On the validity of Epeorella Ulmer, 1939 (Ephemeroptera, Heptageniidae) with general considerations on the Heptageniidae of the Sunda Islands

Abstract The type material of Epeorella borneonia Ulmer, 1939, the sole species of the genus Epeorella Ulmer, 1939 is reinvestigated and a lectotype (male imago) is designated. Based on several morphological structures, the synonymy with Epeorus Eaton, 1881 (Rhithrogeninae) is rejected. Epeorella stat. prop., known only at the winged stages, belongs to the subfamily Ecdyonurinae, and is a probable endemic of the island of Borneo. The newly erected genus Darthus Webb & McCafferty, 2007, also endemic to Borneo and only known by one species at the nymphal stage, is shown to be a junior subjective synonym of Epeorella. The new combination Epeorella vadora (Webb & McCafferty, 2007) is proposed for the species. The distribution of known heptageniid species from the Sunda Islands is discussed.

The genus Rhithrogeniella (type species Rh. ornata Ulmer, 1939, based on winged stages only) had an enigmatic position for a long time, until another species from Vietnam (Rh. tonkinensis Soldán & Braasch, 1986) was described, with the first reference to the nymphal stage. Based on these descriptions, Wang and McCafferty (2004) indicated that Rhithrogeniella was a synonym of Rhithrogena and Rh. ornata was a species of Rhithrogena; the species Rhithrogeniella tonkinensis was transferred to the genus Ecdyonurus. Recently, the nymph of Rh. ornata was described for the first time, and the generic status of Rhithrogeniella revalidated (Sartori 2014a) as a member of Ecdyonurinae.
The monotypic genus Epeorella (type species E. borneonia Ulmer, 1939, known only from the winged stages) was synonymized with Epeorus (Wang and McCafferty 2004) on the basis of similarities in several characters which will be discussed below.
The family Heptageniidae is now divided into three subfamilies which can be broadly characterized as following (Kluge 1989;Webb and McCafferty 2008): • Rhithrogeninae: nymph with a row of setae on the ventral surface of maxillae, with dorsal process of the forefemora projected and narrower than the ventral process, some genera with vestigial paracercus; winged stages with the median depression of the mesothoracic furcasternum convergent anteriorly, and prosternum lacking transverse and longitudinal ridges. • Heptageniinae: nymph with a row of setae on the ventral surface of maxillae, with forefemora without a dorsal projection; winged stages with the median depression of the mesothoracic furcasternum convergent anteriorly, and prosternum with distinct transverse and longitudinal ridges. • Ecdyonurinae: nymph with scattered setae on the ventral side of maxillae; winged stages with the median depression of the mesothoracic furcasternum parallel sided or divergent anteriorly, and prosternum generally lacking transverse and longitudinal ridges.
This study concludes the re-investigation of Ulmer's Heptageniidae from Southeast Asia deposited in the Zoologisches Museum of Hamburg University (ZMH) (Sartori 2014a;b;c;. The type material of Epeorella borneonia Ulmer, 1939 is described, some morphological structures are clarified, the subfamily position is established and a new synonymy is proposed.

Material and methods
The studied material is composed of three pinned specimens. The female imago was rehydrated in a solution of trisodic phosphate 0.35% and then put in alcohol. Pictures were taken with a Visionary Digital Passport II in ZMH, and plates were assembled in Adobe Photoshop CS6.
The male imago was wrongly mentioned as holotype by Weidner (1962). This terminology cannot be accepted because the "holotype" has not been designated by Ulmer and cannot be ascertained by the presence of a single specimen (see also Recommendation 73F. Avoidance of assumption of holotype, ICZN 1999).
The male imago is accordingly designated as LECTOTYPE of the species E. borneonia by present designation.
The three specimens have been adequately described by Ulmer (1939). Only significant morphological characters are mentioned here.
Male imago. Anterior margin of the head not protruding anteriorly (Fig. 1); median depression of mesothoracic furcasternum subparallel, not convergent anteriorly (Fig. 3); mesonotum with a transverse suture (Fig. 2); styliger plate strongly convex, penis lobes minute, rounded and closely tight together (Fig. 4), without apparent sclerites or titillators [a complete analysis of the genitalia will be presented later with the help of non-invasive techniques].

Female imago.
Abdominal patterns similar to the male. Extracted eggs from the rehydrated specimen were unfortunately not in a satisfactory state for chorionic structure examination through SEM.
Female subimago. Similar to the female, except abdominal terga VI-VIII (IV-V to a lesser extent) present the remains of a longitudinal ridge (Figs 5-6).

Discussion
The synonymy of Epeorella with Epeorus was proposed by Wang and McCafferty (2004) based on the following assertions: i) male genitalia and forelegs are similar to those in Epeorus; ii) the vestiges associated with adults indicate that the larvae were two-tailed, and iii) the presence of median tubercles on abdominal segments VI-VIII can also be found in some Epeorus species.
The male genitalia greatly vary in shape among Epeorus species (see Webb and McCafferty 2008, figs 150-154) but are never as found in Epeorella; also, the forelegs of Epeorella are missing (see Ulmer 1939, p. 578) and therefore cannot be compared to Epeorus. The statement that, based on the vestigial paracercus, the nymph was twotailed (hence comparable to Epeorus) is puzzling. "All Heptageniidae have the same vestigial paracercus, which does not allow to distinguish those with two-tailed and three-tailed larvae" (N. Kluge in litt.); the median ridge may be present on some Epeorus species which is true, but the genus Epeorus is absent from most of the Sunda Islands (Edmunds and Polhemus 1990, M. Sartori unpubl. data) as well as the Philippines. It is poorly diversified in Borneo where it is represented by a single species (Epeorus boonsoongi Braasch, 2011) with a large nymph of ca. 15 mm body length, and bifid tergal spines.
The male imago reinvestigated here presents all characteristics of the subfamily Ecdyonurinae, in peculiar the median depression of the mesothoracic furcasternum is not convergent anteriorly. Moreover, the presence of a clear transverse suture on the mesonotum, excludes it from the genus Epeorus. As already suggested by Braasch (2011) the synonymy proposed by Wang and McCafferty (2004) is incorrect and Epeorella is reinstated as Epeorella Ulmer, 1939, stat. prop. One interesting character of Wang and McCafferty (2004) is the presence of remains of median tubercles visible at least on abdominal terga VI-VIII of the female subimago. It has already been demonstrated that the subimaginal stage may retain some nymphal structures, such as gill sockets (Sartori et al. 2008), which may help to link nymphal and winged stages. The presence of vestigial tubercles on the terga thus indicates that the nymph possesses a median single ridge on the abdomen. Among Ecdyonurinae, two genera are known to hold such structures, Notacanthurus Tshernova 1974 known from East Palaearctic, the Himalaya region and Southeast Asia, and Darthus Webb and McCafferty 2007, only known from Borneo (Figs 7-8). Male imagos of Notacanthurus possess the anterior margin of the head distinctly produced (Braasch 1986;Webb and McCafferty 2008), and genitalia possess clearly visible apical and lateral sclerites. Darthus is only known from the nymphal stage from the same island as Epeorella. Although distant by ca 500 km, the type locality of Epeorella borneonia and that of Darthus vadorus Webb & McCafferty, 2007 belong to the Dipterocarpaceae forest of lowland altitudes (Sartori et al. 2003). Moreover, the size of the mature nymphs of D. vadorus (5.5-8.5 mm) is compatible with the adult size of E. borneonia (5.0-5.5 mm) knowing that alate stages are generally smaller than mature nymphs. Therefore, it is likely that Darthus represents in fact the nymphal stage of Epeorella, and is considered as a subjective junior synonym of Epeorella syn. n. The species Epeorella vadora comb. n. is retained as a valid species because it exhibits a different colour pattern of the abdomen.

General considerations on the Heptageniidae of the Sunda Islands
The three genera described by Ulmer (1939) have been put in synonymy by different authors but it is now demonstrated that they represent three groups of species deserving generic rank (Sartori 2014a;b, present study). They all belong to the subfamily Ecdyonurinae which is the most diversified in the studied area. Table 1 summarizes our current knowledge about the Heptageniidae of the Sunda Islands.
The Rhithrogeninae are known by only two species belonging to two widespread and speciose genera: Rhithrogena with more than 150 species and Epeorus with almost 100 species. Although mainly Holarctic, these two genera are present in the Oriental Region with 12 and 32 species respectively. The genus Epeorus is well represented in Indochina, where at least 13 species are known, but was unknown from the Sunda Islands until Braasch (2011) described E. boonsoongi from Borneo. The genus is not   recorded from Java, Sumatra and other Sunda Islands despite numerous samples over the last century. Rhithrogena exhibits an opposite trend in its distribution, being known from Sumatra, Java, Lombok, and possibly Bali, but seems to be absent from Borneo.
The only genus of the subfamily Heptageniinae present on the Sunda Islands is Trichogenia Braasch & Soldán, 1988, a Southeast Asian genus with one species in Sulawesi, T. hubleyi Webb & McCafferty, 2006 and two species in Sumatra, T. nasuta (Ulmer, 1939) and T. ulmeri Braasch & Webb, 2006, the latter also recorded from Borneo (Webb et al. 2006). There is a reasonable probability that T. ulmeri, known only at the nymphal stage, is a junior synonym of T. nasuta, known only at the winged stages, the distance between both type localities being less than 100 kilometres.
The subfamily Ecdyonurinae includes four times as many species as the two previous subfamilies combined. This is not surprising since Ecdyonurinae nymphs are among those which can tolerate slow flowing waters and high water temperatures; they have movable gills which is also an advantage when oxygen concentration is not optimal. The only two genera found in tropical Africa (Afronurus and Notonurus) also belong to the Ecdyonurinae. The genus Afronurus is the most diversified in the Oriental Region with 45 described species, but most of them are poorly known or badly described. It is probable that the concept of Afronurus in the Orient is paraphyletic; nevertheless, the genus seems present mainly on Borneo with three species; the species Afronurus javanicus Ulmer, 1939, is only known by adults collected on Java, which fit the current concept of Afronurus (M. Sartori, unpub. data). The genus is not reported from Sumatra, or Sulawesi, but seems present on Sumbawa and Sumba (M. Balke coll.). The two genera Atopopus and Epeorella are only found on Borneo, the former extending its range to the Philippines with two described species. The genera Asionurus and Rhithrogeniella have a distribution restricted to Indochina, extending to Java and Sumatra only. Finally Sulawesi is the most eastern island to have been colonized by Heptageniidae with Compsoneuria, Compsoneuriella and an undescribed genus (M. Sartori unpubl. data, M. Balke coll.). The family is not recorded from Moluccas, as well as Papua New Guinea, where the families Baetidae, Leptophlebiidae and Caenidae are eudominant.
More studies are needed, especially molecular phylogenies, to infer the timing and patterns of distribution of the genera and species in the area, particularly the relative importance of vicariance processes and dispersal events since the Miocene (Lohman et al. 2011).