lable at ScienceDirect

Estuarine, Coastal and Shelf Science 83 (2009) 434–442

Contents lists avai

Estuarine, Coastal and Shelf Science

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

Phosphorus input by upwelling in the eastern Gotland Basin (Baltic Sea)in summer and its effects on filamentous cyanobacteria

Monika Nausch a,*, Gunther Nausch b, Hans Ulrich Lass c, Volker Mohrholz c, Klaus Nagel b,Herbert Siegel c, Norbert Wasmund a

a Leibniz Institute for Baltic Sea Research, Department of Biological Oceanography, 18119 Rostock-Warnemunde, Seestrasse 15, Germanyb Leibniz Institute for Baltic Sea Research, Department of Marine Chemistry, 18119 Rostock-Warnemunde, Seestrasse 15, Germanyc Leibniz Institute for Baltic Sea Research, Department of Physical Oceanography, 18119 Rostock-Warnemunde, Seestrasse 15, Germany

a r t i c l e i n f o

Article history:Received 23 February 2009Accepted 18 April 2009Available online 7 May 2009

Keywords:upwellingphosphorus inputcyanobacteriaparticulate C:N:P stoichiometryBaltic SeaEastern Gotland Basin

* Corresponding author.E-mail addresses: monika.nausch@io-warnemuen

nausch@io-warnemuende.de (G. Nausch), uli.lass@io-volker.mohrholz@io-warnemuende.de (V. Mohrhmuende.de (K. Nagel), herbert.siegel@io-warnemuewasmund@io-warnemuende.de (N. Wasmund).

0272-7714/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.ecss.2009.04.031

a b s t r a c t

In July 2007, phosphorus input by an upwelling event along the east coast of Gotland Island and theresponse of filamentous cyanobacteria were studied to determine whether introduced phosphorus canintensify cyanobacterial bloom formation in the eastern Gotland Basin. Surface temperature, nutrientconcentrations, phytoplankton biomass and its stoichiometry, as well as phosphate uptake rates weredetermined in two transects between the coasts of Gotland and Latvia and in a short grid offshoreof Gotland. In the upwelling area, surface temperatures of 11–12 �C and average dissolved inorganicphosphorus (DIP) concentrations of 0.26 mM were measured. Outside the upwelling, surface temperatureswere higher (15.5–16.6 �C) and DIP supplies in the upper 10 m layer were exhausted. Nitrite and nitrateconcentrations (0.01–0.22 mM) were very low within and outside the upwelling region. Abundances offilamentous cyanobacteria were highly reduced in the upwelling area, accounting for only 1.4–6.0% of thetotal phytoplankton biomass, in contrast to 18–20% outside the upwelling. The C:P ratio of filamentouscyanobacteria varied between 32.8 and 310 in the upwelling region, most likely due to the introduction ofphosphorus-depleted organisms into the upwelling water. These organisms accumulate DIP in upwellingwater and have lower C:P ratios as long as they remain in DIP-rich water. Thus, diazotrophic cyanobacteriabenefit from phosphorus input directly in the upwelling region. Outside the upwelling region, the C:P ratiosof filamentous cyanobacteria varied widely, between 240 and 463, whereas those of particulate material inthe water ranged only between 96 and 224. To reduce their C:P ratio from 300 to 35, cyanobacteria in theupwelling region had to take up 0.05 mmol m�3 DIP, which is about 20% of the available DIP. Thus, a largerbiomass of filamentous cyanobacteria may be able to benefit from a given DIP input. As determined fromthe DIP uptake rates measured in upwelling cells, the time needed to reduce the C:P ratio from 300 to 35was too long to explain the huge bloom formations that typically occur in summer. However, phosphorusuptake rates increased significantly with increasing C:P ratios, allowing phosphorus accumulation within4–5 days, a span of time suitable for bloom formation in July and August.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

In the Baltic Sea, frequent wind driven upwelling events canoccur due to its topographic and geographic characteristics. Thetopography of the sea bottom, marked by sills, islands, and semi-enclosed basins, favours the induction of upwelling in response to

de.de (M. Nausch), guenther.warnemuende.de (H.U. Lass),olz), klaus.nagel@io-warnende.de (H. Siegel), norbert.

All rights reserved.

nearly all wind directions. Due to the predominant southwest winds,upwelling occurs most frequently at the Swedish coast in the BalticProper and in the Bothnian Bay (Gidhagen, 1987). Upwelling is alsoreported at the northern and southern coasts of the Gulf of Finland(Haapala, 1994; Suursaar and Aps, 2007), the Polish coasts (Kowa-lewski and Ostrowski, 2005), and the German islands (Lass et al.,1996). The frequency of upwelling events varies from region toregion (Kowalewski and Ostrowski, 2005), depending on the align-ment of the coastline with respect to the prevailing wind direction.

With changing wind direction, an upwelling area can beconverted to a downwelling one and vice versa (Kowalewski andOstrowski, 2005). At the Swedish coast, upwelling areas up to100 km long and spreading 5–20 km into the sea have been

Fig. 1. Satellite image of the upwelling event at the east coast of the Gotland Island inJuly 2007 and sampling stations.

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442 435

described by Gidhagen (1987). In the summer of 2006, Suursaarand Aps (2007) observed an upwelling event at the Estonian coastthat extended over a length of 360 km. According to Myrberg andAndrejev (2003), the wind has to blow from a favorable winddirection for at least 50–60 h before an upwelling is created.

Temperature is a useful indicator of the presence of upwellingregions. Upwellings can be identified clearly only in summer, whenwarm surface water with temperature of 16–21 �C (Siegel et al.,2008), is separated by a thermocline from the underlying inter-mediate winter water of 4–6 �C. In the center of an upwelling, thesurface temperature can drop by 5 �C when deep and surface waterare mixed (Gidhagen, 1987).

Upwelling water originates from depths of 20–40 m (Gidhagen,1987) and is depleted of inorganic nitrogen compounds but containssignificant phosphate concentrations of up to 0.50 mM (Nausch et al.,2007) which are in the range of the winter surface concentrations.These nutrient conditions have been described in upwelling regionsin the northern Gulf of Finland (Vahtera et al., 2005) and the HelPeninsula at the Polish coast as well (Kowalewski, 2005). Due to thelow molar N/P ratio of the surface water in winter (Nausch et al.,2008a), the phytoplankton spring bloom is nitrogen-limited. Theamount of phosphate remaining after the spring bloom is used tocalculate biomass development and nitrogen fixation by diazo-trophic cyanobacteria (Janssen et al., 2004; Kahru et al., 2007).However, the surplus phosphate remaining after the spring bloomand the changes in internal phosphorus pools do not always explainthe enormous biomass development and huge bloom formation ineach year (Lilover and Laanemets, 2006; Nausch et al., 2008b). Themodel simulations of Laanemets et al. (2006) showed the best fit ofNodularia spumigena blooms when both turbulent mixing andupwelling were taken into account. Using a hydrodynamic model,Ennet et al. (2000) demonstrated for the Gulf of Riga that the springbloom as well as the cyanobacteria summer bloom can be intensi-fied by upwelling. According to Vahtera et al. (2005), nutrient inputby upwelling does not directly effect the development of filamen-tous cyanobacteria; rather, the relaxation of older, warmer waterand its mixing with upwelled cold phosphorus-rich water favourcyanobacteria development after a lag time of 2–3 weeks.

During a cruise with the R/V ‘‘Poseidon’’ in July 2007, we studiedphosphorus input during upwelling events taking place at theeastern coast of Gotland and subsequent phosphorus trans-formations in the surface layer along transects extending as far asthe coast of Latvia. Focusing on filamentous cyanobacteria, wesought to answer the following questions:

- Is cyanobacterial growth promoted by phosphorus input fromupwelling?

- Can cyanobacteria immediately use the phosphate supplied bythe upwelling region, or is there a delay between phosphateinput and cyanobacterial uptake?

- Is phosphorus introduced only as phosphate or as dissolvedorganic phosphorus as well?

- How do cyanobacterial biomass, intracellular phosphorus content,and C:N:P ratios change with distance from the upwelling region?

Answering these questions would resolve the broader question:Does phosphorus supplied by the upwelling induce or intensifycyanobacterial blooms in the Gotland Basin?

2. Material and methods

2.1. Investigation area and sampling

Investigations were conducted in 2007, between July 10th and23rd, in the eastern Gotland Basin. During this period, winds came

predominantly from southwest to northwest directions (200–300�), with wind speeds ranging between 4 and 18 m s�1 (Lasset al., submitted for publication). On 3 days only (July 17, 21, 22), thewind speed was lower, for few hours, and the wind directionchanged to east. Air temperature ranged between 13.5 and 17.4 �C.Global radiation reached maximum values of 850–1000 W m�2.This was one of the typical summer weather patterns in the BalticSea region characterized by westerly circulation with embeddedcyclones. Under these meteorological conditions, an upwelling wasinduced at the southeast coast of the Gotland Island, as indicated bysea surface temperature of about 10 �C as determined froma satellite image (Fig. 1). Water samples were taken on two tran-sects between the coasts of Gotland and Latvia (Fig. 1). The distancebetween sampling stations was 10 n.m.. A filament of upwelledwater was tracked in a third short grid (Fig. 1).

A rosette sampler (Hydrobios) consisting of thirteen 5 l free-flow bottles was used to obtain the water samples. The rosette wasequipped with a Seabird SBE 911, consisting of sensors for depth,conductivity, temperature, and chlorophyll fluorescence, as well asa light sensor to measure photosynthetically available radiation(PAR).

Filamentous cyanobacteria were sampled from the thermoclineup to the surface using a WP2 plankton net with a mesh size of100 mm. To eliminate zooplankton, the samples were resuspendedin GF/F-filtered seawater in a 2 l bottle and two-thirds covered withblack wrapping for 2 h, causing the zooplankton to move to thestill-lighted bottom such that pure filamentous cyanobacteria couldbe collected from the upper portion.

2.2. Analytical methods

2.2.1. Inorganic nutrient analysisNutrient concentrations (dissolved inorganic phosphorus (DIP),

nitrate, nitrite, ammonium, silicate) were determined by standardphotometric methods, as described by Rohde and Nehring (1979)and Grasshoff et al. (1983). Water samples were filtered throughpre-combusted (450 �C, 4 h) Whatman-GF/F filters. Nitrite, nitrate,phosphate and silicate concentrations were determined simulta-neously using an autoanalyzer (Evolution III, Alliance). Ammoniumwas determined manually in unfiltered samples.

2.2.2. Dissolved and particulate organic phosphorusFor the determination of total and dissolved phosphorus, 40 ml

subsamples were stored at �20 �C before and after filtration

16

15141312

10

5

0

-10

-20

-30

-40

-50

-60

-7019 19.5 20 20.5 21

dep

th

[m

]

Longitude [deg]

Fig. 2. Example for temperature distribution (�C) between the coasts of Gotland andLatvia in July 2007 (transect 2).

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442436

through pre-combusted (450 �C, 4 h) Whatman-GF/F filters untilfurther processing. The thawed samples were oxidized with analkaline peroxodisulfate solution (Grasshoff et al., 1983) to convertorganic phosphorus into DIP, which was measured using themanual photometric method. Dissolved organic phosphorus (DOP)was calculated from the difference between dissolved phosphorusand DIP, and particulate organic phosphorus (POP) as the differencebetween total and dissolved phosphorus.

2.2.3. Chlorophyll aWater samples and subsamples from a suspension of isolated

cyanobacteria were filtered onto Whatman-GF/F filters and storedin liquid nitrogen or at �80 �C. Chlorophyll a (Chl a) was extractedwith 96% ethanol for at least 3 h (Wasmund et al., 2006) after whichthe concentration was measured fluorometrically (excitation450 nm, emission 670 nm) using a Turner fluorometer (10-AU-005). The concentration was calculated according to the method ofJeffrey and Welschmeyer (1997).

2.2.4. Particulate organic carbon and nitrogenFor measurements of particulate organic carbon (POC) and

nitrogen (PON), 500 ml water samples or 100 ml of the cyano-bacterial suspensions were filtered onto pre-combusted Whatman-GF/F filters and frozen at �20 �C until analysis in a CHN analyzer(Vario Microcube, Elementar).

2.2.5. Biomass of diazotrophic cyanobacteriaSamples to determine phytoplankton species composition and

bio-volumes were taken at discrete depths (0, 4, and 8 m) of themixed surface layer and preserved with acetic Lugol solution (KI/I2).After 24 h of sedimentation in an Utermohl sedimentationchamber, phytoplankton counts were obtained using an invertedmicroscope (Zeiss) (Utermohl, 1958) at 100� or 400� magnifica-tion, depending on the cell size. Cell volume was calculated fromsize measurements using the appropriate stereometric formula andconverted into wet weight (ww mg m�3), assuming that thedensity of plasma is equal to that of water (w1 mg mm�3).

Special attention was drawn to the filamentous cyanobacteria.The accompanying species were not fully identified as they werenot subject of this paper but served for a rough estimation of thetotal phytoplankton biomass values.

2.2.6. Phosphate uptakePhosphate uptake was measured by the addition of 880 kBq of

[33P]PO4 (Hartmann Analytics, specific activity 110 TBq mmol�1) to150 ml samples (final concentration 50 pM) in polycarbonatebottles. For blanks, 750 ml of formaldehyde was added prior toradiotracers. All samples were incubated under simulated in-situlight conditions and at in-situ temperatures. During the 3-h incu-bation, 5 ml subsamples were withdrawn at specific time intervalsand filtered onto 0.2 mm polycarbonate filters pre-soaked with1 mM KH2PO4. The filters were rinsed with 5 ml particle freeseawater and the radioactivity trapped on the filters was counted ina liquid scintillation counter (Tri-Carb 2800TR, Perkin Elmer) usingLumasafe Plus (Packard) as scintillation cocktail. All analyses weredone in triplicate. Finally, phosphate uptake was calculated fromthe linear slope of increasing counts on the filters.

3. Results

3.1. Hydrographic and hydrochemical parameters

Temperature profiles along a transect between the coast ofGotland and Latvia measured on July 20th are shown in Fig. 2. In theupwelling area, the water temperature at the surface was 11–12 �C.

At a distance of 10–20 n.m. from the upwelling, the surfacetemperature increased to 15.5–16.6 �C in the upper 10 m layer. Thistemperature was in the range of the surface temperatures of thoseareas of the eastern Gotland Basin unaffected by the upwelling.Additionally, as seen in Fig. 2, the thermocline was uplifted. Thus,the water layer between 10 and 15 m was cold and enriched withDIP as far as the middle of the eastern Gotland Basin. Further east inthe basin, the thermocline shifted to a greater depth and warmwater was present at depths of 20–30 m.

In the upwelling area, neither nitrite nor nitrate (NO2/3) wastransported into the surface layer. Average sum of nitrite andnitrate (NO2/3) concentrations of around 0.05 mM (range 0.01–0.21 mM) and average ammonium concentrations of 0.20 mM (range0.04–0.36 mM) from the surface to a depth of 40 m were insignifi-cant at all stations within and outside the upwelling area. Silicateconcentrations ranging between 7.2 and 11.0 mM varied onlyslightly in the whole area (Table 1).

DIP concentrations of 0.26� 0.03 mM were measured in thesurface layer of the upwelling area. At a distance of 10–20 n.m. fromthe upwelling cell, DIP concentrations were near the detection limit(Fig. 3). In parallel to the thermocline, the nutricline declined fromwest to east, down to 20 m and lower.

In the third, short grid, the lowest temperatures (mean 12 �C forthe upper 0–10 m) and highest DIP concentrations (mean 0.29 mMfor the upper 10 m) were measured at the station in the middle ofthe grid (Table 1). At stations closer to the shore, surface DIPconcentrations were in the range of 0.14–0.16 mM.

Based on all of the values recorded for the upper 10 m layer, aninverse relationship between temperature and phosphateconcentration (r2¼ 0.76, n¼ 77) was estimated. This relationshipallows the upwelling area to be distinguished from both a transi-tion zone and an unaffected area. The separation of the areasbecomes especially clear if mean values of the 0–10 m depth areused (Fig. 4).

DOP values typically ranged between 0.21 and 0.29 mM in theupwelling cell and in the areas outside the upwelling as well (Table1). Higher values were measured only at the first transect, atstations near the coast.

POP concentrations in the two transects differed (Table 2). Attransect 1, they increased from 0.19 to 0.37 mM in the middle of theGotland Basin and decreased to 0.19 mM at the coast of Latvia. Attransect 2, the lowest POP value, 0.20 mM, was measured in the

Table 1Hydrographic parameters, inorganic nutrient concentrations as well as total and dissolved organic phosphorus in the upper 10 m surface layer (mean values).

Station Temp. Salinity NO2,3 DIP Si NH4 TP DOP

(�C) (PSU) (mM) (mM) (mM) (mM) (mM) (mM)

Transect 1 Gotland

Latvia

0019 12.84 7.10 0.04 0.26 8.88 0.11 1.02 0.570018 11.61 7.02 0.04 0.11 8.33 n.d. 0.88 0.390017 16.06 6.97 0.05 0.02 10.10 n.d. 0.96 0.500016 16.05 6.94 0.02 0.02 10.97 n.d. 0.91 0.240015 16.08 6.98 0.02 0.01 10.58 n.d. 0.75 0.440014 16.27 6.99 0.01 0.00 10.33 n.d. 0.65 0.280020 16.17 7.00 0.05 0.02 10.68 0.31 0.51 0.240013 15.83 7.05 0.02 0.01 10.43 n.d. 0.48 0.240012 16.08 7.05 0.08 0.01 9.93 n.d. 0.58 0.260011 15.83 6.94 0.04 0.00 10.20 n.d. 0.49 0.270010 15.63 7.04 0.03 0.01 9.92 n.d. 0.49 0.260009 15.48 7.16 0.02 0.01 8.40 0.14 0.45 0.25

Transect 2 Gotland

Latvia

0021 11.47 7.07 0.06 0.24 7.77 0.72 0.67 0.230022 16.77 6.95 0.04 0.01 10.05 n.d. n.d. n.d.0023 15.92 6.98 0.05 0.02 9.12 n.d. 0.52 0.270025 16.43 6.97 0.03 0.01 10.43 n.d. 0.48 0.260027 16.56 7.00 0.04 0.01 9.95 n.d. 0.51 0.260029 16.30 7.13 0.04 0.01 9.43 n.d. 0.58 0.260031 16.53 7.15 0.21 0.02 9.23 n.d. 0.59 0.24

Grid 3 Gotland

offshore

0033 12.86 7.09 0.20 0.18 8.78 n.d. 0.66 0.230034 12.14 7.04 0.11 0.16 9.77 n.d. 0.68 0.240035 10.98 7.15 0.01 0.29 7.28 n.d. 0.72 0.240036 13.81 7.09 0.00 0.16 9.82 n.d. 0.62 0.240037 15.63 7.05 0.01 0.08 8.68 n.d. 0.60 0.28

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442 437

upwelling while concentrations were highest at the easternmoststation. In the third grid, POP concentrations ranged between 0.20and 0.28 mM, without any trend.

3.2. Phytoplankton biomass development in the upper 10 m layer

At the two transects, Chl a concentrations were two-fold higherin the offshore regions than in the upwelling region (Table 2). In thelatter, Chl a concentrations were below 2 mg l�1 (mean1.49� 0.19 mg l�1) whereas in the unaffected area they rangedbetween 2.70 and 6.68 mg l�1 (mean 3.49 mg l�1). In the transitionzone, Chl a values were intermediate between the two.

0.020.050.10.150.2

0.2 0.15 0.1 0.050.3

0.4

0.5

1

2

0

-10

-20

-30

-40

-50

-60

-70

dep

th

[m

]

18.8 19.2Longitude [deg]

19 19.4 19.6 19.8 20 20.2 20.4 20.6 20.8

Fig. 3. Example for distribution of DIP concentrations (mM) between the coasts ofGotland and Latvia in July 2007 (transect 2).

Changes in Chl a were reflected in significant changes inparticulate organic carbon (POC) (r2¼ 0.58, n¼ 44) (Fig. 5).However, the relationship with POP (r2¼ 0.02, n¼ 44) was onlyvery weak, due to increasing C:P ratios (Fig. 5). These increased inthe upper 10 m layer of an area extending from the upwelling celluntil the middle of the Gotland Basin and then declined eastwards,possibly indicating another phosphorus source at the east coast ofthe basin.

Microscopic counting of water samples showed that filamen-tous cyanobacteria constituted only a small proportion of the entirephytoplankton biomass (Fig. 6, Table 3), ranging from 1.4 to 5.6% atupwelling stations to as high as 18.6% further offshore. Thephytoplankton biomass was dominated by small unicellularphytoplankton organisms (cyanobacteria, flagellates, and others).

Four species of filamentous cyanobacteria, Aphanizomenonbaltica, N. spumigena, Pseudanabaena sp., and Anabaena sp.,

0.0

0.1

0.2

0.3

0.4

temperature (°C)

DIP

(µ

M)

8 10 12 14 16 18 20

upwellingtransitionoffshore

Fig. 4. Relationship between temperature and DIP concentrations in the mixed 10 msurface layer (mean values). Stations in the upwelling area can be clearly distinguishedfrom a transition zone and the unaffected area. In the ANOVA-analysis the nullhypothesis is accepted with p¼ 0.0001.

Table 2Concentrations of Chl a and particulate organic phosphorus, nitrogen and carbon as well as ratios between them in the upper 10 m surface layer (mean values).

Station Chl a POP PON POC C:N C:P

(mg l�1) (mM) (mM) (mM)

Transect 1 Gotland

Latvia

0019 1.64 0.19 3.18 26.79 8.4 96.60018 2.07 0.38 3.23 23.81 7.4 59.10017 3.05 0.45 6.53 48.54 7.4 110.90016 3.21 0.65 6.21 49.37 8.0 79.30015 2.90 0.30 5.26 46.76 8.9 137.50014 3.59 0.37 5.45 49.24 9.0 138.70020 3.40 0.25 5.41 42.96 7.9 182.40013 2.92 0.23 5.66 52.80 9.3 224.70012 6.68 0.32 11.59 103.91 17.9 173.40011 3.87 0.22 5.56 48.00 8.6 220.80010 3.68 0.23 5.35 49.87 9.3 214.60009 4.07 0.19 5.23 37.98 7.3 217.0

Transect 2 Gotland

Latvia

0021 1.37 0.20 2.90 22.41 7.8 120.30022 n.d. n.d. n.d. n.d. n.d. n.d.0023 2.70 0.23 5.11 39.68 7.8 175.70025 2.81 0.22 4.23 36.21 8.4 142.00027 3.12 0.24 35.78 47.77 7.7 201.90029 3.52 0.31 6.39 51.69 8.1 166.00031 4.20 0.34 6.74 55.70 8.3 166.4

Grid 3 Gotland

offshore

0033 1.27 0.26 3.46 30.83 8.9 120.90034 3.23 0.28 5.61 50.31 9.0 179.70035 1.33 0.20 2.90 22.61 7.8 116.00036 1.69 0.22 3.64 31.63 8.7 143.80037 2.41 0.25 5.24 46.01 8.8 187.8

1600

2000

g m

-3)

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442438

predominated at all stations albeit with varying proportions.Abundances of Anabaena sp. remained low, accounting for 0.03–1.9% of the total phytoplankton biomass except at station 0009,nearest to the coast of Latvia, where its abundance reached 6.9%.The proportion of Pseudanabaena sp. was in the range of 1.0–4.3%,without significant differences between the upwelling and theother, unaffected regions.

Abundances of A. baltica and N. spumigena differed according tothe particular area (Fig. 6, Table 3). At the two transects through thebasin, the sum of both species increased linearly, ranging from aslow as 0.1–0.5% at the upwelling area to between 10.7 and 16.6% onthe Latvian side of the basin. Neither of these two species pre-dominated in the upwelling area. At central stations of the basin,the two species occurred in about equal proportions. At stationsnearest the coast of Latvia, either N. spumigena or A. baltica was thedominating organism.

0

10

20

30

40

50

60

70

0.0 1.0 2.0 3.0 4.0 5.0Chl a (µg l

-1)

PO

C (µ

M)

0.00.10.20.30.40.50.60.70.80.9

PO

P (µ

M)

POCPOP

Fig. 5. Relationship between Chl a concentrations and POC and POP in the mixed 10 msurface layer. Data from all sampling depths.

3.3. Nutrient stoichiometry

Diazotrophic filamentous cyanobacteria were characterized byseparating them from the other plankton using the proceduredescribed above, such that the sample was then almost exclusivelyenriched in cyanobacteria, especially N. spumigena followed by A.baltica (Table 4). The other filamentous cyanobacteria, Anabaena sp.and Pseudanabaena sp., were present in significant amounts only at

0

400

800

1200

0021 0023 0025 0027 0029 0031station

bio

mass (w

w m

Gotland Latvia

othersPseudanabaenaNodularia

AphanizomenonAnabaena

Fig. 6. Example for the variation of phytoplankton community structure from theupwelling area to the Latvian coast with special reference of the 4 main filamentouscyanobacteria genera.

Table 3Total phytoplankton biomass and the amount of filamentous cyanobacteria species (absolute and % of total biomass) in the upper 10 m surface layer obtained by microscopiccounting.

Station Total autotrophs Aphanizomenonbaltica

Nodularia spumigena Anabaena spec. Pseudanabaenaspec.

Sum filamentouscyanobacteria

(ww mg m�3) (ww mg m�3) (%) (ww mg m�3) (%) (ww mg m�3) (%) (ww mg m�3) (%) (ww mg m�3) (%)

Transect 1 Gotland

Latvia

0019 832.6 2.6 0.3 2.1 0.2 4.4 0.5 13.0 1.6 23.1 2.80020 1702.9 71.4 4.2 67.3 3.9 0.8 0.05 36.7 2.2 184.3 10.80009 848.2 0.7 0.1 90.7 10.7 53.1 6.3 13.6 1.6 175.1 20.6

Transect 2 Gotland

Latvia

0021 1438.3 1.2 0.1 0.0 0.0 3.7 0.3 14.8 1.0 20.0 1.40023 1857.0 29.7 1.6 51.0 2.7 3.9 0.2 68.8 3.7 157.9 8.50025 1786.6 59.8 3.3 47.1 2.6 1.2 0.1 85.4 4.8 199.5 11.20027 1765.4 101.4 5.7 83.3 4.7 0.5 0.0 32.6 1.8 228.3 12.90029 1871.6 165.3 8.8 126.5 6.8 1.3 0.1 22.9 1.2 331.8 17.70031 1891.6 209.3 11.1 103.2 5.5 0.6 0.03 13.3 0.7 343.0 18.1

Grid 3 Gotland

offshore

0033 1201.3 8.8 0.7 37.0 3.1 9.0 0.7 25.4 2.1 84.7 7.10034 1708.0 29.4 1.7 63.1 3.7 2.0 0.1 37.5 2.2 137.5 8.00035 981.3 7.7 0.8 28.4 2.9 4.6 0.5 13.8 1.4 58.7 6.00036 984.6 12.4 1.3 26.5 2.7 18.8 1.9 42.2 4.3 105.6 10.70037 1607.1 30.5 1.9 99.7 6.2 10.2 0.6 58.0 3.6 207.0 12.9

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442 439

three stations in the third, short grid. In most cases, <5% of theenriched biomass consisted of other phytoplankton organisms.

The C:N and C:P ratios of the isolated cyanobacteria differedfrom those in the particulate material of whole water samples. TheC:N ratios of filamentous cyanobacteria ranged between 5.4 and 6.6(mean 6.2), with a slight increase from the upwelling to the centralstations (Table 5). Those of plankton in the whole water sampleswere higher, between 7.6 and 17.9 (mean 8.7), with the lowestratios in the upwelling region (Table 2). These observations indi-cated that filamentous cyanobacteria are able to fix nitrogen moreintensively in upwelling areas than in warmer, DIP-depletedregions.

The C:P ratios of filamentous cyanobacteria varied to a greaterdegree than those of particulate material obtained from the wholewater samples (Figs. 7 and 8). In the upwelling area, lowest C:Pratios of 32.8 and 36.7 were determined (Table 5) in the secondtransect and at the near-shore station of the third grid. At the firsttransect, the C:P ratio was 187.5 while at the station of the thirdgrid with the most intensive upwelling, a very high C:P ratio of310 was measured despite the fact that sufficient DIP wasavailable.

Outside the upwelling region, the C:P ratios of filamentouscyanobacteria increased, ranging between 234 and 462, andexceeded those of the whole water samples, in which themaximum C:P ratio was 224.7. As in filamentous cyanobacteria, C:Pratios of particulate material in whole water samples declined alsoin upwelling stations (Fig. 7). Values between 96.6 and 120.3indicated that not so much phosphorus had been accumulated inthe other phytoplankton organisms compared to the filamentouscyanobacteria. However, low C:P ratios, e.g., of 79.3, were alsodetected at some stations outside the upwelling (Table 2). In bothsample types it was evident that maximum C:P ratios were reachedin the central eastern Gotland Basin. At the far eastern stations, i.e.,those nearest the Latvian coast, C:P ratios declined even though DIPconcentrations were at the detection limit, as was the case in thecentral parts of the basin.

The relationship between C:P ratios and the P content of fila-mentous cyanobacteria, expressed as P per unit Chl a, is shown inFig. 8. For isolated cyanobacteria, the non-linear relationship isdescribed by: y¼ 23.07x�1.111. The data obtained from the watersamples fit this relationship, but with a smaller range.

3.4. Phosphorus uptake

In Table 6, [33P]PO4 turnover rates and the gross uptake rates offilamentous cyanobacteria are compared with those of the wholewater sample for three stations at transect 1. The activity of fila-mentous cyanobacteria was lower than that of the whole watersample, indicating that they take up P more slowly than otherorganisms. Specifically, the gross uptake was 85% lower at station0020, in the central part of the Gotland Basin and 15% lower at theeastern station compared to the uptake of the whole watersamples. These findings are supported by the results obtained atthree other stations not located within one of the describedtransects.

4. Discussion

Summarizing historical records and measurements as well asinvestigations using model systems, Myrberg and Andrejev (2003)and Lehmann and Myrberg (2008) provided an overview of theregional distribution and frequency of upwelling events in theBaltic Sea. According to their findings, a multitude of smallupwelling events can occur in summer. The studies of Kononen andNiemi (1986), Haapala (1994) and Fonselius (1996) have demon-strated that in an upwelling nutrient-poor, warm surface water isdisplaced by nutrient enriched colder water which promoteshigher production levels. However, how the food chain functions isnot well understood and the contribution of upwelling to newproduction is poorly quantified (Lehmann and Myrberg, 2008;Lilover and Stips, 2008; Nausch et al., 2008b). Nonetheless, it isclear that the introduced nutrients fertilize the nutrient-depletedsurface layer in summer. The growth of diazotrophic cyanobacteriais especially encouraged because the upwelling water is depleted innitrate and nitrite but enriched in DIP. The effects of upwelling onthe development of the filamentous cyanobacteria N. spumigenaand A. baltica, the most important diazotrophs in the Baltic Sea, areparticularly noteworthy in this respect.

Compared to other years, cyanobacterial development in thesummer 2007 was moderate and surface accumulation wasobserved only rarely in the entire Baltic Sea (http://www.helcom.fi/stc/files/Publications/Newsletters/Newsletter_2_08_11.pdf). Duringthe investigation period of this study, filamentous cyanobacteria

Table 4Proportion in biomass (%) of the four species of isolated filamentous cyanobacteria obtained by net tows.

Station Aphanizomenon baltica Nodularia spumigena Anabaena spec. Pseudanabaena spec.

(%) (%) (%) (%)

Transect 1 Gotland

Latvia

0019 31.5 60.4 6.9 0.00020 16.0 77.5 0.1 1.00009 59.9 39.8 0.0 0.0

Transect 2 Gotland

Latvia

0021 40.0 39.0 4.0 1.00023 12.5 84.0 0.1 1.60025 12.1 84.7 0.1 0.70027 5.1 93.4 0.0 0.10029 14.1 81.9 0.1 0.30031 40.3 58.2 0.0 0.1

Grid 3 Gotland

offshore

0033 19.4 53.4 23.5 0.00034 20.4 59.9 9.5 0.00035 1.5 93.8 1.5 3.00036 14.7 34.4 29.7 0.00037 6.1 17.6 0.1 70.2

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442440

accounted for a maximum of only 18.6% of the total phytoplanktonbiomass in the eastern Gotland Basin. In contrast, in the hugecyanobacterial bloom that occurred in 2005, cyanobacterialbiomass reached a high of 4515 mg m�3 and comprised 60% of theautotrophs (Nausch and Nausch, 2007). However, the nutritionstate of diazotrophic cyanobacteria indicated by C:P ratios wassimilar to that observed during huge blooms. The absence of hugeblooms in 2008 was may be due to water temperatures remainingbelow 20 �C and the lack of long calm periods.

According to Gidhagen (1987), upwelling water in the Baltic Seaoriginates from depths of 20–40 m and has a typical temperaturedrop of 5 �C. In our investigations, surface temperatures atupwelling stations were 11–12 �C, indicating that the waterupwelled from the intermediate winter water of about 5 �C hadmixed with warmer surface water of 16 �C. The resulting surfacetemperature drop of about 5 �C in the upwelling water comparedwell with characteristic upwelling processes in the Baltic Seaaccording to Gidhagen (1987). The DIP concentration (0.26 mM) wassimilar to that reported for an upwelling in the northern Gulf of

Table 5C:N and C:P ratios of isolated cyanobacteria (net tows) and their P content related toChl a.

Station C:N C:P P/Chl a

nmol mg�1

Transect 1 Gotland

Latvia

0019 5.7 187.5 125.50020 6.7 379.5 41.70009 6.2 239.8 33.6

Transect 2 Gotland

Latvia

0021 5.4 32.8 804.00023 6.2 234.1 32.80025 6.6 300.9 52.40027 n.d. n.d. 26.60029 6.5 462.6 26.20031 6.4 372.4 29.8

Grid 3 Gotland

offshore

0033 5.8 36.7 330.80034 5.6 241.2 97.70035 6.3 310.0 48.90036 5.9 419.4 24.00037 6.5 197.8 63.7

Finland (Vahtera et al., 2005) and at the Hanko peninsula (Haapala,1994).

Chl a concentration dropped below 2 mg l�1, as in otherupwelling regions (Vahtera et al., 2005; Zalewski et al., 2005).Among phytoplankton, especially the biomass of filamentous cya-nobacteria decreased strongly in upwelling regions (by 7- to 12-fold). Vahtera et al. (2005) reported different distributions of cya-nobacteria species: A. baltica dominated in upwelling regionswhereas N. spumigena were more abundant in the surroundingwater. The authors therefore concluded that A. baltica is directlysupported by DIP input whereas N. spumigena benefits fromregenerated phosphorus during relaxation of the upwelling.Nommann et al. (1991) found higher A. baltica abundances in theupwelling center than at more distant stations during an upwellingevent at Hiiumaa Island. We did not observe any trend in therelatedness of A. baltica and N. spumigena abundances or in thedominance of one or the other. The abundances of both organismswere low in the upwelling area and increased outside of it. At somestations, the biomass of N. spumigena exceeded that of A. balticawhereas at others the relation between the two species wasreversed. This pattern was independent of whether the stationswere located in the upwelling area or outside of it. Thus,

0

100

200

300

400

500

0021 0023 0025 0027 0029 0031stations

C:P

ratio

Gotland Latvia

water samplesselected cyanobacteria

Fig. 7. Variation of the C:P ratios of particulate material of water samples and of iso-lated filamentous cyanobacteria in the transect between the coasts of Gotland andLatvia.

y = 23.07x-1.111

r2= 0.81

0.0

0.2

0.4

0.6

0.8

1.0

0 100 200 300 400 500C:P ratio

P p

er C

hl a (µ

mo

l µ

g-1)

selected cyanobacteriawater sample

Fig. 8. Relationship between the C:P ratios and the P content of isolated filamentouscyanobacteria (P per Chl a). The relationship is also shown for the wholephytoplankton.

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442 441

a succession of cyanobacteria species, as described by Vahtera et al.(2005) in the Gulf of Finland, could not be confirmed by ourinvestigations in the eastern Gotland Basin.

In net tows, both species were present but those of N. spumigenawere mostly enriched to a higher degree than A. baltica. Thus, thecellular phosphorus content and C:P ratios of net samples weremainly influenced by the nutrient conditions of N. spumigena; thatis, C:P ratios of 32.8 and 36.7 can be ascribed directly to phosphorusaccumulation by N. spumigena in the upwelling region and did notoccur after a time lag, as observed by Vahtera et al. (2005). Changesin DIP and DOP concentrations indicated that either DOP was notthe preferred phosphorus source or consumption and production ofDOP were balanced.

The large variations in the C:P ratios (32–310) in the upwellingregion may have reflected the introduction of diazotrophic cyano-bacteria scarce in cellular phosphorus (C:P ratios between 250 and400) into the upwelling water. There, cyanobacterial uptake andaccumulation of DIP can result in declining C:P ratios withincreasing cyanobacterial residence time in the water body, finallyleading to ratios that were very low. Thus, DIP input by anupwelling event may have been converted into phosphorus accu-mulation in cyanobacteria, comparable to what was observed in thespring by Larsson et al. (2001). This phosphorus accumulation is thebasis for subsequent development as observed in tank experimentsin which most cyanobacterial growth occurred in surface waterfrom the upwelling region (Wasmund et al., submitted forpublication).

We estimated the amount of DIP necessary to lower the C:P ratioof filamentous cyanobacteria from 300 to 35, and the time neededfor this reduction deduced from DIP uptake rates. A C:P ratio of 300is the average value for filamentous cyanobacteria outside theupwelling and a C:P ratio of 35 is the average of the lowest ratios. Tocalculate the Chl a content of filamentous cyanobacteria in watersamples, a conversion factor between wet weight and Chl

Table 6[33P]PO4 uptake rates and DIP gross uptake rates of the whole water sample and of isola

Station Whole water sample

[33P]PO4 uptake DIP gross uptake C:P rati

(% mg�1 h�1) (nmol mg�1 h�1)

006 0.25 0.16 160.4007 0.99 0.60 149.8008 n.d. n.d. n.d.0019 n.d. n.d. 96.60020 5.41 1.95 182.40009 1.96 0.09 217.0

a (237.3 ww mg m�3 wet weight per 1 mg Chl a) was derived frommeasurements in isolated cyanobacteria. Using the formula shownin Fig. 8, which describes the relationship between the C:P ratio andP content in isolated cyanobacteria, we determined that a C:P ratioof 300 corresponds to a P content per mg Chl a of0.047 mmol mg Chl a�1, and a C:P of 35 is equivalent to0.44 mmol mg Chl a�1. To achieve this low C:P ratio, 0.39 mmol -DIP mg Chl a�1 must be taken up. We estimated that the cyano-bacterial biomass in upwelling water is 32 mg m�3, whichcorresponds to 0.14 mg Chl a. This biomass needs 0.05 mmol m�3

DIP to decrease its C:P ratio from 300 to 35. Consequently, from themeasured DIP content in the upwelling water of 0.26 mmol m�3,about 20% has to be taken up by filamentous cyanobacteria toincrease their internal phosphorus pool. Thus, there is the potentialto supply larger populations of cyanobacteria with phosphorus. If,during successive growth, C:P ratios of 240–480 were reached, an8- to 13-fold increase of cyanobacterial biomass would be possibledue to reduction of intracellular phosphorus.

The question arises how long does it take for filamentous cya-nobacteria to reduce their C:P ratio from 300 to 35. As shown inFig. 8, it is evident that from the 0.39 mmol DIP mg Chl a�1 neededto do so, only 0.09 mmol mg Chl a�1 is necessary to reduce the C:Pratio from 300 to 100. The remaining 0.29 mmol mg Chl a�1 areneeded to further reduce the ratio from 100 to 35. Based on thephosphorus uptake rate determined for cyanobacteria in theupwelling cells (Table 6), it takes 18 days for the first step and 58days for the second step, which is quite long and does not allow themassive bloom formations in July and in August. However, phos-phorus uptake rates increase significantly with increasing C:P ratios(Table 6). Thus, assuming that cyanobacteria with high C:P ratiosand a high activity of 1.53% mg�1 h�1 are mixed into the DIP-richupwelled water, the gross uptake rate would increase to3.8 nmol mg�1 h�1. Under these circumstances, it would take only4–5 days to reduce the C:P ratio. This time span is in agreementwith investigations of Vahtera et al. (2005), who found that primaryproduction increased 5 days after the start of an upwelling.

Comparison of the C:P ratios of isolated cyanobacteria withthose of the whole water sample indicated that filamentous cya-nobacteria are more variable in their intracellular phosphoruscontent than other phytoplankton organisms. Cyanobacteria inupwelling regions accumulate DIP more extensively and, if DIP isdepleted, are able to reduce their intracellular phosphorus contentto a greater extent than other phytoplankton organisms. Since lowC:P ratios has been found also outside the upwelling, it is not clearhow much DIP is accumulated by other plankton organisms thatpopulate upwelling areas. However, the high ability of the cyano-bacteria to enrich phosphorus is in contrast to the lower uptakerates compared with the other phytoplankton.

In light of the questions posed at the beginning of this study, ourresults can be summarized as follows: Filamentous cyanobacteriaare able to accumulate phosphorus directly in upwelling areas andthey become enriched in phosphorus to a greater degree than other

ted filamentous cyanobacteria.

Isolated filamentous cyanobacteria

o [33P]PO4 uptake DIP gross uptake C:P ratio

(% mg�1 h�1) (nmol mg�1 h�1)

0.15 0.21 177.80.26 0.13 183.80.12 0.21 111.70.06 0.14 187.51.53 0.31 379.50.58 0.07 239.8

M. Nausch et al. / Estuarine, Coastal and Shelf Science 83 (2009) 434–442442

phytoplankton organisms; therefore, they preferentially benefitfrom DIP input. Differences in the response times of N. spumigenaand A. baltica were not observed. DIP accumulation may be thebasis for an 8- to 13-fold increase in biomass and for bloomformation, which is accompanied by increased C:P ratios. However,the exact mechanism of phosphorus accumulation remains to bedetermined.

Acknowledgements

We thank the crew of the RV ‘‘Poseidon’’ for technical support onboard. For their valuable work on board and in the laboratory, wethank: B. Sadkowiak, J. Ficker, T. Heene, R. Hansen, I. Topp, D.Setzkorn, A. Welz.. Best thanks to J. Donath and M. Gerth forpreparing some of the figures.

References

Ennet, P., Kuosa, H., Tamsalu, R., 2000. The influence of upwelling and entrainmenton the algal bloom in the Baltic Sea. Journal of Marine Systems 25, 359–367.

Fonselius, S., 1996. The upwelling of nutrients in the central Skagerrak. Deep SeaResearch II 43, 57–71.

Gidhagen, L., 1987. Coastal upwelling in the Baltic Sea - Satellite and in situmeasurements of sea surface temperatures indicating coastal upwelling.Estuarine, Coastal and Shelf Science 24, 449–462.

Grasshoff, K., Ehrhardt, M., Kremling, K., 1983. Methods of Seawater Analysis,second ed. Verlag Chemie, Weinheim, 419 pp.

Haapala, J., 1994. Upwelling and its influence on nutrient concentration in thecoastal area of the Hanko Peninsula, Entrance of the Gulf of Finland. Estuarine,Coastal and Shelf Science 38, 507–521.

Janssen, F., Neumann, T., Schmidt, M., 2004. Inter-annual variability in cyanobac-teria blooms in the Baltic Sea controlled by wintertime hydrographic condi-tions. Marine Ecology Progress Series 275, 59–68.

Jeffrey, S.W., Welschmeyer, N.A., 1997. Spectrophotometric and fluorometric equa-tions in common use in oceanography. In: Jeffrey, S.W., Mantoura, R.F.C.,Wright, S.W. (Eds.), Phytoplankton Pigments in Oceanography. UNESCOPublishing, Paris, pp. 597–615.

Kahru, M., Savchuk, O.B., Elmgren, R., 2007. Satellite measurements of cyano-bacterial bloom frequency in the Baltic Sea: interannual and spatial variability.Marine Ecology Progress Series 343, 15–23.

Kononen, K., Niemi, Å, 1986. Variation in phytoplankton and hydrography in theouter archipelo. Finnish Marine Research 253, 35–51.

Kowalewski, M., 2005. The influence of the Hel upwelling (Baltic Sea) on nutrientconcentrations and primary production – the results of an ecohydrodynamicmodel. Oceanologia 47, 567–590.

Kowalewski, M., Ostrowski, M., 2005. Coastal up- and downwelling in the SouthernBaltic. Oceanologia 47, 453–475.

Laanemets, J., Lilover, M.-J., Raudsepp, U., Autio, R., Vahtera, E., Lips, I., Lips, U., 2006.A fuzzy logic model to describe the cyanobacteria Nodularia spumigena bloomsin the Gulf of Finland, Baltic Sea. Hydrobiologia 554, 31–45.

Larsson, U., Hajdu, S., Walve, J., Elmgren, R., 2001. Baltic Sea nitrogen fixation fromthe summer increase in upper mixed layer total nitrogen. Limnology andOceanography 46, 811–820.

Lehmann, A., Myrberg, K., 2008. Upwelling in the Baltic Sea – a review. JournalMarine Systems 74, S3–S12.

Lilover, M.J., Laanemets, J., 2006. A simple tool for the early prediction of thecyanobacteria Nodularia spumigena bloom biomass in the Gulf of Finland.Oceanologia 48 (S), 213–229.

Lilover, M.J., Stips, A., 2008. The variability of parameters controlling the cyanobac-teria bloom biomass in the Baltic Sea. Journal of Marine Systems 74, 108–115.

Lass, H.-U., Schmidt, T., Seifert, T., 1996. Hiddensee upwelling field measurementsand modelling results. ICES Cooperative Research Report 257, 204–207.

Lass, H.-U., Mohrholz, V., Nausch, G., Siegel, H.. On phospate pumping into thesurface layer of the eastern Gotland basin by upwelling. Journal of MarineSystems, submitted for publication.

Myrberg, K., Andrejev, O., 2003. Main upwelling regions in the Baltic Sea – a statisticalanalysis based on three-dimensional modelling. Boreal and EnvironmentResearch 8, 97–112.

Nausch, G., Feistel, R., Lass, H.-U., Nagel, K., Siegel, H., 2007. Hydrographisch-chemische Zustandseinschatzung der Ostsee 2006. Marine Science Reports,Warnemunde 70, 1–91.

Nausch, M., Nausch, G., 2007. Bioavailable dissolved organic phosphorus andphosphorus use by heterotrophic bacteria. Aquatic Biology 1, 151–160.

Nausch, G., Nehring, D., Nagel, K., 2008a. Nutrient concentrations, trends and theirrelation to eutrophication. In: Feistel, R., Nausch, G., Wasmund, N. (Eds.), Stateand Evolution of the Baltic Sea, 1952-2005. J. Wiley & Sons, Hoboken,pp. 337–366.

Nausch, M., Nausch, G., Wasmund, N., Nagel, K., 2008b. Phosphorus pool variationsand their relation to cyanobacteria development in the Baltic Sea: a three-yearstudy. Journal Marine Systems 71, 99–111.

Nommann, S., Sildam, J., Noges, T., Kahru, M., 1991. Plankton distribution duringcoastal upwelling event off Hiiumaa, Baltic Sea: impact of short-term flow fieldvariability. Continental Shelf Research 11, 95–108.

Rohde, K.H., Nehring, D., 1979. Ausgewahlte Methoden zur Bestimmung vonInhaltsstoffen im Meer- und Brackwasser. Geodadische Geophysikalische Ver-offentlichungen Reihe IV 27, 1–68.

Suursaar, U., Aps, R., 2007. Spatio-temporal variations in hydro-physical and–chemical parameters during a major upwelling event off the southern coast ofthe Gulf of Finland in summer 2006. Oceanologia 49, 209–228.

Siegel, H., Gerth, M., Tschersich, G., 2008. Satellite derived sea surface temperaturefor the period 1990–2005. In: Feistel, R., Nausch, G., Wasmund, N. (Eds.), Stateand Evolution of the Baltic Sea. 1952–2005. John Wiley & Sons, Hoboken,pp. 241–264.

Utermohl, H., 1958. Zur Vervollkommnung der quantitativen Phytoplankton-Methodik. Mitt. Internationale Verhandlungen der theoretischen und ange-wandten Limnologie 9, 1–38.

Vahtera, E., Laanemets, J., Pavelson, J., Huttunen, M., Kononen, K., 2005. Effect ofupwelling on the pelagic environment and bloom-forming cyanobacteria in thewestern Gulf of Finland, Baltic Sea. Journal Marine Systems 58, 67–82.

Wasmund, N., Topp, I., Schories, D., 2006. Optimising the storage and extraction ofchlorophyll samples. Oceanologia 48, 125–144.

Wasmund, N., Nausch, G., Voss, M., Recruitment of cyanobacteria blooms in theBaltic Sea. Marine Ecology Progress Series, submitted for publication.

Zalewski, M., Ameryk, K., Szymelfenig, M., 2005. Primary production and chloro-phyll a concentration during upwelling events along the Hel Peninsula (TheBaltic Sea). Oceanological and Hydrobiological Studies XXXIV (Suppl. 2),97–113.