Molecular phylogenetic analysis of the Amiota taurusata species group within the Chinese species, with descriptions of two new species.

The relationships among six species of the Amiota taurusata Takada, Beppu, & Toda (Diptera: Drosophilidae) species group were investigated based on DNA sequence data of the mitochondrial NADH dehydrogenase subunit 2 ( ND2) gene, using three species of the genus Amiota as outgroups. A mitochondrial gene, cytochrome c oxidase I ( COI), can be used to discriminate between species of the taurusata group. Two new species are described from South China: A. protuberantis Shao et Chen, sp. nov. and A. shennongi Shao et Chen, sp. nov. A key to all the species of the taurusata group based on morphological characters is provided. This is an open access paper. We use the Creative Commons Attribution 3.0 license that permits unrestricted use, provided that the paper is properly attributed.

Introduction

The Amiota taurusata Takada, Beppu, & Toda (Diptera: Drosophilidae) species group was established by Chen and Toda (2001) based on a phylogenetic analysis using 31 adult male morphological characters. Until now, eight species have been reported in this group from East Asia ( Chen and Toda 2001 ; Chen et al. 2004 , 2005; Cao et al. 2008 ): A. aquilotaurusata Takada et al . , 1979, A. asymmetricaChen et Takamori, 2005 ; A. femorataChen et Takamori, 2005 , A. sacculipes Máca et Lin, 1993, A. spinifemora Li et Chen, 2008, A. taurusata Takada et al . , 1979, A. vulnerablaChen et Zhang, 2004 , and yixiangensisChen et Takamori, 2005 . Chen and Toda (2001) regarded the taurusata group as monophyletic on the basis of the hind femur basoventrally with a small, lobe-like flap (ch. 1; Figure 2 D in Chen and Toda 2001 ); hind tibia apicodorsally much extended flap (ch. 2; Figure 2 D in Chen and Toda 2001 ); hind first tarsomere dorsally expanded (ch. 3; Figure 2 D in Chen and Toda 2001 ); fourth tergite laterally broadened and protruded more than others (ch. 4; Figure 1 B in Chen and Toda 2001 ). However, Chen et al. (2004 , 2005) and Cao et al. (2008) found that the ch. 2 and ch. 3 are usually absent in some species; these two characters have been eliminated from the diagnosis criteria of the taurusata group.

Figure 2.

Amiota shennongi Shao et Chen, sp. nov. ♂: (A) Epandrium (epand) and circus, lateral view; (B) surstylus (sur) and tenth sternite (st 10), ventral view; (C) hypandrium and gonopod, ventral view; (D, E) paramere(s), aedeagus, and aedeagal apodeme, ventral and lateral views. Scale bars: 0.1 mm. High quality figures are available online.

Figure 1.

Amiota protuberantis Shao et Chen, sp. nov. ♂: (A) Epandrium (epand) and circus (cerc), lateral view; (B) surstylus (sur) and tenth sternite (st 10), ventral view; (C) hypandrium and gonopod, ventral view; (D) paramere, aedeagus, and aedeagal apodeme, lateral view. Scale bars: 0.1 mm. High quality figures are available online.

Recently, a molecular approach was used to uncover the relationship among the species in Stegana ( Li et al. 2010 ; Lu et al. 2011a , b), Phortica ( He et al. 2009b ; Cao et al. 2011 ), and Paraleucophenga ( Zhao et al. 2009 ), which are from genera of the subfamily Ste-ganinae. However, few related studies have been carried out in the genus Amiota . Chen and Toda’s (2001) phylogenetic analysis of the subgenus Amiota (currently the genus Amiota ) included the three species of this group mentioned above, the taurusata group, which is closely related to the apodemata , the nagatai , and the sinuata groups, but the relationships within this group were not resolved at all. In the present study, two new species of the taurusata group from China are described, and the relationships among the four known and two new species were investigated based on the DNA sequences of the mitochondrial NADH dehydrogenase subunit 2 ( ND2 ) gene. Barcoding information on the mitochondrial cytochrome c oxidase I ( COI ) genes of most of the species is provided.

Materials and Methods

Materials

All materials were collected from tree trunks or around human eyes and preserved in 75% ethanol. A small piece of tissue was removed from the fly abdomen and used for the DNA extraction. The body and terminalia were dried and deposited in the Department of Entomology, South China Agricultural University, Guangzhou, China (SCAU). The definitions of measurements, indices, and abbreviations follow Zhang and Toda (1992) and Chen and Toda (2001) .

The information on the samples used in the molecular phylogenetic analyses is given in Table 1 . Six species of the taurusata group were employed in the molecular phylogenetic analysis, and three Amiota species of the genus Amiota from the apodemata , nagatai , and sinuata groups were used as outgroups.

Table 1.

Data on samples for DNA sequencing and t he accession numbers of the ND2 and COI sequences.

1) He et al. 2009a ; 2)mistaken as EU431907 in He et al. 2009a .

Abbreviations

4c, third costal section between R2+3 and R4+5/M1 between r-m and dm-cu; 4v, M1 between dm-cu and wing margin/M1 between r-m and dm-cu; 5x, ac, third costal section between R2+3 and R4+5/fourth costal section; adf, longest dorsal branch of arista/width of first flagellomere; arb, dorsal branches/ventral branches of arista; avd, longest ventral branch/longest dorsal branch of arista in length; BL, body length; C, second costal section between subcostal break and R2+3/third costal section between R2+3 and R4+5; C3F, length of heavy setation in third costal section/length of the third costal section ch/o, maximum width of gena/maximum diameter of eye; CuA1 between dm-cu and wing margin/dm-cu between M1 and CuA1; dcl, anterior dorsocentral/posterior dorsocentral in length; dcp, length distance between ipsilateral dorsocentrals/cross distance between anterior dorsocentrals; flw, length/width of first flagellomere; FW/HW, frontal width/head width; M, CuA1 between dm-cu and wing margin/M1 between r-m and dm-cu; orbito, distance between proclinate and posterior reclinate orbitals/distance between inner vertical and posterior reclinate orbital; presctl, prescutellar/posterior dorsocentral in length; prorb, proclinate orbital/posterior reclinate orbital in length; rcorb, anterior reclinate orbital/posterior reclinate orbital in length; sctl, basal scutellar/apical scutellar in length; sctlp, distance between ipsilateral scutellars/cross distance between apical scutellars; sterno, anterior katepisternal/posterior katepisternal in length; THL, thorax length; vb, subvibrissal/vibrissa in length; WL, wing length; WW, wing width.

DNA Extraction and Sequencing

Total DNA was extracted using a DNA extraction Kit (TIANGEN, www.tiangen.com ) according to the manufacturer’s protocol. The ND2 and COI fragments were amplified with the primers listed in Table 2 . The PCR reactions consisted of an initial 4 min pre-denaturation at 94°C, followed by 30 cycles (30 sec of denaturation at 94°C, 1 min of annealing at 54°C for ND2 and at 49°C for COI , and 1 min of extension at 72°C), and a final elongation for 5 min at 72°C. When possible, purified amplified products were directly run on an ABI 3730 sequencer; otherwise, they were cloned into the pMD18-T vector (TAKARA, www.takarabio.com ) and then sequenced. The related ND2 sequences of A. natagai , A. planate , and A. sinuata were retrieved from the National Center for Biotechnology Information (NCBI).

Table 2.

Primers used for PCR and sequencing.

Phylogenetic analyses

The sequences were aligned by the Clustal W ( Thompson et al. 1994 ) method in MEGA 4.0 ( Tamura et al. 2007 ) with the default options and then adjusted manually. Because the substitution saturation masked the phylogenetic signal ( Lopez et al. 1999 ; Philippe and Froterre 1999 ), the method of Xia et al. (2003) was used to test the nucleotide substitution saturation in the program DAMBE 5.0.80 ( Xia and Xie 2001 ). The base compositions of these sequences were investigated using PAUP* version 4.0b10 ( Swofford 2001 ), and the c 2 test was used to evaluate the nucleotide composition homogeneity among them. Uncorrected p distance among taxa was estimated by MEGA 4.0 ( Tamura et al. 2007 ).

Phylogenetic relationships were constructed using the Bayesian inferring (BI) method in MrBayes 3.2.1 ( Huelsenbeck and Ronquist 2001 ; Ronquist and Huelsenbeck 2003 ). In the BI analyses, the data were partitioned by locus (1 data partition) and codon positions (3 data partitions). The nucleotide substitution models of BI analyses were selected by Modeltest 3.7 using the hierarchical likelihood ratio test (hLRT) criterion ( Posada and Crandall 1998 ). Two independent runs with 2,000,000 generations were implemented in parallel, and a sampling frequency of every 100 generations was employed. When the average deviation of split frequencies fell well below 0.01, the two runs were stopped. For each run, the 5,000 early-phase samples were discarded, and the remainder of the samples were used. A majority rule tree showing all the compatible partitions was obtained.

Nomenclature

This publication and the nomenclature it contains have been registered in ZooBank. The LSID number is: urn:lsid:zoobank.org:pub:60353BF6-3506-4286-A8E9-FB72847CD3D9. It can be found online by inserting the LSID number after www.zoobank.org/ .

Results

Amiota taurusata species group

Diagnosis

Hind femur with small, lobe-like flap basoventrally; fourth tergite laterally broadened and protruded more than others (modified from Chen and Toda 2001 ; Figures 1B , 2D ). In the new species described, only characters that depart from the universal description (given by Chen and Toda (2001) and Chen et al. (2004 , 2005) for the subgenus Amiota ) are provided for brevity.

Amiota protuberantis Shao et Chen, sp. nov.( Figure 1 )

This species very similar to A . femorataChen et Takamori, 2005 in hind tibia distinctly expanded on subapical part of dorsal surface; aedeagus bifurcated on basal 1/2, submedially slightly curved dorsad, separated from parameres in lateral view ( Figure 1 D).

Description

Only important characters are given. Male and female: Frons, face, and clypeus nearly black. Ventral branches of arista distinctly shorter than 1/3 of dorsals in male, slightly shorter than 1/2 of dorsals in female. Palpus brown. Legs yellow except for dark brown on all femora in female or dark brown on femora of fore-and midlegs and distal half of hind femur in male. Male hindleg apicodorsally much extended like flap on tibia and dorsally slightly expanded on first tarsomere. Epandrium small, constricted more than 1/2 width mid-dorsally, with ca. 17 setae near posterior to ventral margins on each side ( Figure 1 A). Surstylus distally with numerous setae on outer surface and ca. 7 prensisetae ( Figure 1 B). Vertical lobe of gonopod apically round, without any processes. Parameres fused on basal 3/4, slightly sclerotized, as long as aedeagus, with ca. 9 sensilla subbasally ( Figure 1 C, D).

Measurements

BL = 3.08 mm in the holotype (range in 3 ♂ and 2 ♀ paratypes: 2.60–3.16 mm in ♂ , 3.32– 3.40 mm in ♀ ), THL = 1.76 mm (1.56–1.72 mm in ♂ , 1.60–1.64 mm in ♀ ), WL = 2.60 mm (2.24–2.60 mm in ♂ , 2.80–2.88 mm in ♀ ), WW = 1.20 mm (1.00–1.20 mm in ♂ , 1.20– 1.32 mm in ♀ , arb = 6/4 (6/4–5), avd = 0.33 (0.33–0.57), adf = 1.00 (0.88–1.17), flw = 1.83 (1.50–2.00), FW/HW = 0.39 (0.36–0.44), ch/o = 0.08 (0.08–0.10), prorb = 0.94 (0.69– 1.00), rcorb = 0.50 (0.50–0.85), vb = 0.57 (0.57–0.63), dc1 = damaged, presct1 = 0.44 (0.44–0.60), sct1 = 1.20 (1.20–1.26), sterno = 0.78 (0.60–0.95), orbito = 1.00 (1.00–1.40), dcp = 0.38 (0.30–0.41), sct1p = 0.92 (0.72– 1.33), C = 2.05 (1.83–3.08), 4c = 1.73 (0.93– 1.78), 4v = 3.00 (2.50–3.00), 5x = 1.33 (1.16– 2.00), ac = 3.25 (3.25–5.33), M = 0.64 (0.50– 0.77), and C3F = 0.78 (0.78–0.83).

Types

Holotype ♂ (SCAU, No. 121088), CHINA: Mt. Wuliang, Jingdong, Yunnan, 18°41’N, 108°52’E, altitude 2200 m a.s.l., 4.viii.2006, T Li. Paratypes: 3 ♂ , 2 ♀ (SCAU, No. 121089– 93), same data as the holotype.

Etymology

From the Latin word protuberantis, referring to the hindleg tibia expanded on subapical part of dorsoposterior surface.

Distribution

China (Yunnan).

Amiota shennongi Shao et Chen, sp. nov. ( Figure 2 )

This species very similar to A . aquilotaurusata Takada, Beppu et Toda, 1979 in that it has the same shape of the male terminalia. It differs by having the short process of aedeagus longer than 1/2 of long one ( Figure 2 D, E), the paramere thick rod-like, not expanded ( Figure 1 D, E).

Only important characters are given in here. Male: Frons, face, and clypeus nearly dark brown. Ventral branches of arista distinctly shorter than 1/3 of dorsals in male. Palpus brownish yellow. Legs entirely yellow; hind leg: tibia apicodorsally much extended flap, and first tarsomere dorsally expanded ( Chen and Toda 2001 ; Figure 2 D). Epandrium entirely separated into two lateral lobes, with about 15 setae near posterior to ventral margins per site ( Figure 2 A). Surstylus lacking pubescence, with finger-like process at posteroventral corner, and about nine prensisetae on distal margin ( Figure 2 B). Tenth sternite deeply constricted mid-ventrally, but not separated, entirely fused to surstyli laterally ( Figure 2 B). Anterior portion of hypandrium slightly broadened ( Figure 2 C). Aedeagus basally fused to paramere and deeply bifurcated, two processes of aedeagus nearly equilong ( Figure 2 D, E). Parameres slightly longer than aedeagus, round apically and expanded basally ( Figure 2 D, E).

BL = 2.88 mm in the holotype (3.00 mm in 1♂paratype), THL = 1.16 mm (1.27 mm), WL = 2.14 mm (2.44 mm), WW = 0.96 mm (1.24 mm), arb = 7/5 (5/4), avd = 0.29 (0.33), adf = 1.40 (1.20), flw = 2.40 (2.00), FW/HW = 0.46 (0.38), ch/o = 0.34 (0.24), prorb = 1.08 (0.87), rcorb = 0.75 (0.66), vb = 0.43 (0.50), dc1 = damaged (0.6), presct1 = 0.28(0.50), sct1 = 1.23 (1.09), sterno = 1.50 (0.75), orbito = 1.40 (3.30), dcp = 0.32 (0.36), sct1p = 1.33 (1.33), C = 2.57 (1.67), 4c = 1.27 (2.10), 4v = 3.55 (3.40), 5x = 0.63 (0.75), ac = 3.50 (5.25), M = 0.80 (0.80), and C3F = 0.63 (0.59).

Holotype ♂ (SCAU, No. 121094), CHINA: Dajiuhu, Shennongjia, Hubei, 31°29’N, 110°18’E, altitude 1400 m a.s.l., 31.vii.2004, HW Chen. Paratype: 1 ♂ (SCAU, No. 121095), same data as holotype.

Patronym, the name of Yandi, who was a man in an old Chinese story.

China (Hubei).

Key to species of the taurusata group

1. Hind femur ventro-basally with nearly hyaline, small, lobe-like flap; fourth tergite laterally broadened and protruded more than others (the taurusata group). ………………………………………......2 Hind femur without any flap; fourth tergite neither broadened nor protruded more than others….............other Amiota species

2. Ventral branches of arista distinctly shorter than 1/2 of dorsals; all femora dark brown to black……………………….…...…….3 Ventral branches of arista as long as 1/2 of dorsals; all legs yellow….........................5

3. Hind first tarsomere not expanded dorsally………………………………….4 Hind first tarsomere expanded dorsally……….. A. sacculipes Máca et Lin

4. Vertical lobe of gonopod nearly triangular; aedeagus basally with 1 pair of slender processes… A. femorata Chen et Takamori Vertical lobe of gonopod nearly quadrate; aedeagus without any processes…………. ….. A. protuberantis Shao et Chen, sp. nov.

5. Hind tibia apicodorsally much extended like flap; hind first tarsomere expanded dorsally….................................................6 Hind tibia apicodorsally not extended like flap; hind first tarsomere not expanded dorsally………………………………….8

6. Short process of aedeagus shorter than 1/5 of long one……………………………….. …….. A. taurusata Takada, Beppu et Toda Short process of aedeagus slightly shorter or longer than 1/2 of long one..……….........7

7. Short process of aedeagus shorter than 1/2 of long one; paramere expanded to lobe-like….. A. shennong i Shao et Chen, sp. nov. Short process of aedeagus longer than 1/2 of long one; paramere expanded to lobe-like.. ………....A. aquilotaurusata Takada, Beppu et Toda

8.Parameres nearly entirely sclerotized; gonopod nearly triangular………………... ………...A. asymmetrica Chen et Takamori Paramereswith membranaceous part; gonopod nearly quadrate………………...9

9. Paramere entirely separated from aedeagus.. ……………A. yixiangna Chen et Takamori Paramere basally fused to aedeagus………... …………… A . vulnerabla Chen et Zhang

Molecular analysis

Data set analysis

The alignment of the sequences included 1026 base pairs for ND2 and 684 for COI. There were end gaps in the ND2 sequences of A. spinifemorata –GZ (sites 1–36), A. spinifemorata –YN2 (sites 1015–1026), and A. shennongi sp. nov. (sites 1015–1026). End gaps also existed in the ND2 sequences (sites 149–151) of A. femorata –HN, A. femorata – SC2, A. femorata –SX, A. femorata –YN2, and A. protuberantis sp. nov.–YN2 as well as in the ND2 sequences (sites 148–150) of A. femorata –SC1, A. femorata –YN2, A. femorata –XZ, and A. protuberantis sp. nov.–YN2. The COI sequences of some samples ( A. aquilotaurusata , A. asymmetrica –YN2, and A. femorata –HN) were not acquired, and end gaps existed in the COI sequences (sites 1–15) of A. plannata. There were 349 variable sites (of which 211 were parsimony informative sites) for ND2 and 178 variable sites (of which 106 were parsimony informative sites) for COI. The nucleotide composition of ND2 is shown in Table 3 . The sequences contained much higher AT content (82.8%) than GC content, especially at the third codon positions (95.4%). The c 2 test revealed that the nucleotide composition among the taxa was not het-heterogeneous.

Table 3.

Results of nucleotide composition and composition homogeneity test.

Regardless of whether the analysis was performed with the combined or separated codon position data for ND2 , the test of substitution saturation revealed that the observed substitution saturation index ( Iss ) was significantly lower than the corresponding critical substitution saturation index ( Iss.c ) for both the symmetrical and asymmetrical trees, indicating that there was little saturation in these sequences ( Table 4 ).

Table 4.

Results of substitution saturation tests and model selection.

Index of substitution saturation assuming a symmetrical true tree.

Probability of a significant difference between Iss and Iss.cSym (two-tailed test).

Index of substitution saturation assuming an asymmetrical true tree.

Probability of a significant difference between Iss and Iss.cAsym (two-tailed test).

Table 5 shows the uncorrected pairwise divergence for the ND2 and COI sequences in the taurusata group, excluding the p-distance for COI of A. aquilotaurusata , A. asymmetrica –YN1, and A. femorata –HN. The interspecific genetic divergence for ND2 in the taurusata group ranged from 0.0294 ( A. femorata vs. A. protuberantis sp. nov.) to 0.1049 ( A. aquilotaurusata vs. A. spinifemorata ), and for CO1 it ranged from 0.0207 ( A. femorata vs. A. protuberantis sp. nov.) to 0.0841 ( A. asymmetrica vs. A. protuberantis sp. nov.). The intraspecific genetic divergences for ND2 and COI were calculated for A. asymmetrica (0.0021 for ND2 ), A. protuberantis sp. nov. (0.0072 for ND2 , 0.0015 for COI ), A. spinifemorata (0.0041 to 0.0185 for ND2 , 0.0073 to 0.0088 for COI ), and A. femorata (0.0010 to 0.0474 for ND2 , 0.0000 to 0.0336 for COI ).

Table 5.

Uncorrected pairwise p-distance among the ND2 and COI sequences of the taurusata species group. The matrix in the lower left shows the uncorrected pairwise p-distance among the ND2 sequences; the matrix in the upper right shows the uncorrected pairwise pdistance among the COI sequences.

Phylogenetic analysis

The Bayesian tree for ND2 lent good support for the monophyly of the taurusata group with respect to the outgroups (posterior probabilities (PP) = 0.99) ( Figure 3 ). Samples from different geographical areas of A. protuberantis , A. asymmetrica , and A. spinifemorata clustered as a monophyletic lineage, while samples of A. femorata were rendered paraphyletic with respect to A. protuberantis . A. spinifemorata first diverged in the taurusata group and was then followed by A. asymmetrica . The other four species grouped into a robust supported group (PP = 1.00). A . shennongi sp. nov. and A . aquilotaurusata showed a sibling relationship (PP = 1.00) in agreement with their high similarity in morphological characters. Samples of A. femorata diverged into two highly supported clusters. One consisted of A. femorata (HN, SC1, SX, and XZ) (PP = 0.98); the other consisted of A. femorata (SC2 and YN1–2) (PP = 1.00) and clustered with A. protuberantis sp. nov. (PP = 1.00).

Figure 3.

Bayesian tree of the taurusata group deduced from the ND2 sequences using the site-specific mode numbers. These numbers are beside the nodes and are the posterior probabilities. High quality figures are available online.

Discussion

A phylogenetic tree of the taurusata group was constructed using mitochondrial ND2 sequences. As negative results were obtained in the tests of nucleotide composition heterogeneity and substitution saturation, the conclusions of the phylogenetic analyses should be accepted. The monophyly of the taurusata group was strongly supported in the molecular phylogenetic analyses, and the relationships within this group were almost resolved. However, the unstable position of A. asymmetrica and A. spinifemorata was not resolved, even when using a site-specific model for Bayesian inference. To fully resolve the phylogenetic relationship in the taurusata group, multiple loci or more species in the analyses are necessary.

The ND2 divergence matrix was provided for the taurusata group. The interspecific genetic divergence in the taurusata group ranged from 0.0294 to 0.1049, and the intraspecific genetic divergence ranged from 0.0010 to 0.0474. The geographical samples of A. asymmetrica , A. protuberantis sp. nov., and A. spinifemorata formed highly-supported monophyletic groups in the phylogenetic tree, and intraspecific genetic divergence within them was much less than interspecific genetic divergence in the taurusata group. In addition, no diagnostic morphological character was found to distinguish the geographical samples of these species, indicating that they should be considered conspecific. However, A. femorata diverged into two clusters, and classified characters in its morphology were missing. The three haplotypes of A. femorata , i.e., SC2, YN1, and YN2, clustered with A. protuberantis sp. nov. This relationship could be attributed to stochastic lineage sorting and/or hybridization. The genetic divergence for ND2 within the two clusters ranged from 0.0010 to 0.0474, and the mean divergence between them was 0.0392. Assuming the observed divergence range (0.0294 to 0.1049) reflects the real intraspecific variations in the taurusata group, there likely are cryptic species in A. femorata samples.

Recent work suggests that cytochrome c oxidase I ( COI ) might serve as a DNA barcode for the identification of animal species ( Brown et al. 2003 ; Foster et al. 2004; Barrett and Hebert 2005 ; Cardoso and Vogler 2005 ; Hogg and Hebert 2005; Monaghan et al. 2005; Vences et al. 2005 ; Ward et al. 2005 ). This gene region is easily recovered and provides good resolution, as evidenced by the deep sequence divergences among 13,000 closely related pairs of animal species (Hebert et al. 2003b). In this study, a 684 bp region of COI was acquired, and it showed that COI differences between most of the species far exceeded those within species. The interspecific genetic divergence in the taurusata group ranged from 0.0207 to 0.0841, and the intraspecific genetic divergence ranged from 0.0000 to 0.0336. An overlapping area existed between the intraspecific and interspecific genetic divergence. The intraspecific genetic divergences within A. protuberantis sp. nov. (0.0015) and A. spinifemorata (0.0073 to 0.0088) were much lower than the minimum interspecific genetic divergence (0.0207) and the mean intraspecific variability for Diptera (1.3 ± 1.6%) ( Meier 2008 ), indicating that they should be considered conspecific. This result is consistent with the ND2 result and the morphology analysis. The observed divergence of the two clusters of A. femorata was 0.0322, which is greater than the minimum interspecific genetic divergence but lower than the minimum interspecific genetic divergence for Diptera (5.9 ± 4.1%) ( Meier 2008 ). A. protuberantis sp. nov. was identified as the sister-species of A. femorata , and the p-distance ranged from 0.0146 to 0.0263. Because of the limit coming from the number and distribution of samples, there likely are cryptic species in the two clusters, which is consistent with the ND2 result. It is important to include samples from a wider geographical range in future studies to determine if the two clusters represent morphologically cryptic species. Further samples are also needed for an evaluation of the morphological variability revealed in the results.

Biogeographical implications

All the members of A. spinifemorata and A. asymmetrica were found in southwestern China. According to the phylogenetic analyses, the two species diverged from the taurusata group prior to A. shennongi sp. nov. and A. aquilotaurusata , which were found in central and northeast China. It may indicate that the founder of the taurusata group arose in southwestern China, undergoing some differentiation before the expansion into the central and northern areas. In the zones of low and high elevation, A. femorata can be distributed between one cluster ( A. femorata –HN, SC1, SX, excluding A. femorata –XZ) mainly at low elevations (ca. 300–500 m a.s.l.) and another cluster ( A. femorata –SC2, YN1, and YN2) at high elevations (ca. 1700–2700 m a.s.l.) that clusters with A. protuberantis sp. nov. This result may indicate that some individuals underwent heteromorphosis to different extents following the expansion of A. femorata from low elevations into high elevations, and that then A. protuberantis sp. nov. was found.