and C. vicinaria Dognin. All here recognised 26 species are illustrated and the available molecular genetic information of 25 species, including Barcode Index Numbers (BINs) for most of the taxa is provided. The almost threefold increase from 10 to 26 valid species shows that species richness of tropical moths is strongly underestimated even in relatively conspicuous taxa. Callipia occurs from medium to high elevations in wet parts of the tropical and subtropical Andes from Colombia to northern Argentina. The early stages and host plants are still unknown.


Introduction
described Callipia parrhasiata Guenée, 1858 in a new monotypic genus and illustrated this remarkable, vividly pink coloured species. The type was eventually deposited in the Natural History European Journal of Taxonomy 404: 1-54 ISSN 2118ISSN -9773 https://doi.org/10.5852/ejt.2018 www.europeanjournaloftaxonomy.eu 2018 · Brehm G. This work is licensed under a Creative Commons Attribution 3.0 License.

M o n o g r a p h
urn:lsid:zoobank.org:pub:EFD82C30-DBD4-40D0-8FE5-FAE10B7E560D Museum (London). The locality is (according to its label) given as "Bengale" -obviously an error, since all confi rmed records of species of Callipia are from the wet tropical Andes. In this revision, all specimens assigned to C. parrhasiata have been collected in a very limited geographic range in the eastern Andes of Ecuador. Only 23 years later the second species of the genus -C. constantinaria (Oberthür, 1881) -was described (Oberthür 1881). Twelve further taxa were described between 1904 and 1914 by Thierry-Mieg (1904), Warren (1904Warren ( , 1905Warren ( , 1907 and Dognin (1913Dognin ( , 1914, and two by Sperry (1951), refl ecting a typical temporal pattern and peak of taxonomic activity around 1900 also found in many other Neotropical geometrid taxa (Gaston et al. 1995;Brehm et al. 2011;Brehm 2015). In their catalogue of geometrid moths, Parsons et al. (1999) treated 15 of these 16 taxa; only the taxon C. confl uens Warren, 1907 is not listed in the catalogue, it was originally described as a form of Callipia balteata Warren, 1905. Parsons et al. (1999 recognised ten valid taxa and fi ve junior synonyms. Whenever possible, this paper aims to integrate results from external morphology, genitalia morphology and molecular genetics, i.e., from "DNA barcoding" of the Cytochrome Oxidase I (COI) gene. COI barcoding allowed a reliable match between males and females in several cases and eased the sorting of diffi cult species complexes considerably. In the large majority of cases, molecular and morphological results were congruent, but in some cases, DNA barcoding produced splits in taxa that were morphologically otherwise not distinguishable. I generally followed DNA barcoding results and applied a 2% threshold for the separation of species with the exception of the C. milleri -wojtusiaki complex. Barcode Index Numbers (BINs) were assigned to each species when available, a system that has been established by the Barcode of Life Data Systems (www.boldsystems.org) (Ratnasingham & Hebert 2013). Twenty-one different BINs are assigned to species of Callipia in this paper, thus covering 75% of the valid species. In one species complex, I assigned two different BINs not to different species, but to geographically separated subspecies. The BIN system allows a quick and reliable assignment of DNA-barcoded specimens. This is particularly valuable in cryptic and diffi cult species complexes in poorly known faunas (Brehm et al. 2016). Assignment of DNA barcodes to described species was performed by careful comparison of type material with freshly collected material (Brehm 2015;Brehm et al. 2016), and all relevant types and vouchers are illustrated in this paper. DNA-barcoded material also includes six old specimens collected between 30 and 100 years ago in a couple of cases where no fresh material was available (Strutzenberger et al. 2012).
Most historical descriptions of Callipia consist of a few text lines only that describe external characters and the colour patterns of the moths. Today, the study of external and internal morphological characters in combination with genetic characters and colour illustrations provides far better opportunities for taxonomy than ever. However, the wealth of options to describe a new taxon also means a burden for taxonomic work, because the description of species could include the study and illustration of dozens, if not hundreds of characters. If these are met at high standards, descriptions will not only fare better than in the past, but unfortunately, also consume much more time. Given the severe and rapid loss of biodiversity and the need of sound taxonomic information, especially in tropical regions, I here choose a treatment that focuses on diagnostic characters, high resolution illustrations with standardized scale bars and standard gene sequences. On the other hand, I have decided to waive extensive and detailed descriptions of characters that are diagnostically not valuable and / or plesiomorphic (Forum Herbulot 2014;Brehm 2015).

Material and methods
Moths were pinned and dissected following established techniques (e.g., Lafontaine 2004;Hünefeld et al. 2013). Genitalia slides were embedded in Euparal, stained with Chlorazol Black and digitised at least 10 weeks after production using an Olympus dotSlide system with 10 × magnifi cation. Adult moths were photographed on a neutral grey background in raw format using a 60 mm Nikkor macro lens mounted on a Nikon D700 camera. Photos were adjusted and colour plates were mounted using Photoshop and InDesign software (Adobe Systems, San José, USA).
Sequencing of the barcode fragment of the COI gene was carried out at the Canadian Center for DNA barcoding in Guelph, Ontario. Barcode sequences were com pared by nearest neighbour analyses (Kimura 2 parameter), as implemented on the Barcode of Life Data Systems website (Ratnasingham & Hebert 2007). The resulting tree represents preliminary hypotheses of taxa groupings and can form the basis of fu ture phylogenetic work (Brehm 2015) (Fig. 1). For eight old specimens, a six fragments approach was performed that was originally developed for Sphingidae (Lees et al. 2011;Rougerie et al. 2012), but also successfully applied to Geometridae (Strutzenberger et al. 2012). Distribution maps display only such specimens with original coordinate data on their labels, or specimens labelled with localities that could be found with reasonable certainty, e.g., using Google Earth software. Localities were checked using Lamas (1976), and in correspondence with Gerardo Lamas (personal communication, Peru) and Sebastian Herzog (personal communication, Bolivia). In order to keep an overview and to allow easy tracking of specimens for later analyses, every specimen that was investigated and did not already possess a unique museum identifi cation number was labelled with an individual number, ranging from "Callipia-gb 0001" to "Callipia-gb 0432" (abbreviated as C-0001 to C-0432). Individual numbers were not provided for some of the existing type specimens. Genitalia slides were produced for ca 130 specimens. They are labelled for example "Genitalia slide

Results
In this paper, two taxa are revived from synonomy. With the addition of 15 new taxa, Callipia now comprises 26 valid species, one subspecies and 4 junior synonyms.

General short description
With its large size, typically shaped forewings and colourful wing patterns, Callipia forms a distinct group within the Neotropical Larentiinae that made assignment of newly described species to this genus easy. In contrast to most other Neotropical larentiin genera, all described species were originally assigned to the genus they are still assigned to. During visits to various museums holding Neotropical material, I have not come across species that need to be transferred to Callipia, either.
Compared to most other larentiines, Callipia are very large moths, with wing lengths ranging from ca 24 mm up to ca 32 mm. As a comparison, the three most species rich Neotropical genera Eupithecia Curtis, 1825, Eois Hübner, 1818 and Psaliodes Guenée, 1858 only reach wing lengths up to ca 13 to 14 mm.

Wings
Callipia are exceptionally colourful geometrids, most species possess wing patterns with large rosypink, yellow or red blotches on the wings. The function of these conspicuous colours is unknown, because the moths always rest with vertically folded wings, and their colourful patterns remain largely unseen, see Figs 135 and 137. They have this in common with related genera such as Heterusia Hübner, 1831, Erateina Doubleday, 1848 and Hagnagora Druce, 1885 (Brehm 2015). Many butterfl ies and other insects with striking colours during fl ight are well camoufl aged when they rest. It is possible that that the conspicuous patterns are shown when the moths are disturbed, but this has not been observed so far. In other colourful Neotropical geometrid genera such as Heterusia and Erateina diurnal activity or activity at day and night was recorded (Brehm & Sullivan 2005). However, diurnal behaviour has so far to my knowledge not been documented in Callipia, but it is possible that the moths show their colours when disturbed, fl y away and hide again well camoufl aged.
Callipia can also be distinguished from other geometrids by their typically shaped forewings which are "higher than broad", i.e., the outer margin of the fore wing stretches longer than veins 1A+2A. While the fore wing pattern can vary considerably between species, the underside of the hindwings is rather uniform among most species of Callipia: with a dark brown background colour, they are ochreously striated in such a way that it resembles an irregular orb spider web. Similar striae also occur in the apical region of the underside of the fore wing -the only exception being C. rougeriei sp. nov. in which striae are completely reduced. Most species possess a cream white postmedial line on the underside of the hindwing, but it is largely reduced in the constantinaria group and widened to a large white blotch in C. paradisea Thierry-Mieg, 1904.

Colour
The ground colour in all species is dark to medium brown.

Head (illustrated by Viidalepp 2011), thorax and abdomen
Appear not to possess particular characters that can be interpreted as an apomorphy of the genus. Head: proboscis, compound eyes and labial palpi are well developed, and the antennae are fi liform in both sexes. Thorax: the tegulae are often light ochreous, contrasting with the dark brown scutum, the legs are usually BREHM G., Revision of Callipia brown. The tibia spur formula is 0-2-4 as found in most other Geometridae (Scoble 1992). Abdomen: tympanal organs are well developed. The colour of the abdomen in most species consists dorsally (and in some species also laterally) of alternate ochreous and brown rings in each segment, most prominently in the balteata group, but absent in several species, particularly in the constantinaria group.

Male genitalia
The valvae are broad and rounded, hemitranstilla (sensu Viidalepp 2011) are present, the juxta is shield shaped (Viidalepp 2011) and many species possess a spine-like process on the ventral margin of the valvae. The uncus is slender with a sharp or round apex, the sacculus is round. The aedeagus is enclosed in a sclerotized tube -the manica sensu Viidalepp (2011). The shape of the manica varies considerably between species.

Female genitalia
A signum is not visible on the corpus bursae, the lamella antevaginalis is small, sclerotised, often lunular; the ductus bursae is short, narrow and sclerotised; the ductus bursae is without sclerotisations, its shape varying from narrow and elongated to more roundish.

Preimaginal stages and host plants
Preimaginal stages and host plants are unknown. Figure 1 shows a summary tree with each DNA-barcoded species represented by one specimen per species. It visualizes similarities and differences in the COI gene between the different taxa and it was instrumental in differentiating four clades identifi ed within Callipia. One species, paradisea, could not be reliably assigned to one of these clades. The four clades are subsequently treated as balteata group, vicinaria group, constantinaria group and parrhasiata group, named after the oldest available name in each of the clades. Table 1 provides an overview of all taxa, including available molecular genetic information and currently known distribution. Detailed information on all examined species can be found in Supplementary File 2. Further details for 150 specimens, including voucher images and molecular data, are publicly available in BOLD and can be accessed under https://doi.org/10.5883/DS-CALL2017

Distribution, habitats and behaviour
Specimens of Callipia were collected in Ecuador using blacklight and blacklight-blue tubes (Brehm & Axmacher 2005), and in Peru with a newly developed UV LED lamp (Brehm 2017). All specimens illustrated in Figs 131-139 were collected in this way and photographed either on the surface of the 'light tower', or in the vegetation next to the light source. Specimens of Callipia were often shy and escaped when they were disturbed. Figures 131 and 133, therefore, show benumbed specimens. Other investigated adult specimens have usually been collected with lamps as well, but in most cases this is not explicitly documented on the labels. Figure 2 shows the distribution data of those specimens of Callipia with reliable locality information. BREHM G., Revision of Callipia collections have not been sampled in northern Colombia and in Venezuela either, but this might refl ect a sampling artifact. Many regions are still undersampled, such as large ranges of the Andes in northern and central Peru and in southern Bolivia. More sampling will certainly improve the knowledge of the distribution of Callipia and will inevitably reveal further undiscovered taxa. Callipia appears to be restricted to the Andes and is absent from Central America. It is a truly Andean taxon, similarly as, e.g., the Acrotomodes clota group (Ennominae Duponchel, 1845) (Brehm 2005). The most widely distributed clade is the parrhasiata group that ranges from Colombia to northern Argentina, with the highest species richness found in Peru. Similarly, widely distributed is the constantinaria group, but it has a smaller species richness. The other groups are geographically more restricted (Fig. 2).
Although there are records of Callipia from lowland tropical forests (e.g., 200 m in Madre de Dios, Peru and 450 m in Tena, Napo, Ecuador), these cases are exceptional and would need to be confi rmedmislabelling appears very likely. Most tropical species of Callipia are not found below 1500 m a.s.l. Some, such as species belonging to the balteata group, have only been recorded at elevations around or higher than 3000 m a.s.l.

The balteata group
The four known members of the balteata group possess two distinctive broad, deep rosy bands on the forewing; a pattern that is not found in other species of Callipia. The male valvae are round-shaped and do not possess the spine-like processes on the ventral margin that are present in all other groups of Callipia. The uncus is pointed. The aedeagi are bulkier than in other Callipia, the vesica is long and has a small series of cornuti. The species are restricted to high elevations in the wet eastern Peruvian and Bolivian Andes (observed: 1982-3450 m). All are represented by a single or very few specimens in museum collections only. Surprisingly, the balteata group consists of relatively many species, with small, but clear morphological differences besides supporting molecular information. One species was collected in Huánuco Province, Central Peru (C. lamasi sp. nov.), ca 850 km separated from the collection site of C. fi edleri sp. nov. (Peru, Cusco Province), separated by ca 250 km from the collection site of C. balteata (Peru, Puno Province) and separated by another ca 250 km from the collection site of C. jakobi sp. nov. (western Bolivia). The balteata group represents an example of an Andean high elevation clade with a large degree of local endemism. The females are still unknown. Callipia balteata Warren, 1905 Figs 3-5, 9-10 Callipia balteata confl uens Warren, 1907: 243. (Fig. 5). Syn. nov.
No assigned BIN, but assigned 407 bp fragment.

Description
As illustrated. Female unknown.

Distribution
South eastern Andes of Peru, 2000-3000 m.

Diagnosis
Similar to C. balteata, but the two bands on the forewing further apart, the cream white band on the hindwing underside is very narrow. Vesica of the aedeagus is shorter than in C. balteata. COI-barcode: the observed distance to the genetically most similar species (C. balteata) is 3.3%.

Etymology
The species is named in honour of Konrad Fiedler, Vienna, Austria.

Type material
Holotype (

Description
As illustrated. Only a single male is known.

Distribution
Only known from a single high-elevation locality in the south eastern Andes of Peru, 3200-3450 m, ca 250 km north-west of the collection sites of C. balteata.

Diagnosis
The largest member of the balteata group, patterns strikingly contrasting, colour of the bands on the forewing upperside distinctly pink, and not red-brown as in the related species. The uncus is broader than in the other species. COI-barcode: the observed distance to the genetically most similar species (C. jakobi sp. nov.) is 6.0%.

Etymology
The species is named in honour of Gerardo Lamas (MUSM), Lima, Peru.

Type material
Holotype (

Description
As illustrated. Only a single male is known.

Distribution
Only known from a single locality in the Central Eastern Andes of Peru, 3400 m, ca 850 km north west of the collection site of C. fi edleri sp. nov.

Diagnosis
The smallest species of the balteata group, relatively pale and less contrasting patterns of the wings than in the other species. The cornuti of the vesica are much longer than in the other species of the balteata group. COI-barcode: the observed distance to the genetically most similar species (C. balteata) is 4.3%.

Etymology
The species is named in honour of my son Jakob Brehm, Jena, Germany.

Type material
Holotype (

Description
As illustrated. Only a single male is known.

Distribution
Eastern Andes of W Bolivia, 3300 m a.s.l. Distance to the collection site of C. balteata is ca 250 km.

Without assignment to group
Callipia paradisea is known from a few specimens only. Wing patterns and structure of male genitalia not easily fi tting into any of the other groups, but vicinaria group is possibly most closely related. Since no molecular data are available either, C. paradisea is provisionally treated separately.
No assigned BIN.

Diagnosis
Callipia paradisea is unmistakeable due to its unique mixed colour pattern of white, dark brown and extended rosy elements. The taxon C. admirabilis does not show any signifi cant differences from the type specimen of C. paradisea and, therefore, remains in synonymy with C. paradisea. The other four species of the vicinaria group are smaller than C. paradisea. Male genitalia: a spine-like process on the ventral margin of the valvae is present, but short and blunt. The aedeagus is broader than in most other species, the manica is slightly bent, and the vesica does not possess cornuti.

Description
As illustrated. The female is unknown.

Distribution
Eastern Andes of central and south eastern Peru, 1800-3100 m.

Remarks
Only a few specimens of C. paradisea exist in collections. No recently collected material has been available for DNA barcoding and it failed in an old specimen.

The vicinaria group
This group comprises three species. The moths are on average smaller than members of the other groups. Wings are composed of cream white, ochreous, dark brown and rosé elements. The male valvae have more pronounced spine-like processes on the ventral margin than species of the other groups. The vesicae do not possess cornuti. All species show a pronounced sexual dimorphism that is possibly an apomorphy of the group. The females have a dark grey ground colour with ochreous patterns; the female of C. hausmanni sp. nov. more resembles those of the parrhasiata group. Females of this group are therefore illustrated on a separate plate. Species are distributed from Colombia to central Peru but not further in the south (Fig. 2b). Dognin, 1913 Figs 20-23, 30-31, 36, 39 Assigned BIN: BOLD:ACP6822.

Diagnosis
Callipia vicinaria is one of the smallest known species of Callipia. It is closely related to C. walterfriedlii sp. nov. The spine-like processes on the ventral margin of the valvae are relatively short and the uncus is narrower than in the other closely related species. Other diagnostic characters are discussed in those species. COI-barcode: the minimum observed distance to the genetically most similar species (C. walterfriedlii sp. nov.) is 3.3%.

Description
As illustrated. The extent of dark brown markings on the wings is more extended and pronounced in the Ecuadorian specimens (Fig. 23) than in the males from Colombia (Figs 20-22).

Remarks
The female was unillustrated until now. The uncus tends to be broader than in C. vicinaria. COI-barcode: the minimum observed distance to the genetically most similar species (C. vicinaria) is 3.3%.

Etymology
The species is named in honour of Walter Friedli, Schwarzenburg, Switzerland.

Type material
Holotype (

Remarks
A living female is illustrated in Fig. 133, together with the habitat (Fig. 134).

Diagnosis
While males of C. hausmanni sp. nov. look rather similar as the other three species of the complex, the females are easily distinguished: only females of C. hausmanni sp. nov. possess broad cream white margins and a large pale rosy base of the forewing. The males have purely cream white fringes, a relatively intensive rosy basal area of the forewings and a more contrasting pattern of the hindwings than the related species. The male genitalia are characterised by a rectangular shape of the tegumen European Journal of Taxonomy 404: 1-54 (2018) (rounded in the other species) and a narrower uncus. The spine-like processes of the valvae are long and slightly curved. COI-barcode: the minimum observed distance to the genetically most similar species (C. vicinaria) is 5.0%.

Etymology
The species is named in honour of Axel Hausmann (ZSM), Munich, a leading Geometridae expert in acknowledgment of his support for this study.

Type material
Holotype (

Description
As illustrated.

Distribution
Only known from high elevations in a small area of the Eastern Andes of central Peru, 3000-3400 m.

The constantinaria group
The six known members of the constantinaria group are easily recognised by their wing patterns: brown coloured with large deep yellow blotches on the forewing. Male valves are overall similar as in the parrhasiata group; the aedeagus is straight and slender without cornuti. The male genitalia within the group are rather similar and appear to be of little use for the diagnosis of species. C. occulta stat. rev. and C. hiltae sp. nov. form a cryptic species complex that can reliably only be distinguished by DNA barcoding. Species are distributed from Colombia to northern Argentina (Fig. 2c). Oberthür, 1881 Figs 43-44, 59 Assigned BIN: BOLD:AAD6679.

Diagnosis
The yellow blotch on the forewing is the smallest among all species of the constantinaria group. The male genitalia of the holotype are lost, but male genitalia are very similar among all species in the group. COI-barcode: the minimum observed distance to the genetically most similar species (C. brenemanae) is 3.9%. The most reliable identifi cation method appears to be through DNA barcoding.

Description
As illustrated. The female is unknown.

Distribution
Eastern Andes of central and south eastern Peru, 3000 m.

Diagnosis
Unmistakable. Similar as other constantinaria group members, but striae on the underside of the wings completely reduced, and instead with dark brown ground colour (except for the yellow blotch). It has a light ochreous costa on forewing underside and a light ochreous marginal band on the underside of both wings. COI-barcode: the minimum observed distance to the genetically most similar species (C. brenemanae) is 4.7%.

Etymology
The species is named in honour of Rodolphe Rougerie, MNHN, Paris, France.

Type material
Holotype (

Description
As illustrated.

Diagnosis
Callipia occulta stat. rev. was put into synonymy with C. constantinaria by Parsons et al. (1999), following Prout's catalogue in the British Museum (Natural History). However, the yellow blotch on the forewing is much more restricted in C. constantinaria than in C. occulta stat. rev., moreover, C. constantinaria appears to be a bit smaller than C. occulta stat. rev. The comparison of the genitalia is not possible since the fi nal segments of the abdomen are missing in the holotype of C. constantinaria. Barcoding revealed two different BINs of C. constantinaria-like species from Peru. C. occulta stat. rev. and C. hiltae sp. nov. form a cryptic species complex, see C. hiltae sp. nov. COI-barcode: the minimum observed distance to the genetically most similar species (C. aurata) is 3.4%. Other type material (Fig. 48

Description
As illustrated.

Distribution
Eastern Andes of Ecuador and N and Central Peru, 1500-2800 m.

Remarks
The taxon C. hamaria, originally described as a form of C. constantinaria by Sperry (1951), actually very closely resembles C. occulta stat. rev.; both are from Peru. While Parsons et al.(1999) treated C. hamaria as a junior synonym of C. constantinaria, I treat it here as junior synonym of C. occulta stat. rev.

Diagnosis
Costa in C. hiltae sp. nov. largely without striae; on the contrary the holotype of C. occulta stat. rev. shows such striae as well as most of the 10 successfully DNA-barcoded specimens assigned to C. occulta stat. rev. The extent of the yellow blotch on the forewing appears to be rather variable, and the male genitalia do not offer reliable diagnostic characters either. So far, only two specimens can reliably be assigned to C. hiltae sp. nov., but it is possible that the species has a broader distribution. COI-barcode: the minimum observed distance to the genetically most similar species (C. brenemanae) is 5.0%, and it is 5.7% to C. occulta stat. rev.

Etymology
The species is named in honour of Nadine Hilt, Bayreuth, Germany who carried out a moth diversity study along a disturbance gradient in southern Ecuador.

Description
As illustrated. The female is unknown.

Diagnosis
The species with the largest yellow blotch. Only known from Colombia. COI-barcode: the minimum observed distance to the genetically most similar species (C. occulta stat. rev.) is 3.4%.

Type material
Holotype (

Description
As illustrated. The female is unknown.

Diagnosis
The base of forewing veins CuA 2 , CuA 1 and often M 3 are covered with dark brown scales in brenemanae, particularly at the wing base. Only specimens of C. occulta stat. rev. show a similar pattern, but it is weaker and only true for vein CuA 2 . COI-barcode: the minimum observed distance to the genetically most similar species (C. constantinaria) is 3.9%.

Type material
Holotype (

Description
As illustrated. The female is unknown.

Distribution
Eastern Andes of central to south eastern Peru and central Bolivia, 700-2800 m. 404: 1-54 (2018) The parrhasiata group This group comprises most species; I acknowledge here 12 taxa at species level with one of them comprising two geographically separated subspecies (C. wojtusiaki sp. nov.). Most species have conspicuous rosy or pink wing bases, but in four species (distributed from central Peru to northern Argentina) the base colour is more intensive and ranges from orange to red. The rosy / pink species are distributed from Colombia to Bolivia. Unlike in the vicinaria group, there is no pronounced sexual dimorphism, but females are considerably larger than males. Unfortunately, the male genitalia only offer very limited diagnostic characters and DNA barcoding in combination with distributional data will sometimes be the only reliable way of distinguishing closely related species. Guenée, 1858 Figs 67-69, 75, 124 Assigned BIN: BOLD:AAI3927.

Diagnosis
Probably has the most intense pink colouration of all members of the group, and the most extended dark brown area. Callipia intermedia stat. rev. on forewing upperside with a marginal ochreous blotch, this is only present on the underside in parrhasiata. The male vesica has a series of relatively long cornuti. The cream white margin of the underside of the hindwing is very narrow and considerably less pronounced than in related species. COI-barcode: the minimum observed distance to the genetically most similar species (C. milleri sp. nov. and C. wojtusiaki septentrionalis subsp. nov.) is 4.1%, respectively.

Diagnosis
Ground colour a little bit less intense, less pink and more reddish than in parrhasiata. The cream white margin of the hindwing underside is broader than in parrhasiata. Light ochreous costal area of forewing Limbani], and Agualani (USNM). These six specimens are not conspecifi c; the two Ecuadorian specimens (C-0019; C-0357) are assigned to C. wojtusiaki sp. nov. (see there). Specimens from Limbani were indicated by Dognin (1914) as 'type', and therefore a specimen from this locality is selected as lectotype (C-0327 with GS-296, Figs 71, 76; the other C-0329). Two specimens are from from Agualani (C-0328 with GS-283; C-0330).

Distribution
Eastern Andes of south eastern Peru, 1100-3000 m.

Remarks
The species was described as subspecies of C. parrhasiata, but is clearly an independant species. The six syntypes belong to two different taxa (see above).

Diagnosis
Most similar to C. intermedia stat. rev. (sympatric distribution), but ground colour light rosy with an extended light ochreous costal area on the forewing upperside. Similarly as in C. intermedia stat. rev., but different from C. parrhasiata, male aedeagus with a few small cornuti. COI-barcode: the minimum observed distance to the genetically most similar species (C. intermedia stat. rev.) is 2.8%.

Etymology
The species is named in honour of Ole Karsholt (ZMUC), Copenhagen, Denmark, who collected the type specimens in Peru.  Figs 74, 80). Because of the different sampling location in central Peru, the assignment of this specimens to karsholti sp. nov. is provisional until more material can be examined.

Description
As illustrated.

Diagnosis
Callipia rosetta is usually a rather small species. Some large specimens are provisionally included in C. rosetta (same BIN), but might turn out as different species if more data on the life history of the taxa become available. The spine-like process on the ventral margin of the valvae is well developed and larger than in augustae sp. nov. COI-barcode: the minimum observed distance to the genetically most similar (but not similar looking) species (C. fulvida) is 4.4%.

Type material
Originally described with two male as syntypes from Peru, [Pasco], Huancabamba, 1903 (USNM). One of these males is designated as lectotype (C-0290

Distribution
Eastern Cordillera of Ecuador and Peru, 2100-3000 m.

Remarks
A living male is illustrated in Fig. 131 together with its habitat in Ecuador (Fig. 132).

Diagnosis
Callipia augustae sp. nov. is unique in having spotted wing pattern pattern elements, i.e., the fi ne light ochreous striae that are found in most other species of Callipia are mostly widened to small spots. The spine-like processes on the ventral margin of the valvae are very small. COI-barcode: the minimum observed distance to the genetically most similar (but not similar looking) species (C. fl agrans) is 2.6%.

Etymology
The species is named in honour of Augusta Albrecht, Friedrichshafen, Germany.

Type material
Holotype (

Description
Male and female as illustrated.

Distribution
Eastern Andes of southern Ecuador, Peru and Bolivia, 1900-3100 m.

Remarks
Two living males are illustrated in Figs 135 and 137. The conspicuous colours are only displayed when the moths are active. The habitat is shown in Fig. 136. It is remarkable that one of the apparently most common species of Callipia was undescribed until now.

Diagnosis
Callipia sihvoneni sp. nov. is a medium sized species of Callipia that was identifi ed as C. parrhasiata in many museum collections. The costal area of the forewing is conspicuously striated / spotted and the apical dark brown area is usually not divided as in C. wojtusiaki sp. nov. and C. milleri sp. nov. COIbarcode: the minimum observed distance to the genetically most similar species (C. milleri sp. nov.) is 4.1%. Callipia sihvoneni sp. nov. can most reliably be distinguished from the closely related C. wojtusiaki septentrionalis subsp. nov. and C. milleri sp. nov. by DNA-barcoding.

Etymology
The species is named in honour of Pasi Sihvonen, Veikkola, Finland.

Description
Male as illustrated. Female unknown.

Diagnosis
Among the four species with a yellow-red ground colour (in comparison with the deep rosy species around C. parrhasiata), C. fulvida tends to have the most yellowish colour, but this is not a fully reliable diagnostic feature. The extent of yellow (vs dark brown) is most extended in C. fulvida among the four species, and the yellow blotch always reaches the costa at about two thirds from the base, producing an isolated brown spot. Although this is also seen in some individuals of the other species, the combination of these characters in combination with the geographical distribution will usually allow to identify C. fulvida without DNA barcoding or dissection. COI-barcode: the minimum observed distance to the genetically most similar (but not similar looking) species (C. rosetta) is 4.4%.

Description
Male and female as illustrated.

Distribution
Eastern Andes of south eastern Peru and Bolivia, 2700-3000 m.

Diagnosis
The extent of dark brown colour is larger than in C. fulvida, see diagnosis in this species. Callipia levequei sp. nov. has a considerably more intense ground colour than fl agrans. Callipia fl agrans and C. jonai sp. nov. are very similar and can most reliably be distinguished by DNA-barcoding. The ground colour in C. fl agrans is more yellow than in C. jonai sp. nov., and the species appear to be geographically separated (C. fl agrans in Peru, C. jonai sp. nov. in Bolivia). COI-barcode: the minimum observed distance to the genetically most similar (but not similar looking) species (C. augustae sp. nov.) is 2.6%.

Diagnosis
Specimens of C. jonai sp. nov. were frequently assigned to fl agrans in museum collections, but when seen in larger series, C. jonai sp. nov. has a deeper orange-red than fl agrans which is more yellowish. See also diagnosis in C. fl agrans. COI-barcode: the minimum observed distance to the genetically most similar (but not similar looking) species (C. intermedia stat. rev.) is 2.7%.

Etymology
The species is named in honour of my son Jona Brehm, Jena, Germany.

Distribution
Eastern Andes of northern Argentina and Bolivia, 1700 m.

Discussion
Remarkably, the number of species increased substantially even in a rather conspicuous moth taxonfrom 10 to now 26 valid Callipia species. This shows that species richness of tropical moths can be strongly underestimated even in taxa that were regularly sampled and usually not overlooked. The situation in less conspicuous taxa is much worse: higher proportions of species still remain undescribed, and many species of small taxa (such as Eupithecia Curtis, 1825) still need to be sampled, because they are not represented in any museum collection. For example, Brehm et al. (2011) estimated the percentage of undescribed species in the Neotropical Larentiinae genus Eois Hübner, 1818 at 85%. This revision is far from being perfect, but in my view the results represent a signifi cant progress in the taxonomy of Callipia and a small step in the systematics of Geometridae. When I started working for this paper more than twelve years ago, I focused on morphological characters -and got stuck in a dead end. For example, specimens around C. parrhasiata appeared highly variable, but did they all belong to one single, widespread species? Unfortunately, male genitalia of Callipia did not advance the research much further. They mirrored the obvious differences between distantly related species (such as C. balteata and C. fulvida), but they were of little value in the discrimination of closely related taxa. DNA barcoding then boosted the process again. The new data provided many clear clusters, and this in turn allowed re-sorting of the material, and then patterns became visible that were much more convincing than the lumping before. One might argue that certain species splits are not suffi ciently justifi ed and, in some cases, only a single specimen was available for description. However, more data and new material can always shed more light into these matters in the future. If one, two or three of the new taxa will eventually turn out to be synonyms, then they will be -but in the meantime, there is now a framework available to build upon, and hypotheses that can be tested. It was important to me that as many specimens as possible (and all relevant types) are well illustrated in this paper, and that all molecular information and sampling data are well accessible and presented in the most transparent way, e.g., by providing a Google Earth kml fi le that allows to explore the distribution data in detail. Doubtlessly more species of Callipia will be discovered in the Andes in the future and, hopefully, much more biological knowledge will be gathered about these beautiful moths and their habitats will be valued and conserved.
Forschung and by funding in support of the International Barcode of Life Project from Genome Canada through the Ontario Genomics Institute. Data analysis was aided by the Barcode of Life Data System, which is supported by the Ontario Ministry of Research and Innovation. Visits to the NHM in 2011 and 2017 were founded by grants from the SYNTHESYS programme (GB TAF1048 and 6817). Support from DFG grants is acknowledged (Fi 547/10-1 and 10-2, FOR 816, FOR 402, Br 2280/6-1).