Carystus rufoventris Austin & O. Mielke, 2007

Zhang, Jing, Dolibaina, Diego R., Cong, Qian, Shen, Jinhui, Song, Leina, Mielke, Carlos G. C., Casagrande, Mirna M., Mielke, Olaf H. H. & Grishin, Nick V., 2023, Taxonomic notes on Neotropical Hesperiidae (Lepidoptera), Zootaxa 5271 (1), pp. 91-114 : 95-98

Carystus rufoventris Austin & O. Mielke, 2007 is a subspecies of Carystus elvira (Plötz, 1882)

Hesperia elvira Pl ̂tz, 1882 (type locality in South America), currently a valid species of Carystus Hübner, [1819] (type species Papilio jolus Stoll, 1782 ) was described from an unstated number of specimens, one of which was referenced by its number 5277 in MFNB (Pl̂tz, 1882a). Godman (1907) did not find a sufficiently similar specimen among those accessible to him, and commissioned the reproduction of the original drawing t[afel]. 436 by Pl̂tz. The original of this plate is presumed lost, but these Godman’s copies are in BMNH and t. 436 is show in Fig. 3a . It is likely that another copy of the original drawing of H. elvira (rather than a drawing made from the actual specimen) was made and eventually published by Draudt (1921 –1924) ( Fig. 3b ). The two renderings are sufficiently similar and only differ in the amount of detail shown, suggesting that Godman’s watercolor may be more accurate than Draudt’s printed version.

We were not able to find a specimen that looks similar to these illustrations in MFNB, also failing to find a specimen with the number 5277 referenced in the original description of H. elvira . However, we found a record in the collection catalog for this number, reproduced in Fig. 3c . The catalog lists a single specimen of an undetermined (at that time) species from Suriname collected by Hugo. Judging by the handwriting, this catalog entry was made by the Entomology curator in MFNB Carl Heinrich Hopffer who died in 1876, therefore this specimen existed before the description of H. elvira in 1882. It is not clear whether other specimens were known to Pl̂tz at the time of description. The lack of their explicit mention in the description does not necessarily imply the absence of other specimens. Moreover, Pl̂tz listed South America instead of Suriname as the locality for H. elvira , which suggests that there might have been specimens from other localities in South America in addition to 5277 from Suriname.

The only known specimen identified as C. elvira is a female from Suriname in RMNH collection reported by de Jong (1983) ( Fig. 3d ). It differs in wingshape and by the presence of subapical hyaline spots from the H. elvira drawings ( Fig. 3a, b ) and therefore is not the specimen illustrated, which was likely a male. Judging from the round handwritten locality label ( Fig. 3d ) and the pin, the female specimen in RMNH is from the ancient collections and therefore might have been collected as part of the same series with the Surinamese specimen 5277 listed in MFNB catalog. Because both specimens are from Suriname, phenotypically similar to each other, and uniquely different from any other species, we agree with de Jong’s identification of the female as C. elvira .

Carystus rufoventris Austin & O. Mielke, 2007 (type locality in Brazil: Rondônia) was described on the basis of a single specimen, a male ( Fig. 3e ). Austin & Mielke (2007) compared their new species with C. elvira and concluded that they are distinct without inspecting any C. elvira specimens. We sequenced genomes of the holotype male of C. rufoventris and the female of C. elvira reported by de Jong (1983) and suggest that they are conspecific. First, their COI barcodes are 100% identical and they are grouped closely together in the genomic tree ( Fig. 4 ). Second, the two specimens look sufficiently similar in appearance to be a pair ( Fig. 3d, e ). Third, they are from localities that frequently harbor the same species in a complex of allopatric similar-looking taxa ( Suriname and Brazil: Rondônia). However, both specimens differ from the drawings of C. elvira by having subapical hyaline spots, although the spots are reduced in the male. We also see that the male has only one hyaline spot in the forewing discal cell, not two as the female and the drawings, and a male collected in Peru: Madre de Dios has no spots at all ( Fig. 3g ). Moreover, a photographed live male ( Fig. 3f ) lacks subapical spots at least ventrally. Therefore, the hyaline spots and their number may be variable in this species. Austin & Mielke (2007) suggested that the smaller size of C. rufoventris as one of its characters. However, the specimen shown in Fig. 3g that is more similar to C. rufoventris than to C. elvira , is even larger than the original description of C. elvira indicates: nearly 20 mm (the ruler photographed together) vs. 19 mm forewing length in C. elvira .

For these reasons, we propose that C. rufoventris is a subspecies of Carystus elvira (Pl̂tz, 1882): Carystus elvira rufoventris Austin & O. Mielke, 2007 , stat. nov. Being conservative, we refrain from synonymizing the two, because of possible consistent wing pattern differences between populations from the Guianas and Rondônia/ Madre de Dios, as described by Austin & Mielke, (2007). We hope to test this taxonomic hypothesis further when additional specimens become available, in particular, a male from Suriname without apical spots and with two spots in forewing discal cell that would be suitable for C. elvira neotype designation.

Figures

FIGURE 3. Carystus elvira (Plötz, 1882). a. Godman’s copy of unpublished Pl̂tz’s illustration t. 436 [BMNH]. b. Draudt’s published illustration, dorsal and ventral images rearranged to match Godman’s copy (Draudt, 1921–1924). c. Excerpt from the collection catalog for no. 5277 [MFNB]. d. A specimen identified as C. elvira ♀ from Suriname and reported by de Jong (1983) with its label [RMNH]. e. Holotype of Carystus rufoventris Austin & O. Mielke, 2007 ♁ from Brazil: Rondônia [DZUP]. iNaturalist observations from Peru: Madre de Dios, Tambopata: f. 94562583, 27-Sep-2019 © David Geale and g. 94562583 6- Apr-2018 © Joseph See, rotated and brightened. Dorsal and ventral images are on the left and right from each letter, respectively. All images except f are to scale. Photograph a (by N.V.G.) is © of the Trustees of the Natural History Museum London and is made available under Creative Commons License 4.0 (https:// creativecommons.org/licenses/by/4.0/), e by Diego Dolibaina and f, g CC BY-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

FIGURE 4. Nuclear genome tree of selected Carystina and Hesperiina. The tree is constructed from protein-coding regions. Statistical support values are shown by nodes. For each specimen, its name adopted in this work is given first, and a previously used name is listed in square brackets (if different), supplemented with the DNA sample number, type status, general locality, year of collection (when known), and, for primary type specimens, their repository. Synonyms are given in parentheses preceded by “=”, and in addition by “‡” for unavailable names. The type status refers to this synonym, if the synonym name is provided. Green vertical bar cuts the tree into genera, except Vertica, which is characterized by lower evolutionary rate (leaves are closer to the root compared to most others). Yellow shading denotes genera and its width (horizontal dimension) is proportional to the genetic differentiation within each genus. Taxa discussed in the text are shown in different colors: Carystus elvira (orange), Orthos minka, stat. nov. (green), Orthos orthos (purple), Orthos hyalinus, stat. rest. (olive), Oxynthes (blue, subgenus Oeonus labeled in dark blue), Lindra neroides, comb. nov. (red, neotype labeled in margenta), Lindra brasus (cyan), and Decinea (gray).