Evolutionary History of Apocheima Cinerarius (Lepidoptera: Geometridae), a Female ﬂightless Moth in Northern China

Zoologica Scripta

Evolutionary history of Apocheima cinerarius (Lepidoptera: Geometridae), a female ﬂightless moth in northern China

SHUXIAN LIU,NAN JIANG,DAYONG XUE,RUI CHENG,YANHUA QU,XINXIN LI,FUMIN LEI & HONGXIANG HAN

Submitted: 6 May 2015 Liu, S. X., Jiang, N., Xue, D. Y., Cheng, R., Qu, Y. H., Li, X. X., Lei, F. M. & Han, H. X. Accepted: 30 September 2015 (2016). Evolutionary history of Apocheima cinerarius (Lepidoptera: Geometridae), a female doi:10.1111/zsc.12147 flightless moth in northern China. — Zoologica Scripta, 45, 160–174. The alterations in the phylogeographical structures of insects in response to the uplift of the Qinghai–Tibet Plateau and the Quaternary glaciations in eastern Asia, particularly in northern China, remain largely unknown. In this study, we selected Apocheima cinerarius, a moth with flightless females, using molecular data (complete mitochondrial genomes and nuclear data) and ecological niche modelling (ENM) to investigate the effects of paleoclimatic changes on the evolutionary history of insects in the area of northern China. The phylogenetic tree of complete mitochondrial genomes indicated that there were two lineages, the western and eastern lineages. The nuclear gene analyses also detected unique haplotypes in each lineage. Time of the most recent common ancestor (TMRCA) of the two lineages was approximately in Early–Middle Pleistocene. Bayesian skyline plots revealed that the western lineage underwent a population expansion event after the Last Glacial Maximum, whereas the eastern lineage underwent expansion between the Last Interglacial and the Last Glacial Maximum. Our results suggest that A. cinerarius expanded eastward from western sites until the moth was distributed across the entire region of northern China. Then, A. cinerarius underwent contraction into isolated glacial refugia followed by subsequent expansion driven by Pleistocene climate changes, which established a narrow sympatric area. Our results indicate that the Quaternary environmental fluctuations had profound influences on the diversification and demography of an insect in northern China, and the same species in north- western China and north-eastern China have different demographic histories. Corresponding author: Hongxiang Han and Fumin Lei, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing 100101, China. E-mails: [email protected]; [email protected] Shuxian Liu, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China and University of the Chinese Academy of Sciences, Beijing, 100049 China. E-mail: [email protected] Nan Jiang, and Dayong Xue, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China. E-mail: [email protected], xuedy@ ioz.ac.cn Rui Cheng, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China and University of the Chinese Academy of Sciences, Beijing, 100049 China. E-mail: [email protected] Yanhua Qu, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China. E-mail: [email protected] Xinxin Li, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China and University of the Chinese Academy of Sciences, Beijing, 100049 China. E-mail: [email protected] Fumin Lei, and Hongxiang Han, Key Laboratory of Zoological Systematics and Evolution, Institute of Zoology, Chinese Academy of Sciences, Beijing, 100101 China. E-mail: [email protected], hanhx@ ioz.ac.cn Shuxian Lu and Nan Jiang contributed equally to the present paper.

160 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius

Introduction in climatic conditions during the Pleistocene (Haring et al. Animal phylogeography and patterns of distribution related 2007; Zink et al. 2008, 2009; Olsson et al. 2010; Zhang to geological events and climate change in the temperate et al. 2012). In addition, several studies also indicated that a zone of the Northern Hemisphere have long fascinated geographical gap between the eastern and western ranges biologists and naturalists. The uplift of the Qinghai–Tibet of the azure-winged magpie (Cyanopica cyanus) might be the Plateau (QTP) was the largest geological event in the east- subdivision of a former continuous range during the Pleis- ern Asia and caused dramatic climatic and ecological shifts tocene (Cooper 2000; Palomino et al. 2011). Comparative (Zhang 2000). The uplift of the QTP and the Quaternary phylogeography of two widespread magpies (Cyanopica cya- glaciations are key factors that influence the phylogeo- nus and Pica pica) indicate that bird species with different graphical structure of species in eastern Asia (Cao et al. life histories (e.g. habitat preference and breeding beha- 2012; Lei et al. 2014). viour) might respond differently to fluctuations in the envi- Previous studies have shown that the uplift of the QTP ronment (Zhang et al. 2012). from 1.7 to 3.6 Ma changed the surrounding geography Apocheima cinerarius (Erschoff) is a geometrid moth with and environment, which facilitated the speciation of many wingless females and is distributed from Maracanda east- species in eastern Asia (Li et al. 2001; Qu et al. 2009, 2010; ward across the entire region of northern China to the Cao et al. 2012; Lei et al. 2014). The uplift of the QTP Amur district. This area is in an arid climate zone in the increased the desertification of northern China, and this East Asia monsoon region. The adults of A. cinerarius process had a large effect on the biodiversity of the area emerge from overwintered pupae in soils between the end (Shi et al. 1998; Liu et al. 2001). For example, desertifica- of February and April, when the winter monsoon is strong tion shaped the current biodiversity and distribution pat- and the temperature is very low (Chu 1981; Sattler 1991). tern of a perennial herb (Lagochilus ilicifolius) (Meng & The weak ability for movement, wide distribution and Zhang 2011). The uplift of the plateau also triggered the unique life history make this inactive litter-layer insect an onset of the monsoon circulation, which intensified begin- excellent model organism to study the phylogeographical ning 2.6 Ma, particularly within 0.9–1.2 Ma (Sun et al. pattern and to test the demographic hypothesis of the 2003). However, the effects of the monsoon circulation on insects in northern China. the phylogeographical patterns of the local fauna remain In this study, we combined the mitochondrial genomes unknown. (13 PCGs + 2 rRNAs) and two nuclear genes [translation The Quaternary glaciations led to the extinction or elongation factor 1-alpha (EF1a) as well as the second migration of species, with some surviving in refugia during internal transcribed spacer (ITS2)] and integrated with an the glacial periods and subsequently expanding after the ecological niche model to analyse the phylogeographical glaciation, thus contributing to the current genetic struc- pattern of A. cinerarius. The aims of the study were (i) to ture and distribution (Hewitt 1999, 2000, 2004; Lin et al. investigate the phylogeographical structures of this moth; 2014). Previous studies revealed that the demographic his- (ii) to infer the possible origins, migration routes and tories of most European species are ‘post-LGM expansion’ expansion force of this moth; and (iii) to test whether the mode (Hewitt 1999, 2000, 2004; Kvist et al. 2003). How- demographic history of this moth corresponds to the ‘pre- ever, East Asia, including China, Japan, the Korean Penin- LGM expansion’ mode like other eastern Asian animals. sula and Mongolia, was not completely covered with a widespread ice sheet, which is different from North Amer- Materials and methods ica and Europe (Rost 1994; Ju et al. 2007; Igea et al. 2013). Specimen sampling Many East Asian birds have been hypothesized to undergo We searched for adult specimens on the appropriate host a pre-glacial population expansion, thus considered to fol- plants (Populus, Salix and Elaeagnus). In total, 211 adult low a ‘pre-LGM expansion’ mode (Dai et al. 2011; Qu specimens of A. cinerarius from 31 sampling sites were col- et al. 2011; Zhao et al. 2012; Wang et al. 2013a). For spe- lected across the entire range of this species in China cies that were widely distributed on the Eurasian continent, (Table S1). The specimens were preserved in anhydrous multiple isolated glacial refugia occurred in their area of ethanol at À20 °C and were deposited at the Institute of distribution during the Quaternary glaciations. These refu- Zoology, Chinese Academy of Sciences, Beijing, China gia would have facilitated the differentiation of species (IZCAS). (Qian & Ricklefs 2000; Hu et al. 2013). Previous studies revealed that many bird species distributed across the Eura- DNA extraction, PCR amplification and sequencing sian continent have concordant patterns of ‘western vs. The DNA was extracted using the Qiagen DNeasy Blood eastern lineage divergence’ that resulted from the changes & Tissue Kits (Qiagen, Beijing, China). In all, 27 mitogen-

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 161 Evolutionary history of Apocheima cinerarius S. Liu et al. omes from representative sampling sites were obtained with Bank (Accession numbers listed in Table S2). We obtained polymerase chain reactions (PCR) for the short and long 2795 bp of combined mitochondrial genes from 211 sam- fragments. The primers and PCRs were described in Liu ples from 31 sampling sites (populations) for A. cinerarius, et al. (2014). which included the COI (592 bp), Cytb (934 bp) and ND5 Based on the 27 mitochondrial genomes of A. cinerarius, (1269 bp) genes. For the nuclear data, a 903-bp EF1a frag- we calculated the number of parsimony informative sites of ment was sequenced from 171 individuals, and a 334-bp 15 genes (13 PCGs: ND2, COI, COII, ATP8, ATP6, COIII, ITS2 fragment was obtained from 63 individuals (2–3 indi- ND3, ND5, ND4, ND4L, ND6, Cytb, ND1; 2 rRNAs: 12s viduals from each sampling site) (GenBank accession num- rRNA and 16s rRNA) using the DnaSP 5.0 (Librado & bers showed in Table S5). Rozas 2009). The results indicate that three protein coding genes: cytochrome oxidase I (COI), cytochrome b (Cytb) Population genetic analysis and NADH dehydrogenase subunits 5 (ND5) had the most The numbers of haplotypes (H), and the values of haplo- parsimony informative sites in the mitogenomes. Subse- type diversity (h) and nucleotide diversity (p) (Nei 1987) quently, the three protein coding genes were chosen for for every sampling site, each lineage and all populations further phylogeographical analyses. were calculated in DnaSP 5.0 (Librado & Rozas 2009). We The PCRs of these three genes were run for all samples used two types of neutrality tests, Tajima’s D and Fu’sF. using the following primer pairs: COI with LepF1/LepR1 An analysis of molecular variance (AMOVA) (Excoffier et al. (Hebert et al. 2004), Cytb with CP1/TRs (Simon et al. 1992) was performed to investigate the genetic structure of 1994; Sezonlin et al. 2006) and ND5 with LIU-ND5-F/ the populations. These methods were implemented in Arle- LIU-ND5-R (This study) (Table S4). The PCRs were per- quin 3.5 (Excoffier & Lischer 2010). formed in a 25-lL volume with 0.125 lL Taq DNA polymerase (5 U/lL, Takara), 1 lL of DNA, 2.5 lL109 Taq Phylogenetic analysis buffer (plus Mg2+), 2 lL dNTPs and 10 pmol of each pri- The phylogenetic relationships within A. cinerarius were mer. The PCRs were performed using the following proce- reconstructed using 27 complete mitogenomes (13 dure: 94 °C for 2 min; 5 cycles of 40 s at 94 °C, of 40 s at PCGs + 2 rRNAs) using Bayesian inference (BI) and maxi- 45 °C and of 1.5 min at 72 °C; and 35 cycles of 40 s at mum likelihood (ML) inference. Biston panterinaria (Yang 94 °C, of 40 s at 51 °C and of 1.5 min at 72 °C, followed et al. 2013) and Phthonandria atrilineata (Yang et al. 2009) by 72 °C for 10 min (Hebert et al. 2004). were used as out-groups. The overall Bayesian analysis was The maternal mode of mitochondrial inheritance means performed using MrBayes v3.1.2 (Ronquist & Huelsenbeck that mtDNA-based phylogeography only reflects the his- 2003). The data were divided into 15 partitions (13 PCGs torical processes in females (Ma et al. 2012). The nuclear and 2 rRNA genes) and 41 partitions (13 PCGs partitioned genes are biparental inheritance, reflecting historical pro- by codon position and 2 rRNA genes partitioned by gene). cesses in males and females. Therefore, we amplified the Different models were used for different partitions, which nuclear genes for a subset of individuals with the following were selected from 88 models using the Akaike information primer pairs: EF1a with EF1aLepF2/EF1aR (Kawakita criterion (AIC) using jModelTest 0.1 (Posada 2008). An et al. 2004; Yamamoto & Sota 2007); ITS2 with MCMC analysis was run for 10 000 000 generations, fol- BMRRN58S/CAS28sB1d (Ji et al. 2003) (Table S4). The lowing a burn-in series of 1000 generations. A maximum PCRs were performed using the following procedure: likelihood (ML) analysis was conducted using the program 94 °C for 4 min; 35 cycles of 30 s at 94 °C, of 1 min at RAxML v.7.0.4 (Stamatakis 2006) with 1,000 bootstraps 58 °C and of 1.5 min at 72 °C, followed by 72 °C for conducted. The GTRGAMMA model was used for all 10 min. The PCR products were detected with 1% agarose genes and partitions. gel electrophoresis and were directly sequenced with ABI To determine the geographical origin of A. cinerarius, PRISM 3730 xl capillary sequencers. we constructed a BI phylogenetic tree based on the COI The complete mitogenomes were assembled from the gene using MrBayes v3.1.2, with the European Apocheima overlapping short sequences with ChromasPro (www.tech- hispidaria (Denis & Schiffermuller),€ selected as the out- nelysium.com.au/ChromasPro.html), and the open reading group, as it is the only other species in the genus Apoc- frames (ORFs) were identified using the ORF Finder heima. Only COI was used in the BI tree as it was the only (http://www.ncbi.nlm.nih.gov/gorf/orfig.cgi) at NCBI. The sequence available from the NCBI (GU580757.1, nucleotide sequences were aligned with the known mito- GU654879.1) for A. hispidaria (Wahlberg et al. 2010; genome sequences of closely related species using CLUS- Hausmann et al. 2011). TAL X 1.83 (Thompson et al. 1997). About 27 complete A median-joining (MJ) network was implemented in mitogenomes have been obtained and deposited in Gen- Network 4.6 (Bandelt et al. 1999) based on the COI gene,

162 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius as well as concatenated set of COI + Cytb + ND5 and Ecological niche modelling nuclear gene EF1a. The ancestral node of the network of Data for the occurrence of A. cinerarius were gathered the COI gene was determined by rooting the network with from sample records and documents. A total of 100 locali- the out-group A. hispidaria. The ‘external rooting: active’ ties were used for the final analysis. Bioclimatic variables parameter was chosen for the MJ network calculation. were downloaded from WordClim (version 1.4; http:// www.worldclim.org/). The environmental layer of the Last Estimation of divergence times Glacial Maximum (LGM) was built based on the Commu- Time of the most recent common ancestor (TMRCA) of nity Climate System Model (CCSM), at a resolution of the two clades was estimated using a molecular clock in 2.5-min. The environmental layer of the Last Interglacial BEAST v 2 (Bouckaert et al. 2014) based on the COI gene (LIG) was transformed to a resolution of 2.5-min with under the GTR + I substitution model selected by jModel- ARC GIS 9.3 software. test, with A. hispidaria as the out-group. Although the use We predicted the suitable habitats and potential distribu- of a molecular clock as the only way for calibrating phylo- tions of A. cinerarius in three periods: the present day, the genetic trees is controversial, it does provide a method for Last Glacial Maximum and the Last Interglacial. Ecological estimating approximate divergence times when no other niche models were performed for A. cinerarius in MAX- calibration information, such as fossil or geological evi- ENT (Phillips et al. 2006). We set 80% of the sites to train dence, is available (Maekawa et al. 2001; Lopez-L opez et al. the model and 20% of the sites to test the model predic- 2015). Thus, in this study, the proposed conventional tions. We set the number of maximum iterations to 2,000, mutation rates for insect mitochondrial COI gene 2.3% and the number of replicates was 10. The assessment index and 3.54% per million years were used (Brower 1994; of the ecological niche model prediction included the area Papadopoulou et al. 2010). The exponential relaxed clock under the receiving operator characteristics curve (AUC) was selected for the clock model. Coalescent constant pop- and the binomial probabilities. A value of AUC﹥0.5 indi- ulation model was selected for the tree prior. MCMC was cates that the predicted results of the model are better than set for 200 000 000 steps, sampling every 20,000 steps. those predicted stochastically. The values of AUC﹥0.9 The output was tested in Tracer v 1.6 (Rambaut et al. indicate excellent predictive power. 2013). The effective sample size (ESS) of each parameter was required to be higher than 200. The final tree was Results summarized using the program TreeAnnotator included in Population genetic diversity BEAST v 2. We found 86 parsimony informative sites and 91 haplotypes using the concatenated set of COI + Cytb + ND5 Historical demography genes. The eastern lineage (66) had more haplotypes than The history of the population dynamics of the two lineages the western lineage (30). The haplotype diversity was high of A. cinerarius (see Results) was estimated using Bayesian among all populations (h = 0.971), and the nucleotide skyline plots (BSP) (Drummond et al. 2005) in BEAST v 2 diversity was low (p = 0.00852). The values of haplotype (Bouckaert et al. 2014). The mtDNA alignment for diversity (h) and nucleotide diversity (p) were 0.886 and COI + Cytb + ND5 was used for the historical demography 0.00471 for the western lineage (Clade A), respectively, analyses. We used conventional mutation rates for insect and were 0.984 and 0.00209 for the eastern lineage (Clade mitochondrial COI gene 2.3% and 3.54% per million years B), respectively. The highest nucleotide diversity was (Brower 1994; Papadopoulou et al. 2010). The mtDNA found at sites ZY (0.00693) and JC (0.00623), and the alignment for COI + Cytb + ND5 genetic distance was lowest was from sites JQ (0.00010), HT (0.00012), DH 0.736-fold the COI genetic distance. We multiplied the (0.00021) and QM (0.00022) (Table S3). Nucleotide substitution rate for COI of 0.0115 and 0.0177 per site per diversity was likely highest at sites ZY and JC because the million years by 0.736 for a substitution rate of 0.00846 sympatric populations of the two lineages in central Gansu and 0.013 per site per million years for the combined elevated the diversity values. The AMOVA analysis (Table 1) mtDNA alignment. The substitution model was selected showed that 84.07% of the variation was between the two from 24 models using the Akaike information criterion lineages (FCT = 0.84074, P < 0.0015) and that the varia- (AIC) using jModelTest 0.1 (Posada 2008). The chain tion among populations within each lineage was only length was set to be 200 million generations, with a sample 5.27% (Table 1). frequency of 20,000, and we discarded the first 10% as We derived 26 haplotypes and 11 parsimony informative burn-in. We chose exponential relaxed clock model for the sites for the EF1a gene. The nucleotide and haplotype mtDNA alignment. The demographic history was recon- diversities averaged 0.00171 and 0.753, respectively. Based structed using Tracer v 1.6 (Rambaut et al. 2013). on the nuclear genes, the AMOVA analysis (Table 1) indi-

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 163 Evolutionary history of Apocheima cinerarius S. Liu et al.

Table 1 Analysis of molecular variance (AMOVA) results based on the mitochondrial genes and nuclear genes

Population groups Source of variation Fixation indices P-value Variance components Percentage of variation

COI+Cytb+ND5 Among groups FCT = 0.84074 0.00000 18.02362 Va 84.07% Among populations FSC = 0.33090 0.00489 1.12978 Vb 5.27% Within groups Within populations FST = 0.89344 0.00000 2.28446 Vc 10.66% Total 21.43786 EF1a Among groups FCT = 0.08094 0.00000 0.06791 Va 8.09% Among populations FSC = 0.13778 0.00000 0.10624 Vb 12.66% Within groups Within groups FST = 0.20756 0.00000 0.66484 Vc 79.24% Total 0.83899

cated that 79.24% (FST = 0.20756, P < 0.0015) of the the north of the Altai Mountain ranges and to the south of variation was within each lineage and only 8.09% of the the Altun Mountain region, including the Xinjiang Uygur variation was between the two lineages. Autonomous Region and western Gansu Province (ALT, EM, MNS, DBG, DH, HM, HT, KE, KS, QM, TL, YN, Phylogenetic relationship and phylogeographical structure WX, JQ, ZY and JC). The distribution of the eastern lin- The BI and ML methods indicated approximately the same eage included northern and north-eastern China (BC, BJ, topologies for the phylogenetic analyses based on 27 com- HLE, HX, JMS, LL, LQ, QH, SX, SY, SZ, WF, XX, YC, plete mitogenomes (12 954 bp). Apocheima cinerarius was ZLT, ZY and JC). The two lineages formed a narrow sym- divided into two lineages (Fig. 2, Fig. S2): the western lin- patric area in central Gansu at the ZY and JC sites (Fig. 1). eage and the eastern lineage. The genetic distance between The lack of suitable areas in south of the QTP and north the two lineages was 1% in the complete mitochondrial (desert) of Gansu Province created a linear distribution in genomes. The distribution of the western lineage was to the Hexi Corridor.

Fig. 1 Pie graphs indicating the 31 sampling sites. The size of the pie graph stands for the number of mitochondrial DNA haplotype frequencies. Blue indicates the western lineage, yellow for the eastern lineage. Pictures of adults are shown in the lower left corner. The distribution range of A. cinerarius is shown with the yellow line in the lower right corner.

164 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius

Two distinct lineages for A. cinerarius were also recog- mutations were identiﬁed between the western and eastern nized in the MJ network based on COI + Cytb + ND5 lineages of A. cinerarius. Based on the MJ network with (Fig. S1), which was concordant with the divergence of the COI + Cytb + ND5 (Fig. S1), Clade A (western lineage) western and eastern lineages in the phylogenetic trees was subdivided into three groups, and these groups were based on the 27 mitogenomes (Fig. 2). Thirty ﬁxed base correlated with their geographical distributions. Clade A1

Fig. 2 BI phylogenetic tree partitioned by gene based on the 27 complete mitogenomes. Bayesian posterior probabilities are indicated above tree branches and ML bootstrap values below.

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 165 Evolutionary history of Apocheima cinerarius S. Liu et al. was located north of the Tianshan Mountains, Clade A2 For the nuclear data, EF1a and ITS2, the BI and ML was widespread, with the distribution including south of phylogenetic trees and the MJ network do not indicate two the Tianshan Mountains and western Gansu, and Clade A3 distinct lineages as the mtDNA structures (Fig. 4). The was located at the JQ and ZY sites in central Gansu. These western and eastern lineages shared 6 haplotypes, and the clades were also revealed in the phylogenetic trees (Fig. 2). haplotypes from the lineages were admixed with no fixed We found that haplotypes were not shared between Clades base mutations. However, the western and eastern lineages A1 and A2. The haplotype groups of Clade B (eastern lin- also contained unique haplotypes within each lineage. eage) were indicated with star-like shapes in the MJ network based on COI + Cytb + ND5 (Fig. S1). Based on the Estimation of divergence time COI gene with A. hispidaria as the out-group, the BI phylo- Time of the most recent common ancestor (TMRCA) for genetic tree showed that the western lineage was at the Clades A and B was estimated as 1.2 Ma, with a 95% con- base of the tree (Fig. 3a). The MJ network based on the fidence interval of 0.5–2.4 Ma (Fig. 3), using the rate of COI gene (Fig. 3b) established that the haplotypes from 2.3% per million years. TMRCA was estimated as 0.95 Ma the western group were more closely related to the out- with a 95% confidence interval of 0.43–1.8 Ma (Fig. 3), group A. hispidaria, which is found in Europe, than the using the rate of 3.54% per million years. TMRCA for haplotypes of the eastern lineage. A. cinerarius and A. hispidaria was approximately 2.5–

Fig. 3 (A) BI phylogenetic tree based on COI gene with A. hispidaria as out-group. Bayesian posterior probabilities and the time to most recent common ancestor (TMRCA), with the 95% conﬁdence interval, are indicated at each node. The unit of the time is million years ago (Ma) (B) Median-joining network based on COI, A. hispidaria as outgroup. Circles stand for haplotypes, the size of circles roughly represents the haplotype frequency, and different colours indicate different lineages. Black spots denote unsampled haplotypes. The number of vertical bars is on behalf of the number of mutation sites.

166 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius

Fig. 4 Median-joining network based on EF1a. Circles stand for haplotypes, the size of circles roughly represents the haplotype frequency, and different colours stand for the lineages. The number of vertical bars is on behalf of the number of mutation sites.

3.0 Ma, with a 95% confidence interval of 1.76–3.31 Ma group, the distribution area contracted to all of the north- and 2.6–5.2 Ma. ern China plain in the LGM. Therefore, the western and eastern lineages were isolated in the glacial periods. Historical demography The tests of Fu’s Fs were negative for most of the sampling Discussion sites. Fu’s Fs and Tajima’s D of the eastern lineage were Phylogeographical structure significantly negative, which indicated past population The uplift of the QTP had profound effects on species expansion or selection pressure. The BSP consequences diversification and the phylogeographical patterns of Asian based on the two mutation rates showed similar demo- fauna (Lei et al. 2014; Favre et al. 2015). In southern graphic history. The population of western lineage China, many species originated from the south-eastern expanded rapidly after 7.5 Ka. However, the eastern lin- margin of the QTP, which then dispersed eastward and eage experienced continuous demographic expansion from underwent west-east divergence, such as occurred with 80 Ka (Fig. 5). greenish warbler (Phylloscopus trochiloides) and long-tailed tit (Aegithalos caudatus) (Irwin et al. 2001, 2005; Dai et al. Historical changes in distribution 2010). However, relevant research on the origin of species The ecological niche models for A. cinerarius had excellent in northern China is limited, and our study of the mtDNA predictive power, with an average AUC of 0.935 (Swets data may supply evidence for the geographical origin of 1988). The predicted suitable distribution in the LIG was A. cinerarius. The divergence time of A. cinerarius and similar to the current distribution and was larger than that A. hispidaria was traced to approximately 2.5–3.0 Ma, in the LGM (Fig. 6). For the western group, there were which coincided with the uplift of the QTP (1.7–3.6 Ma) several different isolated and suitable habitats in Xinjiang (Li et al. 2001; Lei et al. 2014); therefore, the uplift of the and western Gansu in the LGM. South of Xinjiang, partic- QTP might be the main mechanism behind the speciation ularly in the south-eastern Tarim Basin, the area was not of A. cinerarius. Moreover, the uplift of the QTP was con- suitable for the survival of A. cinerarius. For the eastern sidered to facilitate the differentiation of other species in

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 167 Evolutionary history of Apocheima cinerarius S. Liu et al.

Fig. 5 Bayesian skyline plots representing historical demographic trends of the western lineage and the eastern lineage of A. cinerarius. The x-axis indicates the time (million years ago). The y-axis indicates the effective population size. The shaded areas correspond to 95% highest posterior densities (95% HPD). LGM represents the Last Glacial Maximum.

Asia, such as the ring-necked pheasant (Phasianus colchicus), expansion of A. cinerarius to the east. The monsoon circu- the black redstart (Phoenicurus ochruros) and the daurian lation was likely triggered and intensiﬁed because of the partridge (Perdix daurica) (Qu et al. 2009, 2010; Cao et al. uplift of the QTP, particularly 2.6 Ma and 0.9–1.2 Ma 2012). Additionally, the western lineage of A. cinerarius was (Sun et al. 2003). The monsoon circulation greatly inﬂu- at the base of the COI phylogenetic tree and was also enced the environment and the distribution of the species genetically closer to A. hispidaria in the MJ network analy- in northern China because it seriously affected the land- sis (Fig. 3). The ancestral haplotypes are in the root of the form of this area with the accumulation of loess after phylogenetic tree and are widely distributed, whereas the 2.6 Ma (An et al. 2001; Li et al. 2001). The time of adult derivative haplotypes are at the tips of the phylogenetic emergence of A. cinerarius is from the end of February to tree (Schaal et al. 1998). Thus, these results imply that the April, which coincides with the time of the prevailing win- western lineage of A. cinerarius was the ancestral population ter monsoon. The adults emerge and search the uncovered and that the western sites might be the areas of ancestral surface of the earth for the nearest trees for oviposition distribution. Similar results were also obtained in a study of (Wang 2008). The adults are easily blown to the east by the house mice subspecies (Mus musculus musculus) (Jing the strong Asian winter monsoon winds, and the adults et al. 2014). tend to climb trees that are to the east of their point of Compared with birds and mammals, A. cinerarius has hatching (Yin et al. 2004; Wang 2008). With this beha- weak dispersal ability. Because the populations of viour, A. cinerarius likely expanded to the east generation A. cinerarius spread widely from their origin eastward to by generation when the entire northern of China region the Amur district, external forces must have prompted the was under the pressure of this winter monsoons. We there-

168 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius

azure-winged magpie (Cooper 2000; Palomino et al. 2011). Additionally, the Badain Jaran Desert began to expand during the Quaternary (Sun et al. 2008), which might have facilitated the divergence of the lineages. In conclusion, we propose that the climate changes caused by glaciers and desertification were the major factors that caused the divergence between the two lineages. A sympatric area of the two lineages occurs in Zhangye (ZY) and Jinchang (JC) of central Gansu. These two populations shared haplotypes from the western and the eastern lineages (Fig. 1); these shared haplotypes at each sampling site (Zhangye and Jinchang) provide evidence to support the hypothesis that the two lineages experienced isolated evolution and then postglacial recolonization. The divergence and the subsequent sympatry of the two lineages were most likely caused by the glacial-induced climate changes. Our molecular data and ENM analysis support the refuge theory of Haffer (Haffer 1997; Haffer & Prance 2001), which postulates that areas of sympatry are formed Fig. 6 Ecological niche models for A. cinerarius in northern China because of the separation and then expansion of the popu- in the present day, at the Last Glacial Maximum (LGM) and the lations triggered by climatic fluctuations, as the East Asian Last Interglacial (LIG). Black circles represent the localities used to build the ecological niche models. Warmer colours show areas black-throated tit (Aegithalos concinnus) (Dai et al. 2011). with better predicted conditions. Different colours correspond to The ENM analysis also indicated that the sympatric area different fitting indices, with high in red and low in blue. might not have been suitable for the survival of A. cinerarius in the LIG and the LGM, although the current habitat fore hypothesize that the Asian winter monsoon might is suitable (Fig. 6). have been the driving force for the expansion of A. cinerar- We found that geographical barriers, such as mountains ius, and possibly of other litter-layer insects with weak dis- and deserts, were also important factors in shaping the persal ability. However, the possibility of wind-aided larval phylogeographical structure of A. cinerarius. The diver- dispersal is not excluded. gence pattern is generally concordant with previous studies on other animal species, such as mongolian gerbil (Meriones Pleistocene climatic impact on the lineage divergence meridianus) (Wang et al. 2013b), green toad (Bufo viridis) Previous studies indicate that changes in the Pleistocene cli- (Zhang et al. 2008b) and red deer (Cervus elaphus) (Mahmut mate had profound effects on the divergence of East Asian et al. 2002), whose divergences in population genetics were bird species (Dai et al. 2011; Qu et al. 2011; Zhao et al. triggered by the Tianshan Mountains. However, A. cinerar- 2012; Wang et al. 2013a,b). In our study, based on the com- ius has further subdivisions, such as Clades A1 and A2 bined mtDNA data, A. cinerarius was divided into two lin- within Clade A (Fig. 2). The reason why Clades A1 and eages: the western lineage and eastern lineage. The time of A2 shared no haplotypes might be that the Tianshan divergence of the two lineages was in accordance with the Mountain Range formed a geological barrier between the glaciation in the Early–Middle Pleistocene period (Ma & populations north and south of the Tianshan Mountains Gao 2004) when the climate underwent rapid changes and (Fig. S1). The populations of Clades A1 and A2 might also low temperature was persistent. These climate changes had have had different refugia during the glacial periods, like enormous influences on the distribution and genetic diver- B. viridis (Zhang et al. 2008b). For Clade A2, A. cinerarius sity of organisms (Shi et al. 2005). Therefore, we propose was distributed along the edge of the Taklamakan desert that Pleistocene climatic fluctuations likely affected the dis- (Fig. 1). The dominant and ancestral haplotypes H 11 and tribution pattern and species diversification of A. cinerarius. H 14 were in the centre of the network and found in most The two lineages of A. cinerarius might have contracted into populations on the southern side of the Tianshan Moun- isolated glacial refugia, like the black-spotted frog (Pelophy- tains (Fig. S1); the basis of this conclusion is that the lax nigromaculata) during a glacial period (Zhang et al. ancestral haplotypes are always in the centre of the network 2008a). The divided ranges of the western and eastern with the highest haplotype frequencies, while the derivative lineages of A. cinerarius might also be the subdivision of a haplotypes are always found around the network away from former continuous range during the Pleistocene, like the the centre (Kingman 1980; Templeton 1993). The low

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 169 Evolutionary history of Apocheima cinerarius S. Liu et al. levels of haplotype and nucleotide diversity found in south- the refuge of the eastern lineage of A. cinerarius. Several ern Xinjiang sites such as Hetian (HT) (0.00012 and 0.333, other species with distributions in north-eastern China also respectively) and Qiemo (QM) (0.00022 and 0.423, respec- expanded southward in the glacial periods, such as the tively) indicate that a recent and rapid colonization black-spotted frog (Yang et al. 2004; Zhang et al. 2008a). occurred during postglacial periods. The arid environment The woolly mammoth (Mammuthus primigenius) underwent of the Tarim Basin might have caused such an influence. two southward dispersal events, reaching as far as the The uplift of the QTP and the monsoon-driven expansion Shandong Peninsula during the Late Pleistocene (Jin et al. of the desert occurred at a large scale after 0.8 Ma, and the 1998). The distribution range of the eastern lineage shrank, Tarim Basin located on the north side of the plateau then although the population size increased in the LGM, based became more arid (Fang et al. 1999, 2002). The arid envi- on the BSP analysis. This may be because the entire north- ronment would have been unfavourable for A. cinerarius ern China plain was very suitable for the survival of (Wu et al. 2002; Head & Gibbard 2005). Compared with A. cinerarius, with the result that the population size did Clade A, Clade B had high values of haplotype (h) and not decrease, although the area of distribution contracted nucleotide diversity (p), most likely because the entire (Fig. 6). Similar results were found for coniferous trees in Northern China Plain was the refugium, which retained northern China (Chen et al. 2008). more haplotypes. Different rates of evolution or sex-biased dispersal? Demographic history The high divergence that was found between the two lin- The climatic fluctuations of the Quaternary had significant eages of A. cinerarius in the combined mtDNA analysis effects on the evolutionary history and contemporary distri- indicated restricted gene flow and long-term isolation, at bution patterns of the current species in northern China, least in the female populations, which might be caused by although the area was not completely covered with ice geographical barriers as environmental impediments and sheets (Mahmut et al. 2002; Chen et al. 2008; Cao et al. the low dispersal ability of the species (Avise et al. 1987). 2012; Wang et al. 2013a; Jing et al. 2014). The demo- The two lineages shared haplotypes based on MJ network graphic history of A. cinerarius was largely affected by gla- with nDNA. We speculated that the reason for the homo- cial-induced climate change. The BSP analyses indicated geneity of the nuclear genes might be due to their lower that the western and eastern lineages had different demo- evolutionary rate or winged male-biased gene flow (Spinks graphic histories. The western lineage began a rapid post- et al. 2010; Toews & Brelsford 2012; Dong et al. 2013; glacial expansion approximately 10 thousand years ago, DeBiasse et al. 2014; Ye et al. 2014). However, unique hap- which was different from the ‘pre-LGM expansion’ mode lotypes were also found within each lineage; thus, some found from East Asian birds (Fig. 5) (Zhao et al. 2012; degree of genetic differentiation in the nuclear genes must Wang et al. 2013a). The climate in China between 10.4 have occurred between the two lineages, and it is possible and 3.1 Ka was warm, immediately after the MIS (Marine that the two lineages were not interbreeding much. To test Isotope Stage) IV stage of the last glacial period (Yi et al. this hypothesis, nuclear microsatellite markers will be used 2005). For the eastern lineage, the BSP analysis showed in the next study to examine whether the male-biased gene that continuous population expansion occurred during the flow or a lower evolutionary rate for causing this pattern. LIG and the transition period from the LIG to the LGM (80 Ka) (Fig. 5), which was similar to the demographic his- Conclusions tory of Asian populations and in contrast to the post-LGM Our molecular results revealed that A. cinerarius most likely expansion mode of European populations of the great tit originated from the west, with the event closely related to (Parus major) (Kvist et al. 2003; Zhao et al. 2012). These the uplift of the Qinghai–Tibet Plateau. The species results suggested that the species in north-western China expanded eastward and became distributed across the entire might have different phylogeographical and demographic region of northern China, most likely because of the winter histories from species in north-eastern China. Determining monsoon. Apocheima cinerarius was subdivided into two dis- whether the northern China plain is a suitable habitat for tributions of western and eastern lineages, from a former the long-term survival of some insects should be further continuous range, because of the effects of Quaternary cli- tested. The star-like shapes of the eastern haplotype groups matic fluctuations. Both molecular data and ecological suggested that A. cinerarius experienced population expan- niche models indicated that the current distribution of sion in this area, which was in complete agreement with A. cinerarius was formed by range expansion from two the BSP analysis. The ecological niche models indicate that independent refugia where two lineages were developed; the suitable habitats of A. cinerarius underwent a south- however, the lineages now occur within a narrow sympatric ward contraction to the northern China plain, which was area in central Gansu Province. The respective demo-

170 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius graphic history of the two lineages indicates the popula- Cooper, J. H. (2000). First fossil record of Azure-winged Magpie tions in north-western China have different demographic Cyanopica cyanus in Europe. Ibis, 142, 150–151. histories and periods of expansion from the populations in Dai, C., Chen, K., Zhang, R., Yang, X. J., Yin, Z. H., Tian, H. J., Zhang, Z. M., Hu, Y. & Lei, F. M. (2010). Molecular phyloge- north-eastern China in the glacial cycling periods. In gen- fl netic analysis among species of Paridae, Remizidae and Aegithalos eral, the Quaternary environmental uctuations had pro- based on mtDNA sequences of COI and cyt b. Chinese Birds, 1, fl fi found in uences on the diversi cation and demography of 112–123. an insect in northern China. Dai, C. Y., Zhao, N., Wang, W. J., Lin, C. T., Gao, B., Yang, X. J., Zhang, Z. W. & Lei, F. M. (2011). Profound climatic effects Acknowledgements on two east Asian black-throated tits (Ave: Aegithalidae), The authors sincerely thank Xuejian Wang, Jing Li, Chao revealed by ecological niche models and phylogeographic analy- Yang, Fuqiang Chen and Le Cui for collecting the speci- sis. PLoS ONE, 6, e29329. DeBiasse, M. B., Nelson, B. J. & Hellberg, M. E. (2014). Evaluat- mens; thank all the local forestry authorities who offered ing summary statistics used to test for incomplete lineage sort- huge assistance for specimen collections. Special thanks are ing: mito-nuclear discordance in the reef sponge Callyspongia given to Anthony Galsworthy, the Natural History vaginalis. Molecular Ecology, 23, 225–238. Museum, London, for polishing the language. This work Dong, L., Heckel, G., Liang, W. & Zhang, Y. Y. (2013). Phylo- was supported by the National Science Foundation of geography of Silver Pheasant (Lophura nycthemera L.) across China (No. 31272288, 31372176), the Ministry of Science China: aggregate effects of refugia, introgression and riverine – and Technology of China (No. 2014FY210200), the barriers. Molecular Ecology, 22, 3376 3390. Drummond, A. J., Rambaut, A., Shapiro, B. & Pybus, O. G. National Science Fund for Fostering Talents in Basic (2005). Bayesian coalescent inference of past population dynam- Research (No. J1210002) and a grant from the Key Labo- ics from molecular sequences. Molecular Biology and Evolution, 22, ratory of the Zoological Systematics and Evolution of the 1185–1192. Chinese Academy of Sciences (No. O529YX5105). Excoffier, L. & Lischer, H. E. (2010). Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses References under Linux and Windows. Molecular Ecology Resources, 10, 564– An, Z., Kutzbach, J. E., Prell, W. L. & Porter, S. C. (2001). 567. fi Evolution of Asian monsoons and phased uplift of the Himala- Excof er, L., Smouse, P. E. & Quattro, J. M. (1992). Analysis of yan Tibetan plateau since Late Miocene times. Nature, 411,62– molecular variance inferred from metric distances among DNA 66. haplotypes: application to human mitochondrial DNA restriction – Avise, J. C., Arnold, J, Ball, R. M., Bermingham, E., Lamb, T., data. Genetics, 131, 479 491. Neigel, J. E., Reeb, C. A. & Saunders, N. C. (1987). Intraspeci- Fang, X. M., Li, J. J. & Van der Voo, R. (1999). Paleomagnetic/ fi fic phylogeography: the mitochondrial DNA bridge between rock-magnetic and grain size evidence for intensi ed Asian population genetics and systematics. Annual Review of Ecology and atmospheric circulation since 800 kyrs. Earth and Planetary – Systematics, 18, 489–522. Science Letters, 165, 129 144. Bandelt, H. J., Forster, P. & Rohl, A. (1999). Median-joining net- Fang, X. M., Lu, L. Q., Yang, S. L., Li, J. J., An, Z. S., Jiang, works for inferring intraspecific phylogenies. Molecular Biology P. & Chen, X. L. (2002). Loess in Kunlun Mountains and and Evolution, 16,37–48. its implication on desert development and Tibetan Bouckaert, R., Heled, J., Kuhnert,€ D., Vaughan, T., Wu, C. H., Plateau uplift in west China. Science in China (Series D), 45, – Xie, D., Suchard, M. A., Rambaut, A. & Drummond, A. J. 289 299. € (2014). BEAST2: a software platform for Bayesian evolutionary Favre, A., Pakert, M., Pauls, S. U., Jahnig, S. C., Uhl, D., Micha- analysis. PLOS Computational Biology, 10, e1003537. lak, I. & Muellner-Riehl, A. N. (2015). The role of the uplift of Brower, A. V. Z. (1994). Rapid morphological radiation and con- the Qinghai-Tibetan Plateau for the evolution of Tibetan biotas. – vergence among races of the butterfly Heliconius erato inferred Biological Reviews, 90, 236 253. from patterns of mitochondrial DNA evolution. Proceedings of the Haffer, J. (1997). Contact zones between birds of southern Amazo- – National Academy of Sciences of the United States of America, 91, nia. Ornithological Monographs, 48, 281 305. 6491–6495. Haffer, J. & Prance, G. (2001). Climatic forcing of evolution in Cao, M., Jin, Y., Liu, N. & Ji, W. (2012). Effects of the Qing- Amazonia during the Cenozoic: on the refuge theory of biotic – hai–Tibetan Plateau uplift and environmental changes on phy- differentiation. Amazoniana, 16, 579 607. logeographic structure of the Daurian Partridge (Perdix Haring, E., Gamauf, A. & Kryukov, A. (2007). Phylogeographic dauuricae) in China. Molecular Phylogenetics and Evolution, 65, patterns in widespread corvid birds. Molecular Phylogenetics and – 823–830. Evolution, 45, 840 862. Chen, K., Abbott, R. J., Milne, R. I., Tian, X. M. & Liu, J. (2008). Hausmann, A., Haszprunar, G. & Hebert, P. D. N. (2011). DNA Phylogeography of Pinus tabulaeformis Carr. (Pinaceae), a domi- barcoding the Geometrid fauna of Bavaria (Lepidoptera): suc- nant species of coniferous forest in northern China. Molecular cesses, surprises, and questions. PLoS ONE, 6, e17134. Ecology, 17, 4276–4288. Head, M. J. & Gibbard, P. L. (2005). Early-Middle Pleistocene Chu, H. F. (1981). Geometridae. Iconographia Heterocerorum Sinico- transitions: the land-ocean evidence. Geological Society Special – rum, 1, 112–131. Publication, 247,1 326.

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 171 Evolutionary history of Apocheima cinerarius S. Liu et al.

Hebert, P. D. N., Penton, E. H. & Burns, J. M. (2004). Ten spe- Librado, P. & Rozas, J. (2009). DnaSP v5: a software for compre- cies in one: DNA barcoding reveals cryptic species in the hensive analysis of DNA polymorphism data. Bioinformatics, 25, neotropical skipper butterfly Astraptes fulgerator. Proceedings of the 1451–1452. National Academy of Sciences of the United States of America, 101, Lin, A. Q., Csorba, G., Li, L. F., Jiang, T. L., Lu, G. J., Vu, D. 14812–14817. T., Soisook, P., Sun, K. P. & Feng, J. (2014). Phylogeography Hewitt, G. M. (1999). Post-glacial recolonization of European of Hipposideros armiger (Chiroptera: Hipposideridae) in the Ori- biota. Biological Journal of the Linnean Society, 68,87–112. ental Region: the contribution of multiple Pleistocene glacial Hewitt, G. M. (2000). The genetic legacy of the Quaternary ice refugia and intrinsic factors to contemporary population genetic ages. Nature, 405, 907–913. structure. Journal of Biogeography, 41, 317–327. Hewitt, G. M. (2004). Genetic consequences of climatic oscilla- Liu, X. D., Li, L. & An, Z. S. (2001). Tibetan plateau uplift and tions in the Quaternary. Philosophical Transactions of the Royal drying in Eurasian interior and northern Africa. Quaternary Society of London Series B-Biological Sciences, 359, 183–195. Science, 21, 114–122. Hu, Z. J., Zhang, Y. L., Liu, L. S. & Yu, H. B. (2013). Refugia Liu, S. X., Xue, D. Y., Cheng, R. & Han, H. X. (2014). The com- and their identification methods: a review. Chinese Journal of plete mitogenome of Apocheima cinerarius (Lepidoptera: Ecology, 32, 3397–3406. Geometridae: Ennominae) and comparison with that of other Igea, J., Aymerich, P., Fernandez-Gonzalez, A. A., Gonzalez-Este- lepidopteran insects. Gene, 547, 136–144. ban, J., Gomez, A., Alonso, R., Gosalbez, J. & Castresana, J. Lopez-L opez, A., Abdul Aziz, A. & Galian, J. (2015). Molecular (2013). Phylogeography and postglacial expansion of the endan- phylogeny and divergence time estimation of Cosmodela (Coleop- gered semi-aquatic mammal Galemys pyrenaicus. BMC Evolu- tera: Carabidae: Cicindelinae) tiger beetle species from Southeast tionary Biology, 13, 115. Asia. Zoologica Scripta, 44, 437–445. Irwin, D. E., Bensch, S. & Price, T. D. (2001). Speciation in a Ma, Z. J. & Gao, Q. H. (2004). Climate changes of China in the ring. Nature, 409, 333–337. quaternary period and analysis on drought disasters of north Irwin, D. E., Bensch, S., Irwin, J. H. & Price, T. D. (2005). Speci- China in the future. Quaternary Sciences, 24, 245–251. ation by distance in a ring species. Science, 307, 414–416. Ma, C., Yang, P.C., Jiang, F., Chapuis, M.P., Shali, Y., Sword, Ji, Y. J., Zhang, D. X. & He, L. J. (2003). Evolutionary conserva- G.A. & Kang, L. (2012). Mitochondrial genomes reveal the glo- tion and versatility of a new set of primers for amplifying the bal phylogeography and dispersal routes of the migratory locust. ribosomal internal transcribed spacer regions in insects and Molecular Ecology, 21, 4344–4358. other invertebrates. Molecular Ecology Notes, 3, 581–585. Maekawa, K., Kon, M., Araya, K. & Matsumoto, T. (2001). Phy- Jin, C. Z., Xu, Q. Q. & Zheng, J. J. (1998). On the dispersal logeny and biogeography of wood-feeding cockroaches, genus events of Mammuthus during the late Late Pleistocene. Vertebrate Salganea Stal (Blaberidae: Panesthiinae), in southeast Asia based Palastatica, 36,47–53. on mitochondrial DNA sequences. Journal of Molecular Evolution, Jing, M., Yu, H. T., Bi, X., Lai, Y. C., Jiang, W. & Huang, L. 53, 651–659. (2014). Phylogeography of Chinese house mice (Mus musculus Mahmut, H., Masuda, R., Onuma, M., Takahashi, M., Nagata, J., musculus/castaneus): distribution, routes of colonization and geo- Suzuki, M. & Ohtaishi, N. (2002). Molecular phylogeography of graphic regions of hybridization. Molecular Ecology, 23, 4387– the red deer (Cervus elaphus) populations in Xinjiang of China: 4405. comparison with other Asian, European, and North American Ju, L., Wang, H. & Jiang, D. (2007). Simulation of the Last Gla- populations. Zoological Science, 19, 485–495. cial Maximum climate over East Asia with a regional climate Meng, H. H. & Zhang, M. L. (2011). Phylogeography of Lagochi- model nested in a general circulation model. Palaeogeography, lus ilicifolius (Lamiaceae) in relation to Quaternary climatic oscil- Palaeoclimatology, Palaeoecology, 248, 376–390. lation and aridification in northern China. Biochemical Systematics Kawakita, A., Takimura, A., Terachi, T., Sota, T. & Kato, M. and Ecology, 39, 787–796. (2004). Cospeciation analysis of an obligate pollination mutual- Nei, M. (1987). Molecular Evolutionary Genetics. New York & ism: have Glochidion trees (Euphorbiaceae) and pollinating Epi- Guildford, England: Columbia University Press. pp.i–x, 1–512. cephala moths (Gracillariidae) diversified in parallel? Evolution, Olsson, U., Alstrom,€ P., Svensson, L., Aliabadian, M. & Sundberg, 58, 2201–2214. P. (2010). The Lanius excubitor (Aves, Passeriformes) conundrum Kingman, J. F. C. (1980). Mathematics of genetic diversity. CBMS- – taxonomic dilemma when molecular and non-molecular data NSF Regional Conference Series in Applied Mathematics, 1, 7,0. tell different stories. Molecular Phylogenetics and Evolution, 55, Kvist, L., Martens, J., Higuchi, H., Nazarenko, A. A., Valchuk, O. 347–357. P. & Orell, M. (2003). Evolution and genetic structure of the Palomino, D., Carrascal, L. M. & Potti, J. (2011). Distribution of great tit (Parus major) complex. Proceedings of the Royal Society B- Azure-winged Magpies Cyanopica cooki in Spain: both local and Biological Sciences, 270, 144–1454. large-scale factors considered. Acta Ornithologica, 46,71–82. Lei, F. M., Qu, Y. H. & Song, G. (2014). Species diversification Papadopoulou, A., Anastasiou, I. & Vogler, A. P. (2010). Revisiting and phylogeographical patterns of birds in response to the uplift the insect mitochondrial molecular clock: the mid-aegean trench of the Qinghai-Tibet Plateau and Quaternary glaciations. Cur- calibration. Molecular Biology and Evolution., 27, 1659–1672. rent Zoology, 60, 149–161. Phillips, S. J., Anderson, R. P. & Schapire, R. E. (2006). Maximum Li, J. J., Fang, X.M., Pan, B. T., Zhao, Z. J. & Song, Y. G. (2001). entropy modeling of species geographic distributions. Ecological On the intensified uplift of Qinghai-Tibet Plateau and its impact Modelling, 190, 231–259. on the peripheral environment in late Cenozoic. Quaternary Posada, D. (2008). jModelTest: phylogenetic model averaging. Research, 21, 381–390. Molecular Biology and Evolution, 25, 1253–1256.

172 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 S. Liu et al. Evolutionary history of Apocheima cinerarius

Qian, H. & Ricklefs, R. E. (2000). Large-scale processes and the Swets, J. A. (1988). Measuring the accuracy of diagnostic systems. Asian bias in species diversity of temperate plants. Nature, 407, Science, 240, 1285–1293. 180–182. Templeton, A. R. (1993). The ‘Eve’ hypotheses: a genetic critique Qu, J. Y., Liu, N. F., Bao, X. K. & Wang, X. L. (2009). Phylo- and reanalysis. American Anthropologist, 95,51–72. geography of the ring-necked pheasant (Phasianus colchicus)in Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & China. Molecular Phylogenetics and Evolution, 52, 125–132. Higgins, D. G. (1997). The CLUSTAL_X windows interface: Qu, Y., Lei, F., Zhang, R. & Lu, X. (2010). Comparative phylo- flexible strategies for multiple sequence alignment aided by qual- geography of five avian species: implications for Pleistocene evo- ity analysis tools. Nucleic Acids Research, 25, 4876–4882. lutionary history in the Qinghai-Tibetan plateau. Molecular Toews, D. P. L. & Brelsford, A. (2012). The biogeography of Ecology, 19, 338–351. mitochondrial and nuclear discordance in animals. Molecular Qu, Y., Luo, X., Zhang, R., Song, G., Zou, F. & Lei, F. (2011). Ecology, 21, 3907–3930. Lineage diversification and historical demography of a montane Wahlberg, N., Sn€all, N., Viidalepp, J., Ruohom€aki, K. & Tam- bird Garrulax elliotii – implications for the Pleistocene evolution- maru, T. (2010). The evolution of female flightlessness among ary history of the eastern Himalaya. BMC Evolutionary Biology, Ennominae of the Holarctic forest zone (Lepidoptera: 11, 174. Geometridae). Molecular Phylogenetics and Evolution, 55, 929–938. Rambaut, A., Drummond, A. J. & Suchard, M. (2013). Tracer v1. Wang, X. J. (2008). A Study on the Molecular Phylogenetics and 6. Available via: http://beast.bio.ed.ac.uk/Tracer. Phylogeography of Apocheima cinerarius (Erschoff). Beijing: Insti- Ronquist, F. & Huelsenbeck, J. P. (2003). MRBAYES 3: Bayesian tute of Zoology Chinese Academy of Sciences. phylogenetic inference under mixed models. Bioinformatics, 19, Wang, W., McKay, B. D., Dai, C., Zhao, N., Zhang, R., Qu, Y., 1572–1574. Song, G., Li, S., Liang, W., Yang, X., Eric, P. & Lei, F. Rost, K. T. (1994). Paleoclimatic field studies in and along the (2013a). Glacial expansion and diversification of an East Asian Qinling Shan (Central China). GeoJournal, 34, 107–120. montane bird, the green-backed tit (Parus monticolus). Journal of Sattler, K. (1991). A review of wing reduction in Lepidoptera. Bul- Biogeography, 40, 1156–1169. letin of the British Museum (Natural History) Entomology, 60, 243– Wang, Y., Zhao, L. M., Fang, F. J., Liao, J. C. & Liu, N. F. 288. (2013b). Intraspecific molecular phylogeny and phylogeography Schaal, B. A., Hayworth, D. A., Olsen, K. M., Rauscher, J. T. & of the Meriones meridianus (Rodentia: Cricetidae) complex in Smith, W. A. (1998). Phylogeographic studies in plants: prob- northern China reflect the processes of desertification and the lems and prospects. Molecular Ecology, 7, 465–474. Tianshan Mountains uplift. Biological Journal of the Linnean Soci- € Sezonlin, M., Dupas, S., Le RU, B., Le Gall, P., Moyal, P., Cala- ety, 110, 362–383. tayud, P. A., Giffard, I. I., Faure, N. & Silvain, J. F. (2006). Phy- Wu, G., Pan, B., Guan, Q. & Gao, H. (2002). The maximum logeography and population genetics of the maize stalk borer glaciation and desert expansion in China during MIS16. Journal Busseola fusca (Lepidoptera, Noctuidae) in sub-Saharan Africa. of Glaciology and Geocryology, 24, 544–549. Molecular Ecology, 15, 407–420. Yamamoto, S. & Sota, T. (2007). Phylogeny of the Geometridae Shi, Y. F., Li, J. J. & Li, B. Y. (1998). Uplift and Environmental and the evolution of winter moths inferred from a simultaneous Changes of Qinghai-Tibetan Plateau in the Late Cenozoic. analysis of mitochondrial and nuclear genes. Molecular Phyloge- Guangzhou, China: Guangdong Science and Technology netics and Evolution, 44, 711–723. Press. Yang, Y. H., Zhang, D. X. & Li, Y. M. (2004). Mitochondrial Shi, Y. F., Cui, Z. J. & Su, Z. (2005). The Quarternary Glaciations DNA diversity and preliminary biogeographic inference of the and Environmental Variations in China. Shijiazhuang, China: evolutionary history of the black-spotted pond frog Rana nigro- Hebei Science and Technology Publishing House. maculata populations in China. Acta Zoologica Sinica, 50, 193– Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook, 201. P. (1994). Evolution, weighting, and phylogenetic utility of Yang, L., Wei, Z. J., Hong, G. Y., Jiang, S. T. & Wen, L. P. mitochondrial gene sequences and a compilation of conserved (2009). The complete nucleotide sequence of the mitochondrial polymerase chain reaction primers. Annals of the Entomological genome of Phthonandria atrilineata (Lepidoptera: Geometridae). Society of America, 87, 651–701. Molecular Biology Reports, 36, 1441–1449. Spinks, P. Q., Thomson, R. C. & Bradley, S. H. (2010). Nuclear Yang, X. S., Xue, D. Y. & Han, H. X. (2013). The complete mito- gene phylogeography reveals the historical legacy of an ancient chondrial genome of Biston panterinaria (Lepidoptera: Geometri- inland sea on lineages of the western pond turtle. Emys Mar- dae), with phylogenetic utility of mitochondrial genome in the morata in California. Molecular Ecology, 19, 542–556. Lepidoptera. Gene, 515, 349–358. Stamatakis, A. (2006). RAxML-VI-HPC: maximum likelihood- Ye, Z., Zhu, G., Chen, P., Zhang, D. & Bu, W. (2014). Molecular based phylogenetic analyses with thousands of taxa and mixed data and ecological niche modelling reveal the Pleistocene his- models. Bioinformatics, 22, 2688–2690. tory of a semi-aquatic bug (Microvelia douglasi douglasi) in East Sun, D., An, Z., Su, R., Lu, H. & Sun, Y. (2003). Eolian sedimen- Asia. Molecular Ecology, 23, 3080–3096. tary records for the evolution of monsoon and westerly circula- Yi, C. L., Cui, Z. J. & Xiong, H. G. (2005). Numerical periods of tions of northern China in the last 2.6 Ma. Science in China Quaternary glaciations in China. Quaternary Sciences, 25, 609– Series D: Earth Sciences, 46, 1049–1059. 619. Sun, Q. F., Chen, F. H. & Li, X. Z. (2008). Review and discussion Yin, F. C., Qin, L. F. & Yu, H. L. (2004). Observation and Con- about the progress of Quaternary research of the Badain Juran trol of Emergemce (Eclosion) of Spring Cankerworm (Apocheima Desert, China. Arid Zone Research, 25, 304–310. Cinerius Erschoff). Xinjiang Agricultural Sciences, 41, 385–387.

ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174 173 Evolutionary history of Apocheima cinerarius S. Liu et al.

Zhang, D. (2000). Eco-environmental effects of the Qinghai-Tibet Supporting Information plateau uplift during the quaternary in China. Environmental Additional Supporting Information may be found in the – Geology, 39, 1352 1358. online version of this article: Zhang, H., Yan, J. & Zhang, G. (2008a). Phylogeography and Fig. S1. Median-joining network based on COI + Cytb + demographic history of Chinese black-spotted frog populations (Pelophylax nigromaculata): evidence for independent refugia ND5. Circles stand for haplotypes, the size of circles expansion and secondary contact. BMC Evolutionary Biology, 8, roughly represents the haplotype frequency, and different 21. colours stand for the lineages. Black spots denote unsam- Zhang, Y. J., Stock,€ M., Zhang, P., Wang, X. L., Zhou, H. & Qu, pled haplotypes. Vertical bars were on behalf of mutation L. H. (2008b). Phylogeography of a widespread terrestrial verte- sites. brate in a barely-studied Palearctic region: green toads (Bufo vir- Fig. S2. BI phylogenetic tree partitioned by gene as well idis subgroup) indicate glacial refugia in Eastern Central Asia. as codon position based on the 27 complete mitogenomes. Genetica, 134, 353–365. Zhang, R., Song, G., Qu, Y., Alstrom,€ P., Ramos, R., Xing, X. Y., Table S1. The 31 sampling sites (exact sites longitude, Ericson, P. G. P., Fjeldsa, J., Wang, H. T., Yang, X. J., Kristin, latitude) were listed. A., Shestopalov, A. M., Choe, J. C. & Lei, F. M. (2012). Com- Table S2. The obtained 27 complete mitochondrial parative phylogeography of two widespread magpies: importance genomes of A. cinerarius in this study and the outgroups: of habitat preference and breeding behavior on genetic structure Biston panterinaria and Phthonandria atrilineata. – in China. Molecular Phylogenetics and Evolution, 65, 562 572. Table S3. Sample size (N), the number of haplotypes Zhao, N., Dai, C., Wang, W., Zhang, R., Qu, Y., Song, G., Chen, (Hap) and phylogenetic variation (p = nucleotide diversity, K., Yang, X., Zou, F. & Lei, F. (2012). Pleistocene climate = changes shaped the divergence and demography of Asian popu- Hd haplotype diversity) were estimated from each sam- lations of the great tit Parus major: evidence from phylogeo- pling sites and the parameters related to demography (Taji- graphic analysis and ecological niche models. Journal of Avian ma’s D and Fu’s Fs) were estimated based on the Biology, 43, 297–310. mitochondrial DNA. Zink, R. M., Pavlova, A., Drovetski, S. & Rohwer, S. (2008). Mito- Table S4. Primers used to amplify and sequence the ﬁ chondrial phylogeographies of ve widespread Eurasian bird mitochondrial genes COI, Cytb, ND5 and the nuclear genes species. Journal of Ornithology, 149, 399–413. EF1a, ITS2 of A. cinerarius. Zink, R. M., Pavlova, A., Drovetski, S., Wink, M. & Rohwer, S. (2009). Taxonomic status and evolutionary history of the Saxicola Table S5. Genbank Accession numbers with sample torquata complex. Molecular Phylogenetics and Evolution, 52, 769– information for individuals used in this study. 773.

174 ª 2015 Royal Swedish Academy of Sciences, 45, 2, March 2016, pp 160–174