Phylogeography of the ring-necked pheasant (Phasianus colchicus) in China

Molecular Phylogenetics and Evolution 52 (2009) 125–132

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev

Phylogeography of the ring-necked pheasant (Phasianus colchicus) in China Jiangyong Qu, Naifa Liu *, Xinkang Bao, Xiaoli Wang Institute of Zoology, School of Life Sciences, Lanzhou University, 730000 Lanzhou, Gansu, China

a r t i c l e

i n f o

Article history: Received 3 October 2008 Revised 7 March 2009 Accepted 10 March 2009 Available online 26 March 2009 Keywords: Phasianus colchicus Mitochondrial DNA control-region Phylogeography

a b s t r a c t The ring-necked pheasants (Phasianus colchicus) are widely distributed in China. We used mitochondrial DNA control-region data to investigate the origin and past demographic changes in 139 ring-necked pheasants (P. colchicus) sampled from the species’ distribution range. A total of 1078 nucleotides from the control region of mitochondrial DNA were sequenced, and 88 polymorphic positions defined 102 haplotypes. High level of genetic diversity was detected in all populations studied which could be associated with the wide ecological distributions and niche variation. Phylogenetic analyses of all haplotypes identified five major clades. The haplotypes of Gray-rumped Pheasants existed in the three clades: A (western clade), B (eastern clade) and C (Sichuan Basin clade). Two haplotypes of Kirghiz Pheasants were in the clade B, and the rest haplotype of Kirghiz Pheasants formed the clade D. Only one haplotype from White-winged Pheasants made up clade E. The results of AMOVA showed a low gene flow (Nm = 0.44) and significant genetic differentiation (Fst = 0.31, P < 0.001) among all populations. Based on the divergence time, we speculate that the divergence of the ring-necked pheasant occurring in the late Pleistocene may have resulted from three events: (1) the uplift of Qinghai-Tibet Plateau, (2) the existence of Qinling Mountains and Liupan Mountains, (3) the isolation of Sichuan Basin. Demographic population expansion was strongly confirmed by the non-significant mismatch distribution analysis. The described subspecies of the ring-necked pheasant could not be supported by the phylogeographical structuring. Ó 2009 Elsevier Inc. All rights reserved.

1. Introduction Geological, ecological and biogeographic histories have played important roles in shaping the current biodiversity over the world through their influences on regional differences in speciation, extinction and migration (Willis and Whittaker, 2002; Qian et al., 2005). These contributions can be evaluated by both paleontological and phylogenetic approaches. A similar approach is widely used to elucidate population histories based on the molecular footprints of the current populations at intrapopulation levels (Hewitt, 1996, 2000). Cyclical Pleistocene climatic oscillations are presumed to have shaped the geographical distribution, demographic history and ultimately patterns of genetic diversification of many plant and animal species in the Palaearctic (Avise, 2000; Hewitt, 2000). Molecular markers provide important measures for population genetic structures and geographic differentiations, which especially widely used in analyses of intraspecific phylogeographical patterns (Avise, 1994, 2000). Several characters of animal mtDNA made it particularly suitable for examining geographic distributions of evolutionary lineages within species (Avise, 1994). The noncoding control region (CR) has been the part of the mitochondrial genome

* Corresponding author. Fax: +86 931 8912561. E-mail address: [email protected] (N. Liu). 1055-7903/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2009.03.015

that is most frequently targeted, as it generally provides sufficient variation for studies at intraspecific level (Brown et al., 1979; Moritz et al., 1987). It is a particularly useful approach for clarifying contemporary geographical patterns of evolutionary subdivision with species (Arbogast and Kenagy, 2001) and allows the assessment of the impact of historical events on the current genealogical relationships among different populations (e.g. Aurelle and Berrebi, 2001; Recueroc et al., 2006). The ring-necked pheasant (Phasianus colchicus) inhabits open country in parks and farmland, scrubby wastes, along edges of woods, grassy steppes, hills and lower mountain slops with relatively sparse trees. They also live in oases of deserts and in reed beds or riverside thickets of tamarisks, poplars, wild olives or of other trees and shrubs along rivers and lakes, swamps, edges of rice fields, and in open forest up to 3965 m above sea level in western China (Johnsgard, 1999). The ring-necked pheasant (P. colchicus) was widely distributed throughout Eurasia: from southeastern Europe in the west to the Kamchatka Peninsula in the east and from eastern Siberia in the north to Indo-China and Afghanistan in the south (Hill and Robertson, 1988). In addition, this species was introduced to North America, Hawaii, Japan, Australia, New Zealand and Europe (Johnsgard, 1999). The ring-necked pheasant include many subspecies and the nineteen described subspecies distribute in China, P.c. tarimensis Pleske, P.c. shawii Elliot, P.c. mongolicus J.K. Brandt, P.c. satscheuensis Pleske, P.c. edzinensis Sushkin, P.c. pallasi Rothschild, P.c.

126

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

suehschanensis Bianchi, P.c. torquatus Gmelin, P.c. kiangsuensis Buturlin, P.c. decollatus Swinhoe, P.c. karpowi Buturlin, P.c. strauchi Przevalski, P.c. elegans Elliot, P.c. sohokotensis Burturlin, P.c. vlangalli Przevalski, P.c. alaschanicus Alpheraky and Bianchi, P.c. rothschildi La Touche, P.c. formosanus Elliot, and P.c. takatsukasae Delacour (Zheng, 1978). The first three subspecies belonged respectively to Tarim Pheasant, White-winged Pheasants (the principalis-chrysomelis group), and Kirghiz Pheasants (the mongolicus group), and distribute in Sinkiang. The last sixteen subspecies belong to Gray-rumped Pheasants (the torquatus group) and are distributed widely in China, except Sinkiang, Hainan and Qiangtang Plateau in Tibet. The region with Gray-rumped Pheasants includes areas of two types of climate: arid and semi-arid climate in the west and temperate, warm-temperate and subtropic monsoon region in the east. Current taxonomy and subspecies distinctions in the ring-necked pheasant are based exclusively on the morphological characters, such as the variation of the body size and color, the presence of the white neckring and the white eyebrows (Johnsgard, 1999; Liu and Sun, 1992). Molecular methods have now often been used in phylogenetic studies and taxonomy of a taxa is often changed with genetic data (i.e. Alectoris magna Liu et al., 2004). No study can be found in the ring-necked pheasant taxonomy studies using molecular methods. Since Pleistocene, habitats of the ring-necked pheasant have undergone a series of changes due to glacial–interglacial effect and the uplift of the Qinghai-Tibetan Plateau. The aim of this study is to: (1) assess whether climatic oscillations in Pleistocene have affected the present distribution of the ring-necked pheasant; (2) reconstruct phylogeographical relationships of the ring-necked pheasant on the background of Pleistocene habitat changes; (3) infer the evolutionary and past demographic processes that probably affected the current population structure.

2. Materials and methods 2.1. Sample collection, DNA extraction, amplification and sequencing A total of 139 samples from 19 populations of the ring-necked pheasants were collected during gaming seasons (Fig. 1). Green pheasant (P. versicolor) was used as outgroup (GenBank Accession No. AY376866). Liver samples were taken from birds and stored in 95% ethanol immediately after removal. DNA was extracted from the samples using the ethanol sedimentation procedure described by Randi and Lucchini (1998). Two oligonucleotide primers, PHDL (50 -AGGACTACGGCTTGAAAAGC-30 ) and PHDH (50 -CATCTTGGCATCTTCAGTGCC-30 ) (Randi and Lucchini, 1998), were used to amplify and sequence about 1000 nucleotides of a mitochondrial DNA segment. There was 2.5 U of Taq DNA polymerase per 50 ll of reactants. The final concentrations were 10 mmol/L Tris–HCl (pH 8.3), 50 mmol/L KCl, 1.5 mmol/L MgCl2, 150 lmol/L dNTP, 10 pmol/L primers and about 100 ng DNA template. The PCR conditions were as follows: 94 °C for 2 min, 94 °C for 15 s, 55 °C for 15 s, 72 °C for 1 min (30 cycles), and 72 °C for 10 min. PCR products were purified and sequencing of both direction performed with each of the PCR primers on an ABI 373 automated sequencer. The sequences were deposited in GenBank and the Accession Nos. are from EU908575 to EU908676. 2.2. Sequence analysis and phylogenetic structure Sequences were aligned by clustal procedure (Thompson et al., 1997) and refined manually. DnaSP4.0 (Rozas et al., 2003) was used to define the haplotypes and to estimate average and population haplotype diversity (h), mean number of pairwise differences (k), nucleotide diversity (p). The population differentiation index (Fst) and gene flow (Nm) were determined using the Arlequin2.0

Fig. 1. Distribution of four pheasant groups of the ring-necked pheasant (Phasianus colchicus) in China (Zheng, 1978) and sampling localities (numbered black circles) in this study which correspond to the ‘‘label” in Table 1.

127

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

analyses of molecular variance (AMOVA) (Schneider et al., 2002). Arlequin2.0 was also used to compute the goodness of fit, time since population expansion (s), relative population size before (h0) and after (h1) expansion and mismatch distributions. Bayesian phylogenetic analyses were performed using MRBAYES version 3.1.2 (Huelsenbeck and Ronquist, 2001). The best-fit model was selected using MRMODELTEST (Nylander, 2004). For Bayesian analyses, four simultaneous Monte Carlo Markov chains (MCMCs) were run for 5  106 generations, sampling a tree every 100 generations. After summarizing the substitution model parameter values, we would make sure that the Potential Scale Reduction Factor (PSRF, a convergence diagnostic) was reasonably close to 1.0 for the parameters.These parameters included the total tree length (TL), the six reversible substitution rate (r (Ah-iC), r (Ah-iG), r (AhiT), r (Ch-iG), r (Ch-iT) and r (Gh-iT)), the four stationary state frequencies (pi (A), pi (C), pi (T) and pi (G)), the shape of the gamma distribution of rate variation across sites (alpha), and the proportions of invariable sites (pinvar). The posterior probabilities (PP) were calculated with the last 37,500 sampled trees after the loglikelihood values had stabilized. Maximum-parsimony (MP) analyses were performed in PAUP*4b10 (Swofford, 1998) with 1000 bootstrap replicates. Gaps were treated as a fifth character, and searches were heuristic with random addition of taxa (five replicates). The shortest trees were retained and zero length branches were collapsed. A 50% strict consensus tree was computed for the 1000 bootstrap trees. The phylogenetic tree was rooted using mtDNA control-region sequence of P. versicolor. Networks were processed first by the median-joining method (Bandelt et al., 1999) and then by the maximum-parsimony Steiner method (Polzin and Daneschmand, 2003). Networks were constructed using the Network 4.5.0.0 (www.fluxus-engineering.com). 2.3. Estimation of mutation rate, expansion time and divergence time Mutation rate and expansion time of the mtDNA control-region haplotypes in the ring-necked pheasant were computed following Rooney et al. (2001). We estimated the number of nucleotide substitutions per site (d) using the formula: d = (tv + tvR)/m (tv = the number of transversions between P. colchicus and P. versicolor; R = transition/transversion ratio in the ring-necked pheasant; m = length of the sequence). The rate of nucleotide substitution per site per lineage per year is k = d/2T, where T is the divergence time between the ingroup and the outgroup species. The fossil re-

cord suggested that P. colchicus and P. versicolor diverged in the early Pleistocene (Hou, 1993), which spanned 2.05–2.22 million years ago (Cao, 1996a). The mutation rate per nucleotide site per generation is l = kg, where g is the generation time (g = 2.00 years in the ring-necked pheasant, Liu, unpul.). The mutation rate per haplotype is u = ml, and the expansion time in generations is t = s/2u (Rogers and Harpending, 1992). This method was also used by Randi et al. (2003). Divergence times between clades were estimated using the program MDIV (Nielsen and Wakeley, 2001). MDIV applies an isolation-with-migration model to the data and uses a Bayesian approach to simultaneously approximate the posterior distribution of three parameters: divergence time between populations (T = tMDIV/2Ne, where Ne is the effective population size), the migration rate between populations since divergence (M = 2Nem), and the population parameter theta (h = 4Nel, where l is the mutation rate per site per generation). The program was first run using default search settings and default priors (for the parameters of interest, h and T). Then, we set our prior value for T to equal 10 and M to equal 1, because it produced consistent and well-behaved posterior distributions (Spellman and Klicka, 2007). MDIV analyses were run for 5 million generations following a burn-in period of 500,000 generations, and repeated three times to ensure convergence upon the same posterior distributions for each of the parameter estimates. 3. Results 3.1. Haplotype and genetic diversity The mtDNA control-region sequence (1078 nucleotides) alignment of 139 individuals showed 102 different haplotypes (73.38% of all samples), defined by 88 polymorphic sites including 86 substitutions (68 transitions and 18 transversions), and six insertion/deletion. The number of observed haplotype within populations ranged from one in KS to 13 in AHQ (Table 1). The percentages of unique haplotypes per population were calculated by dividing the number of unique haplotypes by the same size. Within each population, this percentage varied from 0.00 in DF and KS to 100% in CX, AX, MQ, GEM, AHQ and ZS. A total of 87 haplotypes (85.29% of all the haplotypes) were unique to the 19 populations. The most common haplotype was C22 shared among eight individuals from five sampling sites (XY, AY, HJQ, DF and YX). The haplo-

Table 1 Number of total haplotypes and of unique haplotype found within each population, pairwise difference (K), nucleotide diversity (p) and haplotype diversity (h) of the 19 populations of P. colchicus. The group name (A, B, C, D and E) corresponds to the clade name in Fig. 2. Subspecies names from Zheng (1978). Population (label, code)

Subspecies name

Group name

Latitude

Kashi (1, KS) Zhaosu (2, ZS) Anxi (3, AX) Jinta (4, JT) Zhangxian (5, ZX) Chengxian (6, CX) Hangjinqi (7, HJQ) Yan’an (8, YA) Weiyuan (9, WY) Dafang (10, DF) Libo (11, LB) Yuexi (12, YX) Xinyang (13, XY) Anyang (14, AY) Aohanqi (15, AHQ) Xianghai (16, XH) Geermu (17, GEM) Minqin (18, MQ) Wuwei (19, WW)

P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c. P.c.

E B, D A A A A, B A A, B C B B B B A, B B B A A A

39° 43° 40° 39° 34° 33° 39° 36° 29° 27° 25° 30° 32° 36° 42° 45° 36° 38° 37°

shawii mongolicus satscheuensis edzinensis strauchi strauchi kiangsuensis kiangsuensis suehschanensis elegans decollatus torquatus torquatus torquatus karpowi pallasi vlangalii sohokotensis strauchi

27.610 9.480 31.760 59.010 50.780 44.560 50.550 35.500 31.390 08.780 24.860 51.390 7.320 5.550 17.050 1.840 24.940 37.320 55.490

Longitude

Sample size

Total haplotypes

Unique haplotypes

K

p

h

75° 59.270 81° 7.650 95° 47.270 98° 53.750 104° 27.890 105° 43.040 108° 43.490 109° 29.750 104° 39.560 105° 36.490 107° 52.690 116° 21.450 114° 3.600 114° 20.490 119° 54.000 122° 24.340 94° 53.480 103° 5.180 102° 37.890

3 3 7 9 10 9 9 10 5 2 4 12 7 10 13 4 2 10 10

1 3 7 4 6 9 9 10 4 2 2 11 4 8 13 3 2 10 6

0 3 7 3 4 9 7 8 3 0 1 7 2 3 13 1 2 10 4

0 14.000 7.903 2.000 3.956 9.444 5.889 9.222 3.400 3.000 1.000 3.682 2.571 4.267 9.727 1.667 7.000 5.933 3.667

0 0.0130 0.0073 0.0019 0.0037 0.0088 0.0055 0.0086 0.0032 0.0028 0.0009 0.0038 0.0024 0.0040 0.0090 0.0016 0.0065 0.0055 0.0034

0 1.000 1.000 0.583 0.778 1.000 1.000 1.000 0.900 1.000 0.500 0.970 0.810 0.933 1.000 0.833 1.000 1.000 0.867

128

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

Fig. 2. Bayesian tree based on GTR + I + G model of the ring-necked pheasant (P. colchicus, 102 haplotypes). The phylogenetic tree was rooted using P. versicolor. Values of posterior probability (first numbers) are shown for the key node more than 0.80, and the bootstrap percentage values computed in maximum-parsimony (second numbers) trees. The numbers in the bracket after the sample name stand for the sample size. Abbreviations of the populations as in Table 1. (A) Western clade; (B) eastern clade; (C) Sichuan Basin clade; (D) Zhaosu clade; (E) Kashi clade. ‘‘–” means that bootstrap values were less than 50%.

type C4 was shared by three populations (ZX, HJQ, YA), C44 and C47 by two populations (YX, AY), C64 by two populations (DF, XH), and C65 by two populations (YA and AY) (Fig. 2). The mean haplotype diversity and nucleotide diversity in all populations were 0.992 ± 0.0027 and 0.0106 ± 0.00536 respectively for the ring-necked pheasant sampled in China. There were 87 unique haplotypes found in nineteen populations, indicating

high levels of genetic variation among the ring-necked pheasant populations in China. Nucleotide diversity among the populations varied from 0.0000 (KS) to 0.0130 (ZS), however, haplotype diversity ranged from 0.000 (KS) to 1.00 (CX, AX, HJQ, MQ, GEM, AHQ, DF, YA and ZS). Intrapopulation pairwise divergence was the lowest (k = 0.000) in KS population and the highest (k = 14.000) in ZS population.

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

3.2. Phylogeographic structure and estimation of mutation rate and divergence time 3.2.1. Phylogeographic structure The Bayesian tree was computed with the best-fit model (GTR + I + G) and was rooted using CR sequence from P. versicolor. The convergence was obtained by a convergence diagnostic (PSRFTL = PSRFr(Ah-iG) = 1.000, PSRFr(Ah-iT) = 1.001, PSRFr(Ah-iC) = 1.006, 1.000, PSRFr(Ch-iG) = 1.000, PSRFr(Ch-iT) = 1.001, PSRFr(Gh-iT) = 1.008, PSRFpi(A) = 1.000, PSRFpi(T) = 1.000, PSRFpi(C) = 1.000, PSRFpi(G) = 1.000, PSRFalpha = 1.000, PSRFpinvar = 1.000. The phylogenetic tree grouped all ring-necked pheasant haplotypes into five clades: clade A, clade B, clade C, clade D and clade E (Fig. 2). Clade A (western clade), with an average intrapopulation nucleotide diversity of 0.00616 and an average intrapopulation haplotype diversity of 0.987, encompassed haplotypes that are present in AX, JT, WW, MQ, GEM, CX, ZX, HJQ and YA. These nine localities were located in the arid and semi-arid climate region in western China, with exception of the haplotype C65 was shared by AY (one sample, located in the monsoon climate area) and YA. Clade B (eastern clade), with an average intrapopulation nucleotide diversity of 0.00426 and an average intrapopulation haplotype diversity of 0.977, contained the haplotypes mostly found in individuals from XH, AHQ, AY, XY, YX, DF and LB (located in the monsoon region), as well as three haplotypes from CX (one sample) and YA (two samples) and two haplotypes from ZS (two samples). The haplotype C22 was shared by DF (one sample), AY (three samples), XY (two samples), YX(one sample) and HJQ (one sample) in clade B. Clade C (Sichuan Basin clade, located in the monsoon region), which could be monophyletic, included haplotypes that were only present in Sichuan Basin. Clade D included only one haplotype from ZS, and clade E also contained one haplotype from KS, which were not supported by posterior probability (Fig. 2). The MP procedure produced the similar topology as the bayesian tree. The result of the haplotype network (Fig. 3) showed that the distribution of haplotypes was consistent with the phylogenetic tree. 3.2.2. Estimation of mutation rate and divergence time Based on the equation: d = (tv + tvR)/m, we computed the average number of nucleotide substitutions per site between P. colchicus and P. versicolor haplotypes and obtained d = 0.083 (tv = 18, R = 4.0, m = 1078). The rate of nucleotide substitution per site per

129

lineage per year (k) ranged from 1.9  108 (T = 2.22 million years) to 2.0  108 (T = 2.05 million years), and the mutation rate per generation (l) ranged from 3.8  108 to 4.0  108. MDIV estimates of divergence time (using our estimates of the mutation rate) suggest that the basal split between western clade and eastern/Sichuan Basin clades occured from 2.1  105 years to 2.2  105 years ago (h = 41.69, t = 0.44). This was followed by the divergence time of the eastern and Sichuan Basin clades from 1.7  105 years to 1.8  105 years ago (h = 22.45, t = 0.64). All estimates of migration rates between clades from the MDIV analyses indicated that fewer than 0.2 (range 0.00001–0.02) individuals migrated between clades as they diverged, and many of the posterior distributions for the estimate of M included zero. 3.3. Gene flow The results of AMOVA showed a low level of gene flow (Nm = 0.44) and significant genetic differentiation (Fst = 0.31, P < 0.001) among all populations. Moreover, most pairwise Fst-values were large and significant which indicated restricted gene flow among these populations, and significant pairwise Fst-values between geographic localities ranged from 0.068 (Anyang-Yuexi) to 0.946 (Kashi-Xianghai) (Table 2). The remaining pairwise Fst-values were not significant which suggests high levels of gene flow. The pairwise population Fst-values were positive correlated with geographic distances between populations (R = 0.296, N = 71, P < 0.001). 3.4. Mismatch distributions and expansion time The mismatch distribution for the total samples, Gray-rumped Pheasants, western and eastern groups was bell-shaped as expected under the sudden expansion model (Fig. 4). The observed distribution was not ragged, and the goodness-of-fit statistics were not significant (Table 3). Thus, a fitting of the observed to the expected distribution under the sudden expansion model cannot be rejected. The estimated values of h after the expansion (h1) was higher than that before the expansion (h0), which were similar in the ring-necked pheasant from western group, eastern group and Sichuan Basin (Table 3). The mismatch analyses of Zhaosu and Kashi were not estimated because of small sample sizes.

Fig. 3. Haplotype network estimated from the P. colchicus control-region data. Small black circles indicated missing haplotypes that were not observed. Distances between linked haplotypes corresponded to one mutation, except that with the number of mutations shown. The size of circles denotes the haplotype frequency, and the color of circles is shown in Fig. 2.

130

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

Table 2 Population pairwise Fst-value (above the diagonal) and gene flow (blow the diagonal) for comparisons between populations. DF DF LB WY CX ZX WW YX XY AY HJQ YA AHQ XH GEM MQ AX JT ZS KS

LB

WY

CX a

7.200 0.162 0.169 0.510 0.168 Inf. 15.050 Inf. 0.432 1.156 9.037 8.000 0.217 0.202 0.402 0.102 Inf. 0.033

0.065 0.755 0.800c 0.124 0.139 0.210 0.357 0.418 0.138 0.200 1.043 0.168 0.653 0.128 1.212 0.218 0.307 0.348 0.650 0.574 1.078 0.174 1.692 0.111 0.132 0.158 0.162 0.204 0.279 0.376 0.090 0.149 2.684 0.531 0.035 0.083

ZX c

0.747 0.782b 0.704c 4.333 5.166 0.159 0.137 0.212 1.459 1.971 0.158 0.129 0.391 0.831 3.429 0.746 0.415 0.120

0.495 0.583c 0.544c 0.103a 3.339 0.295 0.325 0.401 2.234 3.465 0.293 0.337 1.573 3.418 11.620 1.481 1.024 0.301

WW

YX a

0.749 0.022 0.783c 0.324c 0.714c 0.748c 0.088a 0.759c 0.130c 0.629c 0.746c 0.170 0.137 5.435 0.218 6.851 1.708 0.325 2.071 0.555 0.164 1.123 0.132 0.918 0.360 0.151 0.779 0.166 2.851 0.256 0.686 0.141 0.426 2.220 0.117 0.104

XY

AY

0.032 0.098 0.434b 0.292a 0.796c 0.670c 0.785c 0.703c 0.606c 0.555c 0.785c 0.697c 0.084 0.068a 0.018 Inf. 0.293 0.429 0.586 0.779 1.380 0.967 0.503 0.855 0.122 0.211 0.164 0.210 0.254 0.338 0.105 0.192 1.742 3.017 0.063 0.144

HJQ

YA a

0.536 0.620c 0.589c 0.255c 0.183c 0.226c 0.606c 0.630c 0.538c Inf. 0.286 0.278 0.649 0.735 10.33 2.986 1.312 0.235

0.302 0.435b 0.466c 0.202c 0.126a 0.194c 0.473c 0.460c 0.391b 0.040 0.494 0.607 1.364 0.956 9.094 3.352 3.527 0.405

AHQ 0.052 0.317c 0.741b 0.759c 0.630c 0.753c 0.308c 0.266c 0.341c 0.636c 0.503c 1.929 0.202 0.177 0.250 0.150 1.154 0.121

XH 0.059 0.228 0.818b 0.795b 0.597c 0.791c 0.353c 0.498c 0.369c 0.642c 0.452c 0.206a 0.115 0.152 0.259 0.081 1.692 0.028

GEM 0.697 0.791 0.760 0.561a 0.241 0.581a 0.768a 0.804b 0.704a 0.435c 0.268c 0.712a 0.812a 0.584 0.947 0.182 1.019 0.067

MQ

AX a

0.712 0.756c 0.711c 0.376c 0.128a 0.391c 0.751c 0.753c 0.704c 0.405c 0.343c 0.738c 0.766c 0.461a 1.837 0.502 0.410 0.148

JT a

0.554 0.641c 0.571c 0.127a 0.041 0.149b 0.662c 0.663c 0.596c 0.046 0.052 0.667c 0.659c 0.345a 0.214b 4.633 1.054 0.258

ZS c

0.830 0.180 0.847c 0.157 0.771c 0.485b c 0.401 0.546a 0.252c 0.328b 0.422c 0.540a 0.780c 0.184 0.827c 0.223a 0.723c 0.142 0.143c 0.276a c 0.130 0.124 0.770c 0.302c 0.860b 0.228 0.733a 0.329 0.499c 0.549c 0.097a 0.322b 0.582c 0.359 0.067 0.389

KS 0.939 0.934a 0.857 0.807b 0.624c 0.810c 0.827c 0.887b 0.777b 0.680b 0.552c 0.805c 0.946b 0.881b 0.771c 0.659b 0.881c 0.562

Abbreviations as in Table 1. Inf., the value of gene flow is infinity. a 0.01 < P < 0.05. b 0.001 < P < 0.01. c P < 0.001.

We computed s = 2 ut = 4.003 in the ring-necked pheasant from west, a value that can be used to estimate an expansion time of t = s/2u = 4.003/(2  4.0  108  1078) = 46,417 generations (92,834 years) to t = s/2u = 4.003/(2  3.8  108  1078) = 48,860 genera-

tions (97,720 years). The expansion time of east was from 52,250 generations (104,500 years) to 55,000 generations (110,000 years), and the expansion time of Sichuan Basin ranged from 52,134 generations (104,268 years) to 54,877 generations (109,754 years).

Fig. 4. Mismatch distributions for the ring-necked pheasant. (A) All samples; (B) Gray-rumped Pheasants; (C) western group; (D) eastern group.

131

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132 Table 3 The statistics of mismatch analyses. Group

s

h0

h1

SSD

P-value

RI

P-value

All samples West East Sichuan Basin

2.904 4.003 4.506 4.496

12.545 2.577 0.207 0.000

99999.000 99999.000 47.812 99999.000

0.0104 0.00178 0.00220 0.0796

0.300 0.504 0.459 0.197

0.00316 0.00855 0.0189 0.310

0.700 0.545 0.438 0.375

SSD, sum of squared deviations; RI, raggedness index.

4. Discussion High levels of genetic diversity were found in the ring-necked pheasant populations. We hypothesize two main factors that are the most likely reasons for genetic structure of this species. First, P. colchicus distributes widely in China with wide ecological differentiations. The pheasant inhabits grassland, shrubby areas, farmland and the edge of forest. High levels of genetic diversity in natural populations can be associated with a wide range ecological types and niche variation (Prentice et al., 1995). Secondly, habitat fragmentation and deterioration resulting from Quaternary glaciations and following the arid trend and human activities of the northern china are most likely factors leading to high genetic diversity and high population differentiation in this species (Table 2). The phylogenetic tree indicated the existence of a major genetic gap between White-winged Pheasants and Gray-rumped Pheasants and Kirghiz Pheasants (Fig. 2). Such a disjunction suggests that populations of the species from these regions have remained isolated throughout the Pleistocene irrespective climatic oscillations as well as the desertification of north China. No haplotype was shared among all 19 populations. The high level of interpopulation differentiation recorded in P. colchicus is most likely due to populations being separated by geographical barriers among different regions. Thus, the Taklimakan Desert, Bardain Jaran Desert, Gurbantonggut Desert and Tengger Desert, which began to expand during Quaternary (Zhang and Men, 2002; Sun et al., 2008; Shi et al., 2006a; Wu et al., 2002), were likely to have imposed significant geographical barriers to gene flow (Nm = 0.59, Fst = 0.46, P < 0.0001) between White-winged Pheasants and the other two pheasant groups. In the phylogenetic tree, two haplotypes from Zhaosu (Kirghiz Pheasants) was found in clade B whose haplotypes belonged to Gray-rumped Pheasants. It is not surprising that high level of gene flow (Nm = 17.24, Fst = 0.03, P > 0.05) was shown between Grayrumped Pheasants and Kirghiz Pheasants. The explanation on the phylogeographical structure of the two pheasant groups should be traced back to the early Pleistocene. In the early Pleistocene, northwestern China (Sinkiang, Chaidamu Basin and western Gansu) was covered with warm-temperate sparse forest-steppe (Wang, 1997) which was livable for the ring-necked pheasant. However, because of the uplift of Qinghai-Tibet Plateau obstructing the warm-wet airflow northward, Sinkiang become more and more arid and several deserts emerged (Wang, 1997), which resulted in the isolation of the pheasant so that Kirghiz Pheasants evolved independently in Zhaosu. Within the Gray-rumped Pheasants group, all populations sampled had no strong geographical pattern except that of Weiyuan from Sichuan Basin (Fig. 2). However, all samples of Gray-rumped Pheasants could be roughly divided into three groups: western group, eastern group and Sichuan Basin group based on the phylogenetic tree (Fig. 2) and the haplotype network (Fig. 3). The phylogeographical structure could be concordant with the uplift of Qinghai-Tibet Plateau, Qinling Mountains and Liupan Mountains. According to the estimated divergence time of CR haplotypes, the divergence between west and east and Sichuan Basin occurred in the forepart of the late Pleistocene (0.15–0.24 million years, the

forth cold period) when the climate in western China was dry and cold (Cao, 1996a). At that time, Qinghai-Tibet Plateau had achieved the current altitude (Li et al., 1979; Li and Fang, 1998) and obstructed northward flow of Indian Ocean warm-wet air. The existence of Liu Pan Mountain (north and south alignment) and Qinling Mountains (west and east alignment) obstructed the westward flow of the wet southeastern monsoon which aggravated the aridity of western China (Liu et al., 1988) and resulted in the formation of the arid and semi-arid region and the monsoon region. Moreover, the vegetation in western China was mostly dry grassland shrub (Cao, 1996b; Li, 1998; Liu et al., 1999; Zhang and Shang, 1991) which was livable for the pheasant. However, in eastern China, affected by wet southeastern monsoon with warm and wet climate, the forest were well developed which resulted in barriers for pheasants distributions. Hence, the differences in habitats and selective pressure resulted in the pheasant evolving independently in eastern and western China. In the monsoon region, the pheasant can be roughly divided into two groups: Sichuan Basin and eastern groups. The divergence time between the two groups was from 0.17 to 0.18 Ma, i.e. in the onset of Riss glaciations (200,000–135,000 years ago). In China the formation of glaciations was mostly mountainous glacier. We can speculate that Sichuan Basin was the refuge for the ring-necked pheasant during the last glaciations. Sichuan Basin was also surrounded by mountains and a mass of forests (Liu, 1983) which restricted the gene flow between Sichuan Basin and eastern populations (Nm = 0.40, Fst = 0.56, P < 0.001). The pheasants were isolated in basins and evolved independently. Overall, this study proves for the first time that a deep genetic divergence (Table 2) is present for a species widely distributed across China, which arose through the continuous interruption of gene flow by the higher mountains (e.g. Qinling Mountains) in the central china during most of the Pleistocene. This finding confirmed the important role of the global climate cooling during the late Tertiary/Early Pleistocene in the divergence of a few species over the North Hemisphere. The west-east genetic boundary found in the present study may also exist for other plants or animals distributed over China because they should have experienced the same continuous cooling period and gene flow barrier through the Pleistocene. In Gray-rumped Pheasants group, low nucleotide diversity, high haplotype diversity and unimodal mismatch distributions (Sichuan Basin group was not shown) in western, eastern and Sichuan Basin groups are consistent with recent population expansions in these three groups (Tables 1 and 3) (Fig. 4). The recent expansion hypothesis is also supported by the common haplotype (C22) shared between geographically distinct regions (within clade B) and by the non-significant mismatch distribution analysis indicating a demographic population expansion. However, the shared haplotypes could be due to retention of ancestral haplotypes. Coalescent theory (Crandall and Templeton, 1993) predicts that the most frequent and widespread haplotypes are ancestral. The most frequent haplotype (C22) would result from the retention of ancestral haplotype. Based on the expansion time, the expansion of the ring-necked pheasant sampled from west, east and Sichuan Basin in China occurred in the metaphase of the late Pleistocene (the

132

J. Qu et al. / Molecular Phylogenetics and Evolution 52 (2009) 125–132

forth warm period) (Cao, 1996a) with the occurrence of Riss-Würm interglaciation when the climate was warm and wet in China. Western China was covered with the forest grassland which was livable for the ring-necked pheasant. In Sichuan Basin, the forests ascended along the hillside and the shrub and grass developed in the low altitude where the pheasant could inhabit. Therefore, it is not surprising that the analyses of demographic expansion indicated that the samples from west and Sichuan Basin experienced the stage of strong population growth. Comparatively, the samples from east experienced slower population growth resulting from the better development of the woods than the steppes (Cao, 1996b). The observed phylogeographical structuring failed to support the described subspecies of the ring-necked pheasant (Zheng, 1978), except for P.c. suechschanensis (sampled from Weiyuan) and P.c. shawii (sampled from Kashi), which maybe distinctive enough for subspecies. Further studies are required to verify the distinctiveness of other subspecies. Acknowledgments We thank Dr. Jicheng Liao, Dr. Bing Tian and Dr. Dongrui Jia for help in amending the figures in this study. We are especially grateful to the two anonymous reviewers for helpful comments on a previous version of this manuscript and Dr. Weihong Ji (College of Sciences, Massey University) for help in improving the English. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ympev.2009.03.015. References Arbogast, B., Kenagy, G.J., 2001. Comparative phylogeography as an integrative approach to history biogeography. J. Biogeogr. 28, 819–825. Aurelle, D., Berrebi, P., 2001. Genetic structure of brown trout (Salmotrutta L.) populations from south-western France: data from mitochondrial control region variability. Mol. Ecol. 10, 1551–1561. Avise, J.C., 1994. Molecular Markers, Natural History and Evolution. Chapman and Hall, New York. Avise, J.C., 2000. Phylogeography: The History and Formation of Species. Harvard University Press, Cambridge. Bandelt, H.J., Forster, P., Rohl, A., 1999. Median-joining networks for inferring intraspecific phylogenies. Mol. Biol. Evol. 16, 37–48. Hewitt, G.M., 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biol. J. Linn. Soc. 58, 247–276. Brown, W.M., George, M., Wilson, A.C., 1979. Rapid evolution of animal mitochondrial DNA. Proc. Natl. Acad. Sci. USA 76 (4), 1967–1971. Cao, X.S., 1996a. The division of Quaternary climate in Gansu Province. Arid Zone Res. 13, 28–40. Cao, X.S., 1996b. The Quaternary paleogeographic environment of Gansu Province. Acta Geol. Gansu 5, 13–26. Crandall, K.A., Templeton, A.R., 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny construction. Genetics 134, 959–969. Hewitt, G.M., 2000. The genetic legacy of the Quaternary ice ages. Nature 405, 907– 913. Hill, D., Robertson, P., 1988. The Pheasant Ecology. Management and Conservation. BSP Professional Books, Oxford. Hou, L.H., 1993. Avian fossils of Pleistocene from Zhoukoudian. Mem. Inst. Vert. Palaeont. Palaeoanthr. Acad. Sinica 19, 165–297. Huelsenbeck, J.P., Ronquist, F., 2001. mrbayes: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755. Johnsgard, P.A., 1999. The Pheasants of the World. Smithsonian Institution Press. Li, J.J., Wen, S.X., Zhang, Q.S., 1979. A discussion on the period, amplitude and type of the uplift of the Qinghai-Xizang Plateau. Sci. Sinica 22, 1314–1328.

Li, J.J., Fang, X.M., 1998. The research on the uplift of Qinghai-Xizang Plateau and the changes of environment. Sci. Sinica 43, 1569–1574. Li, W.Y., 1998. The Vegetation and Environment of Quaternary. Science Press, Beijing. Liu, N.F., Huang, Z.H., Wen, L.Y., 2004. Subspecies divergence of Przewalski’s rock partridge (Alectoris magna) description of a new subspecies. Acta Zootaxon. Sinica 29, 600–605. Liu, N.F., Sun, H.Y., 1992. The geographic variance and classification of Phasianus colchicus strauchi. J. Lanzhou Univ. 28, 135–139. Liu, T.S., Ding, Z.L., Rutter N., 1999. Comparison of Milankovitch periods between continental loess and deep sea records over the last 2.5 Ma. Quaternary Sci. Rev. 18, 1205–1212. Liu, N.F., Wang, X.T., Luo, W.Y., Chang, C., 1988. Avifauna study of the Liupan Mountain in Ningxia. J. Lanzhou Univ. (Natural Science Edition) 24, 63–75. Liu, X.S., 1983. Quaternary of Sichuan Basin. Sichuan Science Press, Sichuan, China. Moritz, T., Dowling, T.E., Brown, W.M., 1987. Evolution of animal mitochondrial DNA: relevance for population biology and systematics. Annu. Rev. Ecol. Evol. S. 18, 269–292. Nielsen, R., Wakeley, J., 2001. Distinguishing migration from solation: a Markov chain Monte Carlo approach. Genetics 158, 885–896. Nylander, J.A.A., 2004. MRMODELTEST, Version 2.0. Program distributed by the author. Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden. Available from: . Polzin, T., Daneschmand, S.V., 2003. On Steiner trees and minimum spanning trees in hypergraphs. Oper. Res. Lett. 31, 12–20. Prentice, H.C., Lonn, M., Lefkovitch, L.P., Runyeon, H., 1995. Associations between alleles frequencies in Festuca ovina and habitat variation in the alvar grasslands on the Baltic island of Oland. J. Ecol. 83, 391–402. Qian, H., Ricklefs, R.E., White, P.S., 2005. Beta diversity of angiosperms in temperate floras of east Asia and east North America. Ecol. Lett. 8, 15–22. Randi, E., Lucchini, V., 1998. Organization and evolution of the mitochondrial DNA control-region in the avian genus Alectoris. J. Mol. Evol. 47, 449–462. Randi, E., Tabarroni, C., Rimondi, S., Lacchini, V., Sfougaris, A., 2003. Phylogeography of the rock partridge (Alectoris graeca). Mol. Ecol. 12, 2201–2214. Recueroc, E., Martinez-Solano, I., Parra-Olea, G., Garcia-Paris, M., 2006. Phylogeography of Pseudacris regilla (Anura:Hylidae) in western North America, with a proposal for a new taxonomic rearrangement. Mol. Phylogenet. Evol. 39, 239–304. Rogers, A.R., Harpending, H., 1992. Population growth makes waves in the distribution of pairwise genetic differences. Mol. Biol. Evol. 9, 552– 569. Rooney, A.P., Honeycutt, R.L., Derr, J.N., 2001. Historical population size change of bowhead whales inferred from NDA sequence polymorphism data. Evolution 55, 1685–1687. Rozas, J., Sanchez-DelBarrio, J.C., Messeguer, X., Rozas, R., 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19, 2496–2497. Schneider, S., Roessli, D., Excoffier, L., 2002. Arlequin Version 2.0. A Software for Population Genetics Data Analysis. Genetics and Biometry Laboratory, Department of Anthropology and Ecology, University of Geneva, Geneva, Switzerland. Shi, Z.T., Song, Y.G., An, Z.S., 2006a. Evolution of Gurbantunggut Desert recorded by Tianshan Loess. J. Desert Res. 26, 675–679. Spellman, G.M., Klicka, J., 2007. Phylogeography of the white-breasted nuthatch (Sitta carolinensis): diversification in North American pine and oak woodlands. Mol. Ecol. 16, 1729–1740. Sun, Q.F., Chen, F.H., Li, X.Z., 2008. Review and discussion about the progress of Quaternary research of the Badain Juran Desert, China. Arid Zone Res. 25, 304– 310. Swofford, D.L., 1998. PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods), Version 4b10. Sinauer Associates, Sunderland, MA. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 24, 4876– 4882. Wang, J.W., 1997. Paleogeographical processes in the formation of China arid land. J. Arid Land Resour. Environ. 11, 1–8. Willis, K.J., Whittaker, R.J., 2002. Species diversity-scale matters. Science 295, 1245– 1248. Wu, G., Pan, B., Guan, Q., Gao, H., 2002. The maximum glaciation and desert expansion in China during MIS16. J. Glaciol. 24, 544–549. Zhang, H.C., Shang, Y.Y., 1991. The partition of loess time and the character of ancient climate. J. Lanzhou Univ. (Nat. Sci.) 27, 131–138. Zhang, H.Y., Men, G.F., 2002. Startigraphic subdivision and climatic change of the Quaternary of the center Taklimakan Desert. Xinjiang Geol. 20, 256–261. Zheng, Z.X., 1978. Fauna Sinica. Aves. Galliformes, vol. 4. Science Press, Beijing, China.