A new species of Dicranocentrus (Collembola, Entomobryidae) from China with comments on the systematic position of the genus

Abstract Dicranocentrus liuae sp. n. is described from the northern subtropical region of China. The new species is most similar to D. wangi Ma & Chen, 2007, but differs from it in the relatively shorter Ant. V, the 1+1 central macrochaetae on Abd. III, the number of chaetae on tenaculum, and the absence of dental spines. The systematic position of Dicranocentrus is also discussed. Present evidence, particularly S-chaetotaxy, indicates that the genus is closer to Heteromurus than to the unscaled species of Orchesella and Orchesellides.


Introduction
The genus Dicranocentrus was erected by Schött (1893) for D. gracilis. It is characterized by 6-segmented antennae, the ratio between abdominal segments IV/III less than 2.0, scales present on antennae, legs, body, manubrium and ventral side of dens, eyes 8+8, postantennal organ absent, and mucro bidentate with a basal spine. Mari-Mutt (1979) published an excellent revision of the taxonomy, biology, and geographical distribution. Mari-Mutt (1980) divided the Orchesellinae into four tribes mainly based on the number of antennal segments, with Orchesellini (Dicranocentrus included), Heteromurini and Corynothrichini having 6, 5, and 4 segments, respectively. Soto-Adames (2008) did not change Mari-Mutt's taxonomical framework, but simply added two new small tribes. However, recent molecular phylogeny of the Entomobryidae (Zhang et al. 2014a) placed Dicranocentrus together with Heteromurus in a separate clade, apart from the unscaled taxa (Orchesella/Orchesellides).
So far, three Dicranocentrus species have been reported from China: D. indicus Bonet, 1930from Taiwan, D. chenae Ma, Chen & Soto-Adames, 2006from Guangxi, and D. wangi Ma & Chen, 2007 from Gangdong (Fig. 1). Here, we describe a new species from the northern subtropical region of China, compare it with other orchesellids, and discuss the systematic position of the genus.

Materials and methods
Specimens were mounted in Marc André II solution after clearing in lactic acid and were studied using a Nikon E600 and SMZ-1000 microscope. Photographs were en- hanced with Photoshop CS2/PC (Adobe Inc.). The number of macrochaetae is given by half-tergite in the descriptions. Dorsal cephalic chaetotaxy and interocular chaetae follow Mari-Mutt (1979, 1986. Types are deposited in the collections of the Department of Entomology, College of Plant Protection, Nanjing Agricultural University (NJAU), P. R. China.
Etymology. Named after the former member Ms L. Liu in our lab, who initiated the study of the genus in China.
Ecology. In litter or on leaves of forest floor. Remarks. This new species belongs to sundanensis-group according to Mari-Mutt(1979). It is the only member with 1+1 inner mac on Abd. III in sundanensisgroup. It is most similar to D. wangi in labrum, cephalic chaetotaxy, trochanteral organ, tergal chaetotaxy of thorax and Abd. I-II, ventral tube, and claw structure. It differs from the latter in having a shorter Ant. V, 9 posterior cephalic mac, 1+1 inner mac on Abd. III, 2-4 chaetae on tenaculum, and the absence of dental spines (Table 1). S-chaetotaxy is also described in the genus for the first time: ms 1, 0|1, 0, 1, 0, 0 and sens 2, 2|1, 3, 3, ?, 4. -Mutt (1979) considered that the closest relatives of Dicranocentrus were Orchesella and Dicranorchesella because the three genera shared 6-segmented antennae. None questioned the systematic position of Dicranocentrus before the work of Zhang et al. (2014a), who also discussed the disputable use of secondary and unstable structures (such as number of antennal segments and number of chaetae on trochanteral organ) during development in modern taxonomy. Taking no account of antennae, Dicranocentrus shares most characters with Heteromurus: the presence of the same type of body scales, pigment reduced or scattered on the body, relatively fewer tergal macrochaetae, dental spines often present, and 3 ordinary S-chaetae on Abd. II/III (see also H. nitidus, Szeptycki 1979). An additional middle ordinary S-chaeta compared to those on species belonging to the Entomobryini/Willowsiini implies that both Dicranocentrus and Heteromurus are possibly closer to the Entomobryinae sensu Szeptycki, 1979, than previously thought. The pattern of four S-chaetae on Abd. V in Dicranocentrus is also similar to that of Heteromurus with the latter lacking the middle one. Compared to Dicranocentrus/Heteromurus, Orchesella/Orchesellides have much more abundant macrochaetae on each tergum and S-chaetae (usually >5) on Abd. II, III and V. The idea that the presence of body scales is a synapomorphy in Dicranocentrus/Heteromurus was strongly supported by molecular phylogeny, although body scales cannot be assumed to be a synapomorphy of the scaled genera (Willowsiini) of Entomobryinae by Zhang et al. (2014b). Body scales have been used successfully to define many groups, such as the Tomoceridae, Oncopoduridae, Seirinae, Lepidocyrtinae, and Cyphoderinae. Mari-Mutt (1979) proposed that Dicranocentrus originated from Orchesella via an intermediate stage represented by Dicranorchesella (with short ciliated chaetae and scales present). Mari-Mutt (1979) stated that Dicranorchesella, which has abundant cephalic and tergal macrochaetae, is quite close to Orchesella. However, of the pointed and fusiform scales of Dicranorchesella indicate that it represents a lineage independently derived from Dicranocentrus; its relationship with Orchesella possibly resembles that of Willowsia/Entomobrya as shown by Zhang et al. (2014a, b). A systematic review and phylogeny based on larger samples would ultimately resolve the systematic position of Dicranocentrus.