Rhyothemis variegata phyllis comb. nov.

The reason is the considerable difference in males from the Indian subcontinent, doubtlessly occupied by R. variegata variegata, as illustrated by Fig. 2a and the key by Fraser (1936), and those from Indochina (Fig. 3a–c) and Sundaland (at least partly) (Fig. 2c). The males from the Indian subcontinent as a rule have dark spots at the forewing triangles and at least at vein IR 3 and optionally also at vein M (our phenotype C) (Fig. 2a), in accordance with Fraser’s key. (As usual, the rich data from iNaturalist (2025) provides us with some rare exceptions, like males with only traces of spots at the triangles and without spots at IR 3 and M, formally corresponding to phenotype A but with a broad basal hindwing area: e.g. observations 273864365 and 182448723). The Thai, Indochinese and Sondaic males very rarely have those additional dark spots at the fore wing triangle (Fig. 3b) or even veins IR 3 and M (Fig. 3c), while the yellow tint of their wing membrane is variable, as a rule much fainter than in the Indian males, and is often missing. We find the overwhelming predominance of the ‘pure phyllis’ phenotype A in Thailand and Indochina (but we miss information from the north of these regions) and from elsewhere in the presumed range of the currently recognised subspecies R. phyllis phyllis sufficient to consider this area to be occupied by the subspecies with the same subspecific name but in combination with another specific name, R. variegata phyllis. We have to stress that the common ‘ variegata -females’, that is our phenotype D, should be attributed to this subspecies in this area as well, in line with our above proposal to abstain from defining subspecific characters in the female sex of this species. Rare males with phenotypes B and C occurring in the presumed range of R. variegata phyllis should be attributed to this subspecies as well, although formally the same morphs predominate in R. variegata variegata.

In view of the same but very broad type locality indicated at the original descriptions of R. variegata and R. phyllis, we have to avoid taking them into account but have to refer to the currently persisting notions of the geographic ranges of ‘ R. variegata ’ and ‘ R. phyllis ’, outlined in Introduction.

The northern border of the presumed range of R. variegata phyllis remains rather unclear for we have no data from northern Thailand, Laos and Vietnam. Anyway, the more northerly areas, namely southern China, Taiwan and northern Vietnam, are occupied by a remarkable subspecies R. variegata arria (Drury, 1773) (type locality unknown) (Tsuda 2000; Zhang 2019). While its females (at least most of them, as follows from iNaturalist 2025) represent our phenotype D, its males (at least a remarkable majority of them, see the same source) show another distinct phenotype characterised by the most coloured wings, which could be named ‘phenotype E’: it resembles phenotype D but the fore wings are now coloured to tips as having broad blackish apical and subapical spots, usually over a yellowish background (Fig. 7). On the other hand, this phenotype is a heavily melanised version of phenotype C, with all dark spots on the wings strongly enlarged. Following our principle of basing subspecies of R. variegata on male characters, we naturally assume the subspecies R. variegata arria as existing, because of such peculiar males. Curiously, the same phenotype E have been once registered in a male far from the range of R. variegata arria in Similan Islands, Thailand (iNaturalist 2025: observation 69022483), that is in the presumed range of R. variegata phyllis. Anyway, this phenotype occurred in that area as not more than a very rare aberration, fitting the above considered concept of subspecies. The subspecies R. variegata imperatrix Selys, 1887 from Ryu Kyu looks like a more melanistic version R. variegata arria, with blackish areas extended in both sexes (Ozono et al. 2012).

In sum, the three subspecies of R. variegata considered and accepted in this section are assumed to differ as follows:

Rhyothemis variegata variegata —overwhelming majority of males are of phenotype C (Fig. 2a: with broad dark and yellow basal area, dark tips and some small dark spots at the forewing triangle and hind wing veins IR 3 and M), their wing membrane usually more or less yellowish.

Rhyothemis variegata phyllis —overwhelming majority of males are of phenotype A (Fig. 2c: wings without dark spots between the basal and apical areas), their wing membrane usually colourless.

Rhyothemis variegata arria —overwhelming majority of males are of phenotype E (Fig. 7: wings with large dark spots throughout; wing membrane yellowish).

The geographic limits of R. variegata variegata, R. variegata phyllis and R. variegata arria are far from clear and demand further studies. We attempted to outline them very putatively in Fig. 1a. It should be stressed that these putative limits are not based on some new or original data but just reflect the current notions of the ranges of R. variegata arria and R. phyllis phyllis in their current treatments (see e.g. Lieftinck 1953; Tsuda 2000; Kalkman et al. 2020; Dow et al. 2024). Most probably these subspecies are separated by broad transitional zones, challenging their very existence. The status of numerous subspecies proposed in R. phyllis deserves a special comprehensive study but there is no doubt that those which will be considered as existing should be re-attributed to the species R. variegata.