Telegonus, Hubner, 1819, Hubner, 1819

Zhang, Jing, Cong, Qian, Shen, Jinhui, Song, Leina & Grishin, Nick V., 2025, Advancing butterfly systematics through genomic analysis, The Taxonomic Report of the International Lepidoptera Survey 12 (5), pp. 1-201 : 119-122

A preliminary taxonomic list of Telegonus (Rhabdoides Scudder, 1889) species from the clade analyzed in this work

Phylogenetic trees constructed from protein-coding regions in genomic sequences reveal that the species of the subgenus Rhabdoides Scudder, 1889 (type species Eudamus cellus Boisduval & Le Conte, [1837] ) in the genus Telegonus Hübner, [1819] (type species Papilio talus Cramer, 1777 ) from the clade (Li et al. 2019) that we analyzed in this study partition into six major subclades that we define as species groups ( Fig. 89 ). These are all Rhabdoides , excluding the clades with Telegonus anaphus (Cramer, 1777) and Telegonus cellus (Boisduval & Le Conte, [1837]) : Rhabdoides species not shown in the list below belong to the outgroup and should be placed after the last entry given in this list. We use the name of the species with the oldest valid name in each group as the group name. Although the attribution of species to these groups is the same according to the trees constructed from autosomes and the Z chromosome, the topology between and within the groups differs somewhat (compare Fig. 89a and b ). Each topology is supported by confident statistics ( Fig. 89 ), suggesting complexities in the early evolution of these species, such as incomplete lineage sorting or gene exchange. These complexities are further corroborated by the mitochondrial genome tree, which reveals a third topology differing from the nuclear tree in the placement of many species and species groups ( Fig. 89 ).

In the list below, we attempt to order species to maximize the phenotypic similarity and geographic proximity of the list neighbors but without disrupting phylogenetic orders given in both genomic trees ( Fig. 89 ): i.e., a strongly supported clade in the trees is a continuous segment in the list. We were guided by the following considerations. We start by ordering species groups. First, Telegonus parmenides ( Stoll, 1781) , stat. rest. was historically regarded as a junior subjective synonym of Telegonus creteus (Cramer, 1780) due to phenotypic similarity and its type locality ( Suriname). However, assuming that our identification of these species is correct (see above, we follow Steinhauser’s identification before designation of neotypes), these two species belong to two different species groups. Therefore, we place the creteus and parmenides species groups next to each other in the list. Second, this adjacent position of these two groups necessitates that the elorus group is the neighbor of the creteus group (on the other side from the parmenides group) and the alector species group is the neighbor of the elorus group (according to both trees, not to disrupt their phylogenetic order); while the latimargo group is the neighbor of the parmenides group (on the other side from the creteus group) and the galesus group is the neighbor of the parmenides group (according to the Z chromosome tree). These two considerations fix the order of the species groups as: alector , elorus , creteus , parmenides , latimargo , and galesus , or a reverse of this order. Third, we choose to start the list from the alector group because of the phenotypic similarity between Telegonus alector (C. Felder & R. Felder, 1867) and the Telegonus fulgerator (Walch, 1775) group (basal area of ventral hindwing by the costa is white and forewings are with a central pale band at least in some species), the latter being placed in the list before the species analyzed in this work.

We use similar considerations to order species within each species group. When the order is inconsistent between the autosome and the Z chromosome trees, we select the Z chromosome order. Phylogenetic trees constructed from genes encoded in the Z chromosome typically correlate better with species trees due to less introgression and gene exchange involving the Z chromosome. One challenge that we met was the presence of both yellow-margined (e.g., T. chiriquensis vs. T. fulvimargo sp. n.: both species have extensive yellow scaling in the submarginal area of the ventral hindwing, giving an appearance of a marginal yellow band) and brown-margined (e.g., T. creteus vs. T. parmenides : both species possess mostly brown margins) species in the creteus and parmenides groups. Because it is not possible to satisfy placing both wing pattern categories next to each other in the list without violating phylogenetic constraints, only one of them must be chosen. We chose to place brown-margined species ( T. creteus and T. parmenides with their closest relatives) adjacent in the list, thus placing the creteus and the parmenides groups near each other due to the similarity between them and the resulting confusion in the literature about these species. Consequently, the yellow-margined species ended up on opposite sides of their corresponding groups. Compensating for this by adjusting the order of species in other groups, we made these yellow-margined species adjacent to pale- or yellow-margined species in species groups that are closest to them in the list (the elorus group is adjacent to the creteus group, and the latimargo group is next to the parmenides group). We note that any phylogenetic arrangement precludes placing all pale-margined species together in the list, because they are distributed among four species groups, three of which include brown-margined species. Moreover, one species ( T. weymeri ) is variable in the expression of yellow submarginal overscaling, and its sister species ( T. perumazon sp. n.) is brown-margined. Further refinements of the list order are encouraged.

In the resulting arrangement below, species of Rhabdoides excluding the clades with Telegonus anaphus (Cramer, 1777) and Telegonus cellus (Boisduval & Le Conte, [1837] are given. The list also includes species discovered by Steinhauser (1987) (“four new species will be added to the group”) that fall within these species groups but remain unpublished, shown in gray font. Type localities (general area only: state, region, department, or county) are in gray font. New taxa described in this study and the category of taxonomic change are in red font. Taxonomic treatment before this work (for valid names) or the category of synonym (for synonyms) and comments are shown in smaller font following a vertical bar | after the type locality; an equal sign = precedes synonyms given in their original genus combination; and a double dagger ‡ marks unavailable names. The list covers 50 valid taxa comprising 44 species (18 newly proposed here and 4 yet undescribed) and 6 additional subspecies (1 new): i.e., 27 previously known and 23 undescribed before this work. Our study follows the trend to reveal approximately as many new Hesperiidae taxa as previously described ones in nearly every genus under revision (Austin and Mielke 1998; Austin and Mielke 2008; Medeiros et al. 2019; Siewert et al. 2020).

Figures

Fig. 89. Phylogenetic trees of Telegonus species analyzed in this study inferred from protein-coding regions in: a) the nuclear genome (autosomes), based on 9,184,581 positions, b) the Z chromosome, based on 319,194 positions, and c) the mitochondrial genome. Ultrafast bootstrap (Minh et al. 2013) values are shown at nodes. Different species groups are colored differently: alector group (green), elorus group (purple), creteus group (blue), parmenides group (orange), latimargo group (cyan), and galesus group (magenta). New taxa proposed in this work are labeled in red, and those with taxonomic changes, such as subspecies-to-species or synonym-to-species status or transfer of a subspecies between species (changes indicated in brackets) are labeled in blue.