Phylogeny and Biogeography of Alyssum (Brassicaceae) Based on Nuclear Ribosomal ITS DNA Sequences

c Indian Academy of Sciences

RESEARCH ARTICLE

Phylogeny and biogeography of Alyssum (Brassicaceae) based on nuclear ribosomal ITS DNA sequences

YAN LI 1,2,3, YAN KONG1,3, ZHE ZHANG1,3, YANQIANG YIN1,3,BINLIU1∗, GUANGHUI LV2 and XIYONG WANG1

1Key Laboratory of Biogeography and Bioresource in Arid Land, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi 830011, People’s Republic of China 2College of Resource and Environment Science, Xinjiang University, Urumqi 830046, People’s Republic of China 3University of Chinese Academy of Sciences, Beijing 100049, People’s Republic of China

Abstract The genus Alyssum consists of about 195 species native to Europe, Asia and northern Africa. All species were assigned to six sections. Previous molecular phylogeny studies indicate that Alyssum is polyphyletic. However, the divergence time and dispersal of the genus are not well studied. In this study, the phylogenetic relationships within the genus Alyssum were studied with nrDNA ITS sequences obtained from five sections. The divergence time was estimated by fossil calibration and the biogeography was examined by spread analysis. The phylogeny indicated two main lineages: lineage 1 includes the section of Alyssum, Gamosepalum and Psilonema; lineage 2 includes the section of Odontarrhena, Meniocus and Clypeola.The phylogenetic relationship was not congruent with the previous sectional classifications. The age of Alyssum was dated to the upper Miocene. Molecular data suggested the diversification of Alyssum in Mediterranean areas and wide-ranging distribution such as North Africa, eastward into Central Asia and immigration into North America. Climatic aridification and arid/semiarid areas established in the Pliocene/Pleistocene could have provided favourable conditions for the migration and diversification of Alyssum.

[Li Y., Kong Y., Zhang Z., Yin Y., Liu B., Lv G. and Wang X. 2014 Phylogeny and biogeography of Alyssum (Brassicaceae) based on nuclear ribosomal ITS DNA sequences. J. Genet. 93, 313–323]

Introduction united with Alyssum (Warwick et al. 2008). Moreover, A. klimesii is most closely related to Ptilotrichum canescens ∼ The genus Alyssum consists of 195 species (Warwick et al. (Al-Shehbaz 2002), and it should belong to the tribe 2006) that are native to Europe, Asia and Northern Africa Arabideae (Al-Shehbaz et al. 2006). However, A. klimesii (Alshehbaz 1987). Species richness and diversity are con- formed the sister group with Crucihimalaya with a super fined to the Mediterranean and Turkey, and only a few network analysis of several genes (Koch and Matschinger species are distributed in North Africa, Central Asia, Siberia 2007). Warwick et al.(2008) suggested that A. klimesii and North America (Dudley 1964a, b). The genus comprises should be placed in Camelineae. German and Al-Shehbaz annual and perennial herbaceous plants and (rarely) small (2010) classified A. klimesii in the tribe Crucihimalayeae. shrubs, with oblong–oval leaves and yellow or white flowers The species of Alyssum are divided into six sections: (pink and purple in a few species), stellate, stalked or ses- Meniocus (Desv.) Hook, Psilonema (C. A. Meyer) Hook, sile trichomes, and dehiscent silicle fruit (Ball and Dudley Alyssum, Gamosepalum (Hausskn.) Dudley, Tetradenia 1964). (Spach) Dudley, and Odontarrhena (C.A.Meyer)Koch. Recent phylogenetic analyses indicate that Alyssum is Infrageneric phylogeny studies of section Odontarrhena polyphyletic (Al-Shehbaz et al. 2006; Warwick et al. 2008). have been conducted (Mengoni et al. 2003; Cecchi et al. Morphological evidence and internal transcribed spacer 2010), which demonstrate the monophyletic character of (ITS) sequencing data demonstrate that Clypeola should be section Odontarrhena. However, these investigations were mostly concentrated on the European taxa and focussed on the evolutionary dynamics of nickel hyperaccumulation. The inclusion of taxa from Asia and North America was ignored ∗ For correspondence. E-mail: [email protected]. in these studies, even though the information about these Keywords. Brassicaceae; phylogeny; ITS; divergence time; biogeography; Alyssum.

Journal of Genetics, Vol. 93, No. 2, August 2014 313 YanLietal. groups is important for the biogeography and global dispersal 45 s annealing at 52◦C, and 1 min elongation at 72◦C; and of the genus Alyssum. a final elongation of 10 min at 72◦C. PCR products were The divergence time and migration of Alyssum are checked for length and concentrations on a 1.2% agarose gel still unclear. As indicated by Cecchi et al.(2010), the in TAE buffer. The gel was stained with nontoxic SYBR section Odontarrhena had a younger divergence time in Green. Before sequencing, the PCR products were purified Europe, with a mean ITS genetic distance of ∼2%, but using Sigma GenElute PCR Clean-up Kit (Sigma-Aldrich, there is no further investigation for the divergence time St Louis, USA) following the manufacturer’s protocol. of this genus (Cecchi et al. 2010). Section Odontarrhena Sequencing was performed on the automated ABI 3730 DNA is characterized by the prevalence of species that are dis- Analyzer, and the original amplification primers were used tributed on serpentine soils. They accumulate high concen- as double-strand sequencing primers. tration of nickel (>1000 μg/g dry mass), hitherto known as nickel hyperaccumulators. Previous study suggests that section Odontarrhena spread from the origin centre of Phylogenetic analysis Turkey based on the highest number of nickel hyperaccu- The ITS sequencing data were aligned using Clustal W mulators (Brooks and Radford 1978). However, the phylo- (Thompson et al. 1994) as implemented in MEGA 5.0 genetic relationships depicted by Mengoni et al.(2003) con- (Tamura et al. 2011). The maximum parsimony (MP) tree flicted with Brooks’ hypothesis. Mengoni argued that the was constructed using MEGA 5.0 with the tree-bisection- present area of distribution of the species might reflect histor- reconnection (TBR) search method. Gaps were treated as ical phenomena such as spreading immediately for the more missing data. Bootstrap values were calculated for analy- cold-sensitive species after the Ice Age. sis sampling trees (BEAST) 1000 replicates. The Bayesian In the present study, nuclear ITS of ribosomal DNA phylogenetic tree was constructed using the Bayesian evo- (include ITS1, 5.8S RNA and ITS2) with a larger collection lutionary software package BEAST ver. 1.7.4 (Drummond of Alyssum from five of six sections and broader geographic et al. 2012). Two independent runs of 45 million genera- distribution areas were used to (i) examine the phylogenetic tions were undertaken by sampling every 1000th generation. status of Alyssum and compare it with the traditional sec- Each run was checked using Tracer ver. 1.5 (Drummond tional classification; (ii) determine the infrageneric relation- and Rambaut 2007) for having reached a sufficient effective ship; and (iii) draw some conclusions of the Alyssum crown sample size (ESS over 100, as suggested by the authors) and stem age. in relevant statistics. The two runs (log-files and trees-files) were then combined with LogCombiner ver. 1.7.4. The maximum clade credibility (MCC) tree was produced from 81,002 Materials and methods post-burn-in trees using TreeAnnotator ver. 1.7.4. Finally, Taxon sampling FigTree ver. 1.5 (http://tree.bio.ed.ac.uk/) was used to display and print the molecular phylogenetic tree. A total of 90 ITS sequences from 63 species were obtained The best fitting nucleotide substitution models were cho- for phylogenetic analysis. A total of 78 accessions from 53 sen using the program MrModeltest ver. 2.3 (Nylander Alyssum species, five Crucihimalaya taxa and four Clype- 2004). The model selected by Bayesian information criterion ola taxa were included in the sequences. ITS region for five (BIC) in the program was chosen for the BEAST analysis accessions representing three species of Alyssum from China and for the MP analysis, the criterion applied was the Akaike was sequenced. Aethionema grandiflorum was selected as information criterion (AIC) following Koch et al.(2010). the outgroup following studies on Brassicaceae (Galloway et al. 1998;Kochet al. 2001; Mathews and McBreen 2008; Couvreur et al. 2010; Franzke et al. 2011). Detailed sequence Divergence time estimates information, geographic origin and GenBank accession num- Based on the markers and calibration points used divergence- bers were provided in table 1. time estimates in Brassicaceae have been discussed widely. In the present study, divergence time was estimated by BEAST using estimated Brassicaceae age, macrofossil DNA extraction, PCR amplification and sequencing calibration and mutation rate of ITS. The primary BEAST Total DNA was extracted by the CTAB protocol (Doyle estimation was based on the age of Brassicaceae which was 1987). PCR amplifications of ITS region was performed in a dated 54.3 Mya (95% highest probability density (HPD) volume of 50 μL, containing 0.2 mM dNTPs, 0.5 μM of each interval: 45.2–64.2) by Beilstein et al.(2010). The secondary primer, 0.5 μL DMSO, 1.5 mM MgCl2 and2UofTaq poly- BEAST estimation used the macrofossils of Clypeola from merase (Takara Bio, Shanghai, China). The ITS region was upper Miocene dated 5–11 Mya (Schulz 1936).Therateof amplified using the primers ITS5 and ITS4 designed by mutation of ITS sequences for calculation was adopted in White et al.(1990). The amplifications were conducted under this research. The mutation rates for ITS region were vari- the following conditions: 5 min initial denaturation at 95◦C; able among plants and the value was between 1.72 × 10−9 35 cycles of amplification with 30 s denaturation at 95◦C, and 1.71 × 10−8 mutations per site per year in herbaceous

314 Journal of Genetics, Vol. 93, No. 2, August 2014 ITS for the phylogenic relationship of the genus Alyssum

Table 1. List of studied taxa including voucher geographical origin and GenBank accession numbers.

Genus Species Location Section Accession number

Alyssum A. alpestre Italy Odontarrhena AY237957.1 A. alyssoides New Zealand Psilonema AF401114.1 A. alyssoides Wyoming, USA Psilonema EF514595.1 A. alyssoides Canada Psilonema EF514596.1 A. americanum Alaska, USA Odontarrhena EF514597.1 A. americanum Yukon, Canada Odontarrhena EF514598.1 A. anatolicum Turkey Odontarrhena AY237956.1 A. argenteum Piedmont Molette, Italy Odontarrhena GQ284854.1 A. aureum Israel Meniocus EF514601.1 A. baldaccii Sterea Ellas, Greece Odontarrhena GQ284858.1 Fourka A. baumgartnerianum Israel Gamosepalum EF514602.1 A. bertolonii Italy Odontarrhena AY237954.1 A. bertolonii Tuscany Mt. Ferrato, Italy Odontarrhena GQ284859.1 A. biovulatum South Siberia Odontarrhena AY237953.1 A. caricum Turkey Odontarrhena AY237952.1 A. chalcidicum Elbasan Perrenjas, Albania Odontarrhena GQ284869.1 Macedonia Drosia A. chalcidicum Central, Greece Odontarrhena GQ284876.1 A. condensatum Syria Odontarrhena AY237951.1 A. corsicum Corsica, France Odontarrhena AY237949.1 A. corymbosoides Western Greece Odontarrhena GQ284878.1 Macedonia, Mt.Vourinos A. cypricum Cyprus Odontarrhena AY237948.1 A. dasycarpum China Psilonema KF512501.1 A. dasycarpum Kazakhstan Psilonema EF514604.1 A. davisianum Turkey Odontarrhena AY237947.1 A. densistellatum Sterea Ellas Evvia, Greece Alyssum GQ284879.1 A. desertorum China Alyssum KF512497.1 A. desertorum China Alyssum KF512498.1 A. desertorum China Alyssum KF512499.1 A. desertorum Oregon, USA Alyssum EF514605.1 A. desertorum Alberta, Canada Alyssum EF514606.1 A. desertorum Nevada, USA Alyssum EF514607.1 A. desertorum Korce Bitincka, Albania Alyssum GQ284881.1 A. euboeum Sterea Ellas Evvia, Greece Odontarrhena GQ284882.1 A. fallacinum Crete Odontarrhena AY237946.1 A. ﬂoribundum Turkey Odontarrhena AY237945.1 A. fragillimum Crete Lefka Ori, Greece Odontarrhena GQ284883.1 A. heldreichii Western Macedonia, Greece Odontarrhena GQ284884.1 Mt.Vourinos A. huber-morathii Turkey Odontarrhena AY237943.1 A. klimesii Jammu and Kashmir, India Camelineae DQ518387.1 A. klimesii Ladak, India Camelineae EF514608.1 A. klimesii Xizang, China Camelineae EF514609.1 A. lenense Kazakhstan Alyssum EF514610.1 A. lesbiacum Lesvos Odontarrhena AY237942.1 A. linifolium China Meniocus KF512500.1 A. linifolium Kazakhstan Meniocus EF514611.1 A. malacitanum Spain Odontarrhena AY237941.1 A. meniocoides Syria Meniocus EF514612.1 A. minus Turkey Alyssum AY237939.1 A. montanum South Europe Alyssum AY237938.1 A. montanum Germany Alyssum EU559700.1 A. murale Georgia, USA Odontarrhena EF514613.1 A. murale Alberta, Canada Odontarrhena EF514614.1 A. murale Blagoevgrad, Bulgaria Odontarrhena GQ284885.1 Anatolic, Greece Odontarrhena GQ284886.1 A. murale Mt. Pangeon, Macedonia Piedmont, Italy Odontarrhena GQ284887.1 A. murale Naturalized near Casteldelﬁno A. nebrodense Sicily Odontarrhena AY237935.1 A. obovatum Siberia, Russia Odontarrhena EF514620.1 A. obovatum Karaghanday, Kazakhstan Odontarrhena EF514619.1 A. orbelicum Blagoevgrad, Bulgaria Odontarrhena GQ284888.1 A. oxycarpum Turkey Odontarrhena AY237934 A. peltarioides Turkey Odontarrhena AY237932 A. pinifolium Turkey Odontarrhena AY240871

Journal of Genetics, Vol. 93, No. 2, August 2014 315 YanLietal.

Table 1 (contd)

Genus Species Location Section Accession number A. pintodasilvae Portugal Odontarrhena AY237929 A. pterocarpum Turkey Odontarrhena AY237931 A. repens subsp. trichostachyum Georgia, USA Alyssum EF514622.1 A. robertianum Sardinia, Italy Odontarrhena GQ284889.1 A. saxatile New Zealand Meniocus AF401115.1 A. serpyllifolium Portugal Odontarrhena EF514623.1 A. serpyllifolium subsp. serpyllifolium France Odontarrhena AY237923.1 A. sibiricum Sterea Ellas Evvia, Greece Odontarrhena GQ284890.1 Austria Alyssum EF514624.1 A. simplex Epirus Mt., Greece Odontarrhena GQ284891.1 A. smolikanum Smolikas A. tenium Tinos Odontarrhena AY237926.1 A. tortuosum Turkmenistan Odontarrhena EF514625.1 A. tortuosum Macedonia prilep Odontarrhena GQ284892.1 Mts. Troodos, Cyprus A. troodi Mt. Khionistra Odontarrhena GQ284893.1 A. virgatum Turkey Odontarrhena AY237925.1 A. wulfenianum Italy Alyssum AY237924.1 Clypeola C. aspera Iran EF514642.1 C. cyclodontea Algeria EF514643.1 C. jonthlaspi Turkmenistan EF514644.1 C. lappacea Iran EF514645.1 Crucihimalaya C. himalaica China AY662283.1 C. mollissima Afghanistan AF137551.1 C. mongolica n.d. FJ187923.1 C. stricta Pakistan AF137554.1 C. wallichii Afghanistan JX146960.1 Aethionema A. grandiflorum − AM905710.1 A. grandiflorum − DQ249867.1 A. grandiflorum − AM905711.1

−Data not available. plants (Kropf 2002;Kayet al. 2006). Mutation in the Bras- Table 2. Locations assigned and the coordinates (latitudes and sicaceae was assumed to be 5–20 × 10−9 substitutions per longitudes). site per year (Koch et al. 2003). To accommodate for calibra- Location Latitudes N/S Longitudes E/W tion uncertainty, we ﬁnally ﬁxed the mutation rates to a range −9 −8 of 1.72 × 10 to 1.71 × 10 mutations per site per year Alaska 65.3 −145.0 as done by Koch et al.(2010). The most appropriate evolu- Albania 40.8 20.2 Alberta 59.3 −128.3 tionary model was estimated using the program MrModel- Algeria 30.8 3.5 test 2.3, which is based on the BIC. BEAST analyses were Austria 47.2 14.6 Bulgaria 42.5 24.8 carried out under the lognormal uncorrelated relaxed clock Canada 64.2 −138.1 (Gernhard 2008) and the Yule speciation priori (Yule 1925) China 45.4 84.7 Corsica 42.1 8.9 as the relaxed clock allowed substitution rates to vary among Crete 35.2 24.8 different lineages (Drummond et al. 2006) and the calcu- Cyprus 35.1 33.1 France 47.2 4.4 lated Bayes factors (Nylander et al. 2004) favoured their use Georgia 31.7 −82.6 over a strict clock and a pure birth model. The sequence Greece 39.2 22.2 Iran 31.3 53.4 distance between taxa was estimated using MEGA 5.0 with Israel 30.6 34.7 K2P model. Italy 42.0 12.5 Kazakhstan 48.0 69.7 Lesvos 39.2 26.2 Biogeography Macedonia 41.4 21.3 Nevada 40.0 −116.7 New Zealand −41.1 175.1 Spread is a software that analyses and visualizes Bayesian Oregon 44.2 −120.1 phylogeographic reconstructions incorporating spatial diffu- Portugal 39.2 −8.4 Siberia 56.2 93.2 sion. This software maps phylogenies annotated with spa- Sicily 37.2 12.8 tial information and can export high-dimensional posterior South Europe 48.3 11.4 Spain 39.5 −4.1 summaries to a keyhole markup language (KML) for ani- Syria 34.4 38.5 mation of the spatial diffusion in virtual globe software Tinos 39.2 21.2 (Bielejec et al. 2011). To construct a spatial diffusion pro- Turkey 38.2 33.8 Turkmenistan 39.5 59.9 jection of Alyssum, the location annotated MCC tree was Tuscany 43.5 10.9 constructed with BEAST ver. 1.7.4 as above. The present Wyoming 43.0 −107.0 − study was focussed on phylogeography of genus Alyssum; Yukon 64.0 136.0

316 Journal of Genetics, Vol. 93, No. 2, August 2014 ITS for the phylogenic relationship of the genus Alyssum thus, the species A. klimesii that was excluded from Alyssum the geographic origin of the sequences. The assigned loca- and Crucihimalaya were not included in the analyses. tions and coordinates (latitudes and longitudes) are sum- Alyssum sequences were assigned to 35 locations based on marized in table 2. The MCC tree was converted into a

Figure 1. Strict consensus Bayesian phylogenetic tree based on the nrDNA ITS data set. Bootstrap values (ﬁrst mentioned) and posterior probability values (second mentioned) appear above branches; sections classiﬁcation by Dudley are indicated. Taxa distribution are indicated following the taxon label: taxa from Europe, Turkey, Central Asia, Africa, New Zealand and North America are marked with a black spot, a black cross, a gray spot, a black triangle, an inverted triangle and a black square, respectively.

Journal of Genetics, Vol. 93, No. 2, August 2014 317 YanLietal.

KML file and visualized in Google Earth (http://earth.google. Discussion com). Intrageneric relationship of Alyssum The present study clearly demonstrated that Alyssum has Results split into two main clades. One clade comprised the sections Alyssum, Gamosepalum and Psilonema. The other clade Phylogenetic analysis included members of the sections Odontarrhena, Meniocus The ITS alignment comprised 90 sequences and 778 charac- and Clypeola. Phylogenetic analysis suggested that these ters of which 240 were constant. Of the 488 variable charac- findings agreed only to a small degree with the previous ters, 364 were parsimony informative. The annotated align- sectional classification. Section Odontarrhena is a mono- ment is available upon request. The general time reversible phyletic lineage (64 BS, PP = 1.00). Section Odontarrhena (GTR) model and gamma-distributed substitution rate het- was characterized by perennial, uniovulate silicle locules and erogeneity were chosen as best fit model and applied to the nonmucilaginous seeds (Dudley 1964a, b). Owing to these Bayesian analysis. The most parsimonious tree length was unique characteristics, this section was first described and 1361. The consistency index (CI) was 0.50, the retention classified as a separate genus Odontarrhena C. A. Meyer that index (RI) was 0.82, and the composite index was 0.45 for was distinct from Alyssum (Brooks et al. 1979). The sections all sites and 0.41 for parsimony informative sites. Meniocus, Alyssum and Psilonema were nonmonophyletic. Phylogenetic trees resulting from the MP analyses were Sections Alyssum, Gamosepalum and Psilonema formed a almost congruent with that from the Bayesian analyses highly supported clade (99 BP, PP = 1.00), but neither section (figure 1). The species of Alyssum clustered into two major Alyssum nor Psilonema is monophyletic. A. baumgartneri- clades. One clade (clade 3) comprised members of anum (section Gamosepalum), A. lenense (section Alyssum) section Alyssum (A. desertorum, A. lenense, A. montanum, A. and A. dasycarpum (section Psilonema) formed a strongly repens, A. simplex, A. wulfenianum and A. densistellatum), supported subclade (99 BP, PP = 1.00). A. alyssoides (section section Gamosepalum (Hausskn.) Dudley (A. baumgart- Psilonema) forms a clade with A. minus, the latter are being nerianum) and section Psilonema (C. A. Mey.) Hook. f. a member of section Alyssum (97 BP, PP = 1.00 figure 1). (A. alyssoides and A. dasycarpum) (BP = 99, PP = 1.00). The section Gamosepalum contains only one species and The second clade consisted of clade 1 and group 2 (BP = 61, thus monophyly of this section was unclear. Results indi- PP = 1.00). Clade 1 comprised exclusively the taxa from cated that traditional classification of Alyssum based on mor- section Odontarrhena (BP = 64, PP = 1.00). Group 2 phological characters was somewhat artificial and did not included section Meniocus (A. meniocoides, A. linifolium and reflect phylogenetic relationships. Such artificiality was elu- A. saxatile) and the four Clypeola species. cidated by molecular studies on Cochlearia (Koch et al. 1999a, b), Thlaspi (Mummenhoff et al. 1997), Arabis (Koch et al. 1999a, b, 2001), Arabidopsis (O’Kane and Al-Shehbaz Divergence time estimates 2003), Draba (Koch and Al-Shehbaz 2002)andIsatis Results of the divergence-time calculation are shown in (Moazzeni et al. 2010). table 1. The estimates calculated using the age of Brassica- ceae were more than six times higher than those calculated Divergence time and biogeography with the calibration point of the Clypeola macrofossils. But the divergence time estimates based on the simple mean pair- Estimates of divergence times varied greatly depending wise distances were largely congruent with those calculated on the approaches applied (Franzke et al. 2011). Sec- with the Clypeola macrofossils. ondary callibrations based on the estimated Brassicaceae age

Table 3. Summary statistics of the BEAST analyses showing the various divergence time estimates (tMRCA) from the ITS data set and the mean ITS sequence distances estimated from K2P.

Calculation based on Calculation based on the age of Brassicaceae, the macrofossils of Clypeola, Calculation based on mean value (95% CI) mean value (95% CI) the mutation rate

Log likelihood −8260.456 −8261.50 Nod a (Mya) 44.66 (28.76–60.06) 7.28 (0.67–15.35) 5.5–11.1 Nod b (Mya) 25.94 (13.48–40.04) 4.21 (0.36–9.20) 2.8–5.6 Nod c (Mya) 36.81 (22.04–51.44) 6.00 (0.57–12.61) 3.5–7.0 Nod d (Mya) 30.75 (18.43–44.50) 5.03 (0.49–10.36) 3.25–6.5 Nod e (Mya) 23.67 (13.67–34.52) 3.90 (0.34–8.24) 2.9–5.8

Node coding of ‘a’ to ‘e’ followed the code used in ﬁgure 2.

318 Journal of Genetics, Vol. 93, No. 2, August 2014 ITS for the phylogenic relationship of the genus Alyssum have resulted in high divergence time values. Estimates of ancestor (tMRCA) of Alyssum parviﬂorum and Clypeola Alyssum dated at 44.66 Mya (95% CI, 28.76–60.06) were aspera was dated at 13.15 Mya by the estimates of Beilstein six times older than that of Clypeola macrofossils, 7.28 Mya (Beilstein et al. 2010), this value was among the 95% CI of (95% CI, 0.67–15.35) (table 3). The K2P sequence-distance- 0.67–15.35 Mya resulted from Clypeola macrofossil-based based calibration gave an estimate of 5.5–11.1 Mya for age calibration. of Alyssum. With these results, the origin of Alyssum would The Mediterranean area was the present generic cen- be dated at upper Miocene (5.5–15.35, with mean value tre of maximum endemism, proliferation and multiplicity of 7.28 Mya). However, the time to most recent common of Alyssum (Brooks and Radford 1978). The split time of

Figure 2. Divergence time estimates obtained from the BEAST analysis. Node (node a to e) ages are indicated. Node age and 95% CI intervals are listed in table .

Journal of Genetics, Vol. 93, No. 2, August 2014 319 YanLietal. section Odontarrhena which is mainly comprised of species Based on the ITS data, it could be concluded that Alyssum from Europe and Turkey, was roughly dated to 2.9–5.8 Mya, could have undergone a rapid radiation and the migration under mean ITS distance of 5.8%. This value was corre- into North America might have occurred in the Pleistocene. sponding to the BEAST estimation of 3.9 Mya. This sug- Similar rapid radiation and migration during the Pleistocene gested a Pliocene diversification in southern Europe and from Asia into Europe or North America were observed in Turkey. Besides, the diversification in Europe and Turkey, Noccaea (Koch and Al-Shehbaz 2004)andArabis (Koch migration events of Alyssum to other parts of the world had et al. 2010). Since large parts of the Arctic and Subarc- occurred, implied by the spread analysis (figure 2). tic in Northwest America and eastern Siberia (i.e. Beringia) were never glaciated (Hultén 1937; Hopkins 1967; Frenzel 1968), these northern areas were also thought to have served Migration into Central Asia and North America as an important refuge for arctic plants during Pleistocene glaciations (Hultén 1937). Several studies demonstrated that A. alyssoides, A. linifolium, A. desertorum and A. tortuosum Beringia was one of refuges during Pleistocene glaciations are common between Europe and Central Asia. This distri- (Redenbach and Taylor 1999; Tremblay and Schoen 1999; bution pattern would indicate dispersal among these regions Abbott et al. 2000; Fedorov and Stenseth 2002). The Bering and which might occur in the Pleistocene (figure 3). Such bridge connected Asia and North America several times dispersal was also observed in Lepidium (Mummenhoff throughout the late Tertiary and Pleistocene (Parrish 1987), et al. 2001)andSolms-laubachia s.l. (Yue et al. 2009). and this land bridge probably served as an immigration North America distributed species of Alyssum (A. obovatum, route for several Brassicaceae, including Thlaspi (Payson A. desertorum, A. alyssoides and A. simplex)wereintro- 1926), Stroganowia Karelinet Kirilov (Rollins 1982), Draba duced (Dudley 1968). A. obovatum are distributed in Cen- (Koch and Al-Shehbaz 2002)andArabis (Koch et al. 2010). tral Asia and north Asia (Kazakhstan, northwest China, Therefore, migration into North America via the Bering land Mongolia and south Siberia) and North America (Alaska, bridge during Pleistocene was suggested with respect to Yukon of Canada). The biogeography implied dispersal event the close biogeographic relationship and low ITS sequence between Asia and North America and a close affinity of divergence. that in Asia and North America. This close relationship between the floras of Central Asia and Northwest America were well documented (Parks and Wendel 1990). The aver- age ITS sequence divergence between subclade A (figure 1) Migration into New Zealand was 1.9%, indicating that the first migration from Mediter- Long-distance dispersal into the southern hemisphere in the ranean area into North America occurred at about 1–2 Mya. Pliocene/Pleistocene was strongly indicated for numerous The mean sequence distances within subclade C (A. deser- plant taxa (Raven and Axelrod 1972; Raven 1973;Barlow torum) and subclade B (A. americanum and A. obovatum) 1994). It is suggested that the New Zealand flora was were 0.5% and 0.2%, respectively, indicating a least migra- entirely from long-distance dispersal (Pole 1994). A. saxatile tion into North America could be dated to 0.1–0.5 Mya. (syn. Aurinia saxatilis and A. alyssoides are native to Eurasia

Figure 3. Spatial diffusion of Alyssum reconstructed from the location-coordinated MCC tree. Maps are based on satellite pictures made available in Google Earth (http://earth.google.com).

320 Journal of Genetics, Vol. 93, No. 2, August 2014 ITS for the phylogenic relationship of the genus Alyssum

(Europe and Asia), and an introduced species in New Al-Shehbaz I. A., Beilstein M. A. and Kellogg E. A. 2006 Systemat- Zealand (Allan 1961). However, their dispersal routes were ics and phylogeny of the Brassicaceae (Cruciferae): an overview. not clear. Two migration routes to New Zealand appear fea- Plant Syst. Evol. 259, 89–120. Ball P. W. and Dudley T. R. 1964. Alyssum L. In Flora of Euro- sible: (i) from Asia to southern hemisphere as suggested pea, vol. 1 (ed. T. G. Tutin) pp. 359–369 Lycopodiaceae to by Raven (Raven 1973); and (ii) direct dispersal from the platanaceae. Cambridge University Press, New York, USA. northern hemisphere, e.g. North America (Wardle 1978; Barlow B. 1994 Phytogeography of the Australian region. Aust. Vanhouten et al. 1993). In fact, some birds were known to be Vegetation 2, 3–36. regular migrants between Australia/New Zealand and North Beilstein M. A., Nagalingum N. S., Clements M. D., Manchester S. R. and Mathews S. 2010 Dated molecular phylogenies indicate a America (Wardle 1978) and between Australia/New Zealand Miocene origin for Arabidopsis thaliana. Proc. Natl. Acad. Sci. and East Asia. USA 107, 18724–18728. According to our present understanding and research, one Bielejec F., Rambaut A., Suchard M. A. and Lemey P. 2011 can speculate that speciation of northern taxa was mostly SPREAD: spatial phylogenetic reconstruction of evolutionary affected by climatic fluctuations during Pleistocene glacia- dynamics. Bioinformatics 27, 2910–2912. Brooks R. R. and Radford C. C. 1978 Nickel accumulation by Euro- tion and deglaciation cycles. For the western Mediterranean pean species of genus Alyssum. Proc. R. Soc. Lond. Ser. B. 200, taxa, climatic fluctuations dating back to the Miocene or 217–224. Pliocene might be more important to explain the present Brooks R. R., Morrison R. S., Reeves R. D., Dudley T. R. and diversification and distribution patterns (Santos-Gally et al. Akman Y. 1979 Hyper-accumulation of nickel by Alyssum Lin- 2012). During late Miocene (ca. 6.5–5.3 Mya), the Mediter- naeus (Cruciferae). Proc. R. Soc. Lond. Ser. B. 203, 387–403. Cecchi L., Gabbrielli R., Arnetoli M., Gonnelli C., Hasko A. and ranean Sea repeatedly desiccated partly or almost entirely Selvi F. 2010 Evolutionary lineages of nickel hyperaccumulation (Rodriguez-Sanchez et al. 2008). This period was accom- and systematics in European Alysseae (Brassicaceae): evidence panied by climatic aridification (Krijgsman 2002)andwas from nrDNA sequence data. Ann. Bot. 106, 751–767. followed by the establishment of the Mediterranean climate Couvreur T. L. P., Franzke A., Al-Shehbaz I. A., Bakker F. T., Koch in the Pliocene and subsequent climatic oscillation during M. A. and Mummenhoff K. 2010 Molecular phylogenetics, tem- poral diversification, and principles of evolution in the mustard Pleistocene (Yokoyama et al. 2000). As a consequence of family (Brassicaceae). Mol. Biol. Evol. 27, 55–71. climatic aridification, arid/semiarid areas were established in Cox C. B. and Moore P. D. 2010 Biogeography: an ecological and Mediterranean area, South Africa, North and South Amer- evolutionary approach, 8th edition. John Wiley, Hoboken, USA. ica, Central Asia and Australia (Mummenhoff et al. 1992; Doyle J. J. 1987 A rapid DNA isolation procedure for small Cox and Moore 2010). The distribution of Alyssum mainly quantities of fresh leaf tissue. Phytochem. Bull. 19, 11–15. Drummond A. J. and Rambaut A. 2007 BEAST: Bayesian evo- in these arid/semiarid areas would indicate that the Mediter- lutionary analysis by sampling trees. BMC Evol. Biol. 7, ranean climate establishment and aridification of the cli- 214. mate provided favourable conditions for the diversification Drummond A. J., Ho S. Y., Phillips M. J. and Rambaut A. 2006 in Europe and Turkey and the worldwide dispersal. Cli- Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, matic aridification and arid/semiarid areas established in the e88. Drummond A. J., Suchard M. A., Xie D. and Rambaut A. 2012 Pliocene/Pleistocene could have provided favourable condi- Bayesian phylogenetics with BEAUti and the BEAST 1.7. Mol. tions for the migration and diversification of Alyssum. Biol. Evol. 29, 1969–1973. Dudley T. 1964a Synopsis of the genus Alyssum. J. Arnold Arbor. 45, 358–372. Dudley T. 1964b Studies in Alyssum: Near Eastern representatives Acknowledgements and their allies. I. J. Arnold Arbor. Harv. Univ. 45, 57–100. Dudley T. 1968 Alyssum (Cruciferae) introduced in North America. This research was supported by the 100 Talents Projects of Chinese Rhodora 70, 298–300. Academy of Sciences. The authors would like to thank Xue Liang, Fedorov V. B. and Stenseth N. C. 2002 Multiple glacial refugia in Xiaoyan Shen and Xueliang Zhang for help in sample extraction. the North American Arctic: inference from phylogeography of The authors also thank the anonymous reviewers for helping to the collared lemming (Dicrostonyx groenlandicus). Proc. R. Soc. revise the paper. Lond. Ser. B. 269, 2071–2077. Franzke A., Lysak M. A., Al-Shehbaz I. A., Koch M. A. and Mummenhoff K. 2011 Cabbage family affairs: the evolutionary history of Brassicaceae. Trends Plant Sci. 16, 108–116. References Frenzel B. 1968 The Pleistocene vegetation of northern Eurasia. Science 161, 637–649. Abbott R. J., Smith L. C., Milne R. I., Crawford R. M., Wolff K. Galloway G. L., Malmberg R. L. and Price R. A. 1998 Phylogenetic and Balfour J. 2000 Molecular analysis of plant migration and utility of the nuclear gene arginine decarboxylase: an example refugia in the Arctic. Science 289, 1343–1346. from Brassicaceae. Mol. Biol. Evol. 15, 1312–1320. Allan H. H. 1961 Flora of New Zealand, vol. I, Owen, New Zealand. German D. A. and Al-Shehbaz I. A. 2010 Nomenclatural novelties Alshehbaz I. A. 1987 The genera of Alysseae (Cruciferae, Brassi- in miscellaneous Asian Brassicaceae (Cruciferae). Nordic J. Bot. caceae) in the Southeastern United-States. J. Arnold Arbor. 68, 28, 646–651. 185–240. Gernhard T. 2008 New analytic results for speciation times in Al-Shehbaz I. A. 2002 New species of Alyssum, Aphragmus, Ara- neutral models. Bull. Math. Biol. 70, 1082–1097. bis,andSinosophiopsis (Brassicaceae) from China and India. Hopkins D. M. 1967 The Bering land bridge. Vol. 3, Stanford Novon 12, 309–313. University Press, Redwood City, USA.

Journal of Genetics, Vol. 93, No. 2, August 2014 321 YanLietal.

Hultén E. 1937 Outline of the history of arctic and boreal Nylander J. 2004 MrModeltest v2. Program distributed by the biota during the quaternary period. Bokförlags Aktiebolaget author. Evolutionary Biology Centre, Uppsala University 2, Thule, Stockholm. Sweden. Kay K., Whittall J. and Hodges S. 2006 A survey of nuclear Nylander J. A., Ronquist F., Huelsenbeck J. P. and Nieves-Aldrey ribosomal internal transcribed spacer substitution rates across J. 2004 Bayesian phylogenetic analysis of combined data. Syst. angiosperms: an approximate molecular clock with life history Biol. 53, 47–67. effects. BMC Evol. Biol. 6, 36. O’Kane S. L. and Al-Shehbaz I. A. 2003 Phylogenetic position and Koch M. and Al-Shehbaz I. A. 2002 Molecular data indicate com- generic limits of Arabidopsis (Brassicaceae) based on sequences plex intra- and intercontinental differentiation of American Draba of nuclear ribosomal DNA. Ann. Mo. Bot. Gard. 90, 603–612. (Brassicaceae). Ann. Mo. Bot. Gard. 89, 88–109. Parks C. R. and Wendel J. F. 1990 Molecular Divergence between Koch M. and Al-Shehbaz I. A. 2004 Taxonomic and phylogenetic Asian and North-American species of Liriodendron (Magnoli- evaluation of the American ‘Thlaspi’ species: identity and rela- aceae) with implications for interpretation of fossil floras. Am. J. tionship to the Eurasian genus Noccaea (Brassicaceae). Syst. Bot. Bot. 77, 1243–1256. 29, 375–384. Parrish J. T. 1987 Global palaeogeography and palaeoclimate of Koch M., Mummenhoff K. and Hurka H. 1999a Molecular phy- the late cretaceous and early tertiary. In The origins of logenetics of Cochlearia (Brassicaceae) and allied genera based angiosperms and their biological consequences (ed. E. M. Friis, on nuclear ribosomal ITS DNA sequence analysis contradict tra- W. G. Chaloner and P. R. Grane) pp. 51–73. Cambridge Univer- ditional concepts of their evolutionary relationship. Plant Syst. sity Press, Cambridge. Evol. 216, 207–230. Payson E. 1926 The genus Thlaspi in North America. Univ. Wyo. Koch M., Bishop J. and Mitchell-Olds T. 1999b Molecular system- Pub. Sci. Bot. 1, 145–163. atics and evolution of Arabidopsis and Arabis. Plant Biol. 1, 529– Pole M. 1994 The New-Zealand flora – entirely long-distance 537. dispersal. J. Biogeogr. 21, 625–635. Koch M., Haubold B. and Mitchell-Olds T. 2001 Molecular system- Raven P. H. 1973 Evolution of subalpine and alpine plant groups in atics of the Brassicaceae: evidence from coding plastidic matK New Zealand. New Zeal. J. Bot. 11, 177–200. and nuclear Chs sequences. Am. J. Bot. 88, 534–544. Raven P. H. and Axelrod D. I. 1972 Plate tectonics and australasian Koch M. A. and Matschinger M. 2007 Evolution and genetic dif- paleobiogeography. Science 176, 1379–1386. ferentiation among relatives of Arabidopsis thaliana. Proc. Natl. Redenbach Z. and Taylor E. B. 1999 Zoogeographical implications Acad. Sci. USA 104, 6272–6277. of variation in mitochondrial DNA of Arctic grayling (Thymallus Koch M. A., Dobeš C. and Mitchell-Olds T. 2003 Multiple hybrid arcticus). Mol. Ecol. 8, 23–35. formation in natural populations: concerted evolution of the inter- Rodriguez-Sanchez F., Perez-Barrales R., Ojeda F., Vargas P. and nal transcribed spacer of nuclear ribosomal DNA (ITS) in North Arroyo J. 2008 The Strait of Gibraltar as a melting pot for plant American Arabis divaricarpa (Brassicaceae). Mol. Biol. Evol. 20, biodiversity. Quaternary Sci. Rev. 27, 2100–2117. 338–350. Rollins R. C. 1982 A new species of the asiatic genus Stroganowia Koch M. A., Karl R., Kiefer C. and Al-Shehbaz I. A. 2010 Coloniz- (Cruciferae) from North-America and its biogeographic implica- ing the American continent: Systematics of the genus Arabis in tions. Syst. Biol. 7, 214–220. North America (Brassicaceae). Am. J. Bot. 97, 1040–1057. Santos-Gally R., Vargas P. and Arroyo J. 2012 Insights into Krijgsman W. 2002 The Mediterranean: Mare Nostrum of Earth Neogene Mediterranean biogeography based on phylogenetic sciences. Earth Planet Sci. Lett. 205, 1–12. relationships of mountain and lowland lineages of Narcissus Kropf M. 2002 Vergleichende Biogeographie europäischer (Amaryllidaceae). J. Biogeogr. 39, 782–798. Gebirgspflanzen: Molekulare und morphometrische unter- Schulz O. 1936 Cruciferae. In Die natürlichen pflanzenfamilics suchungen an der montan-subalpinen’Anthyllis montana’L. (ed. A. Engler and H. Harms), vol 17B, edition 2, pp. 227–658. (’Fabaceae’) und der hochalpinen ’Pritzelago alpina’(L.) Verlag Von Wihelm Engelmann, Leipzeg, Germany. O. Kuntze (’Brassicaceae’). Doctorate dissertation, Mainz Tamura K., Peterson D., Peterson N., Stecher G., Nei M. and Kumar University. Cuvillier, Göttingen, Germany. S. 2011 MEGA5: molecular evolutionary genetics analysis using Mathews S. and McBreen K. 2008 Phylogenetic relationships of B- maximum likelihood, evolutionary distance, and maximum par- related phytochromes in the Brassicaceae: Redundancy and the simony methods. Mol. Biol. Evol. 28, 2731–2739. persistence of phytochrome D. Mol. Biol. Evol. 49, 411–423. Thompson J. D., Higgins D. G. and Gibson T. J. 1994 CLUSTAL Mengoni A., Baker A. J. M., Bazzicalupo M., Reeves R. D., W: improving the sensitivity of progressive multiple sequence Adiguzel N., Chianni E. et al. 2003 Evolutionary dynamics of alignment through sequence weighting, position-specific gap nickel hyperaccumulation in Alyssum revealed by ITS nrDNA penalties and weight matrix choice. Nucleic Acids Res. 22, 4673– analysis. New Phytol. 159, 691–699. 4680. Moazzeni H., Zarre S., Al-Shehbaz I. A. and Mummenhoff K. 2010 Tremblay N. and Schoen D. 1999 Molecular phylogeography of Phylogeny of Isatis (Brassicaceae) and allied genera based on Dryas integrifolia: glacial refugia and postglacial recolonization. ITS sequences of nuclear ribosomal DNA and morphological Mol. Ecol. 8, 1187–1198. characters. Flora 205, 337–343. Vanhouten W. H. J., Scarlett N. and Bachmann K. 1993 Nuclear- Mummenhoff K., Hurka H. and Bandelt H. J. 1992 Systematics of DNA markers of the Australian tetraploid Microseris-scapigera Australian Lepidium species (Brassicaceae) and implications for and its North-American diploid relatives. Theor. Appl. Genet. 87, their origin–evidence from Ief analysis of Rubisco. Plant Syst. 498–505. Evol. 183, 99–112. Wardle P. 1978 Origin of the New-Zealand mountain flora, with spe- Mummenhoff K., Franzke A. and Koch M. 1997 Molecular phy- cial reference to Trans-Tasman relationships. New Zeal. J. Bot. logenetics of Thlaspi sp (Brassicaceae) based on chloroplast 16, 535–550. DNA restriction site variation and sequences of the internal tran- Warwick S. I., Francis A. and Al-Shehbaz I. A. 2006 Brassicaceae: scribed spacers of nuclear ribosomal DNA. Can. J. Bot. 75, 469– Species checklist and database on CD-Rom. Plant Syst. Evol. 482. 259, 249–258. Mummenhoff K., Bruggemann H. and Bowman J. L. 2001 Chloro- Warwick S. I., Sauder C. A. and Al-Shehbaz I. A. 2008 Phyloge- plast DNA phylogeny and biogeography of Lepidium (Brassi- netic relationships in the tribe Alysseae (Brassicaceae) based on caceae). Am. J. Bot. 88, 2051–2063. nuclear ribosomal ITS DNA sequences. Botany 86, 315–336.

322 Journal of Genetics, Vol. 93, No. 2, August 2014 ITS for the phylogenic relationship of the genus Alyssum

White T. J., Bruns T., Lee S. and Taylor J. 1990 Ampliﬁca- Yue J. P., Sun H., Baum D. A., Li J. H., Al-Shehbaz I. A. and Ree tion and direct sequencing of fungal ribosomal RNA genes R. 2009 Molecular phylogeny of Solms-laubachia (Brassicaceae) for phylogenetics. In PCR protocols: a guide to methods s.l., based on multiple nuclear and plastid DNA sequences, and applications (ed. M. A. Innis, D. H. Gelfand, J. J. Sninsky and its biogeographic implications. J. Syst. Evol. 47, 402– and T. J. White), pp. 315–322. Academic Press, New York, USA. 415. Yokoyama Y., Lambeck K., De Deckker P., Johnston P. and Fiﬁeld Yule G. U. 1925 A mathematical theory of evolution, based on the L. K. 2000 Timing of the Last Glacial Maximum from observed conclusions of Dr. J C Willis, FRS. Phil. Trans. R. Soc. Lond. B. sea-level minima. Nature 406, 713–716. Biol. Sci. 213, 21–87.

Received 21 November 2013, in revised form 21 January 2014; accepted 26 February 2014 Published on the Web: 5 July 2014

Journal of Genetics, Vol. 93, No. 2, August 2014 323