Phylogeny and evolutionary histories of Pyrus L. revealed by phylogenetic trees and networks based on data from multiple DNA sequences.

YMPEV 4969

No. of Pages 12, Model 5G

31 July 2014 Molecular Phylogenetics and Evolution xxx (2014) xxx–xxx 1

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev 6 7

Phylogeny and evolutionary histories of Pyrus L. revealed by phylogenetic trees and networks based on data from multiple DNA sequences

3 4 5 8 9 10 11 12 13 15 14 16 1 1 8 3 19 20 21 22 23 24 25 26 27 28 29 30

Q1

Xiaoyan Zheng a,b,1, Danying Cai a,1, Daniel Potter c, Joseph Postman d, Jing Liu a, Yuanwen Teng a,⇑

Q2

a Department of Horticulture, The State Agricultural Ministry Key Laboratory of Horticultural Plant Growth, Development and Quality Improvement, Zhejiang University, Hangzhou, Zhejiang 310058, China b Institute of Horticulture and Landscape, College of Ecology, Lishui University, Lishui, Zhejiang 323000, China c Department of Plant Sciences, University of California, Mail Stop 2, Davis, CA 95616-8780, USA d National Clonal Germplasm Repository, USDA-ARS, Corvallis, OR 97333, USA

a r t i c l e

i n f o

Article history: Received 2 December 2013 Revised 10 July 2014 Accepted 17 July 2014 Available online xxxx Keywords: Pyrus Phylogenetic network Rapid radiation Reticulation DNA sequences

a b s t r a c t Reconstructing the phylogeny of Pyrus has been difficult due to the wide distribution of the genus and lack of informative data. In this study, we collected 110 accessions representing 25 Pyrus species and constructed both phylogenetic trees and phylogenetic networks based on multiple DNA sequence datasets. Phylogenetic trees based on both cpDNA and nuclear LFY2int2-N (LN) data resulted in poor resolution, especially, only five primary species were monophyletic in the LN tree. A phylogenetic network of LN suggested that reticulation caused by hybridization is one of the major evolutionary processes for Pyrus species. Polytomies of the gene trees and star-like structure of cpDNA networks suggested rapid radiation is another major evolutionary process, especially for the occidental species. Pyrus calleryana and P. regelii were the earliest diverged Pyrus species. Two North African species, P. cordata, P. spinosa and P. betulaefolia were descendent of primitive stock Pyrus species and still share some common molecular characters. Southwestern China, where a large number of P. pashia populations are found, is probably the most important diversification center of Pyrus. More accessions and nuclear genes are needed for further understanding the evolutionary histories of Pyrus. Ó 2014 Elsevier Inc. All rights reserved.

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

48 49

1. Introduction

50

The genus Pyrus L. belongs to subtribe Malinae (designated Pyrinae in Potter et al., 2007 and corresponding to the former Maloideae) of the family Rosaceae. Based on fossil evidence, Pyrus has been thought to be of Tertiary or earlier origin, later than the genera Malus Mill. and Crataegus L., in the mountainous regions of western and southwestern China (Rubtsov, 1944). Pyrus species are geographically divided into occidental pears and oriental pears. The former are distributed in Europe, northern Africa, Asia Minor, Iran, Central Asia and Afghanistan, while the latter occur from Tian-Shan and Hindu Kush Mountains in Central Asia eastward to Japan (Rubtsov, 1944). The identification and description of Pyrus species have relied on traditional morphological characters, mainly from fruits and leaves. Most occidental species have five carpels with a persistent

51 52 53 54 55 56 57 58 59 60 61 62 63

⇑ Corresponding author. 1

E-mail address: [email protected] (Y. Teng). X. Zheng and D. Cai contributed equally to this work.

calyx, while most oriental ones display two to five carpels with a deciduous calyx (Challice and Westwood, 1973). However, absence of natural barriers for self-incompatibility and inter-specific hybrid fertility facilitate inter-specific hybridization in Pyrus (Bell and Hough, 1986), leading to diverse wild types with intermediate morphologies. Wide distributions and extensive hybridization make it difficult to identify and distinguish Pyrus species by morphological characters; thus, the number of species recognized in Pyrus has varied among taxonomists, from 60 to 21 (Rubtsov, 1944), and more than 900 Pyrus species names have been recorded (http://www.ipni.org). However, the number of primary (i.e., not of hybrid origin) species has been relatively consistent among different treatments, and about 20 putative primary species were revised by Teng et al. (2004). For the oriental group, 13 species native to China have been accepted by most Chinese taxonomists (Yu, 1979), among which five are putative primary species, five were putative inter-specific hybrids with their parentages uncertain, and three were newly reported species (Yu and Kuan, 1963). Besides, there are four other primary species from Japan, Korean Peninsula and Taiwan Island (Table 1). The occidental species have

http://dx.doi.org/10.1016/j.ympev.2014.07.009 1055-7903/Ó 2014 Elsevier Inc. All rights reserved.

Please cite this article in press as: Zheng, X., et al. Phylogeny and evolutionary histories of Pyrus L. revealed by phylogenetic trees and networks based on data from multiple DNA sequences. Mol. Phylogenet. Evol. (2014), http://dx.doi.org/10.1016/j.ympev.2014.07.009

64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83

YMPEV 4969


31 July 2014 2

X. Zheng et al. / Molecular Phylogenetics and Evolution xxx (2014) xxx–xxx

Table 1 Taxon information, cpDNA haplotypes and LFY2int2 copy types for each accession. Taxona Oriental group P.bretschneideri Chinese white pear

P. pyrifolia*

P.sinkiangensis P. ussuriensis*

P. pashia*

P. dimorphophylla*

P. calleryana*

P. fauriei* P. koehnei* P. betulaefolia*

P.serrulata P. pseudopashia# P. xerophila#

P.hopeiensis P.phaeocarpa P. hondoensis* Occidental group European species P. communis* P. caucasica*

P. pyraster*

Accession

trnL-Fb

accD-psaIb

LFY2int2c

Origin

Leaf sourced

‘Guanli’ ‘Jinhuali’ ‘Yali’ ‘Dangshansuli’ ‘Cili’ ‘Fenhongxiao’ ‘Dongguoli’ ‘Chojuro’ ‘Nijisseiki’ ‘Imamuraaki’ ‘Huobali’ ‘Yiwulizi’ P. pyrifolia1 P. pyrifolia2 P. pyrifolia3 ‘Korlaxiangli’ P. ussuriensis1 P. ussuriensis2 ‘Xiaoxiangshui’ ‘Daxiangshui’ ‘Jianbali’ ‘Nanguoli’ ‘BeiJingbaili’ ‘Yaguang’ ‘Manyuanxiang’ ‘Ruanerli’ P. pashia1 P. pashia2 P. pashia3 P. pashia4 P. pashia5 P. pashia6 P. pashia7 P. pashia8 P. pashia9 P. dimorphophylla4 P. dimorphophylla5 P. dimorphophylla6 P. calleryana1 P. calleryana2 P. calleryana3 P. calleryana4 P. fauriei P. koehnei P. betulaefolia1 P. betulaefolia2 P. betulaefolia3 P. betulaefolia4 P. serrulata1 P. serrulata2 P. pseudopashia1 P. xerophila1 P. xerophila2 P. xerophila3 P. hopeiensis1 P. hopeiensis2 P. phaeocarpa P. hondoensis

tH1 tH5 tH4 tH4 tH5 H tH3 tH4 tH4 tH4 tH5 tH4 tH3 tH5 tH4 tH5 tH1 tH5 tH1 tH4 tH5 tH5 tH5 tH5 tH5 tH5 tH5 tH7 tH4 tH4 tH5 tH7 tH5 tH7 tH5 tH1 tH3 tH1 tH5 tH5 tH5 tH5 tH4 tH5 tH3 tH5 tH2 tH6 tH3 tH3 tH5 tH5 tH5 tH5 tH1 tH1 tH1 tH5

aH8 aH9 aH1 aH1 aH9 aH1 aH2 aH1 aH1 aH1 aH12 aH1 aH2 aH11 aH2 aH9 aH8 aH9 aH8 aH1 aH9 aH9 aH9 aH9 aH9 aH9 aH10 aH12 aH12 aH1 aH10 aH10 aH14 aH10 aH13 aH5 aH2 aH6 aH8 aH8 aH8 aH8 aH1 aH4 aH7 aH8 aH15 aH3 aH2 aH2 aH8 aH8 aH8 aH8 aH8 aH8 aH8 aH9

LN LN, LD2 Del2 LN, LD2 N LN, LD2, LS LN LN LN LN LN LN LN LN LN LN, LD2 LN,LI8 LN LI8 LI8 LN LN, LI8 LN LN LN LN LN LN LN LS LN LN LN LN LN LN LN LN LN, LD2 LS LN, LD2 LS, LD2 LS LN LN LN LN LS LN LN LN, LD2 LN LN LN LN, LI8 LN LN LN, LI8

Hebei, China Sichuan, China Hebei, China Anhui, China Shandong, China Gansu, China Gansu, China Kanagawa Pref. Japan Chiba Pref. Japan Kochi Pref. Japan Yunnan, China Zhejiang, China South China Yunnan, China Yunnan, China Xinjiang, China Liaoning, China Liaoning, China Liaoning, China Liaoning, China Liaoning, China Liaoning, China Beijing, China Liaoning, China Liaoning, China Gansu, China Yunnan, China Yunnan, China Yunnan, China Nepal Yunnan, China Yunnan, China Yunnan, China Yunnan, China Yunnan, China Mie Pref. Japan Mie Pref. Japan Mie Pref. Japan South China Zhejiang, China Zhejiang, China Zhejiang, China Korea South China, Taiwan Gansu, China Ningxia, China Northeastern China Henan, China Hubei, China Hubei, China Yunnan, China Gansue, China Gansu, China Gansu, China Hebei, China Hebei, China North China Middle Japan

Hebei, China CPGR TU ZZFI CPGR Gansu, China Gansu, China TU TU TU CPGR CPGR TU Yunnan, China Yunnan, China CPGR CPGR CPGR CPGR Jilin, China TU TU Beijing, China Liaoning, China Liaoning, China Gansu, China HRIYN HRIYN HRIYN TU HRIYN Yuannan, China Yuannan, China Yuannan, China Yuannan, China TU TU TU HRIYN Zhejiang, China Zhejiang, China Zhejiang, China TU TU CPGR CPGR TU Henan, China CPGR Hubei, China HRIYN GPI GPI GPI Hebei, China Hebei, China CPGR TU

P. P. P. P. P. P. P.

tH8 tH13 tH8 tH8 tH5 tH8 tH8

aH18 aH17 aH18 aH18 aH8 aH30 aH17

LN LN LN LN LN LN LN

unknown Russia, wild Armenia Russia Ukraine Romania Iran

NCGR NCGR, NCGR, NCGR, NCGR, NCGR, NCGR,

communis caucasica694 caucasica2816 caucasica687 caucasica684 pyraster1671 pyraster1288

PI = 440631 PI = 638008 PI = 337437 PI = 322285 PI = 506380 PI = 132094


YMPEV 4969


31 July 2014 3

X. Zheng et al. / Molecular Phylogenetics and Evolution xxx (2014) xxx–xxx Table 1 (continued) Taxona P. nivalis

*

P. cordata*

West Asian species P. elaeagrifolia*

P. spinosa* (P. amygdaliformis)

P. regelii*

P. salicifolia*

P. syriaca*

North African species P. cossonii* (P. longipes)

P. gharbiana*

P. mamorensis*

Outgroup species M. sieboldii M. rockii M. domestica M. domestica subsp. chinensis Li Y. N.

Accession

trnL-Fb

accD-psaIb

LFY2int2c

Origin

Leaf sourced

P. P. P. P. P. P. P. P. P.

nivalis1196 nivalis1590 nivalis725 nivalis1714 nivalis256 cordata1588 cordata739 cordata745 cordata750

tH8 tH10 tH8 tH8 tH8 tH8 tH8 tH12 tH12

aH17 aH27 aH17 aH18 aH17 aH31 aH17 aH17 aH17

LN, LI3 LN LN LN LN LN LN LN LN

Netherlands Yugoslavia Europe, Pure Uzebekistan Netherland Turkey Turkey France, France

NCGR, PI = 541864 NCGR, PI = 541867 NCGR, PI = 541859 NCGR, PI = 541869 NCGR, PI = 312144 NCGR,PI = 541590 NCGR, PI = 541571 NCGR, PI = 617506 NCGR, PI = 541580

P. P. P. P. P.

elaeagrifolia768 elaeagrifolia1490 elaeagrifolia904 elaeagrifolia1711 elaeagrifolia2817

tH10 tH10 tH8 tH8 tH8

aH28 aH22 aH17 aH18 aH17

LI3 N LI3 LN LN

Turkey Turkey glabrous Wild Hungary Uzbekistan hybrid Armenia

NCGR, NCGR, NCGR, NCGR, NCGR,

PI = 541613 PI = 617525 PI = 312154 PI = 541630 PI = 638009

P. P. P. P. P. P. P. P. P. P. P. P. P. P. P. P. P.

spinosa1598 spinosa1608 spinosa1610 spinosa1615 regelii890 regelii2513 regelii2587 salicifolia869 salicifolia2382 salicifolia2797 salicifolia2849 salicifolia2720 syriaca908 syriaca914 syriaca920 syriaca2716 syriaca1011

tH8 tH8 tH16 tH8 tH15 tH15 tH15 tH8 tH8 tH8 tH8 tH9 tH14 tH8 tH8 tH14 tH8

aH31 aH32 aH17 aH17 aH16 aH18 aH18 aH17 aH17 aH18 aH17 aH17 aH18 aH23 aH24 aH18 aH29

LN LN LN LN LN LN LN LN LN LN LN LN LN LN LN LN LN

Sardinia, Italy Macedonia, Greece Turkey Macedonia Russia Kazakhstan Kazakhstan uncertain artificial hybrid Armenia Georgia Russia Israel Israel Armenia Wild Israel Iran

NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR,

PI = 540934 PI = 540937 PI = 540939 PI = 349023 PI = 541945 PI = 617589 PI = 641279 PI = 541950 PI = 617559 PI = 638000 PI = 657921 CPYR = 2720 PI = 541975 PI = 541977 PI = 312149 PI = 641281 PI = 541797

P. P. P. P. P. P. P. P. P. P. P. P.

cossonii753 cossonii828 cossonii829 cossonii 830 gharbiana787 gharbiana794 gharbiana789 gharbiana790 mamorensis834 mamorensis835 mamorensis837 mamorensis1622

tH8 tH8 tH8 tH8 tH10 tH10 tH10 tH10 tH8 tH11 tH8 tH8

aH21 aH21 aH25 aH25 aH26 aH26 aH26 aH26 aH20 aH19 aH20 aH20

LN, LI3 LN LN LN LN LN LN LN, LI3 LN LN LN LN

Morocco uncertain uncertain uncertain Morocco Morocco Morocco Morocco Morocco Morocco Morocco Uncertain

NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR, NCGR,

PI = 316552 PI = 541592 PI = 541593 PI = 541592 PI = 541662 PI = 541666 PI = 541663 PI = 541664 PI = 541843 PI = 541844 PI = 541846 PI = 541855

tH17 tH17 tH19 tH18

aH33 aH33 aH34 aH35

/ / / /

Yunnan, China Yunnan, China USA North China

HRIYN HRIYN GPI CPGR

M. sieboldii M. rockii ‘Ralls’ ‘Nai’

a,*

Primary species. Newly identified species. b Haplotypes of trnL-F (tH) and accD-psaI (aH) for each accession. c Different copy types of LFY2int2. ‘LN’: LFY2int2-N, ‘LD2’: LFY2int2-Del2, ‘LI3’: LFY2int2-Ins3, ‘LI8’: LFY2int2-Ins8, ‘LS’: LFY2int2-S. d TU: Tottori University, Japan; CPGR: China Pear Germplasm Repository, Xingcheng, Liaoning Province; ZZFI: Zhengzhou Fruit Institute, Chinese Academy of Agriculture Science, Zhengzhou, Henan Province, China. GPI: Gansu Pomology Institute, Gansu Academy of Agricultural Science, Gansu Province, China; HRIYN: Horticultural Research Institute, Yunnan Academy of Agricultural Sciences, Kunming, Yunnan Province, China; NCGR: National Clonal Germplasm Repository, Oregon, USA. #

84 85 86 87 88 89 90 91 92 93 94

been geographically divided into three subgroups – West Asian species, European species and North African species (Challice and Westwood, 1973), and 12 primary species were proposed. As shown in Table 1, this study covered most of the primary species and putative hybrids for oriental group and all of the primary species for occidental group. During the last 50 years, scientists have not stopped their efforts to understand evolutionary relationships among Pyrus species. Challice and Westwood (1973) conducted phylogenetic studies of Pyrus using combined data of a number of chemical and botanical characters, but five species were not included in the final

analyses due to lack of data, and some of the findings revealed by combined data and chemical data were controversial. Pollen structure (Westwood and Challice, 1978; Zou et al., 1986) and isozyme Q3 markers (Jang et al., 1992) have been used for characterizations and identification of hybrids in some species, but these data were not informative or consistent enough to be used reliably for phylogeny reconstruction. DNA markers such as RFLP (Iketani et al., 1998; Kawata et al., 1995), RAPD (Teng et al., 2001, 2002), SSR (Bao et al., 2007; Cao et al., 2012; Erfani et al., 2012; Tian et al., 2012; Yao et al., 2010) and AFLP (Bao et al., 2008; Monte-Corve et al., 2000) were also applied for revealing genetic diversity of


95 96 97 98 99 100 101 102 103 104 105

YMPEV 4969


31 July 2014 4


149

some pear groups, but these markers are too variable to study interspecific relationships among Pyrus species. During the last two decades, noncoding DNA regions of chloroplast and nuclear genes have been successfully applied to phylogenetic studies of plants at lower taxonomic levels. Maternal inheritance of cpDNA in Rosaceae taxa has been confirmed (Corriveau and Coleman, 1988), indicating cpDNA regions are useful to reveal the maternal progenitors of putative hybrids. Six noncoding cpDNA regions were found to be highly conserved in Pyrus, among which trnL-F and accD-psaI displayed the highest levels of variation (Kimura et al., 2003; Katayama et al., 2007). Recently, high rates of structural mutations of accD-psaI were detected among Pyrus species (Hu et al., 2011; Katayama et al., 2011). For nuclear genes, our previous phylogenetic study of Pyrus species (mainly from East Asia) based on internal transcribed spacer (ITS) data resulted in a complex and poorly resolved phylogeny due to low sequence divergence and non-concerted evolution of nrDNA arrays, and suggested that some species might have experienced extensive ancient inter-specific hybridization events (Zheng et al., 2008). Therefore, we explored the phylogenetic utility of two low copy nuclear genes (LCNGs), alcohol dehydrogenase gene (Adh) and LEAFY, among Pyrus and Malus species. As a result, the second intron of LFY2 (LFY2int2) was determined to be the best nuclear gene region for studying inter-specific relationships in Pyrus, owing to its highest sequence divergence and less possibility of incomplete lineage sorting as concluded in our previous study (Zheng et al., 2011). Evolution of species has often been presumed to be a branching process, traditionally represented by a tree topology. However, it is widely acknowledged that true evolutionary relationships are often reticulate rather than strictly tree-like. Reticulate evolution such as hybridization, horizontal gene transfer, homoplasy and genetic recombination cannot be appropriately represented by the tree topology (Linder and Rieseberg, 2004; Makarenkov et al., 2006). Therefore, phylogenetic network approaches, which can reveal incompatible phylogenetic information caused by reticulations, have been increasingly applied in phylogenetic studies (Hassanzadeh et al., 2012; Makarenkov et al., 2006). In this study, we constructed traditional phylogenetic trees and phylogenetic networks based on two noncoding regions of cpDNA (trnL-F and accDpsaI) and one LCNG (LFY2int2) in order to reconstruct the phylogeny of Pyrus, reveal the origin of putative inter-specific hybrids and understand evolutionary processes underlying diversification of the genus.

150

2. Material and methods

151

2.1. Taxon sampling

152

163

Fifty-eight accessions representing 13 oriental Pyrus species and 51 accessions representing 12 occidental Pyrus species were sampled (Table 1), which covered most recognized Pyrus taxa at the specific level according to Challice and Westwood (1973) and Yu (1979). As shown in Table 1, Most of oriental accessions were collected from wild region except some cultivars of P. pyrifolia, P. ussuriensis and P. sinkiangensis and Chinese white pear, while the occidental accessions were from germplasm repository and also originally collected from wild regions in different countries. Four accessions from the genus Malus were selected as outgroups (Table 1). All of the accessions were diploid (2n = 34) except ‘Ruanerli’ (P. ussuriensis), which was identified as a triploid.

164

2.2. DNA extraction, amplification and sequencing

165

Total genomic DNA was extracted from young leaves using the CTAB method (Doyle and Doyle, 1987). Amplifications of the trnL

106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148

153 154 155 156 157 158 159 160 161 162

166

intron plus trnL-F region were performed using the primers c and f (Taberlet et al., 1991), and the accD-psaI region was initially amplified using the primers accD-769F and psaI-75R (Small et al., 1998); subsequently, accD-F2 (CTTATTCGATCCAATCGTACCAC) was developed instead of accD-769F to target a short and variable region. The purified PCR products of the cpDNA regions were directly sequenced using the forward amplification primers. PCRs and sequencing were conducted at least twice for accessions that displayed singleton mutations. The LFY2int2 was amplified using the specific primer pair LFY2F and LFY2-R. Sequencing of LFY2int2 was conducted as described in our previously study (Zheng et al., 2011). At least three clones were sequenced for each accession to recover the different copies as indicated by direct sequencing of purified PCR products. Sequences were determined on an ABI 3730 automatic DNA sequencer (Perkin Elmer) using cycle sequencing protocols of Dye™ 3.1 (Perkin Elmer). All cpDNA sequences are newly reported in this study (GenBank accession numbers JX122456-JX122488, KF486653-KF486845), while some LFY2int2 sequences were from our previous published study (GU991466-GU991522) and the rest are newly reported here (KF486521-KF486652).

167

2.3. Alignment, indel coding and haplotype definition

188

Multiple alignments of DNA sequences were performed with Clustal X (Thompson et al., 1997). Indels of different lengths at the same position were treated as separate characters, and each indel was treated as one character and coded as one substitution (T/C) in final phylogenetic analyses. Haplotypes were determined

189

occidental

oriental

Fig. 1. Bayesian consensus tree based on the combined and reduced cpDNA haplotypes. The numbers on the branch indicates corresponding posterior probabilities values.


168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187

190 191 192 193

YMPEV 4969


31 July 2014 X. Zheng et al. / Molecular Phylogenetics and Evolution xxx (2014) xxx–xxx

5

Fig. 2. Neighbor-Net splitsgraph based on reduced tH (a) and aH (b) haplotypes. The nodes represent taxa. Each edge represents a particular pattern of character change (mutation), each set of parallel edges represents a split (bipartition). The parallel edges indicate conflicting characters between or among the associated nodes. The edges and nodes of the splits between oriental and occidental species were in red or blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2010), evidence of recombination was also sought by recombination networks as implemented in SplitsTree 4.10.

212

197

by DnaSP v5 (Librado and Rozas, 2009). For LFY2int2, all of the indels were excluded from analyses. Sequence divergence of each region was calculated using K-2P distance (Kimura, 1980), implemented in Mega v5 (Tamura et al., 2011).

3. Results

214

198

2.4. Phylogenetic analysis

3.1. cpDNA region

215

199

Bayesian analyses were performed with MrBayes 3.1 (Ronquist and Huelsenbeck, 2003). Markov chains were run for 10,000,000 generations with sampling frequency of 100. The first 25% of the trees were discarded as burn-in. Clade posterior probabilities (PP) were calculated from the combined sets of trees from two simultaneous runs (150,002 trees). The best-fit models of nucleotide substitution were selected by the AIC using Modeltest 3.06 (Posada and Crandall, 1998). Neighbor-Net splits graphs (NN graphs) of both cpDNA and LFY2int2 data were constructed using SplitsTree 4.10 (Huson and Bryant, 2006) with the criterion set to uncorrected p distance (Bryant and Moulton, 2004). Putative recombinant were tested by using seven tests (RDP, GENECONV, MaxChi, SiScan, LARD, 3Seq, DSS) implemented in RDP package (Martin et al.,

3.1.1. Sequence characteristics and haplotype determination Sequence lengths of the trnL intron ranged from 517 to 518 bp, and those of the trnL-F ranged from 393 to 403 bp. These two regions were combined for haplotype definition (abbreviated as tH hereinafter) and phylogenetic analyses. A total of four indels and 13 substitutions were observed, producing 16 tH haplotypes for the ingroup taxa (tH1 to tH16) (Table 1, Supplement 2a), seven for oriental species (tH1 to tH7), and nine for occidental species (tH8 to tH16). The exception was P. caucasica-684 (an occidental species) from Ukraine which carried tH5. The dominant haplotype of the oriental group was tH5, covering ten species, while that of the occidental group was tH8, covering all species but P. regelii and P. gharbiana (Table 1). The mean value of pair-wise sequence

216

194 195 196

200 201 202 203 204 205 206 207 208 209 210 211


213

217 218 219 220 221 222 223 224 225 226 227 228

YMPEV 4969


31 July 2014 6 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248


divergence among the 16 ingroup haplotypes was 0.4%. The four outgroup accessions were assigned to three other haplotypes (tH17 to tH19). Sequence lengths of accD-psaI ranged from 685 bp to 1093 bp. Most indels in this region were duplications of adjacent regions, and some duplications overlapped or occurred in the same regions, making the alignment difficult (Supplement 2). We divided the sequences into small datasets and aligned them separately in order to infer indel events accurately. Moreover, a 229-bp deletion was observed in aH1 and aH2, spanning a region that contained eight substitutions among other haplotypes, and we retained the substitutions by manually filling the gaps for haplotype aH1 and aH2, using their most closely related haplotypes as reference. Finally, 24 indels and 20 substitutions were observed, leading to 32 aH haplotypes, 15 for the oriental group (aH1-aH15) and 17 for the occidental group (aH16-aH17) (Table 1, Supplement 1b). Exceptionally, P. caucasica-684 carried aH8. The mean value of pair-wise sequence divergence among 32 ingroup haplotypes was 0.8%. The four outgroup accessions were assigned to three other haplotypes (aH33-aH35).

Indels shared by taxa from distinct clades in the phylogenetic tree were putatively homoplasious (site 1 of tH in Supplement 1a, site 2 and site 4 of aH in Supplement 1b). We estimated their phylogenetic utility by investigating the characteristics (length, position and types of the indel), the variation level of the adjacent regions and evolutionary information based on other data. Pyrus cossonii (aH21, aH25), and P. betulaefolia (aH3, aH7, aH15) shared the indel at site 2 (displaying none 23-bp duplication in Supplement 2), and a connection between these two species was also observed in the NN split graph based on LFY2int2 data, as described below. Therefore, we judged that this indel was not homoplasious but provided important phylogenetic evidence. On the contrary, we did not find a similar explanation for site 4 which was probably homoplasious. The 1-bp indel of site 1, which occurred at a T-repeat region in trnL-F region, was thought to be highly reversible and judged as homoplasy. After excluding these two indels, the following pairs of haplotypes became equal in the phylogenetic analyses: tH10/tH8, aH7/aH15 and aH17/aH18. The original haplotype names were kept in all of the figures and tables to avoid confusion.

(a)

Fig. 3. Bayesian consensus tree of occidental species (a) and oriental species (b) based on LN sequences with all of the putative recombinants excluded but the three ones in blue. Putative inter-specific hybrids between occidental and oriental species are marked by solid rounds in red. Accessions with polymorphic copies locating in different clades were indexed in green, while species that displayed high level of polymorphism were indexed in blue. Triangles in orange, blue, green stand for clade B1, B2 and A, respectively (a), while triangle in red stands for clade A (b). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)


249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268

YMPEV 4969



7

(b)

Fig. 3 (continued)

269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291

3.1.2. Phylogenetic analyses based on the cpDNA haplotypes The combined cpDNA regions produced 42 ingroup and three outgroup haplotypes, and a Bayesian consensus tree was constructed based on them (Fig. 1). Oriental and occidental haplotypes formed two separate clades, both with high support (1.0 pp), but the relationships within each clade were largely unresolved due to polytomies, especially for the occidental group. NN graphs of tH haplotypes were constructed for oriental and occidental groups separately. For the oriental group, the dominant haplotype tH5 was located at center of the star-like graph, and tH1 formed a separate lineage. For the occidental group, the dominant haplotype tH8/tH10 was located in the center of the star-like graph, while tH15 of P. regelii and tH16 of P. spinosa-1610 were the most closely related to the outgroup (Supplement 3). The NN graphs of aH haplotypes were not strictly star-like, and the dominant haplotypes of the oriental (aH8) and occidental groups (aH18/ aH17) were located at the centers of the graphs, respectively (Supplement 4). When combining the tH haplotypes of occidental and oriental groups, parallel edges showing homoplasy between the two pear groups were observed. For example, tH16 of P. spinosa1610 was connected to all of the oriental haplotypes except tH7, while tH7 of P. pashia was connected to all occidental haplotypes but tH16 (split t1 and t2, Fig. 2a) due to the nucleotide character

of site 17 in Supplement 2a. In the NN graph of aH haplotypes (Fig. 2b), haplotypes of four occidental P. mamorensis, P. cossonii, P. spinosa and P. cordata were connected to three haplotypes of P. betulaefolia. The NN graphs based on combined cpDNA region of oriental group displayed more parallel edges than that of single cpDNA region, suggesting homoplasy between the two cpDNA regions, while that of occidental group was almost star-like (Supplement 5).

292

3.2. LFY2int2-N

301

3.2.1. Sequence characteristics Sequence lengths of LFY2int2 ranged from 490 to 1080 bp. As in our previous study (Zheng et al., 2011), LFY2int2-N (LN) was obtained in most of the accessions, while LFY2int2-Ins8 (LI8), LFY2int2-Del2 (LD2) and newly found LFY2int2-Ins3 (LI3) were obtained in some accessions (Table 1). Each of these additional gene copies formed a monophyletic group in the gene tree (Supplement 6), and all three were considered to be paralogs of LN by recent gene duplication for these accessions. The LS sequences were included in the initial phylogenetic analyses after excluding a large insertion homologous to S gene as described in our previous

302


293 294 295 296 297 298 299 300

303 304 305 306 307 308 309 310 311 312

YMPEV 4969


31 July 2014 8


Fig. 4. Neighbor-Net splitsgraph based on LFY2int2 data. The clades or subclades were marked corresponding to the phylogenetic tree of Fig. 3a and b.

313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328

329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348

study (Zheng et al., 2011). The initial dataset contained 175 ingroup sequences and had an aligned length of 612 bp with 21 indels. Only nine out of 21 indels were informative for LFY2int2, which constituted a much lower proportion of the total mutations than for the cpDNA regions. All of the indels in LFY2int2 were removed from phylogenetic analyses, but substitutions within two large indels were retained by filling in the gaps as described above. Using the seven methods of RDP, we failed to find evidence of recombination in the alignment. However, the minimum number of recombination events (Rm) was 17 as detected by DnaSP, and significant evidence for recombination was also found by the Phi test with a P-value of 0.0015 (data not shown) in the SplitsTree. Therefore, putative recombinants were identified manually by combining substitution patterns and phylogenetic analyses as below. 3.2.2. Phylogenetic analyses We constructed the Bayesian consensus tree based on the initial dataset (Supplement 6). Due to large deletion of LFY2int2 in the four outgroup accessions, the sequence from Malus dometica ‘Pinova’ (DQ535886) was used as outgroup as we did before (Zheng et al., 2011). Sequences that displayed parsimony-informative (PI) characters of two distinct clades usually were separated in the gene tree with long braches and were identified as putative recombinants. The LS sequences together with some putative recombinants formed a separate clade. LI8, LD2 and LI3 formed monophyletic subclades in the tree, respectively. All of the paralogs of LFY2int2 and most of the putative recombinants were excluded from the final dataset. The final phylogenetic analyses were conducted based on a reduced dataset of orthologous LN sequences, which included 127 ingroup sequences and had an aligned length of 543 bp with 87 PI sites and 55 autapomorphic sites, with a mean value of pairwise sequence divergences of 2.5%. Bayesian consensus tree was constructed under the best substitution model of ‘TVM + G’ (nst = 6, rates = gamma).

As shown in Fig. 3a and b, all of the occidental species formed clade A, while most oriental species formed clade B, but some oriental species (subclade B1 and B2) were more closely related to the occidental clade A. For the occidental group, the three geographical subgroups were not clearly separated in the tree. Within the occidental clade A, the majority of subclades received high PP values, but relationships among these subclades were unresolved due to polytomies. There were no apparent geographical subgroups for the oriental species. As described above, except accessions in subclade B1 and subclade B2, all oriental accessions were in the clade B, and the relationships among subclades in the clade B were better resolved than in clade A. Intra-individual and intra-specific polymorphisms were extensively observed. Among the 18 primary species represented by more than one accession, only four were resolved as monophyletic: P. mamorensis and P. gharbiana (subclade A1), P. cossonii (subclade A2), P. regelii (subclade A3) and P. betulaefolia (subclade B2). The NN graph of the LN (Fig. 4) depicted a much higher level of conflict than that of the cpDNA (Fig. 2). It was split into two groups by the longest parallel edges; the oriental clades B1 and B2 belonged to the occidental group, while B3 belonged to the oriental group, which is congruent with the Bayesian tree topology. Rooted recombination networks of ingroup Pyrus taxa showed that large proportion of the edges were overlapped and connected like a ‘net’, suggesting most of Pyrus species have experienced reticulation processes (Supplement 7).

349

4. Discussion

375

4.1. Poor resolution of the phylogenetic tree

376

The relationships among Pyrus species were poorly resolved in the phylogenetic trees based on both cpDNA (Fig. 1) and LFY2int2 data (Fig. 3a and b). Low sequence divergence (0.4% for trnL-F, 0.8% for accD-psaI and 2.5% for LN is likely the main cause, especially for the cpDNA data. DNA sequence variation of Maloideae taxa was found to be more conserved than many other angiosperm

377


350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374

378 379 380 381 382

YMPEV 4969


31 July 2014 X. Zheng et al. / Molecular Phylogenetics and Evolution xxx (2014) xxx–xxx 383 384 385 386 387 388 389 390 391 392 393 394 395 396

taxa, and rapid radiation has been presumed due to ease of animal dispersal of their fleshy fruits (Campbell et al., 2007; Lo et al., 2009). Conflicting signals (homoplasy) as shown by the NN graph partially accounted for the poor resolution, especially for the nuclear gene data. Conflicting signals within each cpDNA region and between two cpDNA regions were probably caused by ancient inter-specific chloroplast recombination (Erixon and Oxelman, 2008) during the speciation or evolutionary rate heterogeneity among cpDNA regions. Extensive conflicting signals were observed by the NN graph and recombination network of LN, suggesting presence of ancestral recombinants derived from hybridization events, which also largely contributed to the poor resolution and complex intra-individual and intra-specific polymorphisms of some species as discussed below.

397

4.2. Intra-individual and intra-specific polymorphism of nuclear data

398

After excluding the paralogous copies of LN and putative recombinants, some of the intra-individual and intra-specific LN copies were polyphyletic in the phylogenetic tree (Fig. 3a and b). Hybridization and incomplete lineage sorting of ancestral polymorphic copies should be the two major causes. Hybridization has been thought to be the major mode of the evolutionary history of genus Pyrus (Rubtsov, 1944). Intra-individual polymorphisms were found in P. caucasica684, ‘Korlaxiangli’ and P. salicifolia2382, which were probably hybrids between occidental and oriental groups, since they contained LN copies from both clade A and clade B. ‘Korlaxiangli’ is a famous cultivar of P. sinkiangensis which has been identified as an inter-specific hybrid of multiple lineages including Chinese white pear, P. pyrifolia, P. ussuriensis, and P. communis (Teng et al., 2001; Zheng et al., 2011). Our data confirmed that its maternal lineage should be an oriental pear with the tH5aH9 haplotype, possibly Chinese white pear or P. ussuriensis, while its paternal lineages might include P. pyraster or its close relatives. Pyrus salicifolia2382 is a probable hybrid with occidental species as maternal parentage and oriental species as paternal parentage, while P. caucasica684 is a hybrid with the oriental species as maternal parentage and occidental species as paternal parentage. Intra-individual polymorphisms were also found in P.phaeocarpa, P. pashia2, P. salicifolia2720, P. salicifolia2849 and P. elaeagrifolia2817, among which only P.phaeocarpa was a putative hybrid between P. betulaefolia and P. ussuriensis (Yu, 1979; Teng et al., 2002). Our results suggested P. ussuriensis with tH1aH8 haplotype must be its maternal parentage. It was uncertain if the other three accessions were inter-specific hybrids. Three other putative inter-specific hybrids included in our analyses – P.bretschneideri, P.serrulata and P.hopeiensis – only one LN copy was found in one individual. It was possible that their parentage did not have allelic variation or one of the parentage copies was lost during hybridization. Anyway, P.hopeiensis, P.bretschneideri must have been maternally derived from some P. ussuriensis accessions by sharing the tH1aH8 haplotype. Pyrusserrulata is a putative hybrid involving P. pyrifolia and P. calleryana, and its maternal parent might be wild P. pyrifolia (tH3aH2), and it was closely related to P. dimorphophylla5 in the LN tree. Intra-specific polymorphisms were also extensively found among the putative primary species represented by more than one individual, of which only five were monophyletic in the LN tree. Our previous phylogenetic study based on ITS sequences showed much more complex relationships among Pyrus species due to the non-monophyly of most species (Zheng et al., 2008). It is possible that some of the putative primary Pyrus species are not primary species but inter-specific hybrids since there are no reproductive barriers among Pyrus species, or individuals of these species have fixed different ancestral polymorphic copies.

399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445

9

However, incomplete lineage sorting of ancestral polymorphic LN copies might partially account for the polymorphism, too.

446

4.3. Relationships among primary species based on the phylogenetic trees

448

Although the incomplete lineage sorting problem could not be eliminated for some species, we still can infer the phylogenetic relationships among some Pyrus species.

450

4.3.1. Oriental group Among the oriental species, nine were considered to be primary species (Table 1). As shown in Fig. 1, cpDNA haplotypes with tH1, including P.bretschnederi, P.hopeiensis and P.phaecarpa, P. dimrphophylla ( 4, 6) and P. ussuriensis, formed well supported subclade, suggesting close relationship among these species maternally. Haplotypes with aH12 which were from P. pashia and ‘Huobali’ formed a moderately supported clade. The relationships among the rest haplotypes were not resolved. Pyrus betulaefolia and P. calleryana have primitive characters of the least number of carpels (2–3), and they also have lobed leaves at juvenile stages. Two of the three P. betulaefolia accessions carried unique cpDNA haplotypes (Table 1) and all of them were in the separated clade B2 in the LN tree (Fig. 3b) which supported that P. betulaefolia could be the earliest diverged species. All P. calleryana accessions carried the dominant tH5aH8 haplotype which is an ancestral haplotype since it was the closest to outgroup in the NN graph. However, most of the LN sequences from P. calleryana were either LS copies or putative recombinant or LD2 (Table 1, Supplement 6), so more LCNG data will be needed to confirm the phylogenetic position of this species. Except P. pashia4 from Nepal, all P. pashia accessions were from Yunnan Province, and they have different unique cpDNA haplotypes and were resolved in different subclades of clade B in the LN tree (Fig. 3b). Pyrus pashia is mainly distributed in mountainous areas of southwestern China, and is morphologically diverged from P. calleryana by more carpels (3–5 vs 2–3) and stamens (25–30 vs 20). Such slight difference makes it difficult to distinguish them in their overlapping distributions when one individual has three carpels. The only LN obtained in P. calleryana1 from Yunnan province was closely related to P. pashia. Our population studies of P. calleryana in Zhejiang province and P. pashia in Yunnan Province showed that the latter has more diversified cpDNA haplotypes (Liu et al., 2012, 2013), and they did share some common haplotypes. Therefore, it was possible P. pashia and P. calleryana were not completely isolated, and P. pashia was more diversified which is also discussed below. Besides P. calleryana, P. dimorphophylla, P. koehnei and P. fauriei are three other ‘pea pears’ with two or three carpels, which were once treated as varieties of P. calleryana (Rehder, 1940). Both cpDNA and nuclear data in this study supported that each pea pear species should be treated as independent species (Challice and Westwood, 1973). Furthermore, they were not as closely related based on nuclear gene tree as found in a previous study (Teng, 2001). Especially, P. koehnei was in the special clade B, sister to P. betulaefolia. Pyrus pyrifolia, P. ussuriensis, P. hondoensis are putative primary species with five carpels. Most of the cultivated P. pyrifolia accessions (Chinese Sand Pear, CSP) carried tH4aH1 haplotype, but ‘Huobali’ native to Yunnan province carried the aH12 haplotype which was restricted to P. pashia, and its LN sequence was also identical to P. pashia5 (clade B5-1). Therefore, it was possible that P. pashia was involved in origin of local pear cultivars. The wild P. pyrifolia accessions carried three other different haplotypes. Most P. ussuriensis accessions carried the tH5aH9 haplotype, and were not monophyletic but related to either P. pyrifolia in clade B5-2 or P. betulaefolia in clade B2. Especially, ‘Ruanerli’, a P. ussuriensis cultivar, was identified to be a triploid (Wang et al., 2004). It carried the typical

453


447

449

451 452

454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508

YMPEV 4969


31 July 2014 10 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573


cpDNA haplotype of P. ussuriensis (tH5aH9), but we only obtained one LN sequence which was in the special clade B1 including the P. caucasica684 (Fig. 3b). Therefore, we suggested that its maternal parent is Ussurian pear while its paternal parent might be occidental species, and more nuclear data need to be applied to confirm its origin. Pyrus hondoensis (Aonashi) was once treated as a variety of P. ussuriensis since both of them have fruits with persistent calyx which is common for occidental species, and we confirmed that P. hondoensis was closely related to typical Ussurian pears by sharing tH5aH9 and LN sequence as well as LI8 copy, which is also congruent with the results based on DNA markers (Bao et al., 2007; Teng et al., 2002). Pyrus xerophila is one of the newly reported species growing in northwestern China (Yu, 1979), mainly in Gansu province where P. communis and its hybrid types with other Chinese pears have been planted for a long history. The three P. xerophila accessions carried the dominant oriental haplotypes (tH5aH8) and two of them located in the special clade B1 and/or B3 (Fig. 3b), suggesting complex genetic background of P. xerophila individuals. It is interesting to investigate more P. xerophila individuals and their relationships with occidental species. 4.3.2. Occidental group All of the occidental species used in study were treated as primary species, but each species have multiple cpDNA haplotypes (Table 1). In the phylogenetic tree of combined cpDNA haplotypes, only four subclades were found, of which two of them received moderate or high PP value (tH15aH18 and tH15aH16 in P. regelii, tH8aH21 and tH8aH25 in P. cossonii), and the relationships among the other haplotypes were unresolved due to low level of variation (Fig. 1). Among West Asian species, P. regelii possesses ancestral characters of dissected adult leaves and ovary with few locules, and it is geographically and ecologically isolated (Rubtsov, 1944). It is the only West Asian species that is monophyletic in the LN tree (Fig. 3a) and it displayed the unique tH15 haplotype which is closest to the outgroup taxa in the NN graph (Supplement 3), supporting that P. regelii is an early-diverging and isolated Pyrus species. Pyrus spinosa is characterized by its oval-lanceolate leaves (Aldasoro and Garmendia, 1996), and might be an old species due to its primitive character of lobed juvenile leaves (Challice and Westwood, 1973). Three of the four P. spinosa accessions were in subclade A7 in the nuclear gene tree (Fig. 3a), while P. spinosa1598 whose leaf is not typical to this species was closely related to two P. cordata accessions in subclade A8, so it was possible that this accession was misidentified. Pyrus salicifolia is characterized by its glabrous styles and lanceolate or elliptical leaves, but it has many diverse forms and has been considered to be synonymous to P. nivalis or a hybrid between P. nivalis and P. communis (Aldasoro and Garmendia, 1996). However, the nuclear data showed that P. salicifolia accessions formed two well supported separate subclades, A4 and A5, related to neither P. nivalis nor P. communis. Four accessions of P. syriaca were in subclade A6, and the rest one was in subclade A4 (Fig. 3a). For P. elaeagrifolia, three accessions were in four different subclades in the LN tree (Fig. 3a), and for the rest two, only LI3 were detected (Table 1), suggesting a complex genetic background of this species, otherwise this species might be not well identified. European species displayed lower levels of cpDNA diversity compared to West Asian species, and they might be the latest derived occidental species according to the geographical spread. The LN gene tree showed that P. nivalis and P. cordata were more related to West Asian species, mainly P. spinosa, while P. caucasica, P. pyraster and P. communis were more closely related to three North African species (Fig. 3a), suggesting they were independently derived from West Asian species and North African species.

The three African species have their own unique cpDNA haplotypes, respectively (Table 1), thus they were three well differentiated species. Pyrus mamorensis was thought to be different from other two (P. gharbiana and P. cossonii), and it was considered related to P. communis due to lack of flavone glycosides (Challice and Westwood, 1973). However, we found that P. cosonii was the different one and the other two were related to European species including P. communis.

574

4.4. Evolutionary histories of Pyrus based on phylogenetic networks

582

The evolution of maternally inherited cpDNA can provided clues of species dispersal and diversification. As described above, the NN graphs based on combined cpDNA haplotypes showed that P. regelii, P. spinosa and P. calleryana were the earliest-diverged species since they were the nearest to the Malus taxa (Supplement 5). Pyrus pashia in southwestern China represents a whole number of morphological characters, and has migrated to more western regions, thus was thought to be an intermediate link between the eastern and western species intermediate link (Rubtsov, 1944). Some P. pashia accessions (tH7) were connected to most of the occidental species, while one P. spinosa accession (tH16) was connected to most of the oriental species (Fig. 2a). Therefore, we supported the previous suggestions and the opinion that southwestern China should be the divergence center of Pyrus. According to Rubtsov (1944), Pyrus cossonii (synonymous to P. longipes), P. cordata and P. betulaefolia were thought to be descendent of primitive stock species, and the former two were relict species characterized by fruits having a deciduous calyx, a primary character similar to East Asian species. These species were interpreted as connecting links between West Asian and European group and East Asian group (Challice and Westwood, 1973). In the NN graph based on accD-psaI haplotypes, these species together with P. spinosa were connected by sharing some characters in accD-psaI region (Fig. 2b), which supported the previous understanding of these species and was probably as a result of ancient hybridization between some stock Pyrus species during the divergence of oriental and occidental species. Especially, P. betulaefolia, P. koehnei and its related species formed a sister clade to the occidental species (Fig. 3b) and were connected to occidental species in the NN network based on LN, thus it is particularly interesting to investigate phylogenetic position of P. betulaefolia and P. koehnei by applying more individuals and gene data. Both cpDNA and nuclear data supported that oriental and occidental groups have been evolving independently for a long time. Low sequence divergence of DNA regions and poor resolution of the phylogenetic trees suggested that rapid radiation is one of the main evolutionary processes of Pyrus, which is especially true for the occidental group due to the star-like structure of NN graph of cpDNA haplotypes and poor resolution of the subclade branching order in the LN tree. As described before, rapid radiation has been presumed for Maloideae taxa due to the ease of animal dispersal of their fleshy fruits (Campbell et al., 2007; Lo et al., 2009). That is also why there are limited morphological characters for Pyrus species identification. Reticulate evolution is increasingly being recognized as a fundamental process in the evolutionary histories of organisms, and is particularly common in plants (Linder and Rieseberg, 2004; Sessa et al., 2012). As described above, only five primary species were found to be monophyletic in the LN tree (Fig. 3a and b), and both incomplete lineage sorting of ancestral polymorphism and intensive hybridization accounted for the intra-individual and intra-specific polymorphisms of the rest of the Pyrus species. Besides, the NN graph and recombination networks also suggested that both oriental and occidental groups might have undergone ancient genetic exchange and recombination which were caused by hybridization

583


575 576 577 578 579 580 581

584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637

YMPEV 4969



663

events. Hybridization, even at the diploid level, can cause rearrangements of genes including gene duplication and gene loss in the hybrids (Kellogg and Bennetzen, 2004). Rubtsov (1944) has claimed that one of the most important factors in speciation and form-genesis in Pyrus is hybridization, and established numerous interspecific hybrids in Caucasus and Crimea. Compared to occidental species, more putative recombinants and paralogs of LN were observed in oriental species, indicating more frequent hybridization events have contributed to the diversification of oriental species. It is the first time that we confirm that rapid radiation and reticulation were two major evolutionary processes for Pyrus based on molecular data by combining phylogenetic trees and networks. The feasibility of rapid radiation and uneven ancestral polymorphism among individuals of most Pyrus species were caused by the absence of geographical isolation and reproductive barriers. However, multiple independent LCNGs are needed to test the existence of incomplete lineage sorting of gene copies and further test the complex evolutionary processes and improve our understanding of phylogenetic relationships among Pyrus species. Recently, the genome sequence of ‘Dangshansuli’ (CWP) has been published (Wu et al., 2013) and a number of EST datasets from Pyrus species are available, and new experimental techniques targeting for orthologous LCNGs compatible with the next-generation sequencing have been explored (McCormack et al., 2013), so it should now be possible and cost-effective to reconstruct phylogeny of Pyrus based on large number of nuclear gene sequences.

664

5. Conclusions

665

687

This study represents the first time that phylogenetic relationships among extensively sampled Pyrus species were studied based on multiple DNA sequence data by constructing both phylogenetic trees and phylogenetic networks. Our results showed that: (1) Five primary species represented by more than one accession (the three North African species, P. regelii and P. betulaefolia) were monophyletic. (2) The major diversification center was most probably in southwestern China, where distributing highly diversified P. pashia. (3) P. cossonii, P. mamorensis, P. spinosa, P. cordata and P. betulaefolia were descendent of primitive stock species and still shared some molecular characters. (4) Both occidental and oriental groups have experienced rapid radiation due to ease of seed dispersal and reticulate evolution caused by intensive hybridization; more hybridization events were inferred for the oriental group, resulting in more paralogs and genetic recombinants than in the occidental group, while rapid radiation was more obvious for the occidental group than for the oriental group. (5) Some inter-specific hybrids between occidental and oriental species were newly observed, and the origins of some putative oriental inter-specific hybrids were elucidated. However, more individuals of each species and multiple independent nuclear genes will be needed to better understand the complex evolutionary histories and relationships among species of Pyrus.

688

Acknowledgments

689

692

Funding was provided by National Natural Science Foundation (Nos. 30871690, 31201592) Zhejiang Natural Science Foundation (LQ13C020001) from China and Research Fund for the Doctoral Program of Higher Education of China (20110101110091).

693

Appendix A. Supplementary material

694

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ympev.2014.07. 009.

638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662

666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686

690 691

695 696

11

References

697

Aldasoro, J.J., Garmendia, F.M., 1996. The genus Pyrus L. (Rosaceae) in South-west Europe and North African. Bot. J. Linn. Soc. 121, 143–158. Bao, L., Chen, K., Zhang, D., Cao, Y., Yamamoto, T., Teng, Y., 2007. Genetic diversity and similarity of pear (Pyrus L.) cultivars native to East Asia revealed by SSR (simple sequence repeat) markers. Genet. Resour. Crop Evol. 54, 959–971. Bao, L., Chen, K., Zhang, D., Li, X., Teng, Y., 2008. An assessment of genetic variability and relationships within Asian pears based on AFLP (amplified fragment length polymorphism) markers. Sci. Hort. 116, 374–380. Bell, R.L., Hough, F.L., 1986. Interspecific and intergenic hybridization of Pyrus. HortScience 21, 62–64. Bryant, D., Moulton, V., 2004. Neighbor-Net: an agglomerative method for the construction of phylogenetic networks. Mol. Biol. Evol. 21, 255–265. Campbell, C.S., Evan, R.C., Morgan, D.R., Dickinson, T.A., Arsenault, M.P., 2007. Phylogeny of subtribe Pyrinae (formerly the Maloideae, Rosaceae): limited resolution of a complex evolutionary history. Plant. Syst. Evol. 266, 119–145. Cao, Y., Tian, L., Gao, Y., Liu, F., 2012. Genetic diversity of cultivated and wild Ussurian pear (Pyrus ussuriensis Maxim.) in China evaluated with M13-tailed SSR markers. Genet. Resour. Crop Evol. 1, 9–17. Challice, J.S., Westwood, M.N., 1973. Numerical taxonomic studies of the genus Pyrus using both chemical and botanical characters. Bot. J. Linn. Soc. 67, 121–148. Corriveau, J.L., Coleman, A.W., 1988. Rapid screening method to detect potential biparental inheritance of plastid DNA and results for over 200 angiosperm species. Am. J. Bot. 75, 1443–1458. Doyle, J., Doyle, J., 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19, 11–15. Erfani, J., Ebadi, A., Adollahi, H., Fatahi, R., 2012. Genetic diversity of some pear cultivars and genotypes using simple sequences repeat (SSR) markers. Plant Mol. Biol. Rep. 5, 1065–1072. Erixon, P., Oxelman, B., 2008. Reticulate or tree-like chloroplast DNA evolution in Sileneae (Caryophyllaceae)? Mol. Phylogenet. Evol. 48, 313–325. Hassanzadeh, R., Eslahchi, C., Sung, W.K., 2012. Constructing phylogenetic supernetworks based on simulated annealing. Mol. Phylogenet. Evol. 63, 738–744. Hu, C., Zheng, X., Teng, Y., 2011. Characterization and phylogenetic utility of noncoding chloroplast regions trnL-F and accD-psaI in Pyrus (in Chinese). Acta Hort. Sin. 38, 2261–2272. Huson, D.H., Bryant, D., 2006. Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23, 254–267. Iketani, H., Manabe, T., Matsuta, N., Akihama, T., Hayashi, T., 1998. Incongruence between RFLPs of chloroplast DNA and morphological classification in east Asian pear (Pyrus spp.). Genet. Resour. Crop Evol. 45, 533–539. Jang, J.T., Tanabe, K., Tamura, F., Banno, K., 1992. Identification of Pyrus species by leaf peroxidase isozyme phenotypes (in Japanese with English summary). J. Jpn. Soc. Hort. Sci. 61, 273–286. Katayama, H., Adachi, S., Yamamoto, T., Uematsu, C., 2007. A wide range of genetic diversity in pear (Pyrus ussuriensis var. aromatica) genetic resources from Iwate, Japan revealed by SSR and chloroplast DNA markers. Genet. Resour. Crop Evol. 54, 1573–1585. Katayama, H., Tachibana, M., Iketani, H., Zhang, S., Uematsu, C., 2011. Phylogenetic utility of structural alterations found in the chloroplast genome of pear: hypervariable regions in a highly conserved genome. Tree Genet. Genome 8, 313–326. Kawata, T., Itai, A., Tanabe, K., Tamura, F., 1995. Genetic variation in Pyrus species by RFLP analysis of genomic of DNA (in Japanese). J. Jpn. Soc. Hort. Sci. 61, 148–149. Kellogg, E.A., Bennetzen, J.L., 2004. The evolution of nuclear genome structure in seed plant. Am. J. Bot. 91, 1709–1725. Kimura, M., 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16, 111–120. Kimura, T., Iketani, H., Kotobuki, K., Matsuta, N., Ban, Y., Hayashi, T., Yamamoto, T., 2003. Genetic characterization of pear varieties revealed by chloroplast DNA sequences. J. Hort. Sci. Biotechnol. 78, 241–247. Librado, P., Rozas, J., 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25, 1451–1452. Linder, C.R., Rieseberg, L.H., 2004. Reconstructing patterns of reticulate evolution in plants. Am. J. Bot. 91, 1700–1708. Liu, J., Zheng, X., Potter, D., Hu, C., Teng, Y., 2012. Genetic diversity and population structure of Pyrus calleryana (Rosaceae) in Zhejiang Province, China. Biochem. Syst. Ecol. 45, 69–78. Liu, J., Sun, P., Zheng, X., Potter, D., Li, K., Hu, C., Teng, Y., 2013. Genetic structure and phylogeography of Pyrus pashia (Rosaceae) in Yunnan Province, China, revealed by chloroplast DNA analyses. Tree Genet. Genome 9, 433–441. Lo, E.Y.Y., Stefanovic´, S., Christensen, K.I., Dickinson, T.A., 2009. Evidence for genetic association between east Asian and western North American Crataegus L. (Rosaceae) and rapid divergence of the Eastern North American lineages based on multiple DNA sequences. Mol. Phylogenet. Evol. 51, 157–168. Makarenkov, V., Kevorkov, D., Legendre, P., 2006. Phylogenetic network construction approaches. Appl. Mycol. Biotechnol. 6, 61–97. Martin, D.P., Lemey, P., Lott, M., Moulton, V., Posada, D., Lefeuvre, P., 2010. RDP3: a flexible and fast computer program for analyzing recombination. Bioinformatics 26, 2462–2463. McCormack, J.E., Hird, S.M., Zellmer, A.J., Carstens, B.C., Brumfield, R.T., 2013. Applications of next-generation sequencing to phylogeography and phylogenetics. Mol. Phylogenet. Evol. 66, 526–538.

698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780


YMPEV 4969


31 July 2014 12 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810


Monte-Corve, L., Cabrita, L., Oliveira, C., Leitao, J., 2000. Assessment of genetic relationships among Pyrus species and cultivars using AFLP and RAPD markers. Genet. Resour. Crop Evol. 47, 257–265. Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14, 817–818. Potter, D., Eriksson, T., Evans, R.C., Oh, S., Smedmark, J.E.E., Morgan, D.R., Kerr, M., Robertson, K.R., Arsenault, M., Dickinson, T.A., 2007. Phylogeny and classification of Rosaceae. Plant Syst. Evol. 266, 5–43. Rehder, A., 1940. Manual of cultivated trees and shrubs, second ed. Macmillan, New York. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574. Rubtsov, G.A., 1944. Geographical distribution of the genus Pyrus and trends and factors in its evolution. Am. Nat. 78, 358–366. Sessa, E.B., Zimmer, E.A., Givnish, T.J., 2012. Reticulate evolution on a global scale: a nuclear phylogeny for New World Dryopteris (Dryopteridaceae). Mol. Phylogenet. Evol. 64, 564–581. Small, R.L., Ryburn, J.A., Cronn, R.C., Seelanan, T., Wendel, J.F., 1998. The tortoise and the hare: choosing between noncoding plastome and nuclear Adh sequences for phylogeny reconstruction in recently diverged plant group. Am. J. Bot. 85, 1301–1315. Taberlet, P., Gielly, L., Bouvet, J., 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Pl. Mol. Biol. 17, 1105–1109. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739. Teng, Y., Tanabe, K., Tamura, F., Itai, A., 2001. Genetic relationships of pear cultivars in Xinjiang, China, as measured by RAPD markers. J. Hort. Sci. Biotechnol. 76, 771–779.

Teng, Y., Tanabe, K., Tamura, F., Itai, A., 2002. Genetic relationships of Pyrus species and cultivars native to East Asia revealed by randomly amplified polymorphic DNA markers. J. Am. Soc. Hort. Sci. 127, 262–270. Teng, Y., Chai, M., Li, X., 2004. A historic retrospect and progress in the taxonomy of the genus Pyrus (in Chinese). J. Fruit Sci. 21, 252–257. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucl. Acids Res. 25, 4876–4882. Tian, L., Gao, Y., Cao, Y., Liu, F., Yang, J., 2012. Identification of Chinese white pear cultivars using SSR markers. Genet. Resour. Crop Evol. 59, 317–326. Wang, T., Zhang, J., Qi, Y., Pang, H., 2004. Advances on polyploidy breeding of fruit crops in China (in Chinese with English summary). J. Fruit Sci. 21, 592–597. Wu, J., Wang, Z., Shi, Z., et al., 2013. The genome of the pear (Pyrus bretschneideri Rehd.). Genome Res. 23, 396–408. Yao, L., Zheng, X., Cai, D., Gao, Y., Wang, K., Cao, Y., Teng, Y., 2010. Exploitation of Malus EST-SSRs and the utility in evaluation of genetic diversity in Malus and Pyrus. Genet. Resour. Crop Evol. 57, 841–851. Yu, D., 1979. Taxonomy of the fruit tree in China (in Chinese). China Agr Press, Beijing. Yu, T., Kuan, K., 1963. Taxa nova Rosacearum Sinicarum (1) (in Chinese). Acta Phytotaxon. Sin. 8, 206–236. Zheng, X., Cai, D., Yao, L., Teng, Y., 2008. Non-concerted ITS evolution, early origin and phylogenetic utility of ITS pseudogenes in Pyrus. Mol. Phylogenet. Evol. 48, 892–903. Zheng, X., Hu, C., Spooner, D., Liu, J., Teng, Y., 2011. Molecular evolution of Adh and LEAFY and the phylogenetic utility of their introns in Pyrus (Rosaceae). BMC Evol. Biol. 11, 255. Zou, L., Zhang, X., Zhang, Z., Sun, B., Guo, S., 1986. Studies on the systematic relationship of some of the species in the genus Pyrus based on pollen grain morphology (in Chinese with English summary). Acta Hort. Sin. 13, 219–224.

811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841


A well-resolved phylogeny of the trees of Puerto Rico based on DNA barcode sequence data.

Phylogeny and Evolution of Multiple Common Carp (Cyprinus carpio L.) Populations Clarified by Phylogenetic Analysis Based on Complete Mitochondrial Genomes.

Folding and unfolding phylogenetic trees and networks.

Computing all hybridization networks for multiple binary phylogenetic input trees.

Calculation of evolutionary trees from sequence data.

Inferring evolutionary histories of pathway regulation from transcriptional profiling data.

The evolutionary history of holometabolous insects inferred from transcriptome-based phylogeny and comprehensive morphological data.

The inference of evolutionary trees from molecular data.

Treelink: data integration, clustering and visualization of phylogenetic trees.

Phylogenetic inference based on matrix representation of trees.

Phylogenetic inferences of Nepenthes species in Peninsular Malaysia revealed by chloroplast (trnL intron) and nuclear (ITS) DNA sequences.

Hedysarum L. (Fabaceae: Hedysareae) Is Not Monophyletic - Evidence from Phylogenetic Analyses Based on Five Nuclear and Five Plastid Sequences.

Molecular phylogeny of Asian Meconopsis based on nuclear ribosomal and chloroplast DNA sequence data.

Phylogenetic Trees and Networks Reduce to Phylogenies on Binary States: Does It Furnish an Explanation to the Robustness of Phylogenetic Trees against Lateral Transfers.

Identification of species and materia medica within Angelica L. (Umbelliferae) based on phylogeny inferred from DNA barcodes.

Reticulate evolutionary history and extensive introgression in mosquito species revealed by phylogenetic network analysis.

SICLE: a high-throughput tool for extracting evolutionary relationships from phylogenetic trees.

Phylogenetic relationships of Sarcocystis species from sheep, goats, cattle and mice based on ribosomal RNA sequences.

Phylogenetic Analysis of a 'Jewel Orchid' Genus Goodyera (Orchidaceae) Based on DNA Sequence Data from Nuclear and Plastid Regions.

Phylogenetic relationships among middle American pocket gophers (genus Orthogeomys) based on mitochondrial DNA sequences.

Phylogeny and evolution of antlered deer determined from mitochondrial DNA sequences.

Origin and phylogeny of Guinea yams as revealed by RFLP analysis of chloroplast DNA and nuclear ribosomal DNA.

Identification of Known and Novel Recurrent Viral Sequences in Data from Multiple Patients and Multiple Cancers.

The phylogenetic position of the Loimoidae Price, 1936 (Monogenoidea: Monocotylidea) based on analyses of partial rDNA sequences and morphological data.