The Psilotreta Banks, 1899 of the Dabie Mountains, east central China, with descriptions of two new species (Insecta: Trichoptera: Odontoceridae)

The authors collected caddisflies in the Dabie Mountains in 2014 and 2015. Three species of Psilotreta Banks, 1899 were found: Psilotreta daidalos Malicky, 2000, P. furcata sp. nov., and P. brevispinosa sp. nov. The species were collected by light sheet traps or hand nets near streams in mountainous forests. All the species are described and illustrated, and the species groups to which they belong are discussed. Some variation within P. daidalos is described and the species is newly recorded from the Anhui and Hubei Provinces.

Members of both the Psilotreta japonica and P. trimeresuri species groups are found in China. The differences between these two species groups are: (1) in the P. japonica group the harpagones are located at the apices of the coxopodites, in the P. trimeresuri group the harpagones are located mesally; and (2) in the P. japonica group there is one pair of spines in the phallic apparatus while in the P. trimeresuri group there are two pairs of spines. However, many species in China possess a combination of characters from these two species groups. For example, Psilotreta spinata P. expers Yuan & both have harpagones located mesally on the coxopodites, but with only one pair of spines in the phallus. Moreover, Psilotreta bicruris Yang, 2011, andP. paulula Yuan &Yang, 2011 have harpagones located at the apices of the coxopodites, but with two pairs of spines in the phallus.  also studied Psilotreta and found the length ratio of the maxillary palp segments and wing veins could be characters useful in delineating different species groups. They also recognized several species groups that were different from Oláh & Johanson's classification , 2011. Afterwards, Yuan & Yang (2013) published data on new species with a key to Psilotreta species groups of China and adjacent regions, they also divided part of this genus into five species groups: the Psilotreta chinensis species group, the P. apiculata species group, the P. monacantha species group, the P. kwantungensis species group and the P. jaroschi species group. The Chinese species mentioned above can be classified well using Yuan & Yang's key. Therefore, the discussion of species groups in this article mainly follows Yuan & Yang (2013).
In this article, we will introduce two new species from the Dabie Mountains, east central China. Descriptions and illustrations are provided; the species groups for the new species are also discussed.

Material and methods
Samples were taken in 2014 and 2015; detailed site data are included in the material examined section. Most adult specimens were collected by light sheet traps at night, although some were collected by hand nets in daytime alongside streams. All specimens are preserved in 95% ethanol. Specimens were examined under a stereo microscope. Abdomens of males were dissected and placed in 10% KOH overnight to remove the non-chitin structures; cleaned abdomens and male genitalia were put in glycerin for observation. An eyepiece with grids was used for illustration on paper with grids. The sketches were scanned and digital illustrations were completed in the software Inkscape (ver. 0.48.4.0). Holotypes are deposited in Nanjing Agriculture University, College of Plant Protection, Specimens Room. Paratypes are deposited in Huazhong University of Science and Technology, College of Life Science and Technology, Institute of Bioenergy and Ecology.

NAU = Nanjing Agriculture University HUST = Huazhong University of Science and Technology
Terminology of male genitalia mainly follows Nielsen (1957), except the terminology of segment X and phallus which follow Parker & Wiggins (1987). Terminology of setae warts on the head follows Oláh & Johanson (2007).

Diagnosis
The male genitalia of this species resemble those of P. trispinosa Schmid, 1965, but can be distinguished by the following characters: (1)
Wings. Forewings each 7.4-8.0 mm (n = 2), yellow, discoidal cell (DC) present, fork I is approximately one third longer than fork II; R 4+5 not fused with R 3 , Cu 2 absent; row of setae located in close proximity to posterior margin of each forewing. Hind wings each present with Sc close to R1 mid-span, DC present (Fig. 1D).

Comments
Compared with Malicky's (2000) descriptions and illustrations, the specimens in Henan Province have the intermediate appendages more strongly curved than those in Anhui and Hubei Provinces. The ventral sclerite of the aedeagus of specimens from Henan Province is slender, not pandurate as we described. The shapes of other structures are similar.

Distribution
The type locality of this species is in Henan Province (Malicky 2000), now we report it from Anhui and Hubei Provinces.

Diagnosis
The male genitalia of this new species resemble those of P. quinlani Kimmins, 1964, but can be distinguished by the following characters: (1) the median dorsal process of segment X is more than 3 × as long as wide in dorsal view (short in P. quinlani, about as long as wide in dorsal view); (2) ventral branch of lateral process of segment X is simple and slender, directly pointed posteriorly in lateral view (in P. quinlani, ventral branch of lateral process of segment X is more complicated, directed ventrally and reduced to a digitate projection curving posteriorly in lateral view); and (3) superior appendages each round at apex in lateral view (superior appendages of P. quinlani each triangular at apex in lateral view).

Etymology
The specific name is derived from the Latin 'furcata', or English 'furcate', referring to the forked median dorsal process of segment X.
Wings. Forewings each 8.9-9.2 mm (n = 2), yellow, round at apex; DC present, fork I about ⅓ longer than fork II, R 4+5 not fused with R 3 , Cu 2 absent; row of setae located close to posterior margin of each forewing. Hind wings with R1 straight, Sc and R2 slightly curved at crossvein Sc-r and r separated, DC absent (Fig. 2D).

Distribution
The type location is in the Hubei Province, but the distributions is otherwise unknown.

Diagnosis
The male genitalia of this new species stand apart from those of other species of Psilotreta species by the huge superior appendages, the absence of an intermediate appendage and the short, anteriorly curved lateral process of segment X.

Etymology
The specific name is derived from the Latin 'brevi-' and 'spinosa', meaning 'short' and 'spines', referring to the pair of short spines on the lateral process of segment X.

Male
Head. Head brown, in anterior view with short dark line in centre of frons (Fig. 3A); frontogenal compact setal warts large, approximately 2 × as long as wide, with dark line connecting ventral end of each setal wart and clypeus. Anterior tentorial pits located very close to clypeus. In dorsal view with dark median dorsal line (Fig. 3B). Median dorsal line branched and extending to bases of antennae, between vertexal mediantennal compact setal warts and vertexal lateroantennal compact setal warts, and present with curved horizontal branches posteriorly. Vertexal lateroantennal compact setal warts smaller than vertexal mediantennal compact setal warts. Occipital setal warts oval, smaller than oval postgenal setal warts. Maxillary palps lost in the dissection process.
Wings. Forewings each 7.7 mm (n = 1), dark brown; DC present, fork I slightly longer than fork II, R 4+5 not fused with R 3 , Cu 2 absent; row of setae located close to posterior margin of each forewing. Hind wing C with hump and row of setae in middle; R 1 disappearing after crossvein Sc-r; DC absent ( Fig. 3C).
Male genitalia. Tergum IX narrow in lateral view (Fig. 3D), fused with median dorsal process of segment X forming long thin process (Fig. 3E); sternite of segment IX truncate, with large anterior process in lateral view (Fig. 3D) and triangular concavity on posterior margin in ventral view (Fig. 3F). Lateral process of segment X with two pairs of short branches: apical branches round, with one short spine postero-ventrally; subapical branches forked into few acute ends in dorsal and lateral views (Fig. 3H). Superior appendages leaf-like in lateral view (Fig. 3D) and digitate in dorsal view (Fig. 3E), longer than dorsal process of Tergite IX. Coxopodites wider basally than apically in lateral and ventral views, with bases strongly sclerotized (Fig. 3D, F); harpagones about ⅓ as long as coxopodites, each with small spines and setae on apices (Fig. 3F). Phallotheca sclerotized, thin and long, slightly curved ventrally in lateral view (Fig. 3G), apex forked in ventral view (Fig. 3I); parameres present, slightly longer than height of phallotheca in lateral view (Fig. 3G), with apices curved posteriorly; endotheca and aedeagus membranes, ventral sclerite of aedeagus strongly curved in lateral view (Fig. 3G) and oval in ventral view (Fig. 3I).

Distribution
This species has been found only in Henan Province, China. The wider distribution is unknown.

Discussion
According to Oláh & Johanson (2010), the three discribed species found in the Dabie Mountains all belong to the Psilotreta japonica species group, as their harpagones are located on the apices of the coxopodites with numerous small spines at the end, and they all have one pair of parameres.
According to Yuan & Yang (2013), Psilotreta daidalos and P. furcata sp. nov. belong to the P. chinensis species group, as the harpagones are not forked, segments 1-2 of each maxillary palp are approximately the same length, and the intermediate appendages are present. Psilotreta brevispinosa sp. nov. cannot be assigned to an existing species group. The fusing of segments IX-X and the absence of intermediate appendages show some similarities with Psilotreta apiculata Yuan & Yang, 2013or P. monacantha Yuan & Yang, 2013; however, the phallus of P. apiculata lacks parameres and the forewings of P. monacantha lack DC, while P. brevispinosa has both parameres and DC present. These are significant differences, so the three aforementioned species cannot be considered to be closely related. More characters are needed to analyse the position of P. brevispinosa sp. nov.