Holarctias

Beljaev, Evgeny A., 2011, Redescription of Holarctias rufinularia (Staudinger, 1901), with notes on the taxonomy and relationships of Holarctias Prout, 1913 (Lepidoptera: Geometridae: Sterrhinae), Zootaxa 3097, pp. 57-67 : 64-66

Relationships of Holarctias

The phylogenetic position of Holarctias was not especially considered in Sihvonen's (2005) phylogenetic research on the Scopulini , but the type species of the genus, H. sentinaria (as Scopula sentinaria ), was included to the phylogenetic analysis. Because of the new data on morphology in the genus, I would like to discuss the possible relationships of this genus in the tribe. Morphological data on the compared genera are taken from the cited paper of Sihvonen and from Hausmann (1994, 2004).

In Sihvonen's phylogenetic tree (loc. cit., Fig. 134) H. sentinaria is placed in the common stalk with Leucoxena ; Stigma Alphéraky, 1883 ; Dasybela Turner, 1908 ; Autanepsia Turner, 1908 ; Oar ; Antilycauges Prout, 1913 ; Cinglis Guenée , [1858]; Pseudocinglis Hausmann, 1994; Zygophyxia Prout, 1916 and Scopuloides Hausmann, 1994 . Cinglis, Pseudocinglis , Scopuloides , Antilycauges and Zygophyxia possess in the male a sternum A8 typical for the Scopulini with well developed mappa and cerata, which are indubitable synapomorphies of most Scopulini . Unlike them, Holarctias shares the general structure of sternum A8 with Oar, Leucoxena , Stigma , Dasybela and Autanepsia , which have lacking or small knob-like cerata and rudimentary mappa. This state of sternum A8 could be considered as a secondary reduction of mappa and cerata, which is supported by the optional presence of cerata in Holarctias . However in Scopula and Glossotrophia , when the cerata are strongly reduced ( Scopula cajanderi ( Herz, 1903) , Scopula arenosaria (Staudinger, 1879) , Scopula frigidaria (Möschler, 1860) , Scopula marginepunctata (Goeze, 1781) , Scopula hanna (Butler, 1878) , Glossotrophia asellaria (Herrich-Schäffer, 1847)) they always keep their lateral position.

Australian Dasybela and Autanepsia share two probable synapomorphies: hook-like cuculli and an aedeagus with a pair of similar cornuti, approximated each other, on portion of vesica. They do not have even rudimentary mappa, and possess cerata just on the posterior margin of the sternum. In these two genera sternum A8 lacks lateral sclerotized bars connecting the base of the cerata with the anterior sclerotized pouch of sternum A 8 in most other Scopulini . Possibly Dasybela and Autanepsia demonstrate a morphologically primitive phase of the formation of the Scopulini type sternum A8, which suggests their basal phylogenetic position in the tribe. Also in the genitalia they do not demonstrate any probable common specialization with Holarctias or with other genera from the considered group.

Leucoxena and Oar (though Oar reaumuraria (Milliére, 1864) ; Oar pratana (Fabricius, 1794) , have what is evidently a secondarily modified posterior margin of sternum A8 with irregular processes on the posterior margin) have small simple knob-like cerata, reminiscent of those in Dasybela and Autanepsia , which are placed sub-laterally on the dorsal surface of sternum A8 close to the posterior margin of the sternum. As in other Scopulini , in these two genera sternum A8 has sclerotized lateral bars, connected cerata with an anterior margin of the sternum. Posteriorly to the cerata a rather narrow expansion of the sternum is present, which is probably homologous to the mappa. This construction of sternum A8 is unique and can be considered as a synapomorphy of Leucoxena and Oar .

Stigma lacks cerata completely, and their sternum A8 is closely similar to that of Holarctias where cerata are totally absent. Both Stigma and Holarctias lack lateral bars of sternum A8, which makes them similar to sternum A8 of the Australian Dasybela and Autanepsia . However Holarctias , as well as Leucoxena and Oar , possesses cerata on the dorsal surface of the sternum, while in Dasybela and Autanepsia they are placed posteriorly. Evidently, in Stigma and Holarctias the reduction of the mappa and cerata started from a state of sternum A8 close to that in Leucoxena and, particularly, Oar .

Central Asian Stigma does not demonstrate specialized characters which could be considered as synapomorphies with any from these three genera (resemblance in the shape of sternum A8 with Holarctias cannot be accepted because this is a common similarity of highly reduced homologous organs). However, the Mediterranean Oar and East African Leucoxena both possess an aedeagus thickened in the anellar region, with a short tubular portion just posterior to the anellus and a long groove-like distal portion, which is similar to that in the H. sentinaria - group, and could be considered as apomorphic in comparison with other Scopulini . East African Leucoxena has large sclerotized bar-like apomorphic socii closely reminiscent of those of H. sentinaria and H. rufinaria . Oar shares with H. sentinaria and H. rufinaria a horn-like cornutus attached near the edge of the distal opening of the aedeagus, so that the cornutus does not evert together with the vesica. It is clearly an apomorphic state. In the female genitala Holarctias and Oar both possess very short anterior apophyses, a large well sclerotized antrum, a small sack-shaped corpus bursae without a narrow membranous ductus bursae and without a signum. All these female characters could be considered as synapomorphies. Besides, Oar has enlarged processes of the juxta, similar to those in H. rufinularia . However Oar sharply differs from Holarctias and Leucoxena by the short weak socii, but this state could be secondary because of the general degeneration of the tegminal portion of the male genitalia in Oar .

This grouping of Holarctias with Leucoxena or (more likely) Oar is different from the Scopulini phylogeny of Sihvonen. In the latter Holarctias is sister to the pair Stigma + Dasybela , but this node is supported by a weak apo-morphy—the presence between male 7th and 8th sternites of two pouches. This variation, consisting of one or two pouches, is present in many groups of Scopulini and in Scopula occurs even in related species, and could be considered as easily homoplastic. Leucoxena is placed by Sihvonen opposite all other genera included in this phylogenetic analysis (loc. cit., Fig. 134). This opposition is supported by 3 apomorphies, present in all genera, excepting Leucoxena : absence of sensilla styloconica (or, rather, they are morphologically unidentifiable, as it is noted by Sihvonen), absence of scales among chaetosemata and absence of signum in the female genitalia. The first (and single) apomorphy was indicated as unique, but out of 13 terminal species, marked by this apomorphy, this character was examined in 6 species only, and in 2 species among them, including H. senitaria , is reversed. The second apomorphy looks as easily homoplastic and is also reversed in the treated group of genera (in Antilycauges ). The third one is easily homoplastic in all Geometridae .

Oar in Sihvonen's phylogeny is sister to Antilycauges from Taiwan. This is supported by three non-unique apomorphies: dorsally bicoloured male antennal segments, multiple rows of ventrolateral sensillae on the proximal part of the male flagellomere, and presence of a posterior ventral lamina of the metathorax metafurca. The first character has only two species with this state almost for the whole Lepidoptera and its phylogenetic value is not important. The second apomorphy on the phylogenetic tree (loc. cit., Fig. 134) occurs in 9 places apart from the discussed node, which demonstrates its low phylogenetic value. The third one, the presence of a posterior ventral lamina of the metathorax metafurca, is also indicated for H. senitaria as a homoplasy.

Thus, characters which indicate phylogenetic relationships between the discussed genera in Sihvonen's publication, demonstrate poor uniqueness and are probably not very reliable for phylogenetic reconstruction in the discussed group of Scopulini genera. The characters used in my analysis (fine structure of sternum A8, sclerotization of the socii, details of aedeagus construction, position of cornutus on the vesica, and enumerated female genitalia characters excluding the absence of signum) were not treated by Sihvonen. They are much less homoplastic, and, probably, better indicate phylogenetic relationships in the discussed group of genera.

Generally, Holarctias in the male genitalia demonstrates a mosaic pattern of probable synapomorphies with Leucoxena and Oar , but with many more characters similar to the last genus. This conclusion agrees with Sterneck's (1941: 262) phylogenetic tree of the Sterrhinae , where Holarctias and Oar are placed on a common branch (but Leucoxena was not treated by Sterneck). Sterneck's phylogeny is based on classical principles of estimation of general similarity, and he did not ground sister relationships between these two genera by any characters, apart from the absence of cerata. Interestingly, he treated lack of cerata in Holarctias and Oar as a primary condition in the Sterrhini ('Sterrhicae' of the author), when for Stigma he supposed reduction of them, evidently, following the close general resemblance of the genitalia of Stigma to those of Scopula . This clarifies the intuitive nature of Sterneck's phylogenetic tree, grounded on detailed morphological investigation of morphology in Palaearctic Sterrhinae and creating a mental image of compact morphological groups. It is significant that applying the analytic synapomorphic method to phylogenetic reconstruction confirms Sterneck's phylogeny, at least for Holarctias and Oar .