Oxyptilus, AND

Alipanah, Helen, Gielis, Cees, Sari, Alireza, Sarafrazi, Alimorad & Manzari, Shahab, 2011, Phylogenetic relationships in the tribe Oxyptilini (Lepidoptera, Pterophoridae, Pterophorinae) based on morphological data of adults, Zoological Journal of the Linnean Society 163 (2), pp. 484-547 : 524-526

OXYPTILUS AND CROMBRUGGHIA

As already mentioned, the genus Oxyptilus (sensu lato) is one of the most problematic genera of Oxyptilini . This genus was introduced by Adamczewski (1951) with seven species: O. pilosellae , O. ericetorum , O. chrysodactyla , O. parvidactyla , O. bohemanni , O. delawaricus , and O. hoffmannseggi . Later, O. hoffmannseggi was synonymized with O. parvidactyla , and O. bohemanni was considered to be a synonym of O. pilosellae . The most additional species were later described by Meyrick during the first half of the 20 th century ( Meyrick, 1905, 1911, 1912 –1916, 1920, 1930–1936, 1936–1937). There are many interspecific variations within the genus Oxyptilus and only these five species are close to each other in male and female genitalia and wing characteristics. These species are considered here as true Oxyptilus species. As far as we know, none of the remaining species is similar to these five species and it seems probable that the other species belong to different genera. For example, Oxyptilus cinctipedalis was placed in the genus Nippoptilia by Arenberger (2006) and considered to be a member of Platyptiliini . In the same paper, he regarded Oxyptilus caryornis as a synonym of Nippoptilia cinctipedalis ( Arenberger, 2006) . Gielis (2006) considered Oxyptilus maleficus as a junior synonym of Leptodeuterocopus neales .

Some new findings are mentioned here. Although owing to the lack of specimens, Oxyptilus regulus and O. vibrans were excluded from the analyses, considering the original descriptions and the illustrations, O. regulus clearly belongs to the genus Nippoptilia rather than Oxyptilus because of its similarity in the genitalia and wing pattern (see Nomenclatural changes). Additionally, O. vibrans seems to be a member of the genus Deuterocopus from the subfamily Deuterocopinae . Oxyptilus causodes was excluded from the analysis owing to the lack of information about females and some doubts about male genitalia characteristics; however, including this species in the analysis placed it out of true Oxyptilus species. Moreover, O. variegatus , O. secutor , and O. anthites were not recovered within true Oxyptilus species ( Figs 16 , 17 ), so the monophyly of Oxyptilus was not supported in this study, but its monophyly not rejected statistically ( Table 3). Hence, as already mentioned above, only O. pilosellae , O. ericetorum , O. chrysodactyla , O. parvidactyla , and O. delawaricus are here considered as true members of Oxyptilus and an attentive revision of the species previously included in this genus is suggested.

There are two probable synapomorphies for true Oxyptilus species. In these species, except O. delawaricus , the shape and characteristics of the eighth abdominal tergite of males are the same (character 69). The related information was missing in O. delawaricus and if this species have the same features, the character can be considered as a real synapomorphy for this group. Moreover, in all of them, except O. chrysodactyla , the basal sclerotized process at the back of the valva is triangular (character 138). In O. chrysodactyla , this character was not clear in the examined male, but if it has the same character, this can be another synapomorphy of true Oxyptilus species.

Here we consider Oxyptilus variegatus as a junior synonym of O. secutor (see Nomenclatural changes). The latter species was not recovered in the same clade with true Oxyptilus species. In all MPTs obtained from both the EW and SAW-RI-based analyses, this species was the sister group of all the species included in clade O ( Fig. 16 ). This can be separated from Crombrugghia and true Oxyptilus species in the striking shape of male and female genitalia, i.e. in O. secutor , the eighth sternite of males is heavily specialized; the tegumen is unilobed; the uncus places above the tegumen, making an angle to it; and the valva is well developed without any valvular lobe. Oxyptilus secutor is close to the members of clade O ( Fig. 16 ) in having a sclerotized structure in the bulbus ejaculatorius of the aedeagus, although this feature is not precisely clear in Sphenarches bilineatus , Oxyptilus anthites , and Eucapperia bullifera . Unlike the Crombrugghia and true Oxyptilus species (clade N, Fig. 16 ), in O. secutor , the anellus is an elongated rectangle and not heart-shaped. This feature may join it to Intercapperia , Capperia , Procapperia , and Paracapperia species. In O. secutor , a series of very short and dark scales is visible on the costal margin of the third lobe of hind wing, which is only present in this species. These differences, together with the presence of five autapomorphies (characters 11, 115, 122, 155, and 161) for this species and its position in the consensus trees obtained from both the EW and SAW analyses may support the idea that O. secutor belongs to a new genus; however, constraining O. secutor and true Oxyptilus species to be monophyletic was not rejected statistically (P = 0.6943 –1.000). Therefore, here we consider it as a member of a new genus (will be described elsewhere).

The same results were obtained for O. anthites . In all MPTs of the EW and SAW-RI-based analyses, this species was the sister group of clade Q, which is composed of all Capperia , Procapperia , Paracapperia , Intercapperia , Eucapperia , Sphenarches , Geina , Tomotilus , and Antarches species ( Fig. 16 ). This species is separated from all the examined Oxyptilus species by the wing shape, position of dark scale teeth, wing venation, and the characteristics of the male and female genitalia, and having five autapomorphies (characters 4, 38, 114, 121, and 144). However, constraining O. anthites and true Oxyptilus species to be monophyletic was not rejected statistically (P = 1.000). Owing to the clear differences between this species and other known Oxyptilus species in both external and genitalia characteristics and bearing in mind that it is placed out of the other examined Oxyptilus species in the MPTs obtained from the EW and SAW-RI-based analyses, here we consider it as a member of a new genus (will be described elsewhere).

On the basis of the present study, Crombrugghia and five true Oxyptilus species in both the EW and SAW-based analyses formed a monophyletic clade that was supported by five synapomorphic characters. Gielis (1993) in his phylogenetic study of the family described the genus Dejongia and demonstrated a close relationship between this genus and both Crombrugghia and Oxyptilus . According to his idea, the presence of a hair brush along the third segment of the labial palpus, double bean-like signa of the bursa copulatrix, acute fore wing lobes, and the absence of the termen could support the monophyly of the clade including these three genera. In the current study, even if we consider true Oxyptilus species as the only representatives of Oxyptilus , they do not share all of these features. We found that a narrow hair brush is also present along the third segment of the labial palpus in Stenodacma and some Trichoptilus species. Moreover, the signum is absent in Dejongia californicus . Additionally, in Dejongia lobidactylus , the signum is oval and has an emery surface; in Oxyptilus pilosellae and Crombrugghia laetus , it is as a pointed denticle; and in Oxyptilus chrysodactyla it can be seen in two shapes: (1) oval and slightly invaginated, and (2) as a small linear projection. However, only in Dejongia species both lobes of the fore wing are pointed apically. In all true Oxyptilus and Crombrugghia species , the first lobe is pointed apically and the second lobe has a clear termen.

Although, as found here, in both Dejongia and the group including true Oxyptilus and Crombrugghia species , the ventral hair brush of the second segment of labial palpus extends along the third segment for three-quarters of its length or more (synapomorphic character), Dejongia comes closer to Stenodacma , Trichoptilus , Megalorhipida , Stangeia , and Buckleria species than to the members of clade K ( Fig. 16 ), especially in the wing shape and the pattern of the fore wing lobes. This genus is the sister group of clade K in the consensus trees obtained from the EW and SAW-RI-based analyses.

As already mentioned, Crombrugghia and true Oxyptilus species are closely related. According to Adamczewski (1951), Oxyptilus and Crombrugghia are close to each other in the strongly specialized eighth tergite; weakly developed eighth sternite; and very specialized, weakly sclerotized armed valva. He also added that in Crombrugghia , scale teeth on the dorsum of the third lobe of the hind wing are very far from the tip, whereas in Oxyptilus they are positioned at the tip of the lobe. Additionally, Crombrugghia is more specialized than Oxyptilus , as the number of generations in the former genus is more than in Oxyptilus ( Adamczewski, 1951) . The difference in the sclerotized plates around the ostium bursae of the female genitalia was claimed by Bigot (1962). These two genera, based on the latter character and positions of the scale teeth in the dorsum of third lobe of hind wing, were considered by Bigot & Picard (1988a) as two subgenera of Oxyptilus . As stated by them, scale teeth in the subgenus Crombrugghia are located at two-thirds of the lobe. Having said this, they clearly mentioned that male genitalia characteristics failed to support this division exactly ( Bigot & Picard, 1988a). Later, they considered all the known Crombrugghia species of France as the members of Oxyptilus ( Bigot & Picard, 1988b, 1991). Zagulajev (1997) did the same for all the known Crombrugghia species of Russia. Bigot et al. (1998), in a study of the Pterophorinae of France, synonymized the genus Crombrugghia with Oxyptilus and divided the latter genus into six sections. It is worth mentioning that Gielis (1993, 1996) had already considered them as two separate genera using the same characters as Adamczewski (1951) and Bigot (1962), which was later followed by Arenberger (2002, 2006). Here we believe that, in spite of some differences in the female genitalia, the main difference between these two genera is the positions of scale teeth in the third lobe of the hind wing; i.e. in true Oxyptilus species , the scale teeth are present at both the costa and dorsum of the third lobe of hind wing, whereas in Crombrugghia species these only present at the dorsum. Furthermore, in all true Oxyptilus species , the distance between the two scale teeth of the dorsum is clearly less than the length of the proximal one, whereas in Crombrugghia species it is equal to or slightly more than the length of the proximal scale tooth. Moreover, in true Oxyptilus species the alternative transverse rows of white and brown scales cover the dorsal surface of the antenna, whereas in Crombrugghia species the antenna is covered by both transverse and longitudinal rows of white and brown scales.

The Crombrugghia and true Oxyptilus species are close to each other in several aspects and all their members feed on different species of Hieracium L. ( Asteraceae ) ( Matthews & Lott, 2005). The low values of tree confidence for the clades including them ( Figs 16 , 17 ) and high tree confidence values for the clade composed of these two clades (clade N, Fig. 16 ), as well as the presence of five synapomorphies for clade N, may support the idea that Crombrugghia and true Oxyptilus species can be synonymized with each other and/or, as stated by Bigot & Picard (1988a), each of them may be a subgenus of a single genus. However, they are considered here as two distinct genera.

In all the consensus trees, Crombrugghia distans and C. tristis were found to be sister groups and C. kollari was sister to them both. Based on the present study, C. distans and C. tristis differ from each other in two male genital characters (characters 74 and 90): in C. distans , the specialized eighth sternite is as long as the eighth tergite and the uncus is trapezoidal, whereas in C. tristis the length of the specialized eighth sternite is less than the length of the eighth tergite and the uncus is dome-shaped. However, they have a common character: the slightly sclerotized eighth abdominal tergite of males has a more sclerotized caudal margin. Bearing in mind the two abovementioned differentiated characters, as well as the low values of tree confidence for the clade including C. distans and C. tristis (Bremer support of 1), we leave them here as two distinct species.

It is worth mentioning that, according to Zagulajev (1997), the differences of C. distans and C. tristis are in their size and concavity of the distal margin of sternite VII in the female genitalia, i.e. C. distans has a greater wingspan and the distal margin of sternite VII in females is more concave. However, we believe that both differences are intraspecific variations. Additionally, based on Arenberger’s (2002) idea, the antrum plate is only present in C. distans and C. reichli , but considering the definition of the antrum plate in the present study, it is assignable to all known Crombrugghia species.

Figures

Figure 16. Strict consensus trees of six equally most parsimonious trees obtained in the initial analysis using equally weighed characters. Numbers below the branches show Bremer support values and the capital letters show the clades included in Oxyptilini and are discussed in the text. Tree length = 943, consistency index = 0.4549, retention index = 0.7976.

Figure 17. Strict consensus trees of six equally most parsimonious trees obtained in the equally weighted analysis (left), and that of 380 most parsimonious trees obtained from successive approximation weighting based on maximum retention index values (right). Numbers above and below (underlined) the branches show bootstrap and jackknife values, respectively.