Ulva Prolifera O.F

Phycological Research 2009; 57: 147–151

RESEARCH NOTE

Identity of the Qingdao algal bloompre_532 147..151

Frederik Leliaert,1* Xiaowen Zhang,2 Naihao Ye,2 Erik-jan Malta,3 Aschwin H. Engelen,3 Frédéric Mineur,4 Heroen Verbruggen1 and Olivier De Clerck1 1Phycology Research Group, Biology Department, Ghent University, Krijgslaan 281 S8, 9000 Ghent, Belgium; 2Yellow Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, 106 Nanjing Road, 266071 Qingdao, China; 3ALGAE Research Group, CCMAR, CIMAR-Laboratório Associado, Universidade do Algarve, FCMA, Campus de Gambelas, 8005-139 Faro, Portugal; and 4School of Biological Sciences, Queen’s University Belfast, 97 Lisburn Road, Belfast BT9 7BL, UK

New York Times) simply did not name the alga, while SUMMARY other news agencies used more graphic terms such as ‘bright green slime’ (South China Morning Post). Other In early July 2008, news agencies worldwide reported press services indicated the bloom as red tide (Interna- on a vast algal bloom that was threatening the upcom- tional Herald Tribune), green tide (China Daily, Reuters), ing Olympic sailing events in Qingdao, China. The or denoted the algae as blue-green algae (The New identity of the culpable alga, however, remained undis- Zealand Herald and The Guardian). The Xinhua news cussed. We have identified the alga that caused the agency was the only one to report a putative species bloom by means of morphological and molecular data, name, ‘Entermorpha prolifera’. including sequence data of the plastid encoded large We aimed to resolve the identity of the bloom and, if subunit ribulose 1,5-bisphosphate carboxylase gene possible, determine its origin by means of detailed (rbcL) and the nuclear encoded rDNA internal tran- morphological examination and phylogenetic analysis of scribed spacer (ITS) region. The bloom-forming alga DNA sequence data. Free-floating algae were collected falls within the morphological limits of the green at two localities in Qingdao (Qingdao beach and Zhan seaweed Ulva prolifera O.F. Müller (‘Enteromorpha pro- Qiao beach). Specimens are deposited in GENT lifera (O.F. Müller) J. Agardh’) but our phylogenetic (FL1196-1199). Morphologically the alga unmistak- analyses show that it forms a clade with representatives ably belongs to the green macro-algal genus Ulva of the Ulva linza-procera-prolifera (LPP) complex. The (including the genus Enteromorpha, Hayden et al. Chinese rbcL sequences are identical to those of speci- 2003). Thalli are characterized by unattached, highly mens collected from Japan, New Zealand, Finland and branched narrow tubes, up to 450 mm in diameter, Portugal, suggesting that the taxon is widely distrib- cells in longitudinal rows, rectangular to polygonal, uted. rDNA ITS sequences showed a close affinity with 10–16 ¥ 14–32 mm, chloroplast with one or rarely two Japanese isolates of the species complex. The Qingdao pyrenoids (Figs 1–4). Thalli fall within the morphologi- bloom is a typical illustration of a green tide, which cal limits of Ulva prolifera O.F. Müller (‘Enteromorpha occurs increasingly along several coasts worldwide. prolifera (O.F. Müller) J. Agardh’) (Tseng 1984; Maggs et al. 2007). Species level identification in the genus, Key words: China, Enteromorpha, green algae, green however, is confounded due to the lack of distinguish- tide, harmful algal bloom, Olympic Games, Ulva. ing characters and a high degree of morphological vari- ability, resulting in vague species boundaries in this genus (Hayden & Waaland 2004). Taxonomic identifi- In early July 2008 China mobilized more than 10 000 cation of bloom forming Ulva is even more proble- people and 1400 boats in the fight to clean up a vast matic, since radical changes in gross morphology have algal bloom threatening the upcoming Olympic sailing been demonstrated in unattached populations under events in the sea area of east China’s Qingdao, just one month before the start of the games. The bloom started in late May and covered 13 000 square kilometre of sea. By 16 July, 1 million tonnes of algae had been cleared, *To whom correspondence should be addressed. China’s state-run news agency Xinhua reported. Email: [email protected] Although the Qingdao algal bloom was well covered in Communicating editor: H. Nozaki. the international press, the identity of the alga that Received 11 August 2008; accepted 3 October 2008. caused the bloom remained illusive. Most papers (e.g. doi: 10.1111/j.1440-1835.2009.00532.x

Figs 1–4. Ulva linza-procera-prolifera (LPP) from Qingdao (China). 1. Piles of Ulva washed ashore at Qingdao end of June 2008 (Photo: J-P Heirman). 2. Branched tubes (Enteromorpha type morphology). 3. Detail of tube showing cells organized in rows. 4. Chloro- plasts with single pyrenoid.

eutrophicated conditions (Malta et al. 1999; Blomster time-reversible model with a proportion of invariable et al. 2002). To determine the identity of the Qingdao sites and gamma distribution (GTR+I+G), as deter- bloom with more confidence, we sequenced the plastid mined by the Akaike Information Criterion in PAUP/ encoded large subunit ribulose 1,5-bisphosphate car- Modeltest 3.6 (Posada & Crandall 1998; Swofford boxylase gene (rbcL) and the nuclear encoded rDNA 2002). BI analyses consisted of two parallel runs of internal transcribed spacer (ITS) region, which has each four incrementally heated chains, and 3 million been proven useful in species level phylogenies of the generations with sampling every 1000 generations. A genus Ulva (Hayden et al. 2003; Shimada et al. 2003; burnin sample of 2000 trees was removed before con- Hayden & Waaland 2004; Hiraoka et al. 2004). DNA structing the majority rule consensus tree. For the ML extraction, polymerase chain reaction (PCR) amplifica- trees, the reliability of each internal branch was evalu- tion and sequencing of both markers was carried out ated based on 1000 bootstrap replicates. The new ITS according to Shimada et al. (2003). rbcL (1356 bp) sequences were included in an alignment of 32 closely and ITS (516 bp) sequences were obtained from four related Ulva sequences, and this dataset was used to isolates from the Qingdao bloom (three from Qingdao construct a statistical parsimony network with TCS beach and one from Zhan Qiao beach). rbcL and ITS 1.21 (Clement et al. 2000). sequences from the four Qingdao isolates were identi- The rbcL tree shows that the bloom forming Ulva cal. Additional rbcL sequences were also obtained from from Qingdao forms a clade with specimens morpho- three Ulva samples from Portugal. GenBank accession logically identified as Ulva procera (K. Ahlner) Hayden, numbers are FJ042888, FM210341-FM210344 Blomster, Maggs, P.C. Silva, M.J. Stanhope et J.R. (rbcL) and FJ026732, FM210345 (ITS). The newly Waaland and procera linza Linnaeus, with the Chinese obtained sequences were aligned with closest National rbcL sequence being identical to those of specimens Center for Biotechnology Information (NCBI) basic local collected from Finland, Portugal, New Zealand and alignment search tool (BLAST) hits and a diversity of Japan (Fig. 5). This clade is referable to the Ulva linza- Ulva sequences. The rbcL dataset was analysed with procera-prolifera complex (LPP clade), as circum- Bayesian inference (BI) and maximum likelihood (ML) scribed by Shimada et al. (2008). It should be noted using MrBayes v3.1.2 (Ronquist & Huelsenbeck 2003) that next to the LPP clade, U. linza and U. prolifera- and PhyML v2.4.4 (Guindon & Gascuel 2003), respec- type morphologies are also found in different other tively. BI and ML trees were computed under a general clades of the Ulva tree, as discussed in Hayden and

Fig. 5. Maximum likelihood tree based Ulva procera New Zealand EF110179 on rbcL sequence data, showing the Ulva procera New Zealand EF109944 position of the species causing the Ulva procera New Zealand EF110255 66 Ulva procera Japan AY422562 Qingdao algal bloom. LPP clade refers to Ulva linza Japan AB097620 LPP clade the Ulva linza-procera-prolifera (LPP) Qingdao bloom China, July 2008 Ulva procera Finland AY255863 complex. Numbers above branches indi- 98 Ulva cf intestinalis Portugal FM210343 1.00 cate maximum likelihood (ML) bootstrap 53 Ulva cf clathrata China DQ813496 .80 Ulva prolifera Portugal FM210342 values, numbers below branches indi- 61 .80 Ulva linza Pacific USA AY255861 cate posterior probabilities from the 66 Ulva prolifera Scotland AY255864 Bayesian phylogenetic analysis (only Ulva tanneri AY255858 86 Ulva californica AY422556 bootstrap values above 50% and poste- 1.00 72 Ulva prolifera AY422554 .90 rior probabilities above 0.8 are 91 Ulva flexuosa AB097619 indicated). .96 Ulva linza Portugal FM210344 67 Ulva prolifera AF499670 1.00 Ulva armoricana AB097632 65 .92 Ulva scandinavica AY255870 Ulva rigida AY422564 100 Ulva clathrata AY255862 99 1.00 1.00 Ulva muscoides AY422563 Ulva fasciata AY422565 75 Ulva ohnoi AB116040 1.00 Ulva taeniata AY422566 67 Ulva spinulosa AB097636 .92 Ulva reticulata AB097635 72 Ulva compressa AB097615 .94 100 Ulva intestinalis AB097617 54 1.00 Ulva intestinalis AY422551 74 Ulva lactuca AB097623 65 1.00 Ulva arasakii AB097621 Ulva lobata AY422550 Ulva pertusa AB097626 100 Umbraulva olivascens AY255876 1.00 Ulvaria obscura var. blyttii AF499673 0.2

Waaland (2004). The ITS dataset revealed the presence Japan (prolifera: AB298309) of at least 13 distinct ITS ribotypes within this LPP clade (Fig. 6). The ITS sequences from the Qingdao Finland (procera: AY260558) bloom were identical to sequences of U. linza from the Sweden (procera: AJ012276) same area and an isolate identiﬁed as U. prolifera from Japan (prolifera: AB298317) Japan (prolifera: AB298308)

Japan (Shimada et al. 2008). The large number of ITS Japan (linza: AB299439) ribotypes in Japan may indicate that this area is the Japan: (prolifera: AB298318) Japan (prolifera: AB298313) Japan (prolifera: AB298311) centre of origin of the LPP complex, but this is possibly Qingdao bloom July 2008 Japan (linza: AB097649) Qingdao (linza: EU888138) due to a sample bias in Japan. A wider geographical Japan (prolifera: AB298316) Japan (prolifera: AB298314) sampling and the use of more variable markers, such as

5S rDNA spacer region (see Shimada et al. 2008), will Japan (linza: AB298636) Japan (linza: AB298633) Japan (procera: AY422521) be needed to elucidate the biogeography of this species Japan (prolifera: AB298315) Australia (compressa: AF099722) complex. Netherlands (linza: AF153491) The Qingdao bloom is a typical illustration of a green Scotland (linza: AJ000203) tide, which is a type of harmful algal bloom caused by Japan (linza: AB299440) Japan (linza: AB298634) Scotland (linza: AJ000204) proliferation of unattached green seaweeds, generally members of the genus Ulva. These blooms are triggered Fig. 6. Ribotype network of rDNA internal transcribed spacer by a combination of different factors such as eutrophi- (ITS) sequences from samples within the Ulva linza-procera- cation, high temperature and irradiance level, and slow prolifera (LPP) clade, showing the position of the species causing turnover of marine waters. Green tides are developing the Qingdao algal bloom. The size of the rounded rectangles is every spring and summer along several coasts world- proportional to the number of sequences of that ribotype. Open wide, and their occurrence has been increasing over circles represent ribotypes not present in the sample. A line recent decades (Fletcher 1996; Blomster et al. 2002; between ribotypes represents one mutational step. Largo et al. 2004). Along the coastal areas of China, algal blooms have also been occurring more and more as a result of increased discharge of nitrogen-rich

© 2009 Japanese Society of Phycology 150 F. Leliaert et al. chemical pollutants, sewage and fertilisers from indus- Energy (Flemish Administration) for providing the algal trial, urban and agricultural activities (Yu et al. 2007; samples. We thank Caroline Vlaeminck, Sofie D’hondt Dong et al. 2008; Wang et al. 2008). Contrary to what and Licinia Gouveia for help with generating the the statements of the authorities and the Olympic sequence data. FL and HV are indebted to the Research Sailing Committee officials suggest, green macroalgal Foundation, Flanders for postdoctoral fellowship grants. blooms are not a new phenomenon for this area. Jin EM and AHE acknowledge a postdoctoral fellowship et al. (2005) mention a bloom of Ulva pertusa Kjellman grant from the Portuguese Science and Technology and U. linza in May 2003 (coinciding with the start of Foundation (FCT). the 2008 bloom). Both species, along with Ulva intestinalis Linnaeus and Ulva lactuca Linnaeus, were also REFERENCES recorded along the Qingdao coast in 2001–2003 (Liu et al. 2007), with U. intestinalis dominating in spring Blomster, J., Back, S., Fewer, D. P. et al. 2002. Novel mor- and U. pertusa occurring the whole year round. phology in Enteromorpha (Ulvophyceae) forming green Correct identification of the species is important as tides. Am. J. Bot. 89: 1756–63. significant differences have been reported in the Butler, D. 2007. Ancient algal mixup sorted. Nature News,20 response of the growth rates among various Ulva spp. to Dec 2007, doi: 10.1038/news.2007.396. differences in light, temperature, salinity and nutrient Clement, M., Posada, D. and Crandall, K. 2000. TCS: a concentrations (Pérez-Lloréns et al. 1996; Malta et al. computer program to estimate gene genealogies. Mol. 1999; Taylor et al. 2001). A correct identification, using Ecol. 9: 1657–60. both morphological and molecular characteristics, is a Dong, Q. X., Lin, J. D., Shang, X., Li, J. and Huang, C. J. prerequisite to draw up an ‘ecological profile’ of the 2008. Water, organic matter, nitrogen and phosphorus species, which can form the basis of a bloom-risk contents in sediment of a large-scale mariculture area in assessment, considering nutrient concentrations, cur- the Zhelin Bay of eastern Guangdong Province, China. Acta rents and environmental data such as temperature and Oceanol. Sinica 27: 33–148. irradiance. This reinforces the need for meticulous sys- Fletcher, R. L. 1996. The occurrence of “green tides”: a tematic research of the genus Ulva, combining tradi- review. In Schramm, W. and Nienhuis, P. H. (Eds) Marine tional taxonomic techniques and molecular phylogenetic Benthic Vegetation: Recent Changes and the Effects of analyses of large taxon sets, including type material Eutrophication. Springer, Berlin, pp. 7–43. (Butler 2007). Guindon, S. and Gascuel, O. 2003. A simple, fast, and accu- Authorities and some scientists casted the 2008 rate algorithm to estimate large phylogenies by maximum Qingdao algal bloom as a harmless natural phenomenon likelihood. Syst. Biol. 52: 696–704. related to seawater temperature and salinity, wind and Hayden, H. S. and Waaland, J. R. 2004. A molecular system- currents (BBC News, 1 July 2008). Other scientists atic study of Ulva (Ulvaceae, Ulvales) from the northeast blamed the bloom on increased rainfall and warmer Pacific. Phycologia 43: 364–82. waters in the Yellow Sea (International Herald Tribune, Hayden, H. S., Blomster, J., Maggs, C. A., Silva, P. C., Stan- 30 June 2008) and Olympic Sailing Committee officials hope, M. J. and Waaland, J. R. 2003. Linnaeus was right believed that the harmful algae were a ‘foreign enemy’, all along: Ulva and Enteromorpha are not distinct genera. having invaded the Qingdao coastal waters from an Eur. J. Phycol. 38: 277–94. unspecified area offshore (Reuters, 7 July 2008). Our Hiraoka, M., Shimada, S., Uenosono, M. and Masuda, M. data helps to eliminate some of these misconceptions. If 2004. A new green-tide-forming alga, Ulva ohnoi Hiraoka (considering the notoriously difficult taxonomy of Ulva) et Shimada sp nov (Ulvales, Ulvophyceae) from Japan. their identifications were correct, U. linza has been Phycol. Res. 52: 17–29. reported to bloom in this area before (Jin et al. 2005; Liu Jin, Q., Dong, S. and Wang, C. 2005. Allelopathic growth et al. 2007) and hence is certainly not a recently intro- inhibition of Prorocentrum micans (Dinophyta) by Ulva duced ‘foreign enemy’. Floating mats of Ulva were first pertusa and Ulva linza (Chlorophyta) in laboratory cultures. detected by fishermen on 31 May 2008, 100 km east of Eur. J. Phycol. 40: 31–7. Dagong Island in the Yellow Sea. Subsequent prolifera- Largo, D. B., Sembrano, J., Hiraoka, M. and Ohno, M. 2004. tion of the green tide was likely triggered by a combina- Taxonomic and ecological profile of “green tide” species of tion of eutrophication and high sea water temperatures. Ulva (Ulvales, Chlorophyta) in central Philippines. Hydro- biologia 512: 247–53. Liu, D., Bai, D., Song, S. et al. 2007. The impact of sewage ACKNOWLEDGMENTS discharge on the macroalgae community in the Yellow Sea coastal area around Qingdao, China. Water Air Soil Pollut. We are grateful to Maurice Hoffmann (Research Insti- 7: 683–92. tute for Nature and Forest, Flanders) and the Secretary- Maggs, C. A., Blomster, J., Mineur, F. and Kelly, J. 2007. General of the Department of Environment, Nature and Ulva. In Brodie, J., Maggs, C. A. and John, D. M. (Eds)

Green Seaweed of Britain and Ireland. The British Phyco- Shimada, S., Yokoyama, N., Arai, S. and Hiraoka, M. 2008. logical Society, London, UK, pp. 80–103. Phylogeography of the genus Ulva (Ulvophyceae, Chloro- Malta, E.-J., Draisma, S. G. A. and Kamermans, P. 1999. phyta), with special reference to the Japanese freshwater Free-ﬂoating Ulva in the southwest Netherlands: species or and brackish taxa. J. Appl. Phycol. 20: 979–89. morphotypes? A morphological, molecular and ecological Swofford, D. L. 2002. PAUP*. Phylogenetic Analysis Using comparison. Eur. J. Phycol. 34: 443–54. Parsimony (*and Other Methods). Version 4. Sinauer Asso- Posada, D. and Crandall, K. A. 1998. Modeltest: testing the ciates, Sunderland, MA. model of DNA substitution. Bioinformatics 14: 817–18. Taylor, R., Fletcher, R. L. and Raven, J. A. 2001. Preliminary Pérez-Lloréns, J. L., Vergara, J. J., Pino, R. R., Hernández, I., studies on the growth of selected “Green Tide” algae in Peralta, G. and Niell, F. X. 1996. The effect of photoac- laboratory culture: effects of irradiance, temperature, climation on the photosynthetic physiology of Ulva curvata salinity and nutrients on growth rate. Bot. Mar. 44: 327– and Ulva rotundata (Ulvales, Chlorophyta). Eur. J. Phycol. 36. 31: 349–59. Tseng, C. K. 1984. Common Seaweeds of China. Science Ronquist, F. and Huelsenbeck, J. P. 2003. MRBAYES 3: Press, Beijing, 316 pp. Bayesian phylogenetic inference under mixed models. Bio- Wang, S. F., Tang, D. L., He, F. L., Fukuyo, Y. S. and Azanza, informatics 19: 1572–4. R. V. 2008. Occurrences of harmful algal blooms (HABs) Shimada, S., Hiraoka, M., Nabata, S., Iima, M. and Masuda, associated with ocean environments in the South China M. 2003. Molecular phylogenetic analyses of the Japanese Sea. Hydrobiologia 596: 79–93. Ulva and Enteromorpha (Ulvales, Ulvophyceae), with Yu, J., Tang, D. L., Oh, I. S. and Yao, L. J. 2007. Response of special reference to the free-ﬂoating Ulva. Phycol. Res. 51: harmful algal blooms to environmental changes in Daya 99–108. Bay, China. Terr. Atmos. Ocean. Sci. 18: 1011–27.