Harmful Algae 27 (2013) 130–137

Discrimination of the common macroalgae (Ulva and Blidingia) incoastal waters of Yellow Sea, northern China, based on restrictionfragment-length polymorphism (RFLP) analysis

Jie Xiao a, Yan Li a,*, Wei Song a,b, Zongling Wang a, Mingzhu Fu a, Ruixiang Li a,Xuelei Zhang a, Mingyuan Zhu a

a The First Institute of Oceanography, State Oceanic Administration, Qingdao 266061, PR Chinab Hunan Agricultural University, Furong District, Changsha 410128, PR China

A R T I C L E I N F O

Article history:

Received 11 March 2013

Received in revised form 14 May 2013

Accepted 14 May 2013

Keywords:

Green tides

Ulva prolifera

Genetic identification

RFLP

Porphyra aquaculture

Yellow Sea

A B S T R A C T

Since 2007, reoccurring large-scale green algae blooms have caused deleterious effects to the estuarine

ecosystem of Yellow Sea, northern China and subsequent economical losses. Previous surveys indicated

the green tides were initiated in the coastal water of southern Jiangsu province where Porphyra farming

was intensively conducted; however, the main ‘seed source’ of floating green algae is still under debate.

Ulva prolifera was confirmed to be the major causative species of green tides. The multiple sympatric

ulvoid species in the natural environment has complicated species identification in both field surveys

and laboratory studies due to their morphological plasticity. Thus, we developed a genetic identification

key based on restriction fragment length polymorphism (RFLP) analysis of the ITS nuclear marker to

discriminate the common Ulva and Blidingia species in the Yellow Sea. Ten genetic lineages (1 in Blidingia,

9 in Ulva) were detected along the coast of China through phylogenetic analysis of ITS sequences. They

can be separated by virtual restriction digestion using the four selected restriction enzymes (BspT107 I,

EcoO109 I, Hin1 I and VpaK11B I). With additional PCR amplification of the 5S spacer region, we were

able to discriminate U. prolifera from Ulva linza. Using this genetic key, we screened macroalgal samples

collected from the coast of the Yellow Sea, and the results indicated 6 common lineages (U. prolifera, U.

linza, Ulva compressa, Ulva pertusa, Clade 6 and Blidingia sp.) in this region, which could be explicitly

distinguished by a single enzyme (BspT107 I) coupled with 5S spacer polymorphism. U. prolifera was

confirmed to be present on the Porphyra aquaculture rafts with seasonal variation in the species

composition. This genetic key will facilitate our long-term field surveys to investigate the origin of the

floating U. prolifera and furthermore to explore its bloom dynamics along the coast of the Yellow Sea. It

also provided a framework for the future inclusion of more Ulva species, which will expand the usage of

this key.

� 2013 Elsevier B.V. All rights reserved.

Contents lists available at SciVerse ScienceDirect

Harmful Algae

jo u rn al h om epag e: ww w.els evier .c o m/lo cat e/ha l

1. Introduction

Large-scale ‘green tides’ reoccurred every year since 2007 alongthe coast of the Yellow Sea, China. It has been claimed to be theworld’s largest ‘green tide’ in terms of affected area and biomass itproduced (Liu et al., 2009, 2010a). In 2008, the accumulatedbiomass of green macroalgae was estimated to be 3 million tonsand covered an area of 1.29 � 104 km during its peak (Sun et al.,2008; Keesing et al., 2011; Lin et al., 2011). Such large-scale blooms

* Corresponding author at: First Institute of Oceanography, State Oceanic

Administration, 6 Xianxialing Road, Qingdao 266061, PR China.

Tel.: +86 532 88967451; fax: +86 532 88967451.

E-mail address: liyan@fio.org.cn (Y. Li).

1568-9883/$ – see front matter � 2013 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.hal.2013.05.003

have caused serious detrimental ecological and social-economicimpacts to the adjacent coastal areas (Sun et al., 2008; Ye et al.,2011). The causative green tide genus was identified to be Ulva, ofwhich U. prolifera was dominant (Leliaert et al., 2009; Liu et al.,2010b; Pang et al., 2010; Duan et al., 2012; Shen et al., 2012).Satellite remote sensing and physical oceanographical modelingindicated that U. prolifera floating mats were repeatedly formed inthe near-shore water of southern Jiangsu province where Porphyra

yezoensis1 aquaculture was most intensive (Sun et al., 2008; Hu,2009; Keesing et al., 2011; Liu et al., 2010a), and transportednorthward by seasonal winds and surface currents (Lee et al., 2011).

1 Porphyra yezoensis was re-classified to Pyropia yezoensis based on a recent

molecular phylogenetic work (Sutherland et al., 2011).

Fig. 1. Map of sample locations along the coast of the Yellow Sea. Dashed line

indicated the area of recent expansion of Porphyra aquaculture. See Table 2 for site

abbreviations.

J. Xiao et al. / Harmful Algae 27 (2013) 130–137 131

However, there is still debate regarding the ‘origin’ or ‘source’ of thefloating U. prolifera.

The taxonomy of the Ulvaceae, especially the genus Ulva, hasfollowed a tortuous history (Tan et al., 1999; Woolcott and King,1999; Hayden and Waaland, 2002; Hayden et al., 2003). Thetubular Enteromorpha and lettuce-like Ulva with distromaticblades were originally recognized as two different generaaccording to their distinct morphological differences. Recentresearch, however, indicated that they were not evolutionarilydistinct entities based on the molecular phylogenetic analyses,and it was suggested that Enteromorpha should be synonymizedwith Ulva (Tan et al., 1999; Hayden et al., 2003. Only Ulva is usedin this paper). Several Ulva species have long been recognized inChina and they have co-existed ubiquitously in the Yellow Seawith no evident harm to the local habitats (Tseng, 1984).Discrimination of these co-occurring macroalgae based on grossmorphology is difficult due to their large intra-specificmorphological variations and relatively few inter-specificdifferences (Mathieson et al., 1981; Blomster et al., 1998,1999, 2002).

Molecular analysis has been applied to distinguish specieswithin Ulva since large-scale blooms occurred in the Yellow Sea(Shimada et al., 2003, 2008; Hiraoka et al., 2011; Duan et al., 2012).The dominating species of macroalgal blooms was confirmed to beUlva prolifera (Leliaert et al., 2009; Wang et al., 2010; Zhao et al.,2012). Previous molecular studies, however, mainly relied onsequencing and phylogenetic analysis which requires tediouslaboratory work and complicated program computation. Thoseassays were not readily accessed and not economically applicablefor large-scale, long-term surveys that usually come along withlarge number of field samples. Thus, the purpose of this study wasto: (1) investigate the number of common macroalgal species (orgenetic lineages) in our regional survey in the Yellow Sea, based onthe available molecular data; (2) develop a reliable and easy to usegenetic identification key based on restriction fragment lengthpolymorphism (RFLP) analysis of the ITS nuclear marker; and (3)analyze the species composition of the attached and floatingmacroalge from Porphyra aquaculture rafts and near-shorehabitats in the Yellow Sea.

2. Materials and methods

2.1. Sample collection

Macroalgae samples were collected from a Porphyra aquacul-ture area along the coast of Jiangsu province and intertidal areasalong the coast of Qingdao (Fig. 1). Attached macroalgae wassampled monthly from the Porphyra rafts at three stations (XA,Xiaoyangkou; GA, Gaoni; NA, Niluosha; Fig. 1) except inNovember of 2011, January, February and August of 2012, dueto adverse weather conditions. Floating samples were collected inJune, July, September and October in the waters near the Porphyra

aquaculture area (Fig. 1). Two additional samples (GS, NS) werecollected from macroalge clumps settled on the muddy flat atGaoni and Niluosha. These macroalge clumps were most likely leftby Porphyra farmers when they cleaned the rafts after thePorphyra harvest season (Liu et al., 2010a; Fan et al., personalcommunication).

For comparison, we also sampled attached macroalgae at theintertidal rocky and sandy beaches of Qingdao in July. Further-more, we noticed that floating Ulva spp. bloomed frequently at amuch smaller scale (usually <100 m2) along the recreationalbeaches in Qingdao in summer. These blooms usually occurredlocally (i.e. not drifting to open area from offshore large-scale U.

prolifera floating mats). To investigate and compare the speciescomposition of these small-scale blooms with the large-scale more

offshore green tides, we sampled these local floating macroalgae inJune as well.

Fresh macroalgal samples were sealed in the plastic bags andtransported to the laboratory in coolers. Individual plants werecarefully separated after the macroalge clumps being soaked indistilled (DI) water. The plants were examined carefully anddivided into several groups based on their gross morphology(Tseng, 1984; Hayden et al., 2003). Thalli of two to five plants foreach morphological group were randomly taken for DNA extrac-tion, and the rest of the sample was weighted, dried at roomtemperature and stored at 4 8C for further research.

2.2. DNA extraction and PCR amplification

Before DNA extraction, individual thalli were rinsed three timeswith DI water, frozen and thawed twice and homogenized withpellet pestles. Then genomic DNA was extracted using E.Z.N.A.TM

HP plant DNA kits (OMEGA Bio-tek Inc., GA, USA) following themanufacture’s protocol.

Polymerase chain reaction (PCR) amplifications of ITS rRNAgene (including ITS1, 5.8S, ITS2 regions) were performed as inLeskinen and Pamilo (1997) and Hayden et al. (2003) using theprimer pair: forward 50-TCGTAACAAGGTTTCCGTAGG-30, reverse50-GCTTATTGATATGCTTAAGTTCAGCGGGT-30. Amplicons werethen electrophoresed on 1% agarose gels and visualized afterethidium bromide staining.

In order to distinguish Ulva prolifera and Ulva linza which couldnot be separated by ITS sequence polymorphism (Leliaert et al.,2009), the 5S spacer regions were amplified following the protocolof Yotsukura et al. (2002) and Shimada et al. (2008) using theprimer pair: forward 50-GGTTGGGCAGGATTAGTA-30 and reverse50-AGGCTTAAGTTGCGAGTT-30. The PCR products were separatedon a 2% agarose gel. Based on the sequence analyses on bothsequences downloaded from GeneBank database and obtained inthis study for U. prolifera and U. linza, length polymorphism wasevident and could be used to differentiate these two taxa (Fig. 2).

Fig. 2. RFLP patterns of ITS genes using the four restriction enzymes (A and B) and

length polymorphisms of the amplified 5S spacer region (C). (A) Lanes 1, 8 and 15,

100 bp ladder with fragment sizes marked aside; lanes 2–7, BspT107 I digestion

patterns of U. prolifera, Clade 6, U. linza, U. compressa, U. pertusa and Blidingia sp.;

lanes 9–14, EcoO109 I digestion patterns of the same six samples as BspT107 I. (B)

Lanes 1, 8 and 15, 100 bp ladder; lanes 2–7 and 9–14 were RFLP patterns for the

same six samples digested by VpaK11B I (lanes 2–7) and Hin1 I (lanes 9–14),

respectively. (C) Lane 1, ladder; lanes 2–5, PCR amplification of 5S spacer regions for

two U. prolifera (lanes 2 and 3) and two U. linza samples (lanes 4 and 5); lane 6,

negative control with ddH2O added to the PCR reaction mixture.

2 U. pertusa was currently considered as a synonym of U. australis (Kraft et al.,

2010).

J. Xiao et al. / Harmful Algae 27 (2013) 130–137132

2.3. Cloning and sequencing

PCR products from the ITS region were directly sequenced atboth directions by Majorbio Bio-pharm Technology Co., Ltd(Shanghai, China). In order to evaluate potential intra-individualnucleotide variations for this multi-copy gene polymorphism ofamplicons, some PCR products were ligated with pMD118-Tvectors (TaKaRa Co., Dalian, China) and transformed into DH5@competent cells (TaKaRa) following the manufacture’s instruc-tions. Two to four clones with inserts were screened andsequenced. For 5S spacer amplicons, the shortest bands wereexcised by E.Z.N.A.TM gel extraction kits (OMEGA Bio-tek Inc.), thencloned and sequenced as above.

2.4. Phylogenetic analysis

The ITS gene sequences of the genera Ulva and Blidingia

identified along the coasts of China were retrieved from theGenebank database. The downloaded sequences and those

obtained from this study were aligned using ClustalW algorithm(Thompson et al., 1994) in MacVector 12.6 (Accelrys, CA, USA).Maximum-likelihood (ML) and neighbor-joining (NJ) analyseswere performed using MEGA 5 (Tamura et al., 2011). NJ trees wereestimated with maximum likelihood distances. The general timereversible model (GTR) and gamma distribution with invariantsites (G+I) was set for the ML tree searches based on the result ofnucleotide substitution model testing. The robustness of nodeswas assessed by bootstrapping for 1000 replicates.

2.5. RFLP analysis

Based on the results of phylogenetic analysis, ITS sequenceswere subject to virtual restriction digestion (VRED) usingrestriction enzymes provided by TaKaRa Bio. Inc. (Dalian, China).Restriction enzymes were selected based on the criteria ofmaximizing the number of separated sequence groups, minimizingintra-group variation and the distinguishableness of the resultedfragments by agarose gel electrophoresis. Priority was given toseparate the genotypes commonly observed in our survey region(Yellow Sea, China).

The restriction fragment length polymorphism (RFLP) patternsof ITS amplicons were assessed by agarose gel electrophoresis.Specifically, the PCR products were digested with the enzymesselected by the method described above following the manufac-tor’s instruction. Digested PCR products were then electrophoresedon 3% agrose gels and visualized under the UV light after EtBrstaining. Samples identified by sequencing were treated asreferences for clade designation. Digestion patterns of each samplewere compared with the references.

3. Results

3.1. Sequence phylogeny of ITS and 5S spacer regions

A total of 176 ITS sequences (Table S1) were detected along thecoast of China, including 138 downloaded from GeneBank (up to10/15/2012) and 38 from this study, formed 11 clades, indicating11 genetic lineages in genera Ulva and Blidingia. All Blidingia

sequences were grouped together and formed a monophyleticclade with 100% support which was a sister to the big Ulva clade.Within the Ulva clade, the sequence GU266276 was located at thebase and formed Clade 7 by itself. A large number of ITS sequences(101) fell into a moderately supported (69% for NJ and 84% for ML)U. prolifera/linza clade, which contained two morphologicallydifferentiated groups Ulva prolifera and Ulva linza. Two morpho-logically distinct taxa, Ulva compressa and Ulva pertusa,2 formedtwo highly supported (100% and 99%) monophyletic clades,respectively. Another well-supported Clade 6 consisted ofsequences originally identified as U. prolifera, Ulva flexuosa or U.

linza. The remaining 14 sequences could be separated into 5 groups(Clades 1–5), which were nested with Clade 6. The taxonomic statusof the 7 unnamed clades (labeled as Clades 1–7) was not fullyresolved.

Supplementary material related to this article found, in theonline version, at http://dx.doi.org/10.1016/j.hal.2013.05.003.

The 38 ITS sequences from this study were distributed in 5clades: U. prolifera/linza, Clade 6, Ulva compressa, Ulva pertusa andBlidingia sp., which was consistent with previous reports (Duanet al., 2012; Shen et al., 2012). Most of the 124 sequences known tobe from the Yellow Sea, including those from this study and otherreports, fell into the five clades listed above. A few other sequencesamples (HM046597, HM031156 and GQ202118), detected from

Fig. 3. Phylogenetic tree determined by analysis of ITS sequences of Ulva and

Blidingia species observed along the coast of China. Numbers at the nodes were

bootstrapping support values larger than 50% after 1000 replicates in neighbor-

joining (above) and maximum-likelihood (below) analyses. The inferred clades

were labeled.

J. Xiao et al. / Harmful Algae 27 (2013) 130–137 133

the coast of Yellow Sea, fell to the outside (Clade 5 and Clade 2) of thefive major clades. The number of these sequences was small and wedid not observe the similar genotypes in our following survey usingthe RFLP assay, indicating the minority of these genotypes in thesurvey region. The sequences presumably assigned as U. prolifera

were allocated to three clades (U. prolifera/linza, Clades 7 and 6) withthe majority in the U. prolifera/linza group. Morphology andmolecular analyses indicated that ITS sequences of U. prolifera weretangled with those from U. linza, and the two taxa were notdifferentiated solely by ITS sequences (Shimada et al., 2008; Leliaertet al., 2009). The ‘U. prolifera’ specimen in Clades 6 and 7 perhapsrepresented the Ulva strains that were morphologically similar butgenetically different from the true U. prolifera.

Phylogeny of the 5S spacer region separated the 72 sequences(including 50 downloaded from the GeneBank and 22 from thisstudy) into two distinct clades with 100% support (Fig. 4), one forUlva prolifera and the other for Ulva linza. The alignment of 5Sspacer sequences for U. prolifera and U. linza revealed an indel ofabout 121 bp in length, which resulted in length polymorphism forthese two groups of sequences. There were also two short indels(53 bp and 9 bp in length, respectively) within the U. prolifera

group of sequences, which led to variation in the U. prolifera

sequences. In general, the shortest 5S spacer fragments of U.

prolifera varied between 455 and 393 bp in length (with primers),while the fragments for U. linza were shorter (about 338 bp inlength). Therefore, we were able to differentiate these two speciesbased on the length polymorphism of 5S spacer amplicons asshown in Fig. 2.

3.2. RFLP analysis of ITS fragments

The ML and NJ analyses grouped the ITS sequences into 11groups indicating 11 genetically differentiated lineages (Fig. 3, alsosee above). Sequences of 10 clades were subject to virtualrestriction digestion (VRED). While the sequence GU266276formed Clade 7 by itself, it was not analyzed by VRED becausewe could not generate the full sequence covering the primerregions due to its short length and unalignableness of both endswith other Ulva ITS sequences. Four restriction enzymes (BspT107I, EcoO109 I, Hin1 I and VpaK11B I) were chosen based on thecriteria addressed above. The VRED patterns of the 10 geneticlineages in Ulva and Blidingia found along the coasts of China wereanalyzed and listed in Table 1. As indicated in Table 1, each of thesix lineages (U. prolifera/linza, Ulva compressa, Ulva pertusa, Clades

1, 3 and 4) exhibited one single pattern across all four enzymes,while intra-clade variations existed for the other 3 clades (Blidingia

sp., Clades 5 and 6). In total, there were 15 alleles from 10 cladesdetected by VRED analysis (Table 1).

Since we surveyed the major green tide area in the Yellow Sea,priority was given to distinguish the 5 lineages involving thecommon alleles detected in this region which were U. prolifera/

linza, U. compressa, U. pertusa, Blidingia sp. and Clade 6. The firstenzyme BspT107 I was able to explicitly discriminate the 5 lineages(marked with b in Table 1). If we included all the alleles detectedalong the coasts of China for the purpose of expanding our surveyregion in the future, ambiguities were found between Allele B vs. I,and among alleles K, L, M and N. These alleles could be furtherseparated by the other three enzymes. Allele B and I could befurther differentiated by EcoO109 I or VpaK11B I, while separationof Allele K, L, M and N required the combination of either EcoO109I + Hin1 I, or EcoO109 I + VpaK11B I after BspT107 I digestion. Ingeneral, a one-step restriction enzyme digestion (BspT107 I) of ITSproducts was sufficient to distinguish the common Ulva andBlidingia macroalgae in the Yellow Sea. However, for those alongthe coasts of China, the use of more enzymes is probably needed(Fig. 4).

Table 1Allele sizes and restriction fragment-length polymorphism (RFLP) patterns after restriction enzyme digestions of amplified ITS fragments for 15 alleles of 10 lineages in Ulva

and Blidingia detected in the coastal water of China.

Clade designation Allele Total size (bp)a RFLP patterns (bp)a

BspT107 I EcoO109 I Hin1 I VpaK11B I

U. prolifera/linza Ab 630 444, 186 101, 529 437, 8, 185 420, 210

U. compressa Bb 638 76, 469, 93 167, 471 77, 26, 535 431, 15, 192

U. pertusa Cb 621 149, 472 621 150, 379, 92 404, 217

Blidingia sp. Db 645 135, 222, 157, 131 212, 433 136, 415, 94 434, 211

E 623 350, 142, 131 205, 418 99, 430, 94 423, 200

Clade 3 F 664 664 76, 99, 489 664 664

Clade 2 G 622 529, 93 134, 15, 396, 77 622 134, 6, 270, 212

H 627 523, 104 129, 15, 395, 88 627 129, 6, 492

Clade 1 I 660 92, 19, 457, 92 660 112, 548 76, 76, 30, 296, 182

Clade 4 J 654 172, 388, 94 654 145, 28, 389, 93 188, 272, 10, 184

Clade 5 K 653 653 576, 77 561, 92 437, 216

L 630 630 135, 417, 78 537, 93 105, 50, 258, 217

Clade 6 Mb 635 635 157, 400, 78 635 113, 315, 207

N 634 634 156, 278, 122, 78 634 114, 313, 207

O 633 122, 511 153, 480 540, 93 170, 256, 207

a Total sizes and RFLP fragments are based on the consensus sequences with primers included for each allele. Band sizes in RFLP patterns were listed according to the order

of restriction sites on the sequences.b The common alleles observed in the Yellow Sea by this survey and the previous reports.

J. Xiao et al. / Harmful Algae 27 (2013) 130–137134

Furthermore, the digestion pattern of BspT107 I for the U.

prolifera/linza lineage was distinct from all the other alleles.Therefore, a one-step digestion by BspT107 I could sort out thesamples of U. prolifera and Ulva linza which were undistinguishableby ITS fragments (Leliaert et al., 2009). Amplification of the 5S

Table 2Species diversity of the green macroalgae collected from the Porphyra aquaculture are

Location Code Sampling

date

Sample

type

Species

U. prolifera

Xiao Yang Kou, Rudong,

Jiangsu province

XA1211 12/2011 Attached

XA0312 03/2012 Attached +

XA0412 04/2012 Attached +

XA0512 05/2012 Attached ++

X1F0612 06/2012 Floating +++

X1F0912 09/2012 Floating +++

XA1012 10/2012 Attached +++

Gao Ni, Rudong,

Jiangsu province

GA1211 12/2011 Attached

GA0312 03/2012 Attached +

GA0412 04/2012 Attached ++

GA0512 05/2012 Attached ++

G0612 06/2012 On muddy flat ++

G1F0612 06/2012 Floating +++

G3F0912 09/2012 Floating +++

G7F0912 09/2012 Floating +++

GA1012 10/2012 Attached

Ni Luo Sha, Rudong,

Jiangsu province

NA1211 12/2011 Attached

NA0312 03/2012 Attached

NA0412 04/2012 Attached +

NA0512 05/2012 Attached ++

N1F0612 06/2012 Floating +++

N3F0612 06/2012 Floating +++

N0612 06/2012 On muddy flat +++

N4F0612 07/2012 Floating +++

NA0712 07/2012 Attached ++

NA0912 09/2012 Attached ++

N3F0912 09/2012 Floating +++

N6F0912 09/2012 Floating +++

NA1012 10/2012 Attached +++

Qingdao, Shandong

province

QDA0712 7/2012 Attached

QDF0712 07/2012 Floating locally

QDF0612 06/2012 Floating +++

+, the type specimen was present but rare (<10%); ++, the type specimen was commo

spacer region was necessary to separate U. prolifera and U. linza.Therefore, we carried out a two-step procedure to identify U.

prolifera, the green-tide causative species, which was PCR-RFLP(BspT107 I) analysis on ITS fragments followed by lengthpolymorphisms of 5S spacer amplicons.

a in the coast water of southern Jiangsu province and from the Qingdao coast.

U. linza Clade 6 U. compressa B. sp. U. pertusa Total # of species

+ +++ 2

++ +++ + ++ 5

++ + ++ 4

++ 2

1

1

++ 2

+++ + 1

++ +++ ++ 4

+++ + + 3

++ ++ 2

++ 2

1

1

1

++ ++ 2

++ 1

++ +++ ++ 3

++ +++ ++ 4

++ ++ 3

1

1

++ 2

1

++ 2

+ ++ 3

1

1

++ 2

++ ++ ++ 3

++ ++ ++ 3

1

n (10–50%); +++, the type specimen was abundant (>50%).

Fig. 4. Neighbor-joining tree determined by analysis of 5S spacer sequences of U.

linza and U. prolifera observed along the China’s coasts including those retrieved

from the GeneBank database (listed as accession numbers) and from this study. The

bootstrapping support values larger than 50% after 1000 replicates were given aside

each node. Sample abbreviations were as indicated in Table S1 and the inferred

species clades were labeled.

J. Xiao et al. / Harmful Algae 27 (2013) 130–137 135

3.3. Species composition in coastal waters of Jiangsu and Qingdao

The species composition of the macroalgal samples collectedfrom the Porphyra rafts along the coast of Jiangsu province, and thefloating green algae in the water nearby (Fig. 1) were assessedusing the method described above. The monthly survey indicatedhigh species diversity of attached macroalgae on the Porphyra raftswhich consisted of Ulva prolifera, Ulva linza, Ulva compressa,Blidingia sp. and an unknown genotype resembling Clade 6 in thephylogenetic tree (Table 2). U. prolifera was absent or rare on thePorphyra rafts in winter at all three locations (XA, GA, NA), whereasits abundance started to increase in spring (March 2012), andbecame common in summer and fall (May–October 2012). U. linza,Clade 6 and Blidingia sp. were commonly present throughout theyear, while U. compressa was relatively rare. By contrast, thefloating samples exhibited extremely low species diversity withone single species, U. prolifera, being dominant.

The macroalgal samples collected from Qingdao coastal watersduring large-scale more offshore green-tide blooms was mainlydominated by Ulva prolifera. The macroalgae floating locally near

beaches were comprised of different Ulva spp. (U. linza, U.

compressa and U. pertusa) compared to those samples from thefurther south along the coast of Jiangsu province and the large-scale green-tide samples. These three species (U. linza, U. compressa

and U. pertusa) were also common in the attached samplescollected from the intertidal zone along the coast of Qingdao. Thespecies discrepancy between these Qingdao local sample andlarge-scale floating mats indicated the un-indigenous origin of thefloating macroalgae, which was consistent with previous reports(Liu et al., 2010b; Duan et al., 2012).

4. Discussion

Based on phylogenetic analysis ITS nrDNA sequences, wedeveloped a PCR-RFLP genetic identification key that candiscriminate Ulva and Blidingia macroalgal species commonlyobserved along the coast of China. The five common lineages (U.

prolifera/linza, Ulva compressa, Ulva pertusa, Blidingia sp. and Clade

6) in the Yellow Sea could be explicitly separated by a one-stepPCR-RFLP analysis on ITS fragments using BspT107 I. Withadditional amplification of 5S spacer region, U. prolifera, the maincausative species of green tides in the Yellow Sea, can besuccessfully differentiated from the sympatric and geneticallyclosely related species, U. linza. Use of additional enzymes(EcoO109 I, Hin1 I and VpaK 11B I) could further separate moreUlva lineages (Clades 1–5). Furthermore, this genetic identificationkey established an expandable framework for future additions ofmore Ulva species to this key. Initial testing using this assay provedits use in identifying the common attached and floating Ulvaceanalgae around the Porphyra farming area in the coastal waters ofsouthern Jiangsu province and Qingdao. This reliable and easy touse assay will facilitate continuing studies of the origin and bloomdynamics of the floating U. prolifera in coastal water of the YellowSea.

Ulva prolifera, a ubiquitous green seaweed along the coast ofChina, was found to be responsible for the successive large-scalegreen tides in the Yellow Sea in recent years (Leliaert et al., 2009;Liu et al., 2010a; Keesing et al., 2011; Zhao et al., 2012). Great efforthas been attributed to identify the ‘seed source’ of the floating U.

prolifera in order to better understand the bloom dynamics andtherefore to propose appropriate management to control ormitigate the blooms. Satellite image analyses and field observa-tions indicated that green tides were consistently initiated fromthe near-shore water of Jiangsu province, southeastern Yellow Sea,and then transported northwest by surface winds and currents(Sun et al., 2008; Liu et al., 2009, 2010a; Keesing et al., 2011; Qiao etal., 2011; Ye et al., 2011). Thus, Liu et al. (2009, 2010a), Hu et al.(2010) and Keesing et al. (2011) ascribed the cause of the large-scale green tides to the unhygienic of waste treatment and recentexpansion of Porphyra aquaculture. Several other surveys, howev-er, found none or few U. prolifera on the Porphyra rafts (Pang et al.,2010; Shen et al., 2012) and they hypothesized that U. prolifera mayoriginate from land-based mariculture ponds, or directly growfrom the somatic cells, vegetative fragments or micro-propagulesin the coastal sediments etc. (Pang et al., 2010; Zhang et al., 2010,2011; Liu et al., 2012). However, we conducted continuous fieldsurveys in the Yellow Sea and found that U. prolifera wascommonly present on Porphyra rafts, and the Ulva biomass startedincreasing in spring and became abundant in summer and fall. Thediscarded macroalgal clumps settling on the muddy flat alsocontained a high proportion of U. prolifera (Liu et al., 2010a; thisstudy), which readily served as a source of floating U. prolifera.Some other studies reported different genotypes between floatingU. prolifera and those attached on Porphyra rafts (Zhao et al., 2011;Zhang et al., 2011). Their conclusion, however, was weakened bythe small sample size compared to the large macroalgae biomass

J. Xiao et al. / Harmful Algae 27 (2013) 130–137136

and high species diversity on Porphyra rafts. And a recent studyconfirmed the presence of the same ‘floating U prolifera genotypes’on the Porphyra rafts (Liu et al., 2013). Above all, a comprehensivesurvey on the attached and floating macroalgae was essential,considering the high species and genetic diversity both on thePorphyra rafts and in environmental waters and sediments, andtherefore the assay developed in this study has helped to completethese tasks.

Additionally, the genetic identification key developed in thisstudy established a framework for future inclusion of more Ulva

species along the coast of China. The phylogenetic analysis of ITSsequences indicated 11 genetically different lineages, with a fewun-designated clades/taxa. The total number of Ulva species alongthe coast of China was probably higher than that we estimatedhere since there is little molecular data for Ulva spp. in the coastalwater of southern China where many more Ulva spp. have beenrecorded (Tseng, 1984; Huang, 1994). Huang (1994) reported 25species in three genera in the family of Ulvaceae along the coasts ofChina (Huang, 1994). More recently, Ding et al. (2011) revised andreported three genera and 14 species in this family. Therefore, tobetter understand the taxonomy and distribution of Ulva spp.along the coast of China, a combination of morphology andmolecular analyses is needed. The genus Urospora (Ulotrichaceae,Chlorophyta) was reported in the Yellow Sea by some othersurveys (Tian et al., 2011; Shen et al., 2012). We excluded it fromour analyses considering the distinct morphological characteristicsand distant genetic relationships with the Ulvacea (Tseng, 1984).Some cultural specimens derived from the coast of Qingdao werefound to be close to Ulva rigida (AJ234319, AY260565) and Ulva

lactuca (AJ000208) in the ITS region (Liu et al., 2010b). They werenot analyzed in this study either, due to unavailability of thesesequences. In vitro digestion of the closest ITS sequences(AJ234319, AY260565, AJ000208) using the four selected restric-tion enzymes revealed distinct RFLP patterns compared to thegenotypes in Table 1, which implied they can also be identified bythe RFLP genetic key. However, this type of specimen was notdetected in our samples from Qingdao, which was likely due to ourinsufficient sampling, or limited knowledge regarding the distri-bution of this Ulva sp. Overall, the genetic identification key wascapable of discriminating common Ulva and Blidingia macroalgaein the Yellow Sea. Expand the usage of this key beyond the region ofYellow Sea, however, should be taken with cautious, especiallywhen we tended to apply it to other countries. With the assistanceof sequencing technology, the taxonomy of ulvoid macroalgaearound the world was undergoing numerous changes and waschallenged by the molecular species concept (Hayden and Waa-land, 2002; O’Kelly et al., 2010; Kraft et al., 2010). Currenttaxonomy of ulvoid macroalgae, which was mostly based on theEuropean flora, was found insufficient to cover all Ulva spp. world-widely, especially in tropical and sub-tropical coasts (O’Kelly et al.,2010; Kraft et al., 2010). To avoid all the confusions from thecontroversial taxonomy and nomenclature of many ulvoid species,we primarily focused on the Ulva spp. in Yellow Sea, northernChina in this study and the species names accepted by mostreferences were adopted here. A detailed phylogenetic analysis onthe Ulva spp. along the coasts of China required more extensivesampling and was in progress in our following project.

The macroalgae grouped in Clade 6 (Fig. 3) were commonlyobserved in the Yellow Sea (Pang et al., 2010; Tian et al., 2011;Duan et al., 2012; this study). Several studies assigned thisspecimen to Ulva flexuosa considering its high sequence similarityin the ITS region with U. flexuosa that was collected in Japan(AB097644–AB097647, Shimada et al., 2003) (Tian et al., 2011;Zhang et al., 2011; Shen et al., 2012; Liu et al., 2012). It was,however, insufficient to support that Clade 6 and the Japanese U.

flexuosa specimen (Shimada et al., 2003) was genetically close to

the European U. flexuosa from the herbarium and re-sampledspecimens at the historical localities (Tan et al., 1999; Mares et al.,2011). Thus, the true identity of Clade 6 in this study needs to befurther investigated. Using the RFLP identification key developedhere, we did not find the specimen resembled the European U.

flexuosa genotypes in our study region, although Zhang and Ye(unpublished) deposited one sequence (HM031156) found inChina which was close to the European U. flexuosa lineage.

Acknowledgements

The authors gratefully acknowledge the valuable help from thefollowing people during the collection and processing of thesamples: Dr Xiao Wang, Mr. Shiliang Fan, Xiangqing Liu, Pei Qu,Jintao Xu and Song Fang, and Ms. Xiaona Wang and Ping Sun. Wealso thank Drs. Lingyun Qu, Minggang Zheng and Fengrong Zhengfor assistance with molecular work. Special thanks to Paul J.Harrison and Caiwen Li for their critiques on the manuscript. Thiswork was supported by the National Basic Research Program ofChina (973 Program) award 2010CB428703, the National NaturalScience Foundation of China award 41276119, the InternationalScience & Technology Cooperation Program of China award2010DFA24340 and Qingdao Municipal Science and Technologyplan project (11-3-1-7-hy).[SS]

References

Blomster, J., Back, S., Fewer, D.P., Kiirikki, M., Lehvo, A., Maggs, C.A., Stanhope, M.J.,2002. Novel morphology in Enteromorpha (Ulvophyceae) forming green tides.American Journal of Botany 89, 1756–1763.

Blomster, J., Maggs, C.A., Stanhope, M.J., 1998. Molecular and morphological analy-sis of Enteromorpha intestinalis and E. compressa (Chlorophyta) in the Britishisles. Journal of Phycology 34, 319–340.

Blomster, J., Maggs, C.A., Stanhope, M.J., 1999. Extensive intraspecific morphologicalvariation in Enteromorpha muscoides (Chlorophyta) revealed by molecularanalysis. Journal of Phycology 35, 575–586.

Ding, L., Huang, B., Xie, Y., 2011. Advances and problems with the study of marinemacroalgae of China seas. Biodiversity Science 19, 798–804 (in Chinese withEnglish abstract).

Duan, W., Guo, L., Sun, D., Zhu, S., Chen, X., Zhu, W., Xu, T., Chen, C., 2012.Morphological and molecular characterization of free-floating and attachedgreen macroalgae Ulva spp. in the Yellow Sea of China. Journal of AppliedPhycology 24, 97–108.

Hayden, H.S., Waaland, J.R., 2002. Phylogenetic systematics of the Ulvacea (Ulvales,Ulvophyceae) using chloroplast and nuclear DNA sequences. Journal of Phycol-ogy 38, 1200–1212.

Hayden, H.S., Blomster, J., Maggs, C.A., Silva, P.C., Stanhope, M.J., Waaland, J.R., 2003.Linnaeus was right all along: Ulva and Enteromorpha are not distinct genera.European Journal of Phycology 38, 277–294.

Hiraoka, M., Ichihara, K., Zhu, W., Ma, J., Shimada, S., 2011. Culture and hybridizationexperiments on an Ulva clade including the Qingdao strain blooming in theYellow Sea. PLoS ONE 6 (5) e19371, http://dx.doi.org/10.1371/journal.-pone.0019371.

Hu, C., 2009. A novel ocean color index to detect floating algae in the global oceans.Remote Sensing of Environment 113, 2118–2129.

Hu, C., Li, D., Chen, C., Ge, J., Muller-Karger, F.E., Liu, J., Yu, F., He, M., 2010. On therecurrent Ulva prolifera blooms in the Yellow Sea and East China Sea. Journal ofGeophysical Research 115, C05017, http://dx.doi.org/10.1029/2009JC005561.

Huang, Z., 1994. Marine Species and Their Distributions in China’s Seas. China OceanPress, Beijing, pp. 222–223 (in Chinese).

Keesing, J.K., Liu, D., Fearns, P., Garcia, R., 2011. Inter- and intra-annual patterns ofUlva prolifera green tides in the Yellow Sea during 2007–2009, their origin andrelationship to the expansion of coastal seaweed aquaculture in China. MarinePollution Bulletin 62, 1169–1182.

Kraft, L.G.K., Kraft, G.T., Waller, R.F., 2010. Investigations into southern AustralianUlva (Ulvophyceae, Chlorophyta) taxonomy and molecular phylogeny indicateboth cosmopolitanism and endemic cryptic species. Journal of Phycology 46,1257–1277.

Lee, J.H., Pang, I.C., Moon, I.J., Ryu, J.H., 2011. On physical factors that controlled themassive green tide occurrence along the southern coast of the ShandongPeninsula in 2008: a numerical study using a particle-tracking experiment.Journal of Geophysical Research 116, C12036, http://dx.doi.org/10.1029/2011JC007512.

Leliaert, F., Zhang, X., Ye, N., Malta, E., Engelen, A.H., Mineur, F., Verbruggen, H.,Clerck, O.D., 2009. Identity of the Qingdao algal bloom. Phycological Research57, 147–151.

Leskinen, E., Pamilo, P., 1997. Evolution of the ITS sequences of ribosomal DNA inEnteromorpha (Chlorophyceae). Hereditas 126, 17–23.

J. Xiao et al. / Harmful Algae 27 (2013) 130–137 137

Lin, H., Jiang, P., Zhang, J., Wang, J., Qin, S., Sun, S., 2011. Genetic and marine cycloniceddy analyses on the largest macroalgal bloom in the world. EnvironmentalScience and Technology 45, 5996–6002.

Liu, D., Keesing, J.K., Dong, Z., Zhen, Y., Di, B., Shi, Y., Fearns, P., Shi, P., 2010a.Recurrence of the world’s largest green-tide in 2009 in Yellow Sea, China:Porphyra yezoensis aquaculture rafts confirmed as nursery for macroalgalblooms. Marine Pollution Bulletin 60, 1423–1432.

Liu, D., Keesing, J.K., Xing, Q., Shi, P., 2009. World’s largest macroalgal bloom causedby expansion of seaweed aquaculture in China. Marine Pollution Bulletin 58,888–895.

Liu, F., Pang, S., Chopin, T., Gao, S., Shan, T., Zhao, X., Li, J., 2013. Understanding therecurrent large-scale green tide in the Yellow Sea: temporal and spatial corre-lations between multiple geographical, aquacultural and biological factors.Marine Environment Research 83, 38–47.

Liu, F., Pang, S.J., Chopin, T., Xu, N., Shan, T.F., Qin, S., Sun, S., 2010b. The dominantUlva strain of the 2008 green algal bloom in the Yellow Sea was not detected inthe coastal waters of Qingdao in the following winter. Journal of AppliedPhycology 22, 531–540.

Liu, F., Pang, S.J., Zhao, X.B., Hu, C.M., 2012. Quantitative, molecular and growthanalyses of Ulva microscopic progagules in the coastal sediment of Jiangsuprovince where green tides initially occurred. Marine Environment Research 74,56–63.

Mares, J., Leskinen, E., Sitkowska, M., Skacelova, O., Blomster, J., 2011. True identityof the European freshwater Ulva (Chlorophyta, Ulvophyceae) revealed by acombined molecular and morphological approach. Journal of Phycology 47,1177–1192.

Mathieson, A.C., Norton, T.A., Neushul, M., 1981. The taxonomic implications ofgenetic and environmentally induced variations in seaweed morphology. Bo-tanical Review 47, 313–347.

O’Kelly, C.J., Kurihara, A., Shipley, T.C., Sherwood, A.R., 2010. Molecular assessmentof Ulva spp. (Ulvophyceae, Chlorophyta) in the Hawaiian island. Journal ofPhycology 46, 728–735.

Pang, S.J., Liu, F., Shan, T.F., Xu, N., Zhang, Z.H., Gao, S.Q., Chopin, T., Sun, S., 2010.Tracking the algal origin of the Ulva bloom in the Yellow Sea by a combination ofmolecular, morphological and physiological analyses. Marine EnvironmentResearch 69, 207–215.

Qiao, F., Wang, G., Lu, X., Dai, D., 2011. Drift characteristics of green macroalgae inthe Yellow Sea in 2008 and 2010. Chinese Science Bulletin 56, 2236–2242.

Shen, Q., Li, H., Li, Y., Wang, Z., Liu, J., Yang, W., 2012. Molecular identification ofgreen algae from the rafts based infrastructure of Porphyra yezoensis. MarinePollution Bulletin 64, 2077–2082.

Shimada, S., Hiraoka, M., Nabata, S., Lima, M., Masuda, M., 2003. Molecularphylogenetic analyses of the Japanese Ulva and Enteromorpha (Ulvales, Ulvo-phyceae), with special reference to the free-floating Ulva. PhycologicalResearch 51, 99–108.

Shimada, S., Yokoyama, N., Arai, S., Hiraoka, M., 2008. Phylogeography of the genusUlva (Ulvophyceae, Chlorophyta), with special reference to the Japanese fresh-water and brackish taxa. Journal of Applied Phycology 20, 979–989.

Sun, S., Wang, F., Li, C., Qin, S., Zhou, M., Ding, L., Pang, S., Duan, D., Wang, G., Yin, B.,Yu, R., Jiang, P., Liu, Z., Zhang, G., Fei, X., Zhou, M., 2008. Emerging Challenges:

Massive Green Algae Blooms in the Yellow Sea. Available from Nature Proceed-ings http://hdl.handle.net/10101/npre.2008.2266.1.

Sutherland, J.E., Lindstrom, S.C., Nelson, W.A., Brodie, J., Lynch, M.D., Hwang, M.S.,Choi, H.-G., Miyata, M., Kikuchi, N., Oliveira, M.C., Farr, T., Neffus, C., Mols-Mortensen, A., Milstein, D., Muller, K.M., 2011. A new look at an ancient order:generic revision of the Bangiales (Rhodophyta). Journal of Phycology 47,1131–1151.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using Maximum Likelihood, evolu-tionary distance, and maximum parsimony methods. Molecular Biology andEvolution 28, 2731–2739.

Tan, I.H., Blomster, J., Hansen, G., Leskinen, E., Maggs, C.A., Mann, D.G., Sluiman, H.J.,Stanhope, M.J., 1999. Molecular phylogenetic evidence for a reversible mor-phogenetic switch controlling the gross morphology of two common genera ofgreen seaweeds, Ulva and Enteromorpha. Molecular Biology and Evolution 16,1011–1018.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. Improving the sensitivity ofprogressive multiple sequence alignment through sequence weighting, posi-tions-specific gap penalties and weight matrix choice. Nucleic Acids Research22, 4673–4680.

Tian, X.L., Huo, Y.Z., Chen, L.P., et al., 2011. Molecular detection and analysis of greenseaweeds from Rudong coasts in Jiangsu Province (in Chinese). Chinese ScienceBulletin (Chinese Version) 56, 309–317 (in Chinese with English abstract).

Tseng, C.K., 1984. Common Seaweeds of China. Science Press, Beijing 316 pp.Wang, J., Jiang, P., Cui, Y., Li, N., Wang, M., Lin, H., He, P., Qin, S., 2010. Molecular

analysis of green-tide-forming macroalgae in the Yellow Sea. Aquatic Botany93, 25–31.

Woolcott, G.W., King, R.J., 1999. Ulva and Enteromorpha (Ulvales, Ulvophyceae,Chlorophyta) in Eastern Australia: comparison of morphological features andanalyses of nuclear rDNA sequence data. Australian Systematic Botany 12,709–725.

Ye, N., Zhang, X., Mao, Y., Liang, C., Xu, D., Zuo, J., Zhuang, Z., Wang, Q., 2011. ‘Greentides’ are overwhelming the coastline of our blue planet: taking the world’slargest example. Ecological Research 26, 477–485.

Yotsukura, N., Kawai, T., Motomura, T., Ichimura, T., 2002. Tandem 5S ribosomalRNA genes and the spacer region sequences of three Japanese Laminaria species.Journal of Applied Phycology 14, 233–239.

Zhang, X., Wang, H., Mao, Y., Liang, C., Zhuang, Z., Wang, Q., Ye, N., 2010. Somaticcells serve as a potential propagule bank of Enteromorpha prolifera forming agreen tide in the Yellow Sea, China. Journal of Applied Phycology 22, 173–180.

Zhang, X., Xu, D., Mao, Y., Li, Y., Xue, S., Zou, J., Lian, W., Liang, C., Zhuang, Z., Wang, Q.,Ye, N., 2011. Settlement of vegetative fragments of Ulva prolifera confirmed asan important seed source for succession of a large-scale green tide bloom.Limnology & Oceanography 56, 233–242.

Zhao, J., Jiang, P., Liu, Z., Wang, J., Cui, Y., Qin, S., 2011. Genetic variation of Ulva(Enteromorpha) prolifera (Ulvales, Chlorophyta) – the causative species of thegreen tides in the Yellow Sea, China. Journal of Applied Phycology 23, 227–233.

Zhao, J., Jiang, P., Liu, Z.Y., et al., 2012. The Yellow Sea green tides were dominated byone species, Ulva (Enteromorpha) prolifera, from 2007 to 2011. Chinese ScienceBulletin, http://dx.doi.org/10.1007/s11434-012-5441-3.