Molecular Phylogenetics and Evolution 56 (2010) 821–839

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution

journal homepage: www.elsevier.com/locate/ympev

Large multi-locus plastid phylogeny of the tribe Arundinarieae (Poaceae: Bambusoideae) reveals ten major lineages and low rate of molecular divergence

Chun-Xia Zeng a,b,c,1, Yu-Xiao Zhang a,b,c,1, Jimmy K. Triplett d, Jun-Bo Yang a,c, De-Zhu Li a,c,* a Key Laboratory of Biodiversity and Biogeography, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming, Yunnan 650204, PR China b Graduate University of Chinese Academy of Sciences, Beijing 100049, PR China c Plant Germplasm and Genomics Center, Germplasm Bank of Wild Species, Kunming, Yunnan 650204, PR China d Department of Biology, National Museum of Natural History, MRC 166, Smithsonian Institution, Washington, DC 20013-7012, USA article info abstract

Article history: The temperate bamboos (tribe Arundinarieae) are notorious for being taxonomically extremely difficult. Received 30 December 2009 China contains some of the world’s greatest diversity of the tribe Arundinarieae, with most genera and Revised 31 March 2010 species endemic. Previous investigation into phylogenetic relationships of the temperate bamboos Accepted 31 March 2010 revealed several major clades, but emphasis on the species-level relationships among taxa in North Available online 8 April 2010 America and Japan. To further elucidate relationships among the temperate bamboos, a very broad sam- pling of Chinese representatives was examined. We produced 9463 bp of sequences from eight non-cod- Keywords: ing chloroplast regions for 146 species in 26 genera and 5 outgroups. The loci sequenced were atpI/H, Arundinarieae psaA-ORF170, rpl32-trnL, rpoB-trnC, rps16-trnQ, trnD/T, trnS/G, and trnT/L. Phylogenetic analyses using China Chloroplast DNA regions maximum parsimony and Bayesian inference supported the monophyly of Arundinarieae. The two major Large sample size subtribes, Arundinariinae and Shibataeinae, defined on the basis of different synflorescence types, were Phylogenetic analysis indicated to be polyphyletic. Most genera in this tribe were confirmed to be paraphyletic or polyphyletic. The cladograms suggest that Arundinarieae is divided into ten major lineages. In addition to six lineages suggested in a previous molecular study (Bergbamboes, the African alpine bamboos, Chimonocalamus, the Shibataea clade, the Phyllostachys clade, and the Arundinaria clade), four additional lineages were recov- ered in our results, each represented by a single species: Gaoligongshania megalothyrsa, Indocalamus sini- cus, Indocalamus wilsonii, Thamnocalamus spathiflorus. Our analyses also indicate that (1) even more than 9000 bp of fast-evolving plastid sequence data cannot resolve the inter- and infra-relationships among and within the ten lineages of the tribe Arundinarieae; (2) an extensive sampling is indispensable for phylogeny reconstruction in this tribe, especially given that many genera appear to be paraphyletic or polyphyletic. Perhaps the ideal way to further illuminate relationships among the temperate bamboos is to sample multiple nuclear loci or whole chloroplast sequences in order to obtain sufficient variation. Ó 2010 Elsevier Inc. All rights reserved.

1. Introduction Graebner, and their herbaceous allies in tribe Olyreae Kunth ex Spen- ner (Clark et al., 1995; GPWG, 2001; Bouchenak-Khelladi et al., 2008; The grass subfamily Bambusoideae (true bamboos) as currently Sungkaew et al., 2009). circumscribed encompasses ca. 80–90 genera and 1000–1500 spe- The temperate bamboo tribe Arundinarieae is primarily distrib- cies distributed in temperate regions to mountains of the tropics uted in the North Temperate Zone or at high elevations in the Old worldwide, with the highest species richness in Asia Pacific and World tropics. There are approximately 32 genera and 600 temper- South America and the least in Africa (Bystriakova et al., 2003a,b). ate bamboo species, most of which are distributed in China and Ja- The subfamily has been resolved as monophyletic (GPWG, 2001), pan (Li, 1999; Ohrnberger, 1999). Arundinarieae is a highly consisting of members from the woody bamboo tribes Bambuseae diversified group with different habits (e.g., erect, arching, scan- s.s (Kunth ex Dumort.) and Arundinarieae Nees ex Ascherson and dent, twining, and decumbent) and complex features of morphol- ogy, including pachymorph or leptomorph rhizomes, solitary to many branches, semelauctant or iterauctant synflorescences, 3–6 * Corresponding author at: Key Laboratory of Biodiversity and Biogeography, stamens, and bacoid, nucoid or basic caryopsis (Keng and Wang, Kunming Institute of Botany, Chinese Academy of Sciences, Kunming, Yunnan 1996; Li et al., 2006; Yang et al., 2008; Yi et al., 2008). Based on 650204, PR China. Fax: +86 871 5217791. morphological and anatomical characters, the temperate bamboos E-mail addresses: [email protected], [email protected] (D.-Z. Li). are usually divided into two subtribes, Arundinariinae and 1 These authors contributed equally to this work.

1055-7903/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2010.03.041 822 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839

Table 1 Classification system of temperate bamboos by Li (1997).

Tribe Bambuseae Subtribe Arundinariinae Acidosasa (Metasasa), Ampelocalamus, Arundinaria (incl. Bashania, Pleioblastus), Chimonocalamus, Drepanostachyum (incl. Himalayacalamus), Ferrocalamus, Gaoligongshania, Gelidocalamus, Indocalamus, Oligostachyum, Pseudosasa, Sasa, Sinarundinaria (incl. Borinda, Yushania), Thamnocalamus (Fargesia) Subtribe Shibatainae Chimonobambusa, Indosasa, Phyllostachys, Qiongzhuea, Semiarundinaria (Brachystachyum), Shibataea, Sinobambusa

Shibataeinae, but with different treatments of genera within these amus jinpingensis (Yi et al., 2007b), and Indosasa jinpingensis (Yi, subtribes (Soderstrom and Ellis, 1987; Dransfield and Widjaja, 2001). Subtribal assignments in the classification of Li (1997; Table 1995; Li, 1997). 1) were adopted for this study. Various molecular data sets, including sequence data from the chloroplast genome (Ní Chonghaile, 2002; Triplett, 2008; 2.2. Taxon sampling Sungkaew et al., 2009) and nuclear GBSSI and ITS regions (Guo et al., 2001, 2002; Guo and Li, 2004; Zhuge et al., 2004; Peng Based on prior studies (Clark et al., 1995; GPWG, 2001; et al., 2008), as well as amplified fragment length (AFLP) data Bouchenak-Khelladi et al., 2008; Triplett, 2008; Sungkaew et al., (Triplett, 2008; Triplett et al., 2010) have been utilized to analyze 2009), Bonia amplexicaulis, Bonia levigata, Neomicrocalamus prainii, the temperate bamboos. These molecular phylogenetic studies Bambusa ventricosa, and Dendrocalamus farinosus of the tribe Bam- support Arundinarieae as a natural group, but relationships within buseae were chosen as outgroups. The ingroup taxa were chosen in the tribe remain unclear. Most recently, phylogenetic analyses order to include representatives for as many taxonomic groups as based on chloroplast DNA regions (including rpoB-trnC, rps16- possible within the tribe Arundinarieae. Type species for genera trnQ, trnD/T, and trnT/L intergenic spacers) provided additional in- and other subdivisions were included whenever material was avail- sights into the relationships within Arundinarieae (Triplett, 2008; able. DNA sequences from a total of 160 taxa in 30 genera were gen- Triplett and Clark, 2010). The tribe was resolved to include six ma- erated for this study from herbarium specimens or silica-dried plant jor lineages: Bergbamboes, the African alpine bamboos, Chimono- material. In addition, sequences of Yushania alpina (K. Schumann) calamus, the Shibataea clade, the Phyllostachys clade, and the Lin were retrieved from GenBank (Triplett and Clark, 2010). Table Arundinaria clade. That study also emphasized species-level 2 lists all species sequenced for this study and their sources. relationships among taxa in Japan and North America, but had relatively limited sampling of Chinese species. 2.3. Choice of markers In total, about 25 genera and 380 species of Arundinarieae occur in China (Keng and Wang, 1996; Li et al., 2006), representing In order to obtain phylogenetic resolution at species and generic approximatively 4/5 and 2/3 of the total number of genera and spe- levels within the ingroup, fast-evolving markers are needed. Triplett cies in the tribe, respectively. Of these taxa, most genera and spe- (2008) and Triplett and Clark (2010) identified twelve chloroplast re- cies are endemic to China. Therefore, the inclusion of these Chinese gions that were useful for the study of the temperate bamboos, taxa in molecular systematic studies of Arundinarieae is indispens- including the intergenic spacers atpI/H, ndhF(30 end), psaA- able for producing a comprehensive, phylogeny-based classifica- ORF170, rpl32-trnL, rpoB-trnC, rps16-trnQ, trnD/T, trnG intron, tion of the tribe. To date, no phylogenetic analysis has been trnH-psbA, trnK-rps16, trnT/L, and trnV-ndhC. After several pilot performed utilizing a comprehensive sampling of these species. studies, we also identified atpI/H, psaA-ORF170, rpl32-trnL, rpoB- In this paper, representatives from all subtribes and most genera trnC, rps16-trnQ, trnD/T, and trnT/L as being especially useful for of Arundinarieae in China according to Li (1997; Table 1) were se- our target group, plus the region trnS/G. Because of the low apparent quenced for eight chloroplast DNA regions (atpI/H, psaA-ORF170, evolutionary rate in the temperate bamboos, there is a substantial rpl32-trnL, rpoB-trnC, rps16-trnQ, trnD/T, trnS/G, and trnT/L inter- advantage in combining chloroplast regions to provide sufficient genic spacers) for a combined phylogenetic analysis. information to supported resolution within genera. Thus, each The aim of this study was to generate and analyze a large multi- accession was sequenced for all eight of these chloroplast regions. gene sequence matrix that includes a better representation of the tribe than previous studies. The main goals were: (1) to identify the major clades and inter-relationships of these clades within tribe 2.4. DNA extraction, amplification and sequencing Arundinarieae, (2) to study molecular variation among different chloroplast regions and to assess their value for phylogenetic studies Total genomic DNA was extracted from silica gel-dried leaves or in this group, and (3) to identify potential sources of information for herbarium specimens using a modified CTAB procedure (Doyle and further clarifying the phylogeny of the temperate bamboos. Doyle, 1987) or DNeasy Plant Mini Kits (Qiagen, Valencia, Califor- nia) following the manufacturer’s protocol, with the following modifications: 40–50 mg dry leaf tissue; 500 ll lysis buffer; 20– 2. Materials and methods 30 min incubation; and a final wash with ice cold 100% EtOH. The primers used for amplification and sequencing are listed in Ta- 2.1. Terminology ble 3. We used several protocols for different regions. The polymer- ase chain reaction (PCR) parameters were as follows: for atpI/H, Most scientific names of species follow Suzuki (1978), Li et al. psaA-ORF170, rpl32-trnL, trnD/T, trnS/G, and trnT/L, initial denatur- (2006) and Clark and Triplett (2007). Those of Bashania, Brachys- ation at 95 °C for 5 min, followed by 35 cycles of 30 s at 95 °C for tachyum, Metasasa, Pseudosasa guanxianensis and P. hirta follow denaturation, primer annealing at 50 °C for 1 min, and 1 min 30 s Keng and Wang (1996), and Pseudosasa nanningensis follows Zhang at 72 °C for DNA extension, followed by a final extension period and Li (in press). Five recently described species were included in of 7 min at 72 °C; for rpoB-trnC and rps16-trnQ, initial denaturation our study: Bashania abietina (Yi and Yang, 1998), Bashania aristata at 95 °C for 2 min, followed by 35 cycles of denaturation at 95 °C (Ren et al., 2003), Bashania qiaojiaensis (Yi et al., 2007a), Indocal- for 1 min, primer annealing at 48 °C for 10 s, followed by a slow Table 2 Voucher information and GenBank accession numbers for taxa used in this study. Voucher specimens are deposited in the following herbaria: KUN = Kunming Institute of Botany, China; SAUD = SiChuan Agricultural University Dujiangyan Campus, China (or SIFS = Sichuan Forestry School Dendrological Herbarium, China). Regions not sampled are indicated by an em dash (—). Description following genus name corresponds to specific characters: L or P = leptomorph or pachymorph rhizomes; S or I = semelauctant or iterauctant synflorescences; two consecutive numbers mean stamens and branches (M = many), respectively.

Taxon Voucher/herbarium Source GenBank No.

atpI/H psaA- rpl32-trnL rpoB-trnC rps16- trnD/T trnS/G trnT/L ORF170 trnQ Arundinarieae Acidosasa (LS63) Acidosasa chinensis C.D.Chu et C.S.Chao Zhang 08035 (KUN) Guangdong, China GU355045 GU355365 GU355525 GU354407 GU354565 GU354725 GU355205 GU354885 Acidosasa chienouensis (Wen) C.S.Chao et Wen Zhang 08065 (KUN) Fujian, China GU355047 GU355367 GU355527 GU354409 GU354567 GU354727 GU355207 GU354887 Acidosasa edulis Wen Zhang 08042 (KUN) Fujian, China GU355056 GU355376 GU355536 GU354418 GU354576 GU354736 GU355216 GU354896 Acidosasa guangxiensis Q.H.Dai et C.F.Huang Huang 07017 (KUN) Guangxi, China GU355016 GU355336 GU355496 GU354378 GU354536 GU354696 GU355176 GU354856 Acidosasa nanunica (MaClure) C.S. Chao et G.Y.Yang Zhang & Zeng 06112 Hunan, China GU355019 GU355339 GU355499 GU354381 GU354539 GU354699 GU355179 GU354859 (KUN) Acidosasa notata (Z.P.Wang et G.H.Ye)S.S.You Zhang 08061 (KUN) Fujian, China GU355049 GU355369 GU355529 GU354411 GU354569 GU354729 GU355209 GU354889 .X ege l oeua hlgntc n vlto 6(00 821–839 (2010) 56 Evolution and Phylogenetics Molecular / al. et Zeng C.-X. Acidosasa purpurea (Hsueh et Yi) Keng f. Zhang 07067 (KUN) Yunnan, China GU355020 GU355340 GU355500 GU354382 GU354540 GU354700 GU355180 GU354860 Ampelocalamus (PS3M) Ampelocalamus actinotrichus (Merr. et Chun) S.L.Chen et al. Zeng & Zhang 06054 Hainan, China GU355081 GU355401 GU355561 — GU354601 GU354761 GU355241 GU354921 (KUN) Ampelocalamus patellaris (Gamble) Stapleton Zhang 07075 (KUN) Yunnan, China GU355082 GU355402 GU355562 — GU354602 GU354762 GU355242 GU354922 Ampelocalamus scandens Hsueh et W.D.Li Zhen-Hua Guo 013 Yunnan, China GU355118 GU355438 GU355598 GU354478 GU354638 GU354798 GU355278 GU354958 (KUN) Arundinaria (LS33) Arundinaria appalachiana Triplett Triplett 184 (ISC) North Carolina, United GU355021 GU355341 GU355501 GU354383 GU354541 GU354701 GU355181 GU354861 States Arundinaria gigantea (Walter) Muhl. Zhang US1025 (KUN) Arkansas, United States GU355135 GU355455 GU355615 GU354495 GU354655 GU354815 GU355295 GU354975 Arundinaria tecta (Walter) Muhl Triplett 173 (ISC) South Carolina, United GU355024 GU355344 GU355504 GU354386 GU354544 GU354704 GU355184 GU354864 States Bashania (LS3(1–3 to M)) Bashania abietina T.P.Yi et L.Yang Zhang 07092 Sichuan, China GU355143 GU355463 GU355623 GU354503 GU354663 GU354823 GU355303 GU354983 Bashania aristata Y.Ren, Y.Li et G.D.Dang Zhang 08080 Shaanxi, China GU355063 GU355383 GU355543 GU354425 GU354583 GU354743 GU355223 GU354903 Bashania fangiana (A. Camus) Keng f. et Wen Lu 071104 (KUN) Sichuan, China GU355017 GU355337 GU355497 GU354379 GU354537 GU354697 GU355177 GU354857 Bashania fargesii (Camus) P.C.Keng et Yi Zhang 08083 (KUN) Shaanxi, China GU355062 GU355382 GU355542 GU354424 GU354582 GU354742 GU355222 GU354902 Bashania qiaojiaensis Hsueh et Yi Zhang 07046 (KUN) Yunnan, China GU355142 GU355462 GU355622 GU354502 GU354662 GU354822 GU355302 GU354982 Bashania qingchengshanensis P.C.Keng et T.P.Yi Zhang 07085 (KUN) Sichuan, China GU355145 GU355465 GU355625 GU354505 GU354665 GU354825 GU355305 GU354985 Bashania spanostachya Yi Zhang 07093 (KUN) Sichuan, China GU355144 GU355464 GU355624 GU354504 GU354664 GU354824 GU355304 GU354984 Brachystachyum ((L or P)I33) Brachystachyum densiflorum (Rendle) Keng Zeng & Zhang 06174 Zhejiang, China GU355127 GU355447 GU355607 GU354487 GU354647 GU354807 GU355287 GU354967 (KUN) Chimonobambusa (PI33) Chimonobambusa macrophylla Wen et Ohrnb Zhang 07091 (KUN) Sichuan, China GU355122 GU355442 GU355602 GU354482 GU354642 GU354802 GU355282 GU354962 Chimonobambusa sichuanensis (T.P.Yi) T.H.Wen Zhang 07084 (KUN) Sichuan, China GU355011 GU355331 GU355491 GU354373 GU354531 GU354691 GU355171 GU354851 Chimonobambusa szechuanensis (Rendle) P.C.Keng Lu 2711 (KUN) Sichuan, China GU355124 GU355444 GU355604 GU354484 GU354644 GU354804 GU355284 GU354964 Chimonocalamus (PS33) Chimonocalamus dumosus Hsueh et Yi Zhang 07061 (KUN) Yunnan, China GU355039 GU355359 GU355519 GU354401 GU354559 GU354719 GU355199 GU354879 Chimonocalamus fimbriatus Hsueh et Yi Zeng et al. 08020 (KUN) Yunnan, China GU355036 GU355356 GU355516 GU354398 GU354556 GU354716 GU355196 GU354876 Chimonocalamus longiusculus Hsueh et Yi Zhang 07064 (KUN) Yunnan, China GU355037 GU355357 GU355517 GU354399 GU354557 GU354717 GU355197 GU354877 Chimonocalamus montanus Hsueh et Yi Zhang 07057 (KUN) Yunnan, China GU355038 GU355358 GU355518 GU354400 GU354558 GU354718 GU355198 GU354878 Chimonocalamus pallens Hsueh et Yi Zhang 07071 (KUN) Yunnan, China GU355059 GU355379 GU355539 GU354421 GU354579 GU354739 GU355219 GU354899 Drepanostachyum (PS3M) Drepanostachyum ampullare (T.P.Yi) Demoly GLM 081860 (KUN) Xizang, China GU355079 GU355399 GU355559 GU354441 GU354599 GU354759 GU355239 GU354919 Drepanostachyum hookerianum (Munro) P.C.Keng DZL 199903 (KUN) Kew, Britain GU355116 GU355436 GU355596 GU354476 GU354636 GU354796 GU355276 GU354956 Fargesia (PS3M) 823 (continued on next page) 824 Table 2 (continued)

Taxon Voucher/herbarium Source GenBank No.

atpI/H psaA- rpl32-trnL rpoB-trnC rps16- trnD/T trnS/G trnT/L ORF170 trnQ Fargesia decurvata J.L.Lu Zhang 07087 (KUN) Hubei, China GU355073 GU355393 GU355553 GU354435 GU354593 GU354753 GU355233 GU354913 Fargesia edulis Hsueh et Yi; Li & Zhang 07051 (KUN) Yunnan, China GU355130 GU355450 GU355610 GU354490 GU354650 GU354810 GU355290 GU354970 Fargesia fungosa T.P.Yi Zhang 07048 (KUN) Yunnan, China GU355129 GU355449 GU355609 GU354489 GU354649 GU354809 GU355289 GU354969 Fargesia macclureana (Bor) Stapleton GLM 082064 (KUN) Xizang, China GU355085 GU355405 GU355565 GU354445 GU354605 GU354765 GU355245 GU354925 Fargesia nitida (Mitford) Keng f. et Yi Zhang KMBG10 (KUN) Sichuan, China GU355120 GU355440 GU355600 GU354480 GU354640 GU354800 GU355280 GU354960 Fargesia qinlingensis Yi et J.X .Shao Zhang 08079 (KUN) Shaanxi, China GU355086 GU355406 GU355566 GU354446 GU354606 GU354766 GU355246 GU354926 Fargesia robusta Yi Zhang 08014 (KUN) Sichuan, China GU355083 GU355403 GU355563 GU354443 GU354603 GU354763 GU355243 GU354923 Ferrocalamus (LS31) Ferrocalamus rimosivaginus Wen Zhang 07068 (KUN) Yunnan, China GU355095 GU355415 GU355575 GU354455 GU354615 GU354775 GU355255 GU354935 Ferrocalamus strictus Hsueh et P.C.Keng Zeng & Zhang SB1 (KUN) Yunnan, China GU355096 GU355416 GU355576 GU354456 GU354616 GU354776 GU355256 GU354936 Gaoligongshania (PS31) .X ege l oeua hlgntc n vlto 6(00 821–839 (2010) 56 Evolution and Phylogenetics Molecular / al. et Zeng C.-X. Gaoligongshania megalothyrsa (Handel-Mazzetti) D.Z.Li, JRX 9401 (KUN) Yunnan, China GU355121 GU355441 GU355601 GU354481 GU354641 GU354801 GU355281 GU354961 Hsueh et N.H.Xia Gelidocalamus (LS3(7–12)) Gelidocalamus rutilans Wen Zeng & Zhang 06183 Zhejiang, China GU355151 GU355471 GU355631 GU354511 GU354671 GU354831 GU355311 GU354991 (KUN) Gelidocalamus sp1. Zeng & Zhang Jing (KUN) Guangdong, China GU355094 GU355414 GU355574 GU354454 GU354614 GU354774 GU355254 GU354934 Gelidocalamus sp2. Zeng & Zhang 06180 Zhejiang, China GU355152 GU355472 GU355632 GU354512 GU354672 GU354832 GU355312 GU354992 (KUN) Gelidocalamus tessellatus Wen et J.Q.Zhang Zeng & Zhang 06200 Guangxi, China GU355155 GU355475 GU355635 GU354515 GU354675 GU354835 GU355315 GU354995 (KUN) Himalayacalamus (PS3M) Himalayacalamus falconeri (Munro) P.C.Keng GLM 081524 (KUN) Xizang, China GU355080 GU355400 GU355560 GU354442 GU354600 GU354760 GU355240 GU354920 Indocalamus (LS31) Indocalamsu aff. latifolius (Keng) McClure Triplett 243 (KUN) Tennessee, United GU355146 GU355466 GU355626 GU354506 GU354666 GU354826 GU355306 GU354986 States Indocalamus barbatus McClure Zeng & Zhang 06198 Guangxi, China GU355093 GU355413 GU355573 GU354453 GU354613 GU354773 GU355253 GU354933 (KUN) Indocalamus bashanensis (C.D.Chu et C.S.Chao) H.R.Zhao et Zhang 07083 (KUN) Sichuan, China GU355148 GU355468 GU355628 GU354508 GU354668 GU354828 GU355308 GU354988 Y.L.Yang Indocalamus decorus Q.H.Dai Zeng & Zhang 06208 Guangxi, China GU355104 GU355424 GU355584 GU354464 GU354624 GU354784 GU355264 GU354944 (KUN) Indocalamus emeiensis C.D.Chu et C.S.Chao Zeng & SD Zhang 07001 Sichuan, China GU355101 GU355421 GU355581 GU354461 GU354621 GU354781 GU355261 GU354941 (KUN) Indocalamus guangdongensis H.R.Zhao et Y.L. Yang Zeng & Zhang 06156 Guangdong, China GU355099 GU355419 GU355579 GU354459 GU354619 GU354779 GU355259 GU354939 (KUN) Indocalamus herklotsii McClure Zeng & Zhang 06147 Zhejiang, China GU355103 GU355423 GU355583 GU354463 GU354623 GU354783 GU355263 GU354943 (KUN) Indocalamus hirsutissimus Z.P.Wang et P.X.Zhang Zhang 07033 (KUN) Guangxi, China GU355147 GU355467 GU355627 GU354507 GU354667 GU354827 GU355307 GU354987 Indocalamus hirtivaginatus H.R.Zhao et Y.L.Yang Yi 06016 (SAUD = SIFS) Shaanxi, China, GU355107 GU355427 GU355587 GU354467 GU354627 GU354787 GU355267 GU354947 Indocalamus jinpingensis T.P.Yi et al. Zeng et al. 08022 (KUN) Yunnan, China GU355042 GU355362 GU355522 GU354404 GU354562 GU354722 GU355202 GU354882 Indocalamus latifolius (Keng) McClure Zeng & Zhang 06128 Zhejiang, China GU355100 GU355420 GU355580 GU354460 GU354620 GU354780 GU355260 GU354940 (KUN) Indocalamus longiauritus Handel-Mazzetti Zeng & Zhang 06211 Guangxi, China GU355150 GU355470 GU355630 GU354510 GU354670 GU354830 GU355310 GU354990 (KUN) Indocalamus pseudosinicus McClure Zeng & Zhang 06068 Hainan, China GU355089 GU355409 GU355569 GU354449 GU354609 GU354769 GU355249 GU354929 (KUN) Indocalamus quadratus H.R.Zhao et Y.L.Yang Zeng & Xing 06001 Guangxi, China GU355098 GU355418 GU355578 GU354458 GU354618 GU354778 GU355258 GU354938 (KUN) Indocalamus sinicus (Hance) Nakai Zeng & Zhang 06081 Hainan, China GU355153 GU355473 GU355633 GU354513 GU354673 GU354833 GU355313 GU354993 (KUN) Indocalamus sinicus (Hance) Nakai Zhang 08034 (KUN) Guangdong, China GU355154 GU355474 GU355634 GU354514 GU354674 GU354834 GU355314 GU354994 Indocalamus tessellatus (Munro) P.C.Keng Zeng & Zhang 06176 Zhejiang, China GU355149 GU355469 GU355629 GU354509 GU354669 GU354829 GU355309 GU354989 (KUN) Indocalamus tongchunensis K.F.Huang et Z.L.Dai Zhang 08075 (KUN) Fujian, China GU355043 GU355363 GU355523 GU354405 GU354563 GU354723 GU355203 GU354883 Indocalamus victorialis P.C.Keng Zeng & Zhang 06152 Zhejiang, China GU355109 GU355429 GU355589 GU354469 GU354629 GU354789 GU355269 GU354949 (KUN) Indocalamus wilsonii (Rendel) C.S.Chao et C.D.Chu Zeng & SD Zhang 07119 Chongqing, China GU355105 GU355425 GU355585 GU354465 GU354625 GU354785 GU355265 GU354945 (KUN) Indocalamus wilsonii (Rendel) C.S.Chao et C.D.Chu Zhang 07088 (KUN) Hubei, China GU355106 GU355426 GU355586 GU354466 GU354626 GU354786 GU355266 GU354946 Indocalamus wuxiensis Yi Zeng & Zhang 06094 Guangdong, China GU355108 GU355428 GU355588 GU354468 GU354628 GU354788 GU355268 GU354948 (KUN) Indosasa (LI63) Indosasa crassiflora McClure Zhang 07014 (KUN) Guangxi, China GU355066 GU355386 GU355546 GU354428 GU354586 GU354746 GU355226 GU354906 Indosasa gigantea (Wen) Wen Zhang & Zeng 06139 Zhejiang, China GU355002 GU355322 GU355482 GU354364 GU354522 GU354682 GU355162 GU354842 (KUN) Indosasa gigantea (Wen) Wen Zhang 08067 (KUN) Fujian, China GU355048 GU355368 GU355528 GU354410 GU354568 GU354728 GU355208 GU354888 Indosasa hispida McClure Zhang 08026 (KUN) Guangdong, China GU355006 GU355326 GU355486 GU354368 GU354526 GU354686 GU355166 GU354846 .X ege l oeua hlgntc n vlto 6(00 821–839 (2010) 56 Evolution and Phylogenetics Molecular / al. et Zeng C.-X. Indosasa jinpingensis Yi Zhang et al. 08021 (KUN) Yunnan, China GU355041 GU355361 GU355521 GU354403 GU354561 GU354721 GU355201 GU354881 Indosasa lipoensis C.D.Chu et K.M.Lan Zhang 07040 (KUN) Guizhou, China GU355008 GU355328 GU355488 GU354370 GU354528 GU354688 GU355168 GU354848 Indosasa patens C.D.Chu et C.S.Chao Zhang 07028 (KUN) Guangxi, China GU355010 GU355330 GU355490 GU354372 GU354530 GU354690 GU355170 GU354850 Indosasa shibataeoides McClure Zhang 07030 (KUN) Guangxi, China GU355012 GU355332 GU355492 GU354374 GU354532 GU354692 GU355172 GU354852 Indosasa sinica C.D.Chu et C.S.Chao Zhang 07001 (KUN) Guangxi, China GU355013 GU355333 GU355493 GU354375 GU354533 GU354693 GU355173 GU354853 Indosasa sp. Zhang & Zeng 06088 Guangdong, China GU355040 GU355360 GU355520 GU354402 GU354560 GU354720 GU355200 GU354880 (KUN) Indosasa spongiosa C.S.Chao et B.M.Yang Zhang 07022 (KUN) Guangxi, China GU355014 GU355334 GU355494 GU354376 GU354534 GU354694 GU355174 GU354854 Indosasa triangulata Hsueh et Yi Zhang 07072 (KUN) Yunnan, China GU355015 GU355335 GU355495 GU354377 GU354535 GU354695 GU355175 GU354855 Metasasa (LS62) Metasasa carinata W.T.Lin Zhang 08031 (KUN) Guangdong, China GU355044 GU355364 GU355524 GU354406 GU354564 GU354724 GU355204 GU354884 Oligostachyum ((L or P)S(3 or 4 (5))3) Oligostachyum gracilipes (McClure) G.H.Ye et Z.P.Wang Zhang & Zeng 06071 Hainan, China GU355018 GU355338 GU355498 GU354380 GU354538 GU354698 GU355178 GU354858 (KUN) Oligostachyum hupehense (J.L.Lu) Z.P.Wang et G.H.Ye Zhang 07089 (KUN) Hubei, China GU355033 GU355353 GU355513 GU354395 GU354553 GU354713 GU355193 GU354873 Oligostachyum lubricum (Wen) Keng f. Zhang & Zeng 06150 Zhejiang, China GU355035 GU355355 GU355515 GU354397 GU354555 GU354715 GU355195 GU354875 (KUN) Oligostachyum oedogonatum (Z.P.Wang et G.H.Ye) Q.F.Zhang Zhang & Zeng 06151 Zhejiang, China GU355022 GU355342 GU355502 GU354384 GU354542 GU354702 GU355182 GU354862 et K.F.Huang (KUN) Oligostachyum paniculatum G.H.Ye et Z.P.Wang Zhang 07011 (KUN) Guangxi, China GU355004 GU355324 GU355484 GU354366 GU354524 GU354684 GU355164 GU354844 Oligostachyum scabriflorum (McClure) Z.P.Wang et G.H.Ye Zhang 07007 (KUN) Guangxi, China GU355034 GU355354 GU355514 GU354396 GU354554 GU354714 GU355194 GU354874 Oligostachyum shiuyingianum (Chia et But) G.H.Ye et DZL 09122 (KUN) Hongkong, China GU355070 GU355390 GU355550 GU354432 GU354590 GU354750 GU355230 GU354910 Z.P.Wang Oligostachyum sulcatum Z.P.Wang et G.H.Ye Zhang 07024 (KUN) Guangxi, China GU355023 GU355343 GU355503 GU354385 GU354543 GU354703 GU355183 GU354863 Phyllostachys (LI32) Phyllostachys edulis (Carriere) Houzeau Zhang KMBG04 (KUN) Yunnan, China GU355134 GU355454 GU355614 GU354494 GU354654 GU354814 GU355294 GU354974 Phyllostachys heteroclada Oliver PSH14 (KUN) Yunnan, China GU355061 GU355381 GU355541 GU354423 GU354581 GU354741 GU355221 GU354901 Phyllostachys nidularia Munro Zhang 08078 (KUN) Yunnan, China GU355060 GU355380 GU355540 GU354422 GU354580 GU354740 GU355220 GU354900 Phyllostachys nigra (Loddiges ex Lindley) Munro Zhang KMBG07 (KUN) Yunnan, China GU355133 GU355453 GU355613 GU354493 GU354653 GU354813 GU355293 GU354973 Pleioblastus (LS3(3–7)) Pleioblastus amarus (Keng) P.C.Keng Zhang 07082 (KUN) Sichuan, China GU355132 GU355452 GU355612 GU354492 GU354652 GU354812 GU355292 GU354972 Pleioblastus argenteostriatus (Regel.)Nakai Zhang 07080 (KUN) Yunnan, China GU355069 GU355389 GU355549 GU354431 GU354589 GU354749 GU355229 GU354909 Pleioblastus gramineus (Bean) Nakai Zhang & Zeng 06157 Zhejiang, China GU355025 GU355345 GU355505 GU354387 GU354545 GU354705 GU355185 GU354865 (KUN) Pleioblastus hsienchuensis var. subglabratus (S.Y.Chen) Zhang 08056 (KUN) Fujian, China GU355054 GU355374 GU355534 GU354416 GU354574 GU354734 GU355214 GU354894 C.S.Chao et G.Y.Yang Pleioblastus intermedius S.Y.Chen Zhang & Zeng 06188 Zhejiang, China GU355026 GU355346 GU355506 GU354388 GU354546 GU354706 GU355186 GU354866 (KUN) Pleioblastus juxianensis Wen et al. Zhang & Zeng 06136 Zhejiang, China GU355046 GU355366 GU355526 GU354408 GU354566 GU354726 GU355206 GU354886

(continued on next page) 825 826 Table 2 (continued)

Taxon Voucher/herbarium Source GenBank No. atpI/H psaA- rpl32-trnL rpoB-trnC rps16- trnD/T trnS/G trnT/L ORF170 trnQ (KUN) Pleioblastus maculatus (McClure) C.D.Chu et C.S.Chao Zhang 07049 (KUN) Yunnan, China GU355057 GU355377 GU355537 GU354419 GU354577 GU354737 GU355217 GU354897 Pleioblastus pygmaeus (Miquel) Nakai Zeng & Zhang 06059 Guangdong, China GU355088 GU355408 GU355568 GU354448 GU354608 GU354768 GU355248 GU354928 (KUN) Pleioblastus sanmingensis S.Y.Chen et G.Y.Sheng Zhang 08074 (KUN) Fujian, China GU355053 GU355373 GU355533 GU354415 GU354573 GU354733 GU355213 GU354893 Pleioblastus solidus S.Y.Chen Zhang & Zeng 06190 Zhejiang, China GU355131 GU355451 GU355611 GU354491 GU354651 GU354811 GU355291 GU354971 (KUN) Pleioblastus sp. Zeng & Zhang 06141 Zhejiang, China GU355087 GU355407 GU355567 GU354447 GU354607 GU354767 GU355247 GU354927 (KUN) Pleioblastus wuyishanensis Q.F.Zheng et K.F.Huang Zhang 08062 (KUN) Fujian, China GU355055 GU355375 GU355535 GU354417 GU354575 GU354735 GU355215 GU354895 Pleioblastus yixingensis S.L.Chen et S.Y.Chen Zhang & Zeng 06134 Zhejiang, China GU355058 GU355378 GU355538 GU354420 GU354578 GU354738 GU355218 GU354898

(KUN) 821–839 (2010) 56 Evolution and Phylogenetics Molecular / al. et Zeng C.-X. Pseudosasa (LS(3(4 or 5))(1–3 to M)) Pseudosasa acutivagina Wen et S.C.Chen Zhang & Zeng 06186 Zhejiang, China GU355007 GU355327 GU355487 GU354369 GU354527 GU354687 GU355167 GU354847 (KUN) Pseudosasa amabilis (McClure) Keng f. Zhang & Zeng 06140 Zhejiang, China GU355064 GU355384 GU355544 GU354426 GU354584 GU354744 GU355224 GU354904 (KUN) Pseudosasa amabilis var. convexa Z.P.Wang et G.H.Ye Zhang 08054 (KUN) Fujian, China GU355050 GU355370 GU355630 GU354412 GU354570 GU354730 GU355210 GU354890 Pseudosasa cantorii (Munro) Keng f. Zhang & Zeng 06045 Guangdong, China GU355071 GU355391 GU355551 GU354433 GU354591 GU354751 GU355231 GU354911 (KUN) Pseudosasa gracilis S.L.Chen et G.Y.Sheng Zhang & Zeng 06107 Guangdong, China GU355029 GU355349 GU355509 GU354391 GU354549 GU354709 GU355189 GU354869 (KUN) Pseudosasa guanxianensis Yi Zhang 07081 (KUN) Sichuan, China GU355005 GU355325 GU355485 GU354367 GU354525 GU354685 GU355165 GU354845 Pseudosasa hindsii (Munro) C.D.Chu et C.S.Chao Zhang 07013 (KUN) Guangxi, China GU355030 GU355350 GU355510 GU354392 GU354550 GU354710 GU355190 GU354870 Pseudosasa hirta S.L.Chen et G.Y.Sheng Peng 07096 (KUN) Jiangxi, China GU355065 GU355385 GU355545 GU354427 GU354585 GU354745 GU355225 GU354905 Pseudosasa japonica (Sieb. et Zucc.) Makino Zhang 07023 (KUN) Guangxi, China GU355027 GU355347 GU355507 GU354389 GU354547 GU354707 GU355187 GU354867 Pseudosasa japonica var. tsutsumiana Zhang & Zeng 06011 Zhejiang, China GU355028 GU355348 GU355508 GU354390 GU354548 GU354708 GU355188 GU354868 (KUN) Pseudosasa longiligula Wen Zhang 07021 (KUN) Guangxi, China GU355067 GU355387 GU355547 GU354429 GU354587 GU354747 GU355227 GU354907 Pseudosasa maculifera J.L.Lu Ha 07094 (KUN) Henan, China GU355068 GU355388 GU355548 GU354430 GU354588 GU354748 GU355228 GU354908 Pseudosasa nanningensis (Q.H.Dai) D.Z.Li et Y.Z.Zhang Zhang 07002 (KUN) Guangxi, China GU355009 GU355329 GU355489 GU354371 GU354529 GU354689 GU355169 GU354849 Pseudosasa orthotropa S.L.Chen et Wen Zhang & Zeng 06161 Zhejiang, China GU355001 GU355321 GU355481 GU354363 GU354521 GU354681 GU355161 GU354841 (KUN) Pseudosasa viridula S.L.Chen et G.Y.Sheng Zhang & Zeng 06164 Zhejiang, China GU355003 GU355323 GU355483 GU354365 GU354523 GU354683 GU355163 GU354843 (KUN) Sasa (LS61) Sasa guangxiensis C.D.Chu et C.S.Chao Zeng & Zhang 06197 Guangxi, China GU355092 GU355412 GU355572 GU354452 GU354612 GU354772 GU355252 GU354932 (KUN) Sasa kurilensis (Ruprecht) Makino et Shibata Triplett 223 (KUN) California, United GU355137 GU355457 GU355617 GU354497 GU354657 GU354817 GU355297 GU354977 States Sasa longiligulata McClure Zeng & Zhang 06123 Guangdong, China GU355090 GU355410 GU355570 GU354450 GU354610 GU354770 GU355250 GU354930 (KUN) Sasa magnonoda Wen et G. L. Liao Yi 06024 (SAUD = SIFS) Jiangxi, China GU355091 GU355411 GU355571 GU354451 GU354611 GU354771 GU355251 GU354931 Sasa oblongula C.H.Hu Zeng & Zhang 06055 Guangdong, China GU355112 GU355432 GU355592 GU354472 GU354632 GU354792 GU355272 GU354952 Sasa oshidensis Makino et Uchida Triplett 161 (KUN) Tennessee, United GU355136 GU355456 GU355616 GU354496 GU354656 GU354816 GU355296 GU354976 States Sasa palmata (Mitford) Camus Triplett 228 (KUN) California, United GU355141 GU355461 GU355621 GU354501 GU354661 GU354821 GU355301 GU354981 States Sasa qingyuanensis (C.H.Hu) C.H.Hu Zeng & Zhang 06182 Zhejiang, China GU355097 GU355417 GU355577 GU354457 GU354617 GU354777 GU355257 GU354937 (KUN) Sasa senanensis (Franchet et Savatier) Rehder Triplett 146 (KUN) Tennessee, United GU355111 GU355431 GU355591 GU354471 GU354631 GU354791 GU355271 GU354951 States Sasa sinica Keng Zeng & Zhang 06175 Zhejiang, China GU355102 GU355422 GU355582 GU354462 GU354622 GU354782 GU355262 GU354942 (KUN) Sasa sp1. Zeng & Zhang 06098 Guangdong, China GU355110 GU355430 GU355590 GU354470 GU354630 GU354790 GU355270 GU354950 (KUN) Sasa sp2. Zeng & Zhang 06092 Guangdong, China GU355140 GU355460 GU355620 GU354500 GU354660 GU354820 GU355300 GU354980 (KUN) Sasa tsuboiana Makino Triplett 133 (KUN) Tennessee, United GU355139 GU355459 GU355619 GU354499 GU354659 GU354819 GU355299 GU354979 States Sasa veitchii (Carriere) Rehder Triplett 126 (KUN) Tennessee, United GU355138 GU355458 GU355618 GU354498 GU354658 GU354818 GU355298 GU354978 States Sasaella (LS61) Sasaella ramosa (Makino) Makino Triplett 140 (KUN) Tennessee, United GU355113 GU355433 GU355593 GU354473 GU354633 GU354793 GU355273 GU354953 States Shibataea (PI3(3 to5)) Shibataea chinensis Nakai PSH s.n. (KUN) Yunnan, China GU355115 GU355435 GU355595 GU354475 GU354635 GU354795 GU355275 GU354955 Shibataea hispida McClure PSH 10 (KUN) Zhejiang, China GU355114 GU355434 GU355594 GU354474 GU354634 GU354794 GU355274 GU354954

Shibataea lanceifolia C.H.Hu Zhang 08064 (KUN) Fujian, China GU355051 GU355371 GU355531 GU354413 GU354571 GU354731 GU355211 GU354891 821–839 (2010) 56 Evolution and Phylogenetics Molecular / al. et Zeng C.-X. Shibataea nanpingensis Q.F.Zheng et K.F.Huang Zhang 08060 (KUN) Fujian, China GU355052 GU355372 GU355532 GU354414 GU354572 GU354732 GU355212 GU354892 Sinobambusa (LI33) Sinobambusa intermedia McClure Zhang & Zeng 06133 Zhejiang, China GU355032 GU355352 GU355512 GU354394 GU354552 GU354712 GU355192 GU354872 (KUN) Sinobambusa tootsik (Sieb.) Makino Zhang & Zeng 06090 Guangdong, China GU355031 GU355351 GU355511 GU354393 GU354551 GU354711 GU355191 GU354871 (KUN) Thamnocalamus (PS3(3 to M)) Thamnocalamus spathiflorus (Trin.) Munro Mcbeath 19901722 Kew, Britain GU355119 GU355439 GU355599 GU354479 GU354639 GU354799 GU355279 GU354959 (KUN) Thamnocalamus spathiflorus (Trin.) Munro GLM 081775 (KUN) Xizang, China GU355074 GU355394 GU355554 GU354436 GU354594 GU354754 GU355234 GU354914 Thamnocalamus spathiflorus var. crassinodus (T.P.Yi) DZL 199902 (KUN) Kew, Britain GU355117 GU355437 GU355597 GU354477 GU354637 GU354797 GU355277 GU354957 Stapleton Thamnocalamus spathiflorus var. crassinodus (T.P.Yi) GLM 081578 (KUN) Xizang, China GU355076 GU355396 GU355556 GU354438 GU354596 GU354756 GU355236 GU354916 Stapleton Thamnocalamus spathiflorus var. crassinodus (T.P.Yi) GLM 081593 (KUN) Xizang, China GU355075 GU355395 GU355555 GU354437 GU354595 GU354755 GU355235 GU354915 Stapleton Thamnocalamus tessellatus (Nees) Soderstrom et Ellis DZL 199901 (KUN) Kew, Britain GU355125 GU355445 GU355605 GU354485 GU354645 GU354805 GU355285 GU354965 Yushania (PS3 (1 to M)) Yushania alpina (K.Schumann) Lin Triplett and Clark (2010) FJ643705 FJ643759 FJ643788 FJ643971 FJ643878 FJ644064 — FJ644215 Yushania baishanzuensis Z.P.Wang et G.H.Ye Zeng & Zhang 06181 Zhejiang, China GU355123 GU355443 GU355603 GU354483 GU354643 GU354803 GU355283 GU354963 (KUN) Yushania basihirsuta (McClure) Z.P.Wang et G.H.Ye Zeng & Zhang 06108 Hunan, China GU355126 GU355446 GU355606 GU354486 GU354646 GU354806 GU355286 GU354966 (KUN) Yushania brevipaniculata (Hand.-Mazz.) Yi Zhang 08005 (KUN) Sichuan, China GU355072 GU355392 GU355552 GU354434 GU354592 GU354752 GU355232 GU354912 Yushania crassicollis Yi Liuj 08032 (KUN) Yunnan, China GU355077 GU355397 GU355557 GU354439 GU354597 GU354757 GU355237 GU354917 Yushania maculata Yi Zhang 08006 (KUN) Sichuan, China GU355084 GU355404 GU355564 GU354444 GU354604 GU354764 GU355244 GU354924 Yushania qiaojiaensis Hsueh et T.P.Yi Zhang 07044 (KUN) Yunnan, China GU355128 GU355448 GU355608 GU354488 GU354648 GU354808 GU355288 GU354968 Yushania yadongensis Yi GLM 081767 (KUN) Xizang, China GU355078 GU355398 GU355558 GU354440 GU354598 GU354758 GU355238 GU354918

Bambuseae Bonia (PI61) Bonia amplexicaulis (L.C.Chia et al.) N.H.Xia Zeng & Zhang SB5 (KUN) Yunnan, China GU355160 GU355480 GU355640 GU354520 GU354680 GU354840 GU355320 GU355000 Bonia levigata (L.C.Chia et al.) N.H.Xia Zeng & Zhang 06076 Hainan, China GU355159 GU355479 GU355639 GU354519 GU354679 GU354839 GU355319 GU354999 (KUN) Bambusa (PI6M) Bambusa ventricosa McClure Zhang KMBG09 (KUN) Yunnan, China GU355158 GU355478 GU355638 GU354518 GU354678 GU354838 GU355318 GU354998 Dendrocalamus (PI6M) Dendrocalamus farinosus (Keng et Keng f.)Chia et H. L.Fung Zhang SB2701 (KUN) Yunnan, China GU355157 GU355477 GU355637 GU354517 GU354677 GU354837 GU355317 GU354997 Neomicrocalamus (PI6M) Neomicrocalamus prainii (Gamble) Keng f. LL07236 (KUN) Xizang, China GU355156 GU355476 GU355636 GU354516 GU354676 GU354836 GU355316 GU354996 827 828 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839

Table 3 Chloroplast DNA primers used for amplification (A) and sequencing (S).

Region Name Primer sequence (50–30) Source Amp/Seq trnS/G trnGUUC GTAGCGGGAATCGAACCCGCATC Shaw et al. (2005) A, S trnSGCU AGATAGGGATTCGAACCCTCGGT Shaw et al. (2005) A, S trnD/T trnD-for ACCAATTGAACTACAATCCC Demesure et al. (1995) A trnT-rev CCCTTTTAACTCAGTGGTA Triplett (2008) and Triplett and Clark (2010) A trnE-for GCCTCCTTGAAAGAGAGATG Doyle et al. (1992) S trnY-rev CTCTTTGCTTTGGATCTAG Triplett (2008) and Triplett and Clark (2010) S rpoB-trnC trnC TGGGGATAAAGGATTTGCAG Yamane and Kawahara (2005) A rpoB ATTGTGGACATTCCCTCRTT Yamane and Kawahara (2005) A Jt400-for CAGGTCCGAACAGCATTA Triplett (2008) and Triplett and Clark (2010) S Jt700-rev CGTAGTAGTAGAATTGCTAG Triplett (2008) and Triplett and Clark (2010) S trnT/L trnTL for CATTACAAATGCGATGCTCT Taberlet et al. (1991) A, S trnTL rev TCTACCGATTTCGCCATATC Taberlet et al. (1991) A, S psaA-ORF170 psaA TCGAAATCGTGAGCATCAGC Saltonstall (2001) A, S ORF170 TCTCAAGTACGGTTCTAGG Saltonstall (2001) A, S rpl32-trnL trnLUAG CTGCTTCCTAAGAGCAGCGT Shaw et al. (2007) A, S rpl32-F CAGTTCCAAAAAAACGTACTTC Shaw et al. (2007) A, S atpI/H atpI TATTTACAAGYGGTATTCAAGCT Shaw et al. (2007) A, S atpH CCAAYCCAGCAGCAATAAC Shaw et al. (2007) A, S rps16-trnQ 1F GCACGTTGCTTTCTACCACA Triplett (2008) and Triplett and Clark (2010) A, S 1574R ATCCTTCCGTCCCAGATTTT Triplett (2008) and Triplett and Clark (2010) A. S 334F CGAGATGGTCAATCCTGAAATG Triplett (2008) and Triplett and Clark (2010) S 628R CTTTTGGTATTCKAGTCGAAG Triplett (2008) and Triplett and Clark (2010) S

ramp of 0.3 °C/s to 65 °C, and primer extension at 65 °C for 5 min. A manually where necessary. Substitution and indels were used as final extension step consisted of 5 min at 65 °C. Alternative proto- equally probable events. Potentially informative indels that were cols for atpI/H, psaA-ORF170, rpl32-trnL, rps16-trnQ, trnD/T, trnS/G, located in regions of unambiguous alignment were scored follow- and trnT/L were 1 cycle of 2 min at 95 °C, followed by 35 cycles of ing the ‘‘simple indel coding” method (Simmons and Ochoterena, 1 min at 95 °C, 10 s at 48 °Cor50°C, ramped to 65 °C at 0.3 °C/s, 2000) and added to the matrix as binary presence/absence charac- and 1 min 30 s at 65 °C, followed by 1 cycle of 10 min at 65 °C; ters. This provides a useful method to utilize indels as phylogenet- for rpoB-trnC 1 cycle of 2 min at 95 °C, followed by 35 cycles of ically information (Kawakita et al., 2003; Guo and Li, 2004). All 1 min at 95 °C, 10 s at 46 °C, ramped to 65 °C at 0.1 °C/s, and data matrices are available in TreeBASE (study accession number 1 min 30 s at 65 °C, followed by 1 cycle of 10 min at 65 °C. Reac- SN4974) and all sequences obtained in the present study have been tions were performed with 0.2 lM of each primer, 10 mM Tris– deposited in Genbank (accession numbers are listed in Table 2).

HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 200 lM of each dNTP, To test the phylogenetic effects of analyzing different regions of 1U Taq DNA polymerase, and 10–50 ng of template DNA per the chloroplast genome, we ranked the eight intergenic spacers 25 ll reaction volume. Reaction products were purified using Wat- according to the proportion of parsimony-informative (PI) charac- son’s purification kit prior to sequencing. ters as a function of analyzed, aligned sequence length for each Double-stranded and purified PCR products were sequenced by gene (Table 4). Based on the PI characters/sequences length, we the dideoxy chain termination method with ABI PRISM Bigdye Ter- analyzed the following data partitions for all data sets: (1) rpl32- minator Cycle Sequencing Ready Reaction Kit with AmpliTaq DNA trnL intergenic spacer; (2) rpl32-trnL and trnT/L intergenic spacers; polymerase FS (Perkin-Elmer). Reactions and programs were cho- (3) rpl32-trnL, trnT/L, and rps16-trnQ intergenic spacers; (4) rpl32- sen according to the recommendations of the handbook, with trnL, trnT/L, rps16-trnQ, and rpoB-trnC intergenic spacers; (5) rpl32- slight modification in some cases. Samples were electrophoresed trnL, trnT/L, rps16-trnQ, rpoB-trnC, and trnD/T intergenic spacers; in an ABI 3700, ABI 3730 or ABI 3730xl automated sequencer. (6) rpl32-trnL, trnT/L, rps16-trnQ, rpoB-trnC, trnD/T, and atpI/H intergenic spacers; (7) rpl32-trnL, trnT/L, rps16-trnQ, rpoB-trnC, 2.5. Sequence alignment and phylogenetic analyses trnD/T, atpI/H, and psaA-ORF170 intergenic spacers; (8) all eight intergenic spacers. * Base determination was complete and unambiguous in all Maximum parsimony (MP) analysis was performed with PAUP cases. DNA sequences were edited with SeqMan (DNASTAR Pack- 4.0b10 (Swofford, 2003). Heuristic tree searches were conducted age), aligned by Clustal X (Jeanmougin et al., 1998), and adjusted with 1000 random addition sequence replicates and TBR branch

Table 4 Properties of data partitions used in this study and statistical characteristics resulting from MP analyses.

Partition Aligned length Variable characters Informative characters (percentage of total) Tree length CI RI RC Indels rpl32-trnL 966 101 60 (6.21%) 125 0.864 0.959 0.829 3 trnT/L 945 89 52 (5.50%) 118 0.780 0.924 0.721 11 rps16-trnQ 1805 162 86 (4.76%) 205 0.834 0.943 0.786 21 rpoB-trnC 1441 109 66 (4.58%) 142 0.817 0.934 0.763 11 trnD/T 1399 106 60 (4.29%) 131 0.855 0.946 0.809 9 atpI/H 999 66 38 (3.80%) 77 0.909 0.959 0.871 5 psaA-ORF170 983 50 35 (3.56%) 65 0.892 0.981 0.875 4 trnS/G 925 56 31 (3.35%) 70 0.800 0.944 0.755 8 Combined 9463 739 428 (4.52%) 1037 0.756 0.912 0.690 72 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 829

Table 5 Best-fitting models and parameter values for separate (trnS/G, trnD/T, rpoB-trnC, trnT/L, psaA-ORF170, rpl32-trnL, atpI/H and rps16-trnQ) and combined datasets in this study.

Region AIC selected model Base frequencies Substitution model (rate matrix) IG A C G T A-C A-G A-T C-G C-T G-T trnS/G TVM+G 0.3457 0.1841 0.1494 0.3209 0.4432 2.2574 0.1300 0.9519 2.2574 1.0000 0 0.4248 trnD/T TVM+I+G 0.3413 0.1757 0.1686 0.3144 0.3145 1.2887 0.1816 0.8289 1.2887 1.0000 0.5924 0.9755 rpoB-trnC TVM+I+G 0.3412 0.1577 0.1518 0.3490 0.7091 1.5560 0.0195 0.6449 1.5560 1.0000 0.5919 0.9620 trnT/L TVM+I+G 0.4014 0.0990 0.1572 0.3424 0.3789 0.8749 0.0986 0.0503 0.8749 1.0000 0.5341 0.9909 psaA-ORF170 K81uf+I 0.3015 0.1841 0.1482 0.3663 1.0000 2.2241 0.4646 0.4646 2.2241 1.0000 0.8356 Equal rpl32-trnL TVM+G 0.3634 0.1445 0.1326 0.3595 1.2315 2.0040 0.3587 1.2344 2.0040 1.0000 0 0.4894 atpI/H K81uf+G 0.3342 0.1386 0.1756 0.3516 1.0000 1.5722 0.3374 0.3374 1.5722 1.0000 0 0.3237 rps16-trnQ K81uf+I+G 0.3781 0.1475 0.1394 0.3350 1.0000 1.4265 0.3677 0.3677 1.4265 1.0000 0.5771 1.1345 Combined TVM+I+G 0.3522 0.1543 0.1522 0.3413 0.6177 1.2584 0.1828 0.5385 1.2584 1.0000 0.6538 1.0687

swapping, multrees option not in effect. All characters in the data highest percentage of informative characters with 6.21%, whereas matrix were unordered and equally weighted and gaps were trea- the trnS/G region has the lowest with only 3.35%. Furthermore, se- ted as missing data. All of the most parsimonious trees were sum- quences between nucleotide sites 376 and 441 of the rpl32-trnL marized into a strict consensus tree. Nodal support was evaluated intergenic spacer contained a pair of 24 bp exact inverted repeats, using the bootstrap method (Felsenstein, 1985) with 1000 repli- which are separated by 6 bases (Fig. 1). Five types of sequences can cates, TBR branch swapping, and multrees option not in effect, sav- be recognized. Type I and II sequences can be converted into each ing 1 tree for each replicate. Bootstrap proportions >70% are other by inversions of sequences bordered by the inverted repeats. considered well supported (Hillis and Bull, 1993). Tree statistics We examined the phylogeny of the individual or combined matrix including consistency index and the retention index were calcu- with or without the inversions, respectively, and the results were lated using PAUP. almost identical. Therefore, in subsequent analyses we retained Bayesian analyses were performed using the program MrBayes the inversions, and introduced gaps aligned with the uninverted version 3.1.2 (Ronquist and Huelsenbeck, 2003). The best-fitting sequences. Those gaps were not scored and treated as missing data. models were determined using the Akaike Information Criterion Each individual cpDNA region provided very low resolution (Posada and Buckley, 2004) as implemented in the program Mod- within the tribe, with most branches unresolved in the eight sepa- eltest 3.7 (Posada and Crandall, 1998). The best-fitting models and rate MP strict consensus trees (data not shown). However, the tem- parameter values are shown in Table 5. All priors were set accord- perate bamboos formed a monophyletic group in every analysis. ing to the chosen model. For example, Bayesian analyses of the Bayesian analyses were also performed for individual DNA re- combined data sets were run with default priors as follows: state- gions. The results (trees not shown) were similar to those obtained freqpr = fixed (0.3522, 0.1543, 0.1522, 0.3413); revmat = fixed using MP. In the Bayesian analyses, the tribe Arundinarieae re- (0.6177, 1.2584, 0.1828, 0.5385, 1.2584, 1.0000); shapepr = fixed ceived 1.00 PP support values for all eight regions. (1.0687); pinvar = fixed (0.6538). The Bayesian analyses started from random trees sampling one tree every 100th generation with 3.2. Analysis of combined data four incrementally heated chains. The Markov chain Monte Carlo (MCMC) algorithm was run for 50,00,000 generations for each data The final combined data matrix, after excluding ambiguously set. Stationarity of the Markov chain was ascertained by plotting aligned regions, consisted of 9463 characters for 161 accessions, likelihood values against number of generations for apparent sta- of which 999 derived from the atpI/H region, 983 from the psaA- tionarity. The first 5000 trees corresponding to the ‘‘burn-in” peri- ORF170 region, 966 from the rpl32-trnL region, 1441 derived from od were discarded, and the remaining trees were used to construct the rpoB-trnC region, 1805 from rps16-trnQ region, 1399 from the majority-rule consensus tree. We consider posterior probabilities trnD/T region, and 925 from the trnS/G, and 945 from the trnT/L re- >0.95 to indicate significant probability for a clade. gion, respectively. The matrix also included 72 binary coded indel Congruence of datasets was assessed using the incongruence characters (Table 4). length difference (ILD) test (Farris et al., 1994) implemented in Maximum parsimony analysis of the combined matrix pro- * PAUP ver. 4.0b10 (Swofford, 2003). We used 1000 homogeneity duced 986 most parsimonious trees of 1037 steps in length, with replicates. Heuristic search with initial trees obtained by random a consistency index of 0.756 and a retention index of 0.912. The addition, NNI (nearest-neighbor interchange) branch swapping, strict consensus of all the maximum parsimony trees is shown in and ten random addition sequences following Sjolin et al. (2005). Fig. 2. The majority-rule consensus of 45,000 trees derived from Uninformative characters were excluded prior to ILD tests as they Bayesian analyses with accompanying posterior probability values may overestimate the level of incongruence (Cunningham, 1997; is presented in Fig. 3. Lee, 2001). Different combinations of datasets were also analyzed. Comparisons among the eight individual chloroplast sequences using the ILD test indicated that none of them was significantly 3.3. Phylogenetic relationships incongruent (P = 0.086–0.675). The ILD value for the entire dataset was 0.675, suggesting congruence among the datasets. The eight The strict consensus MP tree agreed with the Bayesian tree at datasets were accordingly analyzed separately and concatenated. nearly all nodes; only minor differences appeared at some terminal nodes that were not well-supported by bootstrap values. Of the nodes discussed below, only Gaoligongshania megalothyrsa (IX) 3. Results and Yushania alpina (II) differed in placement in the MP phylogeny as compared with the BI tree. Unless otherwise stated, therefore, 3.1. Analysis of individual cpDNA regions discussions of the concatenated phylogeny are based on the topol- ogy of the MP strict consensus tree. A summary of the Maximum Parsimony statistics for each data Many well-supported clades were recovered in this study (e.g., partition is presented in Table 4. The rpl32-trnL sequence had the Clades I–X; Fig. 2). The monophyly of the temperate bamboo tribe 830 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839

Ty p e I

ACTTTTCATAATAGAATCCTCATA TTTTAT TATGAGGATTCTATTATGAAAAGT

Type II

ACTTTTCATAATAGAATCCTCATA ATAAAA TATGAGGATTCTATTATGAAAAGT

Type III

------TCATAATAGAATCCTCATA TTTTAT TATGAGGATTCTATTATGAAAAGT

Type IV

ACTTTTCATAATAGAATCCTCATA ATAAAAATAAAA TATGAGGATTCTATTATGAAAAGT

Type V

ACTTTTCATAATAGAA-----TCCTCATA TTTTAT TATGAGGATTCTATTCTATTATGAAAAAGT

Fig. 1. Five types of sequences of region between nucleotide sites 376 and 441 of rpl32-trnL intergenic spacer. Lines and arrows below sequences indicate inverted repeats. Nucleotides (boldface letters) that are bordered by inverted repeats which have undergone inversions.

Arundinarieae was strongly supported (100% BP; 1.00 PP). Ten ma- Bashania abietina was supported with 0.99 PP; Oligostachyum oedo- jor clades of temperate bamboos (Clades I–X; Fig. 2) were recov- gonatum was indicated to be sister to Pleioblastus sanmingensis ered congruently by both Bayesian and Maximum Parsimony (1.00 PP); Acidosasa notata was nested within a clade with two methods (Figs. 2 and 3). In addition to the six clades (I) Bergbam- accessions of Pleioblastus (0.99 PP); and a subclade with Gelidocal- boes, (II) the African alpine bamboos, (III) Chimonocalamus, (IV) the amus tessellatus and Gelidocalamus sp2. together with four acces- Shibataea clade, (V) the Phyllostachys clade and (VI) the Arundinaria sions of Indocalamus was supported with 0.96 PP. Within this clade corresponding to those obtained earlier in previous study subclade, Indocalamus guangdongensis and Indocalamus herklotsii (Triplett and Clark, 2010), another four clades were newly recog- formed a sister relationship with support of 95% BP and 1.00 PP. nized, and designated (VII) Thamnocalamus, (VIII) Indocalamus In addition, the Bayesian analysis recovered a robust but unre- wilsonii, (IX) Gaoligongshania, and (X) Indocalamus sinicus. solved subclade with 21 species representing Acidosasa, Brachys- (I) Bergbamboes, including only Thamnocalamus tessellatus, (II) tachyum, Indocalamus, Oligostachyum, Phyllostachys, Pleioblastus, the African alpine bamboos, consisting of Yushania alpina, (VIII) Pseudosasa, and Sasa (1.00 PP) (Fig. 3). Indocalamus wilsonii, (IX) Gaoligongshania, a monotypic genus, (VI) The Arundinaria clade. This clade received strong support and (X) Indocalamus sinicus formed distinct lineages, respectively. with 94% BP and 1.00 PP. Four subclades can be identified. These (III) Chimonocalamus. This clade was strongly supported (100% four subclades are designated as the Sasa subclade, the North BP; 1.00 PP), and comprised most of the sampled taxa of Chimono- American subclade, the Medake subclade and the Sinicae subclade. calamus, plus Ampelocalamus actinotrichus. However, Chimonocal- The Sasa subclade (99% BP; 1.00 PP) consisted of Sasa, Sasaella, and amus pallens (the type species of Chimonocalamus) and C. Pleioblastus. The North American subclade (0.99 PP), comprised fimbriatus were not part of this clade, instead nesting within the three accessions of the genus Arundinaria s.s. (N. America). Within Phyllostachys clade. this subclade Arundinaria appalachiana and Arundinaria tecta were (IV) The Shibataea clade. Shibataea, Pseudosasa gracilis, Gelidocal- resolved as sister species with support of 100% BP and 1.00 PP. amus, Indocalamus, Ferrocalamus, and Chinese accessions of Sasa The Medake subclade (100% BP; 1.00 PP) included three accessions subgen. Sasa clustered in this clade with strong support (99% BP; of Pleioblastus s.s. (Japan archipelago), within which were nested 1.00 PP). the Japanese taxa Pleioblastus sp. and Pseudosasa japonica (includ- (V) The Phyllostachys clade. A total of 69 species were encom- ing one variety). This subclade can be divided into two lineages: passed in this clade (1.00 PP), including representatives of Phyllo- one consisting of Pleioblastus gramineus, Pseudosasa japonica and stachys and other 16 genera. Despite the fact is that these species its variety (100% BP; 1.00 PP), and the other consisting of Pleiobla- represent all of the morphological diversity among three subtribes stus argenteostriatus, Pleioblastus sp., and Pleioblastus pygmaeus of Arundinarieae, little resolution or genetic divergence was appar- (99% BP; 1.00 PP). The Sinicae subclade received support of 99% ent. However, several clusters can be identified, including Pseudo- BP and 1.00 PP but was poorly resolved at the generic level. Acido- sasa guanxianensis and Bashania qingchengshanensis (100% BP; 1.00 sasa nanunica and two additional lineages formed a polytomy PP), Indocalamus decorus and Indocalamus longiauritus (89% BP; within this subclade. These two lineages were (1) Pseudosasa 1.00 PP), and a lineage consisting of Pseudosasa amabilis var. contex- orthotropa Pseudosasa cantorii (‘‘” signifies the inclusion of all a and Pseudosasa maculifera (85% BP; 1.00 PP). Bashania fargesii was species between the two on our trees) and (2) Oligostachyum pan- revealed to be more similar to four accessions of Fargesia than to iculatum Pseudosasa longiligula. Within the first lineage, one other accessions of Bashania (87% BP; 1.00 PP). Yushania crassicollis clade (95% BP; 1.00 PP) and two other clades with support of and Yushania baishanzuensis formed a monophyletic lineage with 1.00 PP for both were recovered, each of which was a heteroge- support of 87% BP and 1.00 PP. Drepanostachyum and Himalayacal- neous assemblage of taxa. amus formed a clade with support of 97% BP and 1.00 PP. Further- (VII) Thamnocalamus. The newly-recognized Thamnocalamus more, the following lineages were supported with high posterior clade was represented by five accessions of Thamnocalamus spath- probability values: a subclade consisted of Bashania fangiana and iflorus, including one variety (100% BP; 1.00 PP). Within this clade, C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 831

Fig. 2. Strict consensus of 986 equally most parsimonious trees based on the 8-region cpDNA dataset (rpl32-trnL, trnT/L, rps16-trnQ, rpoB-trnC, trnD/T, atpI/H, psaA-ORF170, trnS/G). Numbers above the branches indicate bootstrap (BP)/posterior probability (PP) support values. Clade and subclade names correspond to those described in previous studies or newly recognized in this study. Alternate accessions of Indocalamus sinicus, Indocalamus wilsonii, Indosasa gigantea, Thamnocalamus spathiflorus and its variety are indicated by voucher numbers. 832 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839

Fig. 3. Phylogram of the majority-rule consensus tree from the Bayesian analysis of the 8-region cpDNA dataset (rpl32-trnL, trnT/L, rps16-trnQ, rpoB-trnC, trnD/T, atpI/H, psaA- ORF170, trnS/G). Numbers above and beside the branches are posterior probability (PP) values. Clade and subclade names correspond to those described in previous studies or newly recognized in this study. Alternate accessions of Indocalamus sinicus, Indocalamus wilsonii, Indosasa gigantea, Thamnocalamus spathiflorus and its variety are indicated by voucher numbers. C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 833 individuals from Tibet and Kew Garden (introduced from Nepal) Life history may significantly affect the rate of molecular evolu- formed a pair with support of 96% BP and 1.00 PP. Likewise, all tion (Smith and Donoghue, 2008). The flowering cycles from dec- accessions of T. spathiflorus var. crassinodus formed a clade with ades to 120 years in temperate bamboos have been reported support of 84% BP. (Janzen, 1976). Long generation times for woody bamboos have been considered as another explanation for poor resolution of molecular phylogenetic tree (Gaut et al., 1997). The effect of flow- 4. Discussion ering cycles on evolutionary rates of temperate woody bamboos is needed to test in the future. The temperate bamboos have long been considered a complex and taxonomically difficult group. With the exception of strong sup- 4.1. The Thamnocalamus group and subtribes Arundinariinae and port for the monophyly of the temperate bamboos (tribe Arundinar- Shibataeinae ieae), phylogenetic relationships within the tribe have remained for the most part enigmatic. Peng et al. (2008) focused on the generic-le- Guo et al. (2002), Guo and Li (2004) concluded that the Thamn- vel phylogenetic relationships and indicated the temperate bam- ocalamus group and its allies were monophyletic based on the phy- boos as highly heterogeneous groups. Triplett and Clark (2010) logenetic analyses of the ITS and GBSSI sequences. With extensive presented a robust multi-locus chloroplast phylogeny of the temper- sampling, Peng et al. (2008) indicated that most species of the ate bamboos. It revealed that the temperate bamboos were resolved Thamnocalamus group and its allies displayed a close relationship to include six major lineages. But the study emphasizes the species- in both the analysis of GBSSI sequences and the combined analysis level relationships among taxa in Japan and North America, with of ITS and GBSSI data. However, the Thamnocalamus group was not relatively limited sampling of Chinese species. With a very broad resolved as monophyletic. In our study, the Thamnocalamus group sampling of Chinese taxa, ten major lineages within the tribe was indicated to be polyphyletic. However, it is interesting to note Arundinarieae were revealed in the current study (Figs. 2 and 3). that Thamnocalamus spathiflorus and its variety are strongly sup- The Chinese species was confirmed as a very complicated and con- ported as a distinct lineage (Fig. 2, clade VII), which is consistent fusing group in accordance with their diverse morphological charac- with previous results (Guo and Li, 2004). ters. Those species were clustered within eight of the total ten According to our results, neither of the two major subtribes of lineages. Some species had closer relationships with taxa from North the temperate bamboos (Arundinariinae nor Shibataeinae; Li, America and Japan than sympatric species (i.e., Arundinaria clade). 1997; Table 1) is indicated to be a monophyletic group. Taxa Species examined in this study assembled into lineages that belonging to these subtribes are distributed randomly among three were largely incongruent with current taxonomic circumscrip- major clades, i.e., the Shibataea clade, the Phyllostachys clade, and tions. Moreover, relationships among ten lineages were repre- the Arundinaria clade (Fig. 2). Unequivocal morphological synapo- sented in unresolved polytomies (Figs. 2 and 3). This poor morphies have not been identified to delimitate these molecular- resolution and the extremely short internal branches may be based clades. This implies that morphological characters used in attributed to low rate of molecular divergence and a complicated the traditional classification of subtribes, such as rhizomes, synflo- evolutionary history possibly involving rapid radiation and reticu- rescences, ovaries, branches and culm sheaths, may not have sig- late evolution. nificant phylogenetic meanings, and that the subdivision of Molecular dating of major clades of the grass family by Arundinarieae should be reevaluated. Bouchenak-Khelladi et al. (2009) indicated that the Arundinarieae was a relatively young group in the Bambusoideae, their origin 4.2. Genera within the tribe Arundinarieae dating back to the Miocene, approximately 15 mya. Within the temperate bamboos, in contrast to other clades, the relationships Peng et al. (2008) suggested that most genera within the tem- among species in the Phyllostachys clade and the Sinicae subclade perate bamboos were highly heterogeneous. This opinion was con- of the Arundinaria clade remain largely unresolved. These two firmed again by Triplett and Clark (2010). In their study, the monophyletic lineages mainly occur in China, encompass about monophyly of 19 temperate genera was tested. Of these, only four 60% of the diversity in the temperate bamboos, and span a wide were supported as monophyletic, while ten were indicated to be range of vegetative and reproductive morphology. This may imply paraphyletic or polypheletic, and the remaining genera were that two episodes of recent or rapid radiation happened indepen- ambiguous due to little resolution. In our study, we sampled 156 dently among the Chinese taxa. According to the phylogeny and accessions representing 26 genera. At least two species per genus distribution of the species in the Phyllostachys clade and the Sinicae were sampled, with the exception of monotypic genera Brachys- subclade of the Arundinaria clade (see above; Fig. 2), we infer that tachyum and Gaoligongshania, and the genera Himalayacalamus, lineage diversification is possibly correlated with the geological Metasasa, and Sasaella, for which only one sample was available. history of Chinese mainland and the adjacent areas, particularly Only four genera (Arundinaria, Drepanostachyum, Ferrocalamus, the uplift of Tibetan Plateau and the subsequent habitat change. and Shibataea) were found to be monophyletic (Figs. 2 and 3), Additional evidence combinated with fossils is needed to confirm although additional taxa should be sampled to fully test the mono- this inference. phyly of Drepanostachyum and Shibataea. Fifteen genera (Acidosasa, On the basis of the recent molecular phylogeny and previous Ampelocalamus, Bashania, Chimonocalamus, Fargesia, Gelidocalamus, morphological and molecular evidence, Triplett and Clark (2010) Indocalamus, Indosasa, Oligostachyum, Phyllostachys, Pleioblastus, suggested that reticulate evolution may be more significant in the Pseudosasa, Sasa, Sinobambusa, and Yushania) were confirmed to temperate bamboos than previously suspected. Furthermore, AFLP be paraphyletic or polyphyletic. The genera Chimonobambusa and data considered in conjunction with morphology and cpDNA haplo- Thamnocalamus were represented in unresolved polytomies (Figs. types provide unequivocal evidence of hybridization among species 2 and 3). of Arundinaria in North America (Triplett, 2008; Triplett et al., 2010) as well as among several genera in East Asia (Triplett, 2008; Triplett 4.3. Major lineages within the tribe Arundinarieae and Clark, 2010). Potential events of the allopolyploidy of the tem- perate bamboos are being analyzed by Zhang et al. (in preparation) Several major lineages within the tribe Arundinarieae are based on the incongruence between separate phylogenies derived strongly supported (Figs. 2 and 3). This study confirms the six from chloroplast and nuclear low-copy gene sequences. clades found earlier (Triplett and Clark, 2010) and delineates 834 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 several other distinct lineages for the first time. While the branch- tivation in the US was found to belong to the Chimonocalamus ing order among these ten major lineages was still unresolved, a clade. possible association was indicated among the African alpine bam- Another surprising result is the position of Ampelocalamus boos, Gaoligongshania megalothyrsa, and the Chimonocalamus clade actinotrichus, the type species of Ampelocalamus, which was re- in the BI tree with support of 1.00 PP (Fig. 3). Moreover, the possi- vealed to be part of the Chimonocalamus clade. Ampelocalamus is ble sister relationships between Indocalamus sinicus and the Phyllo- distinguished by pendulous or scrambling culms, geniculate stachys clade received support of 0.96 PP. branches with a dominant central branch, and semelauctant synfl- We also conducted an approach according to the criteria used orescences clustering as pendulous panicles on leafy or leafless by Triplett and Clark (2010) to select taxa for the 12-region analy- flowering branches. Ampelocalamus was indicated to be a good sis. MP and BI trees (not show) were consistent with the above re- genus in previous studies using nuclear regions (Guo et al., 2002; sults. The inter-relationships of the ten major clades were still Guo and Li, 2004). However, in this study the genus is indicated unclear. To reconstruct the phylogeny of the temperate bamboos to be polyphyletic (Figs. 2 and 3), although the same three species and reveal the deep-level relationships, work is currently under- were used to reconstruct the phylogeny. This may reflect different way to add other more useful characters such as a single-copy nu- evolutionary histories between nuclear and chloroplast genomes. clear gene and whole chloroplast genome (Zhang et al., in In order to confirm and further test our results, additional taxa of preparation). Chimonocalamus and Ampelocalamus (possibly including popula- tion sampling) should be collected and additional molecular mark- ers should be explored. 4.3.1. Bergbamboes (I) and African alpine bamboos (II) Thamnocalamus tessellatus is from South Africa, where it is 4.3.3. Shibataea clade (IV) called Bergbamboes, or mountain bamboo. This species was tenta- The Shibataea clade was considered one of the more surprising tively placed in Thamnocalamus by Soderstrom and Ellis (1982) on results by Triplett and Clark (2010) based on the multi-locus chlo- the basis of morphological features, such as pachymorph rhizomes roplast phylogeny. The reason is the position of Shibataea from East and a condensed synflorescence, while the terminal synflorescenc- Asia and its relationship with morphologically distinctive and geo- es, with a series of spadix-shaped bracts, are more similar to Farge- graphically endemic species from Southeastern China. With wider sia. The position of Thamnocalamus tessellatus was unresolved due taxon sampling, the current molecular results confirm the Shiba- to limited information by Guo and Li (2004) using nuclear ITS and taea clade, currently represented as a polytomy of four monophy- GBSSI. In combined 4-region and 12-region analyses of cpDNA se- letic lineages: (1) Ferrocalamus, (2) Shibataea and Gelidocalamus quences, bergbamboes was revealed to be a divergent lineage tessellatus, (3) Indocalamus and Gelidocalamus sp1., and (4) Sasa, (Triplett and Clark, 2010) and was suggested to be described as a Pseudosasa gracilis, and Indocalamus jinpingensis. As such, this clade new genus. The current molecular results are consistent with pre- represents not only species from Southeastern China, but also spe- vious study. Relationships with other lineages are still unclear. cies from Southwestern China. The monophyly of both Ferrocal- Yushania alpina is from central and east Africa. This species was amus and Shibataea are supported, while the genera Pseudosasa, indicated to have a possible affiliation with Chimonocalamus or Sasa, Indocalamus, and Gelidocalamus are indicated to be polyphy- Gaoligongshania, but with low support values by Guo and Li letic. On the basis of morphology, Shibataea is considered to be (2004) examining in GBSSI or an combined analysis of ITS and close to Phyllostachys. Nonetheless, current data provide that Gelid- GBSSI, respectively. Triplett and Clark (2010) retested the place- ocalamus tessellastus was the close relative of Shibataea. Thus ment of this species, the results indicated that Yushania alpina clus- accessions from two subtribes, Arundinariieae (semelauctant, lep- tered with Chimonocalamus in the 12-region analysis. In our study, tomorph species) and Shibataeinae (iterauctant, leptomorph spe- Yushania alpina was found to cluster with Gaoligongshania and the cies), were nested within this one clade. Although a relationship Chimonocalamus clade and the lineage was supported with support between the Arundinaria clade and this clade was weakly sup- of 1.00 PP. Additional work is required to confirm the phylogenetic ported in this study, more work is necessary to ascertain its posi- affinities among the African bamboos. tion among lineages. Additional work is currently underway (1) to retest the monophyly of Shibataea by increasing taxon sampling, 4.3.2. Chimonocalamus (III) and (2) to collect all Chinese species of Sasa subgen. Sasa in order to Chimonocalamus was established by Hsueh and Yi (1979) and test the monophyly of this group and to discuss the necessity of encompasses 11 species distributed in subtropical to warm tem- describing a new genus, as suggested by Triplett and Clark (2010). perate regions of Southernwest China, the eastern Himalayas, and northern Myanmar. This genus is characterized by pachy- 4.3.4. Phyllostachys clade (V) morph rhizomes, three branches per node, a ring of root thorns After adding more representatives of Chinese taxa, the Phyllo- especially dense at lower nodes, and semelauctant synflorescences stachys clade now comprises about 17 genera and 69 species. This with three stamens. In general it has been recognized as a good clade unites species from different genera and subtribes (Figs. 2 genus (Soderstrom and Ellis, 1987; Dransfield and Widjaja, 1995; and 3; Table 1) with the most diversified morphological features, Keng and Wang, 1996; Li, 1997; Li et al., 2006), except that it such as iterauctant or semelauctant synflorescences, pachymorph was treated as a section of Sinarundinaria (a synonym of Fargesia) or leptomorph rhizomes, various numbers of branches (1, 2 , 3 or by Chao and Renvoize (1989). The monophyly of Chimonocalamus more), and three or six stamens. However, the more surprising as- was strongly supported by Guo et al. (2002) and Guo and Li pect is the low variation of chloroplast DNA regions, which does (2004) based on analyses of GBSSI and ITS sequences. Triplett not correspond to the high morphological heterogeneity in this and Clark (2010) reinforced this result. In our study, we sampled clade. five species of Chimonocalamus, and the results indicate that Although none of the 17 genera were resolved to be monophy- Chimonocalamus is polyphyletic. Chimonocalamus longiusculus, letic, several subclades with moderate to strong support were ob- C. dumosus, and C. montanus were indicated to be a distinct lineage, tained. Pseudosasa guanxianensis and Bashania qingchengshanensis consistent with the Chimonocalamus clade (Triplett and Clark, constitute a small clade with support of 100% BP and 1.00 PP. These 2010). Surprisingly, C. fimbriatus and C. pallens were nested within two shrubby bamboos are both distributed in Dujiangyan or the the Phyllostachys clade. This placement contrasts the results of adjacent area, Sichuan, and share some morphological characters, Triplett and Clark (2010), in which an accession of C. pallens in cul- such as sheath rings with culm sheath remnants and black setae, C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 835 branches 3 to many, and culm sheaths persistent or tardily decid- acters (especially the swollen nodes) a new genus Clavinodum was uous with black setae. Evidence from molecular data, geography proposed with this species as type (Wen, 1984). Finally, this spe- and morphology imply that these two species are closely related, cies was transferred to Oligostachyum (Liang et al., 1987) and and may originate from the same progenitors. In the light of this adopted by Flora Reipublicae Popularis Sinicae (Keng and Wang, result, the taxonomic revision of these species may be needed. 1996) and Flora of China (Li et al., 2006). Acidosasa notata is the only Pseudosasa guanxianensis was treated as a synonym of Indocalamus representative of Acidosasa nested in the Phyllostachys clade in this longiauritus in the Flora of China (Li et al., 2006), however, there are study. It is also notable that Ps. amabilis var. convexa and Ps. ama- several differences to distinguish them, such as the number of bilis (the Arundinaria clade) do not occur in the same major clades. branches (3 to many for the former, and 1–3 for the later). Indocal- We speculate that (1) we made a misidentification of Ps. amabilis amus longiauritus has pale red-brown tomentose rings below the var. convexa, or (2) morphological homoplasy led taxonomists to nodes, while B. qingchengshanensis is characterized by gray-yellow consider these plants as conspecifics. Pseudosasa amabilis var. con- waxy powder below the nodes, and flowers with three stamens vexa was collected around its type locality; therefore, we consider and two stigmas. Many of the features found in B. qingchengshan- the second inference to be more likely. ensis resemble Gelidocalamus. Further study will be necessary to Many of the subclades discussed above have geographic impli- understand morphology in light of our molecular results. cations. Pseudosasa guanxianensis, Bashania qingchengshanensis, A subclade of two accessions of Drepanostachyum and Himalay- four accessions of Fargesia, and B. fargesii are distributed at the acalamus falconeri was supported with 97% BP and 1.00 PP. Drep- eastern edge of Tibetan Plateau; two accessions of Drepanostach- anostachyum is medium size clumping, mountain bamboos with yum and Himalayacalamus falconeri occur in the Himalayan moun- a lax falcate synflorescence and occurs in habitats similar to those tains; and the subclade of 21 species are mainly from eastern of Ampelocalamus, which is regarded as a related genus (Li et al., China. Although they bear different morphology, the same plastid 1996) on the basis of morphology. Their sister relationship was genome donors may be shared among these taxa in the process generally supported by Guo et al. (2002) and Guo and Li (2004) of speciation. using the ITS and GBSSI sequences. Himalayacalamus is clump- Many problems remain in the Phyllostachys clade. In order to forming bamboos of the lower altitudes of the Himalayan Moun- enhance our understanding of its phylogeny, more species from tains. The morphological differences between Drepanostachyum Ampelocalamus, Drepanostachyum, Fargesia, Phyllostachys, Yushania and Himalayacalamus are subtle; for example, Drepanostachyum and so on should be sampled, and the development of markers has many equal branches per node, while Himalayacalamus has with improved resolution at and below the genus level, such as many subequal branches and a central dominant one. A close rela- low-copy nuclear regions or the whole chloroplast genome, is tionship between Drepanostachyum and Himalayacalamus was required. indicated by Triplett and Clark (2010). Here, we retrieved two accessions of Drepanostachyum as a monophyletic group, and 4.3.5. Arundinaria clade (VI) Himalayacalamus was recovered to be sister to Drepanostachyum Within the Arundinaria clade, four major subclades were re- rather than Ampelocalamus. However, the monophyly of Drepanos- solved, which we refer to as the Sasa subclade, the North American tachyum was not supported by Triplett and Clark (2010) based on subclade, the Medake subclade, and the Sinicae subclade. The for- different representative species. These results suggest that exten- mer two subclades were indicated as the Sasa/cane subclade with sive taxon sampling is essential in future studies. moderate support in the previous study, while the latter two cor- A lineage consisting of four accessions of Fargesia plus Bashania respond to the previous results (Triplett and Clark, 2010). There fargesii was recovered with moderate support (87% BP; 1.00 PP). are at least two possible explanations for this soft incongruence: Those taxa are distributed in western Sichuan, western Hubei or (1) different taxa were sampled; (2) the molecular characters sup- Qingling Mountains. Within this lineage, F. nitida was indicated porting the Sasa/cane clade were not present in the cpDNA regions to be a close relative of F. robusta with support of 95% BP and sampled in the current study. Although the Arundinaria clade is 1.00 PP, whereas the previous study (Triplett and Clark, 2010) indi- united by rhizome type (leptomorph), it exhibits high morpholog- cated that F. nitida was sister to Thamnocalamus spathiflorus with ical diversification, for example iterauctant or semelauctant synfl- support of 94% BP and 1.00 PP. This conflict may be caused by a orescences, various numbers of branches (1 or 3), and three or six misidentification of the plants in cultivation in the U.S. as T. spath- stamens. iflorus (see below for details of Thamnocalamus (VII)). The delimitation of the genus Arundinaria based on morphology A surprising subclade consisting of 21 species was recovered has been disputed for a long time. In the broad sense, Arundinaria is only in the BI tree (Fig. 3). This subclade includes at least two gen- usually treated to include some narrowly defined East Asian gen- era from Shibataeinae (Brachystachyum and Phyllostachys), six from era, such as Bashania, Oligostachyum, Pleioblastus, and Pseudosasa, the Arundinariinae (Acidosasa, Indocalamus, Oligostachyum, Pleiob- as well as the North American species (Chao et al., 1980; Clayton lastus, Pseudosasa, and Sasa). A corresponding subclade was also and Renvoize, 1986; Soderstrom and Ellis, 1987; Watson and indicated with support of 71% MPBS (maximum parsimony boot- Dallwitz, 1992 onwards; Yang and Zhao, 1993, 1994). Other strap analysis), 71% MLBS (maximum likelihood bootstrap analy- authors treat the genus as endemic to eastern North America, with sis) and 1.00 PP by Triplett and Clark (2010), although with the East Asian genera as its closest relatives (Suzuki, 1978; Li, different representatives and fewer samples. Within this subclade, 1997; Stapleton, 1997; Judziewicz et al., 1999; Li et al., 2006). In close relationships of four groups, i.e., O. oedogonatum + Pl. san- this study, the three species of Arundinaria s.s. formed a moder- mingensis, A. notata + Pl. wuyishanensis + Pl. solidus, Ps. amabilis ately supported lineage, while its relationship with the other three var. convexa + Ps. maculifera, and I. latifolius + I. victorialis were re- subclades was unresolved. Oligostachyum, Pleioblastus, and Pseudo- vealed. Most of these species occur in eastern China (Fujian and sasa were revealed to be close to Arundinaria s.s., while Bashania Zhejiang) except for Ps. maculifera in Henan. Oligostachyum oedo- was nested in Phyllostachys clade. Acidosasa, Indosasa, Metasasa, gonatum is the only species of Oligostachyum nested in Phyllosta- Sasa s.s. (Japan), Sasaella, and Sinobambusa were clustered within chys clade to date, although one accession tentatively identified the same clade. Therefore, our results conflict significantly with as Oligostachyum sp. (and cultivated in the US as Pleioblastus oleo- existing classifications. We suggest that Arundinaria should be sus) nested in this clade in Triplett and Clark (2010). Oligostachyum treated in the strict sense to consist of three species in eastern oedogonatum was first described as Pleioblastus oedogonatus (Wang North America. Triplett and Clark (2010) tentatively suggested that and Ye, 1981), and then due to its distinctive morphological char- Arundinaria s.s. may in fact be closer to Sasa or Sasamorpha. But this 836 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 relationship did not receive robust support in this study. Further study, T. spathiflorus and its variety, collected from Tibet and Nepal, work will be needed to clarify the phylogeny of the East Asian spe- formed a distinct lineage with support of 96%BP and 1.00 PP. This cies. However, whatever the relationships are between the North result is consistent with those based on ITS and GBSSI sequences American subclade and the other three subclades, it is clear that by Guo and Li (2004), but conflicts with Triplett and Clark there is a disjunction between temperate bamboo species in East- (2010), in which T. spathiflorus was found to be sister to Fargesia ern Asia and North America. nitida, and those two taxa were nested within the Phyllostachys The subclades Medake and Sinicae were monophyletic lineages clade (Figs. 2 and 3). This study did not find an association between that correlated with geography, i.e., the Medake subclade in Japan T. spathiflorus and T. tessellatus, instead suggesting that these spe- and the Sinicae subclade in China. This is consistent with previous cies are part of divergent lineages. results (Triplett and Clark, 2010). Japanese species of Pleioblastus and Pseudosasa were clustered in the subclade Medake in this 4.3.7. Indocalamus wilsonii (VIII), Gaoligongshania (IX) and Indocal study as well as the previous study by Triplett and Clark (2010). amus sinicus (X) The Sinicae subclade is composed of Acidosasa, Indosasa, Metasasa, Indocalamus Nakai was established by Nakai (1925) and is char- Oligostachyum, one species of Pleioblastus (from China), Pseudosasa acterized by semelauctant synflorescences, leptomorph rhizomes, subg. Sinicae, and Sinobambusa. Those species are mainly distrib- solitary mid-culm branches, and large leaves. It encompasses at uted in central, southern and southeastern China. Their vegetative least 23 species. Most species are endemic in China except one spe- characters are so similar that it is quite difficult to distinguish them cies distributed in Japan. This genus is especially prevalent in hills based on vegetative organs or tissues for some species, and some below 1000 meters, with a few species reaching 3000 meters. The misidentifications have been made, such as Acidosasa purpurea, current analysis demonstrates that Indocalamus is a heterogeneous, which was originally described as a member of Indosasa (Hsueh polyphyletic genus, with representatives in the Shibataea clade and and Yi, 1983), then was transferred to Acidosasa due to the semel- the Phyllostachys clade, as well as forming distinct lineages outside auctant synflorescences (Keng, 1986). Pleioblastus maculatus is the of each of the previously recognized clades. Indocalamus sinicus only Chinese representative of Pleioblastus nested in the Arundina- (Hance) Nakai is the type species of this genus, and accessions from ria clade in this study, and it has the broadest distribution among Hainan and Guangdong clustered together as a distinctive lineage. the Chinese species of Pleioblastus, from southwestern to south- It is surprising that Indocalamus wilsonii was also indicated to be a eastern China. The other Chinese species of Pleioblastus were clus- divergent lineage in this investigation. In our study, two accessions tered in the Phyllostachys clade, and are primarily distributed in of I. wilsonii from the provinces Sichuan and Hubei, respectively, eastern China. Representatives of Pseudosasa subgen. Sinicae were formed a lineage that was distinct from all other representatives of separated into three divergent clades, i.e., Shibataea clade, Phyllo- Indocalamus. Indocalamus wilsonii is distinct from other congeneric stachys clade and Arundinaria clade, with most in the Arundinaria taxa by its distribution at altitudes from 1700 to 3000 meters, and clade. Those results imply that we should reevaluate the Chinese usually with two stigmas in the florets (versus three in other species and Japanese Pleioblastus and Pseudosasa. Two species of Sinobam- of Indocalamus). busa were included in this study, one of which is the type species Gaoligongshania is a recently recognized monotypic genus dis- (S. tootsik). These were nested in the Sinicae subclade, a result that tributed in northwest Yunnan (Li et al., 1995). The type species, G. is consistent with Peng et al. (2008), while one unnamed species of megalothyrsa, bears semelauctant synflorescences, pachymorph rhi- this genus clustered in the Phyllostachys clade in the study by Trip- zomes, and solitary mid-culm branches, and is distinguished from all lett and Clark (2010), and S. rubroligula was close to Chimonobam- other temperate genera by its epiphytic habitat. Guo et al. (2002) and busa sichuanensis (a synonym of Menstruocalamus sichuanensis)in Guo and Li (2004) found that Gaoligongshania was unexpectedly Peng et al. (2008). Therefore, more species should be sampled to resolved as sister to Thamnocalamus spathiflorus in the ITS-based ascertain the position of Sinobambusa. analysis, but sister to the African Yushania alpina with low support Nakai (1931) established the genus Sasamorpha as distinct from in GBSSI and combined analyses. In a multi-locus analysis of the Sasa, according to the characters of flat nodes (versus prominent in grass family that included a small subset of temperate species, Sasa), culm sheaths shorter than internodes (versus longer in Sasa), Gaoligonshania was resolved to be sister to Chimonobambusa and auricles and fimbriae undeveloped (versus developed in Sasa). (Bouchenak-Khelladi et al., 2008), suggesting this species may be- Hu (1985) proposed that Sasamorpha should not recognize as a dis- long in the Phyllostachys clade. However, in our analysis, Gaoligongsh- tinct genus, and divided Sasa into Sasa subgen. Sasa and Sasa sub- ania was clearly revealed to be a distinct lineage. Additional gen. Sasamorpha. The subgenera Sasa and Sasamorpha were molecular and morphological studies will be necessary to confirm adopted by FRPS and grass volume of the Flora of China. Based on the positions of Indocalamus sinicus, I. wilsonii, and Gaoligongshania. molecular data, S. qingyuanensis and S. sinica (representatives of Sasa subgen. Sasamorpha), were nested within the Phyllostachys 4.4. Phylogenetic utility of non-coding chloroplast DNA in the tribe clade, three Chinese taxa of Sasa subgen. Sasa (i.e., S. longiligulata, Arundinarieae S. magnonoda, and S. guangxiensis) were embedded within the Shibataea clade, Sasa oblongula cultivated in Guangzhou (but origi- As phylogenetic studies move to focus on lower-level taxo- nal locality unknown) was indicated to be sister to the Japanese nomic groups, it has become apparent that a multi-locus approach taxa of Sasa, Sasaella, and Pleioblastus in the Sasa subclade. These is necessary to obtain a sufficient number of phylogenetically observations indicate that both the monophyly of Sasa and the informative characters, especially when using the relatively slowly classification of subgenera based on morphology are rejected by evolving chloroplast genome. Many recent investigations have molecular results. The placement of Chinese Sasa species should used the combined analysis of several non-coding cpDNA regions be fully examined in future studies. to obtain sufficient characters for phylogenetic resolution (e.g., Perret et al., 2003; Rouhan et al., 2004; Shaw and Small, 2004; Bar- 4.3.6. Thamnocalamus (VII) fuss et al., 2005; Ickert-Bond and Wen, 2006; Smedmark et al., Thamnocalamus is distinguished by pachmorph rhizomes, ini- 2006; Fischer et al., 2007; Bellusci et al., 2008; Egan and Crandall, tially 5 branches in the mid-culm, semelauctant synflorescences 2008; Panero and Funk, 2008; Yuan and Olmstead, 2008; Rex et al., with spathelike bracts, three stamens. There are about 4 species 2009). At lower taxonomic levels, some non-coding cpDNA regions distributed in Bhutan, Northeast India, Nepal, South Tibet, South show sufficient variation for phylogenetic resolution while others Africa and Madagascar (Ohrnberger, 1999; Li et al., 2006). In our do not (e.g., Small et al., 1998; Ohsako and Ohnishi, 2000; Shaw C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 837 and Small, 2004; Shaw et al., 2005; Hughes et al., 2006; Rex et al., et al., 2003; Heath et al., 2008), as it enables a better detection of 2009). The eight chloroplast regions chosen for this phylogenetic multiple substitutions at the same nucleotide site. This helps coun- investigation revealed dissimilar characteristics with respect to teract branch-attraction effects and therefore improves phyloge- the amount of variability present within a particular region (Table netic inference (Hillis, 1996; Graybeal, 1998; Heath et al., 2008). 4). The rpl32-trnL intergenic spacer, with 6.21% parsimony-infor- Some empirical studies have also found that data combination mative characters, showed the highest phylogenetic utility at (i.e., multi-locus approaches) of multiple sequences from the same resolving relationships in the temperate bamboos clade, followed taxon does improve accuracy of phylogenetic inference (e.g., Qiu by the trnT/L (5.50%) and rps16-trnQ (4.76%) regions. et al., 1999; Soltis et al., 1999; Bapteste et al., 2002). In our study, To further assess how many chloroplast intergenic spacers are we conducted a very broad survey of representatives in the tribe needed to resolve relationships within the tribe Arundinarieae, Arundinarieae and emphasized on the Chinese taxa because it ac- we explored the relationships between the number of intergenic counts for the most morphological diverse and most complex spacers and the bootstrap and posterior probability values of the group of this tribe. Compared to the previous studies (Guo et al., gene trees that support the topology or clades shown in Tables 6 2002; Guo and Li, 2004; Bouchenak-Khelladi et al., 2008; Peng and 7. Although additional chloroplast markers were used, and et al., 2008; Triplett and Clark, 2010), there are several conflicts higher resolution and support especially at some terminal levels about the placements of some taxa. One apparent example is the could be obtained, the results demonstrated that regardless of Thamnocalamus group and its allies. Because of limited taxa these methods used, at least four intergenic spacers are needed to obtain bamboos were resolved as a monophyletic group (Guo et al., the identical gene trees of combined matrix and to recover the ma- 2002; Guo and Li, 2004), however, with more species, Peng et al. jor lineages with support of P70% BS and P0.95 PP. These four (2008), Triplett and Clark (2010) and our study illustrated that intergenic spacers, i.e., rpl32-trnL, trnT/L, rps16-trnQ, and rpoB- the Thamnocalamus group and its allies are paraphyletic. This indi- trnC, can be recommended to resolve major relationships within cates that we need to fully test the positions of putative subtribes, the tribe Arundinarieae with reliable support, and can be utilized genera, and species complexes within a broad phylogenetic con- in future studies to test the position of additional species within text in order to obtain an accurate understanding of relationships. the current phylogenetic framework. Moreover, Shibataea chinensis was supported as sister to Thamno- calamus spathiflorus according to Bouchenak-Khelladi et al. 4.5. Effect of taxon sampling in the tribe Arundinarieae (2008), however, Peng et al. (2008) and our results showed that these two species were embedded within two different clades. Increasing taxon sampling might reduce misleading effects or These types of observations indicate that the true phylogeny might systematic bias (Wiens, 1998; Zwickl and Hillis, 2002; Hillis not be reflected if only one or two representatives of a genus were

Table 6 Bayesian posterior probability for key clades under different data combinations.

12345678 Posterior probability values Chimonocalamus (III) 0.89 0.97 1.00 1.00 1.00 1.00 1.00 1.00 Shibataea clade (IV) 1.00 0.93 1.00 1.00 1.00 1.00 1.00 1.00 Phyllostachys clade (V) 0.55 0.29 1.00 1.00 1.00 1.00 1.00 1.00 Arundinaria clade (VI) NP 0.87 0.98 1.00 1.00 1.00 1.00 1.00 VI: North American subclade NP NP 0.94 0.99 0.99 0.99 0.99 0.99 VI: Sasa subclade 1.00 0.99 1.00 1.00 1.00 1.00 1.00 1.00 VI: Medake subclade NP NP 1.00 1.00 1.00 1.00 1.00 1.00 VI: Sinicae subclade NP NP NP 1.00 1.00 1.00 1.00 1.00 VI: Medake subclade sister to Sinicae subclade NP NP 0.99 1.00 1.00 1.00 1.00 1.00 Thamnocalamus (VII) 0.85 0.99 1.00 1.00 1.00 1.00 1.00 1.00 Indocalamus wilsonii (VIII) 1.00 0.99 1.00 1.00 1.00 1.00 1.00 1.00 Bergbamboes (I) sister to Indocalamus wilsonii (VIII) 0.89 0.86 0.99 1.00 0.94 0.88 0.91 0.86 Gaoligongshania (IX) sister to clade II + III NP NP NP 0.87 1.00 1.00 1.00 1.00 Indocalamus sinicus (X) 1.00 0.99 1.00 1.00 1.00 1.00 1.00 1.00

NP, clade not present in tree. ‘‘1–8” means the analysis of different combined dataset according to the criterion mentioned in Section 2.5.

Table 7 Parsimony bootstrap support for key clades under different data combinations.

12345678 Bootstrap values Chimonocalamus (III) NP 57 95 96 98 100 100 100 Shibataea clade (IV) 67 52 77 91 92 94 99 99 Phyllostachys clade (V) NP NP 59 57 52 61 65 68 Arundinaria clade (VI) NP NP NP 71 88 87 92 94 VI: North American subclade NP NP NP 57 65 62 62 60 VI: Sasa subclade 76 93 95 91 93 92 96 99 VI: Medake subclade NP 57 81 98 100 99 100 100 VI: Sinicae subclade NP NP NP 85 92 92 99 99 VI: Medake subclade sister to Sinicae subclade NP NP NP 70 75 73 90 95 Thamnocalamus (VII) NP 64 92 97 100 100 100 100 Indocalamus wilsonii (VIII) 96 99 100 100 100 100 100 100 Bergbamboes (I) sister to Indocalamus wilsonii (VIII) NP NP NP NP NP NP NP NP Gaoligongshania (IX) sister to clade II + III NP NP NP NP NP NP 51 NP Indocalamus sinicus (X) 87 94 100 100 100 100 100 100

NP, clade not present in tree. ‘‘1–8” means the analysis of different combined dataset according to the criterion mentioned in Section 2.5. 838 C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 used to study the generic relationships within the temperate bam- Dransfield, S., Widjaja, E.A., 1995. Plant Resources of South-East Asia, No. 7, boos when a genus has not been established to be monophyletic. Bamboos. Backhuys Publishers, Leiden. Egan, A.N., Crandall, K.A., 2008. Incorporating gaps as phylogenetic characters Extensive taxon sampling is essential for future studies. across eight DNA regions: ramifications for North American Psoraleeae (Leguminosae). Mol. Phylogenet. Evol. 46, 532–546. Farris, J.S., Källersjö, M., Kluge, A.G., Bult, C., 1994. Testing significance of Acknowledgments incongruence. Cladistics 10, 315–319. Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the This study was supported by the National Basic Research Pro- bootstrap. Evolution 39, 783–791. Fischer, G.A., Gravendeel, B., Sieder, A., Andriantiana, J., Heiselmayer, P., Cribb, P.J., gram of China (973 Program, Grant No.: 2007CB411601) and the Smidt, E. de C., Samuel, R., Kiehn, M., 2007. Evolution of resupination in National Natural Science Foundation of China (Grant No: Malagasy species of Bulbophyllum (Orchidaceae). Mol. Phylogenet. Evol. 45, 30770154), also partly supported by the Knowledge Innovation 358–376. Gaut, B.S., Clark, L.G., Wendel, J.F., Muse, S.V., 1997. Comparisons of the molecular Program of the Chinese Academy of Sciences (Grant No.: KSCX2- evolutionary process at rbcL and ndhF in the grass family (Poaceae). Mol. Biol. YW-N-029), and grants from the National Geographic Society Evol. 14, 769–777. (U.S.A.) (7336-02 to Lynn G. Clark and De-Zhu Li) and the National Grass Phylogeny Working Group (GPWG), 2001. Phylogeny and subfamilial classification of the grasses (Poaceae). Ann. Mo. Bot. Gard. 88, 373–457. Science Foundation (U.S.A.) (DEB-0515712 to Lynn G. Clark). We Graybeal, A., 1998. Is it better to add taxa or characters to a difficult phylogenetic are indebted to Professor Yu-Long Ding of Nanjing Forestry Univer- problem? Syst. Biol. 47, 9–17. sity; to Professor Nian-He Xia, Dr. Yun-Fei Deng, and Mr. Ru-Shun Guo, Z.H., Chen, Y.Y., Li, D.Z., Yang, J.B., 2001. Genetic variation and evolution of the alpine bamboos (Poaceae: Bambusoideae) using DNA sequence data. J. Plant Lin of South China Botanical Garden, Chinese Academy of Sciences Res. 114, 315–322. (CAS); to Professor Tong-Pei Yi and Mr. Lin Yang of Sichuan Agri- Guo, Z.H., Chen, Y.Y., Li, D.Z., 2002. Phylogenetic studies on the Thamnocalamus cultural University; and to Professors Shui-Sheng You and Shi-Pin group and its allies (Gramineae: Bambusoideae) based on ITS sequence data. Mol. Phylogenet. Evol. 22, 20–30. Chen of Fujian Agriculture and Forestry University for their essen- Guo, Z.H., Li, D.Z., 2004. Phylogenetics of the Thamnocalamus group and its allies tial and generous help during field work. We would also like to (Gramineae: Bambusoideae): inference from the sequences of GBSSI gene and thank Drs. Zhen-Hua Guo, Han-Qi Yang, Lian-Ming Gao, Jin-Mei ITS spacer. Mol. Phylogenet. Evol. 30, 1–12. Lu, Hong-Tao Li and other colleagues at the Key Laboratory of Bio- Heath, T.A., Hedtke, S.M., Hillis, D.M., 2008. Taxon sampling and the accuracy of phylogenetic analyses. J. Syst. Evol. 46, 239–257. diversity and Biogeography of the Kunming Institute of Botany, Hillis, D.M., 1996. Inferring complex phylogenies. Nature 383, 130–131. CAS for much practical and theoretical help throughout the course Hillis, D.M., Bull, J.J., 1993. An empirical test of bootstrapping as a method for of this research. We also specially thank Dr. Lei Xie at the College of assessing confidence in phylogenetic analysis. Syst. Biol. 42, 182–192. Hillis, D.M., Pollock, D.D., McGuire, J.A., Zwickl, D.J., 2003. Is sparse taxon sampling a Biological Sciences and Biotechnology, Beijing Forestry University problem for phylogenetic inference? Syst. Biol. 52, 124–126. for his valuable suggestions for revising the manuscript. Hsueh, C.J., Yi, T.P., 1979. Two new genera of Bambusoideae from S.W. China, 1: Chimonocalamus Hsueh et Yi. Acta Bot. Yunnanica 1, 74–92. Hsueh, C.J., Yi, T.P., 1983. Two new species of bamboos from China. Acta Phytotax. References Sinica 21, 94–99. Hu, C.H., 1985. A revision of the genus Sasa from China. Bamboo Res. 2, 56–63. Hughes, C.E., Eastwood, R.J., Bailey, C.D., 2006. From famine to feast? Selecting Bapteste, E., Brinkmann, H., Lee, J.A., Moore, D.V., Sensen, C.W., Gordon, P., Durufle, nuclear DNA sequence loci for plant species-level phylogeny reconstruction. L., Gaasterland, T., Lopez, P., Muller, M., Phillippe, H., 2002. The analysis of 100 Phil. Trans. R. Soc. B 361, 211–225. genes supports the grouping of three highly divergent amoebae: Dictyostelium, Ickert-Bond, S.M., Wen, J., 2006. Phylogeny and biogeography of Altingiaceae: Entamoeba, and Mastigamoeba. Proc. Natl. Acad. Sci. USA 99, 1414–1419. evidence from combined analysis of five non-coding chloroplast regions. Mol. Barfuss, M.H.J., Samuel, R., Till, W., Stuessy, T.F., 2005. Phylogenetic relationships in Phylogenet. Evol. 39, 512–528. subfamily Tillandsioideae (Bromeliaceae) based on DNA sequence data from Janzen, D.H., 1976. Why bamboos wait so long to flower. Annu. Rev. Ecol. Syst. 7, seven plastid regions. Am. J. Bot. 92, 337–351. 347–391. Bellusci, F., Pellegrino, G., Palermo, A.M., Musacchio, A., 2008. Phylogenetic Jeanmougin, F., Thompson, J.D., Gouy, M., Higgins, D.G., Gibson, T.J., 1998. Multiple relationships in the orchid genus Serapias L. based on noncoding regions of sequence alignment with Clustal X. Trends Biochem. Sci. 23, 403–405. the chloroplast genome. Mol. Phylogenet. Evol. 47, 986–991. Judziewicz, E.J., Clark, L.G., Londoño, X., Stern, M.J., 1999. American Bamboos. Bouchenak-Khelladi, Y., Salamin, N., Savolainen, V., Forest, F., Van der Bank, M., Smithsonian Institution Press, Washington, DC. Chase, M.W., Hodkinson, T.R., 2008. Large multi-gene phylogenetic trees of the Kawakita, A., Sota, T., Ascher, J.S., Ito, M., Tanaka, H., Kato, M., 2003. Evolution and grasses (Poaceae): progress towards complete tribal and generic level sampling. phylogenetic utility of alignment gaps within intron sequences of three nuclear Mol. Phylogenet. Evol. 47, 188–505. genes in Bumble Bees (Bombus). Mol. Biol. Evol. 20, 87–92. Bouchenak-Khelladi, Y., Verboom, G.A., Hodkinson, T.R., Salamin, N., Francois, O., Keng, P.C., 1986. A preliminary study of the inflorescence type arising from NíChonghaile, G., Savolainen, V., 2009. The origins and diversification of C 4 bamboos and its variation. J. Wuhan Bot. Res. 4, 323–336. grasses and savanna-adapted ungulates. Glob. Change Biol. 15, 2397–2417. Keng, P.C., Wang, C.P., 1996. Flora Reipublicae Popularis Sinicae, vol. 9. Science Bystriakova, N., Kapos, V., Lysenko, I., Stapleton, C.M.A., 2003a. Distribution and Press, Beijing. conservation status of forest bamboo biodiversity in the Asia-Pacific region. Lee, M.S.Y., 2001. Uninformative characters and apparent conflict between Biodivers. Conserv. 12, 1833–1841. molecules and morphology. Mol. Biol. Evol. 18, 676–680. Bystriakova, N., Kapos, V., Stapleton, C.M.A., Lysenko, I., 2003b. Bamboo Li, D.Z., 1997. The Flora of China Bambusoideae project—problems and current Biodiversity: Information for Planning Conservation and Management in the understanding of bamboo taxonomy in China. In: Chapman, G.P. (Ed.), The Asia-Pacific Region. UNEP-World Conservation Monitoring Center and Bamboos. Linnean Society Symposium Series 19. Academic Press, London, International Network for Bamboo and Rattan, Cambridge, UK. United Kingdom, pp. 61–81. Clark, L.G., Triplett, J.K., 2007. Arundinaria. In: Flora of North America Editorial Li, D.Z., 1999. Taxonomy and biogeography of the Bambuseae (Gramineae: Committee (Eds.), 1993+. Flora of North America North of Mexico, 14+ vols. Bambusoideae). In: Rao, A.N., Rao, V.R. (Eds.), Bamboo-Conservation, New York and Oxford, vol. 24. pp. 17–24. Diversity, Ecogeography, Germplasm, Resource Utilization and Taxonomy. Clark, L.G., Zhang, W.P., Wendel, J.F., 1995. A phylogeny of the grass family Proceeding of training course cum workshop, 10–17 May 1998, Kunming & (Poaceae) based on ndhF sequence data. Syst. Bot. 20, 436–460. Xishuangbanna, Yunnan, China. IPGRI-APO, Serdagn, Malaysia, pp. 14–23. Clayton, W.D., Renvoize, S.A., 1986. Genera Graminum, Grasses of the World. Her Li, D.Z., Hsueh, C.J., Xia, N.H., 1995. Gaoligongshania, a new bamboo genus from Majesty’s Stationery Office, London. Yunnan, China. Acta Phytotax. Sinica 33, 597–601. Chao, C.S., Chu, C.D., Hsiung, W.Y., 1980. A revision of some genera and species of Li, D.Z., Stapleton, C.M.A., Xue, J.R., 1996. A new combination in Ampelocalamus and Chinese bamboos. Acta Phytotax. Sinica 18, 20–36. notes on A. patellaris (Gramineae: Bambusoideae). Kew Bull. 51, 809–813. Chao, C.S., Renvoize, S.A., 1989. Revision of species described under Arundinaria Li, D.Z., Wang, Z.P., Zhu, Z.D., Xia, N.H., Jia, L.Z., Guo, Z.H., Yang, G.Y., Stapleton, (Gramineae) in South-east Asia and Africa. Kew Bull. 44, 349–367. C.M.A., 2006. Bambuseae (Poaceae). In: Wu, Z.Y., Raven, P.H., Hong, D.Y. (Eds.), Cunningham, C.W., 1997. Can three incongruence tests predict when data should be Flora of China, vol. 22. Science Press and Missouri Botanical Garden Press, combined? Mol. Biol. Evol. 14, 733–740. Beijing and St. Louis. Demesure, B., Sodzi, N., Petit, R.J., 1995. A set of universal primers for amplification Liang, T.G., Huang, K.F., Zheng, Q.F., Shen, R.Z., Kong, F.S., Xie, Q.M., Lin, Y.Y., 1987. of polymorphic non-coding regions of mitochondrial and chloroplast DNA in Bamboos in Fujian. Fujian Science and Technology Publishing House, Fuzhou. plants. Mol. Ecol. 4, 129–131. pp. 128–131. Doyle, J.J., Davis, J.I., Soreng, R.J., Garvin, D., Anderson, M.J., 1992. Chloroplast DNA Nakai, T., 1925. Two new genera of Bambusaceae, with special remarks on the inversions and the origin of the grass family (Poaceae). Proc. Natl. Acad. Sci. USA related genera growing in eastern Asia. J. Arnold Arbor. 6, 145–153. 89, 7722–7726. Nakai, T., 1931. Gramineae-Bambuseae in Miybe et Kudo, Flora of Hokkaido and Doyle, J.J., Doyle, J.L., 1987. A rapid DNA isolation procedure for small quantities of Saghalin II. J. Fac. Agr. Hokkaido Univ. 26, 180–195. fresh leaf tissue. Phytochem. Bull. 19, 11–15. C.-X. Zeng et al. / Molecular Phylogenetics and Evolution 56 (2010) 821–839 839

Ní Chonghaile, G., 2002. Systematics of the woody bamboos (Tribe Bambuseae). Soltis, P.S., Soltis, D.E., Chase, M.W., 1999. Angiosperm phylogeny inferred from Unpublished Ph.D. thesis. University of Dublin, Trinity College, Ireland. multiple genes as a tool for comparative biology. Nature 402, 402–403. Ohrnberger, D., 1999. The Bamboos of the World: Annotated Nomenclature and Stapleton, C.M.A., 1997. The morphology of woody bamboos. In: Chapman, G.P. Literature of the Species and the Higher and Lower Taxa. Elsevier Science B.V., (Ed.), The Bamboos. Academic Press, Linnean Society of London Symposium Amsterdam. Series, London, pp. 251–267. Ohsako, T., Ohnishi, O., 2000. Intra- and interspecific phylogeny of wild Fagopyrum Sungkaew, S., Stapleton, C.M.A., Salamin, N., Hodkinson, T.R., 2009. Non-monophyly (Polygonaceae) species based on nucleotide sequences of noncoding regions in of the woody bamboos (Bambuseae; Poaceae): a multi-gene region chloroplast DNA. Am. J. Bot. 87, 573–582. phylogenetic analysis of Bambusoideae s.s.. J. Plant Res. 122, 95–108. Panero, J.L., Funk, V.A., 2008. The value of sampling anomalous taxa in phylogenetic Suzuki, S., 1978. Index to Japanese Bambusaceae. Gakken Co. Ltd., Tokyo. * studies: major clades of the Asteraceae revealed. Mol. Phylogenet. Evol. 47, Swofford, D.L., 2003. PAUP . Phylogenetic Analysis Using Parsimony ( And Other 757–782. Methods). Version 4.0b10. Peng, S., Yang, H.Q., Li, D.Z., 2008. Highly heterogeneous generic delimitation within Taberlet, P., Gielly, L., Pautou, G., Bouvet, J., 1991. Universal primers for the temperate bamboo clade (Poaceae: Bambusoideae): evidence from GBSSI amplification of three non-coding regions of chloroplast DNA. Plant Mol. Biol. and ITS sequences. Taxon 57, 799–810. 17, 1105–1109. Perret, M., Chautems, A., Spichiger, R., Kite, G., Savolainen, V., 2003. Systematics and Triplett, J.K., 2008. Phylogenetic relationships among the temperate bamboos evolution of tribe Sinningieae (Gesneriaceae): evidence from phylogenetic (Poaceae: Bambusoideae) with an emphasis on Arundinaria and allies. Ph.D. analyses of six plastid DNA regions and nuclear ncpGS. Am. J. Bot. 90, 445–460. thesis. Iowa State University, Ames, Iowa, USA. Posada, D., Buckley, T.R., 2004. Model selection and model averaging in Triplett, J.K., Clark, L.G., 2010. Phylogeny of the temperate bamboos (Poaceae: phylogenetics: advantages of akaike information criterion and Bayesian Bambusoideae: Bambuseae) with an emphasis on Arundinaria and allies. Syst. approaches over likelihood ratio tests. Syst. Biol. 53, 793–808. Bot. 35, 102–120. Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitution. Triplett, J.K., Oltrogge, K.A., Clark, L.G., 2010. Phylogenetic relationships and natural Bioinformatics 14, 817–818. hybridization among the North American woody bamboos (Poaceae: Qiu, Y.L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis, P.S., Zanis, M., Zimmer, E.A., Bambusoideae: Arundinaria). Am. J. Bot. 97, 471–492. Chen, Z., Savolainen, V., Chase, M.W., 1999. The earliest angiosperms: evidence Wang, Z.P., Ye, G.H., 1981. Miscellaneous notes on Chinese Bambusoideae. J. Nanjing from mitochondrial, plastid and nuclear genomes. Nature 402, 404–407. Univ. Nat. Sci. 1, 91–108. Ren, Y., Li, Y., Dang, G.D., 2003. A new species of Bashania (Poaceae: Bambusoideae) Watson, L., Dallwitz, M.J., 1992. The Grass Genera of the World. CAB International, from Mt. Qingling, Shaanxi, China. Novon 13, 473–476. Wallingford, Oxon, UK. Rex, M., Schulte, K., Zizka, G., Peters, J., Vásquez, R., Ibisch, P.L., Weising, K., 2009. Wen, T.H., 1984. New taxa of Bamboo from China (I). J. Bamboo Res. 3, Phylogenetic analysis of Fosterella L.B. Sm. (Pitcairnioideae, Bromeliaceae) 23–47. based on four chloroplast DNA regions. Mol. Phylogen. Evol. 51, 472–485. Wiens, J.J., 1998. The accuracy of methods for coding and sampling higher-level taxa Ronquist, F., Huelsenbeck, J.P., 2003. MRBAYES 3: Bayesian phylogenetic inference for phylogenetic analysis: a simulation study. Syst. Biol. 47, 381–397. under mixed models. Bioinformatics 19, 1572–1574. Yamane, K., Kawahara, T., 2005. Intra and interspecific phylogenetic relationships Rouhan, G., Dubuisson, J.Y., Rakotondrainibe, F., Motley, T.J., Mickel, J.T., Labat, J.N., among diploid Triticum-Aegilops species (Poaceae) based on base-pair Moran, R.C., 2004. Molecular phylogeny of the fern genus Elaphoglossum substitutions, indels, and microsatellites in chloroplast noncoding sequences. (Elaphoglossaceae) based on chloroplast non-coding DNA sequences: Am. J. Bot. 92, 1887–1898. contributions of species from the Indian Ocean area. Mol. Phylogenet. Evol. Yang, G.Y., Zhao, Q.S., 1993. A revision of the genus Arundinaria Michaux in China (I). 33, 745–763. J. Bamboo Res. 12, 1–6. Saltonstall, K., 2001. A set of primers for amplification of noncoding regions of Yang, G.Y., Zhao, Q.S., 1994. A revision of the genus Arundinaria Michaux in China chloroplast DNA in the grasses. Mol. Ecol. Notes 1, 76–78. (II). J. Bamboo Res. 13, 1–23. Shaw, J., Lickey, E.B., Beck, J.T., Farmer, S.B., Liu, W.S., Miller, J., Siripun, K.C., Winder, Yang, H.-Q., Yang, J.-B., Peng, Z.-H., Gao, J., Yang, Y.-M., Li, D.-Z., 2008. A molecular C.T., Schilling, E.E., Small, R.L., 2005. The tortoise and the hare II: relative utility phylogenetic and fruit evolutionary analysis of the major groups of the of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Am. J. paleotropical woody bamboos (Gramineae: Bambusoideae) based on nuclear Bot. 92, 142–168. ITS, GBSSI gene and plastid trnL-F DNA sequences. Mol. Phylogenet. Evol. 48, Shaw, J., Lickey, E.B., Schilling, E.E., Small, R.L., 2007. Comparison of whole 809–824. chloroplast genome sequences to choose noncoding regions for phylogenetic Yi, T.P., 2001. A new species of the genus Indosasa McClure from Yunnan, China. J. studies in angiosperms: the tortoise and the hare III. Am. J. Bot. 94, 275–288. Bamboo Res. 20, 1–3. Shaw, J., Small, R.L., 2004. Addressing the ‘‘hardest puzzle in American pomology:” Yi, T.P., Shi, J.Y., Ma, L.S., Wang, H.T., Yang, L., 2008. Iconographia Bambusoidearum phylogeny of Prunus Sect. Prunocerasus (Rosaceae) based on seven noncoding Sinicarum. Science Press, Beijing. chloroplast DNA regions. Am. J. Bot. 91, 985–996. Yi, T.P., Shi, J.Y., Ma, L.S., Yang, L., Wang, H.T., 2007a. A report on a new bamboo Simmons, M.P., Ochoterena, H., 2000. Gaps as characters in sequence-based species in Northeast Yunnan, China. J. Sichuan Forest. Sci. Tech. 28, 1–3. phylogenetic analyses. Syst. Biol. 49, 369–381. Yi, T.P., Shi, J.Y., Ma, L.S., Yang, L., Wang, H.T., 2007b. Indocalamus jinpingensis T.P. Yi Sjolin, E., Erseus, C., Kallersjo, M., 2005. Phylogeny of Tubificidae (Annelida, and J.Y. Shi, a new species of the Gramineae from South Yunnan, China. Acta Clitellata) based on mitochondrial and nuclear sequence data. Mol. Phytotax. Sinica 45, 693–695. Phylegenet. Evol. 35, 431–441. Yi, T.P., Yang, L., 1998. A new species of the alpine bamboo from China. J. Bamboo Small, R.L., Ryburn, J.A., Cronn, R.C., Seelanan, T., Wendel, J.F., 1998. The tortoise and Res. 17, 1–3. the hare: choosing between noncoding plastome and nuclear Adh sequences for Yuan, Y.W., Olmstead, R.G., 2008. A species-level phylogenetic study of the Verbena phylogenetic reconstruction in a recently diverged plant group. Am. J. Bot. 85, complex (Verbenaceae) indicates two independent intergeneric chloroplast 1301–1315. transfers. Mol. Phylogenet. Evol. 48, 23–33. Smedmark, J.E.E., Swenson, U., Anderberg, A.A., 2006. Accounting for variation of Zhang, Y.X., Li, D.Z., in press. A new combination in Pseudosasa with a revised substitution rates through time in Bayesian phylogeny reconstruction of description of Indosasa hispida (Poaceae: Bambusoideae). Ann. Bot. Fennici. Sapotoideae (Sapotaceae). Mol. Phylogenet. Evol. 39, 706–721. Zhang, Y. X., Zeng, C. X., Li, D. Z., in preparation. Reticulate evolution of the Smith, S.A., Donoghue, M.J., 2008. Rates of molecular evolution are linked to life temperate woody bamboos (Poaceae: Bambusoideae): evidences from nuclear history in flowering plants. Science 322, 86–89. and plastid DNA sequence data. Soderstrom, T.R., Ellis, R.P., 1982. Taxonomic status of the endemic South African Zhuge, Q., Ding, Y.L., Xu, C., Zou, H.Y., Huang, M.R., Wang, M.X., 2004. Phylogeny of Bamboo, Thamnocalamus tessellatus. Bothalia 14, 53–67. the genus Arundinaria based on nucleotide sequences of nrDNA ITS region. Acta Soderstrom, T.R., Ellis, R.P., 1987. The position of bamboo genera and allies in a Genetica Sinica 31, 349–356. system of grass classification. In: Soderstrom, T.R., Hiu, K.W., Campbell, C.S., Zwickl, D.J., Hillis, D.M., 2002. Increased taxon sampling greatly reduces Barkworth, M.F. (Eds.), Grass Systematics and Evolution. Smithsonian phylogenetic error. Syst. Biol. 51, 588–598. Institution Press, Washington, DC, pp. 225–238.