Content uploaded by Alexandra C Kraberg
Author content
All content in this area was uploaded by Alexandra C Kraberg on Feb 03, 2021
Content may be subject to copyright.
Author's personal copy
Long-term trends in phytoplankton composition in the western and central Baltic Sea
Norbert Wasmund
a,
⁎, Jarno Tuimala
b
, Sanna Suikkanen
c
, Leen Vandepitte
d
, Alexandra Kraberg
e
a
Leibniz Institute for Baltic Sea Research, Seestr. 15, D-18119 Warnemünde, Germany
b
Finnish Red Cross Blood Service, Kivihaantie 7, FI-00310 Helsinki, Finland
c
Finnish Environment Institute, Marine Research Centre, P.O. Box 140, FI-00251 Helsinki, Finland
d
Flanders Marine Institute (VLIZ), Wandelaarkaai 7, B-8400 Oostende, Belgium
e
Biologische Anstalt Helgoland, Alfred-Wegener Institute for Polar and Marine Research, Kurpromenade 201, D-27498 Helgoland, Germany
abstractarticle info
Article history:
Received 5 November 2010
Received in revised form 16 March 2011
Accepted 21 March 2011
Available online 5 April 2011
Keywords:
Long-term changes
Trend breaks
Phytoplankton
Diatoms
Dinoflagellates
Abiotic factors
Baltic Sea (10°E 54°N) (21°E 58°N)
The phytoplankton biomass data of the period 1979–2005 of the Belt Sea area and the Baltic Proper, separated
into spring, summer and autumn data, were checked for trends, together with the relevant abiotic factors
(temperature, salinity, and nutrient concentrations). The Mann–Kendall test was used for detecting monotonic
trends over the whole investigation period or, if trend breaks occurred, over the period before and after the
trend breaks. The relationships between phytoplankton community composition and the environmental
variables were assessed by a redundancy analysis (RDA), which could support some results of the trend
analyses. Water temperature increased but salinity and inorganic nitrogen concentrations decreased in the
southern Baltic Proper. Spring phytoplankton biomass and chlorophyll aconcentrations increased in the Baltic
Proper and decreased in Mecklenburg Bight. The biomass of Diatomophyceae decreased in spring at some
stations but increased in autumn. If the Diatomophyceae spring blooms decreased, the total Dinophyceae
biomass increased. Strong spring blooms of Diatomophyceae occurred in the 1980s and since 2000, but those of
Dinophyceae in the 1990s. These two groups showed alternating oscillations. Trends in most phytoplankton
components were different in the Baltic Proper and the Belt Sea area, confirming that Darss Sill is a biological
border.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
Man lives in a changing world, but his activities provoke
environmental changes which are much stronger than naturally
occurring evolution. To counteract the adverse effects of his activities
is of vital importance for his future welfare. Actions to be undertaken
involve the identification, analysis and mitigation of anthropologically
caused changes. In this paper, we concentrate on the identification of
changes in the marine environment, namely the phytoplankton of the
Baltic Sea, irrespective of whether they are anthropogenic or natural.
The Baltic Sea is heavily impacted by eutrophication, caused by
nutrient input from the densely populated and intensely cultivated
catchment area and from the atmosphere, resulting in an increase in
phytoplankton biomass, primary production and turbidity in the
euphotic zone and oxygen deficit in deep water layers already in the
early 1970s (Elmgren, 2001; Rönnberg and Bonsdorff, 2004). The
riparian countries recognised the increasing environmental problems
and agreed to establish the Baltic Marine Environment Protection
Commission (Helsinki Commission, HELCOM) in 1974. One of its aims
was to investigate long-term trends in trophic conditions by the Baltic
Monitoring Programme (BMP, later the COMBINE programme), which
has been conducted since 1979 according to a coordinated sampling
schedule and binding methods.
The results of the HELCOM monitoring have been analysed in
periodic assessments (e.g. HELCOM, 1996, 2002) or, recently, in
thematic assessments (e.g. on eutrophication, see HELCOM, 2009) and
Indicator Fact Sheets (e.g. Jaanus et al., 2007), published on the
HELCOM web page. They show that not only eutrophication but also
climate is a major trigger for changes in the Baltic ecosystem (BACC,
2008; HELCOM, 2007).
Hickel (1998) and Edwards et al. (2006) claimed that some
phytoplankton bloom events, e.g. in the North Sea, may have been
incorrectly attributed to eutrophication while the real modifier of
change was of climatic or hydrodynamic origin. For several areas
including the Baltic (Wasmund et al., 1998) and Mediterranean Sea
(Goffart et al., 2002), a decrease in diatom and an increase in
dinoflagellate abundance was detected in the late 1980s. In both cases
this was related to the higher stability of the water column in the
winter-spring period due to increasing winter temperatures. These
higher winter temperatures were associated to a high positive North
Atlantic Oscillation (NAO) index (Boyce et al., 2010; Reid et al., 2001).
Significant changes in different trophic levels (phytoplankton,
zooplankton, fish) and in larger areas (both in the central Baltic and
Journal of Marine Systems 87 (2011) 145–159
⁎Corresponding author. Tel.: +49 381 5197212; fax: +49 381 5197440.
E-mail addresses: norbert.wasmund@io-warnemuende.de (N. Wasmund),
jtuimala@gmail.com (J. Tuimala), sanna.suikkanen@ymparisto.fi(S. Suikkanen),
leen.vandepitte@vliz.be (L. Vandepitte), alexandra.kraberg@awi.de (A. Kraberg).
0924-7963/$ –see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmarsys.2011.03.010
Contents lists available at ScienceDirect
Journal of Marine Systems
journal homepage: www.elsevier.com/locate/jmarsys
Author's personal copy
the North Sea) in the late 1980s were termed a regime shift by Alheit
et al. (2005). Likewise, a regime shift occurred in the North Pacific
after a warming in 1988–1989 (Chiba et al., 2008; Hare and Mantua,
2000; Tian et al., 2008). Besides regime shifts in 1977/79 and 1988/89,
indications of a further regime shift were identified in 1998 both in
the North Pacific(Overland et al., 2008) and in the North Sea
(Weijerman et al., 2005). At almost the same time, in 1999, an
increasing trend in chlorophyll concentrations changed to a decreas-
ing trend in the permanently stratified regions of the oceans
(Behrenfeld et al., 2006).
The first step to identify complex alterations of an ecosystem is the
identification of changes in biomass and composition of the key
communities. Phytoplankton, as the basic primary producer in marine
ecosystems, is directly dependent on abiotic variables and is very
sensitive to environmental changes. The availability of consistent
long-term data is the precondition for the detection of shifts in
community composition. HELCOM provides such data which are
stored in the ICES database.
The first statistically comprehensive analysis of the HELCOM long-
term data for changes in phytoplankton composition of the Baltic
Proper and the Kattegat/Belt Sea area was performed by Wasmund
and Uhlig (2003). They analysed the data up to the year 1999 and
found downwards trends for diatoms in spring and summer whereas
dinoflagellates generally increased in the Baltic Proper, but decreased
in the Kattegat. Möllmann et al. (2006) analysed phytoplankton,
zooplankton, fisheries and abiotic data from the period 1974 to 2004
of the central Baltic Sea and the Gulf of Riga and confirmed the known
regime shift between 1987 and 1989, but found another shift between
1993 and 1994. Suikkanen et al. (2007) performed trend analyses of
the phytoplankton composition in combination with abiotic variables
for the years 1979–2003. They restricted their analyses to the summer
period and the northern Baltic Proper and the Gulf of Finland. Their
main results concerning phytoplankton groups were: increase in
biomass of chrysophytes and chlorophytes in both areas, increase in
dinophytes and decrease in euglenophytes biomass in the northern
Baltic Proper, increase in cyanobacteria and decrease in cryptophytes
biomass in the Gulf of Finland.
Henriksen (2009) investigated six stations in the Kattegat and the
western Baltic Sea for periods reaching from 1979–1997 up to 1979–
2006. He detected a decrease in phytoplankton biomass, mainly
diatoms, which correlated with reduced inputs of N to the Danish
straits and with increases in water temperature. Moreover, he
evaluated historical semi-quantitative data which showed that drastic
changes in the dominant phytoplankton species occurred during the
20th century.
Wasmund et al. (2008) compared the earliest quantitative
phytoplankton data of Kiel Bight from the beginning of the 20th
century with recent data and found shifts in species composition and
bloom characteristics. However, now as before the diatoms and
dinoflagellates are the most important components of the phyto-
plankton although the total phytoplankton biomass has roughly
doubled in the course of the last century.
As phytoplankton reacts directly to eutrophication, water acidifi-
cation and climate change, the identification of trends in phytoplank-
ton is of high scientific and political interest. Changes in
phytoplankton, e.g. composition, amplitude and timing of the blooms,
will have considerable consequences for the whole marine food web.
As the latest trend analyses covered only the northern and western
parts of the Baltic Sea, the phytoplankton of the central Baltic Sea still
requires re-evaluation because the earlier trend analysis stopped with
the year 1999. Moreover, errors have been identified in older original
data, which make a validation necessary. This paper provides a new
trend analysis updating that of Wasmund and Uhlig (2003), based on
newly validated and extended data. It not only considers taxonomic
groups but also the key species and their correlations with abiotic
variables.
2. Study area
The Baltic Sea is a shallow intra-continental shelf sea which has
only a small connection to the fully marine North Sea. The mixture of
the freshwater input, mainly from the east, with the salt water input
from the west causes a salinity gradient reaching from approximately
15 PSU in surface water of the Danish Belts to 2–3 PSU in the northern
Bothnian Bay. The topographical structuring into relatively deep
basins further increases variability. Accordingly, the Baltic is divided
into regions, each with its own characteristic properties (Fig. 1). We
restrict our analyses to the Baltic Proper (211 069 km²) and the Belt
Sea (18 273 km²), which are separated by the Darss Sill. If further
subdivided according to HELCOM (2009), our stations represent the
Eastern and Western Gotland Sea, Bornholm Sea, Arkona Sea,
Mecklenburg Bight and Kiel Bight. Our statements will only apply to
these regions (Table 1).
3. Material and methods
3.1. Sampling and analyses
Sampling, microscopic examination, and analyses of nutrient and
chlorophyll aconcentrations were performed as described in the
HELCOM guidelines (HELCOM, 1988), which have been only slightly
modified during the three decades of the running monitoring
programme. The phytoplankton biomass was calculated as described
by Olenina et al. (2006).
3.2. Database
The data originate from the monitoring programme of HELCOM
and from diverse research projects conducted by the Leibniz-Institute
for Baltic Sea Research (IOW), provided that the HELCOM methods
were applied. These data were already used by Wasmund and Uhlig
(2003), but were revised for this publication. The data were updated
by contributions of recent monitoring data by Sweden (Stat. J1, K2,
and K4), Poland (Stat. K1 and K2), Lithuania (Stat. J1), Denmark (Stat.
K2) and by own data (all stations).
After assembling the data they were checked for quality. We
checked 68013 species level data records, and corrected errors if
original information was sufficient for a recalculation. For example,
biomass calculations based on abnormal volumes of the counting
units, sometimes occurring for Aphanothece or Gomphosphaeria
colonies, were corrected. If recalculation was not possible, the data
of the complete sample was deleted. Mixed samples covering depth
intervals deeper than 10 m were excluded. The species or higher taxa
were assigned to classes and most of the “unidentified”taxa could also
be allocated to a class. Heterotrophic (H) species were treated
separately from autotrophic (A) and mixotrophic (M) species.
Stations in close proximity to each other and representing very
similar environmental conditions were combined for statistical
analyses. Stations K4, K5 and K7 were combined to a single station
called K457, and stations M1 and M2 were combined to a station M12.
These stations have a higher priority in our discussion than other
stations with smaller amounts of data.
The monitoring programme covers all seasons but accentuates the
growing season. The different bloom periods (spring, summer, and
autumn) have to be analysed separately because they are charac-
terised by completely different phytoplankton communities. The
complete data set was split into four parts according to the seasons
(Table 2), but winter was excluded from further analyses due to
insufficient data. Note that the seasons differ in the Baltic regions
because the spring bloom starts earlier in the Belt Sea than in the
central Baltic Proper.
146 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
3.3. Statistical analysis
3.3.1. Trend analysis
The Mann–Kendall trend test (McLeod, 2005) was used for
detecting monotonic trends over time. Possible trend break points
were detected by fitting a spline with one degree of freedom into the
data (Sonderekker, 2008; Toms and Lesperance, 2003; Wood, 2000). A
separate spline was fitted for every species and station combination,
but for further analyses, the breakpoints were averaged at the station
level. The presence of monotonic trends before and after the break
point was then separately tested using the Mann–Kendall trend test.
Only taxa that contained at least 10 observations per station and
season were used for the statistical analyses. All trend analyses were
performed in R 2.9.0 (R Development Core Team, 2009). Statistical
tests were two-sided, and considered significant, if the p-value was
less than or equal to 0.05.
3.3.2. Multivariate analysis
Redundancy analysis (RDA), (ter Braak, 1994), was used to assess
relationships between the phytoplankton community composition and
five environmental variables. The analysis was run with the combined
data from all stations, but separately for the three different seasons.
Phytoplankton data was linked to temperature and salinity of the same
sampling occasion, but to nutrient data from the previous sampling
event (max. one month before). All biomass data were log(x+1)-
transformed to stabilize variance and reduce the influence of dominant
taxa on the ordination. RDA wasperformed using CANOCOfor Windows
4.0 (ter Braak and Šmilauer, 1998) as described in Suikkanen et al.
(2007). Due to the limitations of the program, samples containing
missing environmental data had to be omitted.
4. Results and discussion
4.1. Changes in environmental drivers
The output of the trend analyses are summarized in Fig. 2 and
showed very complex patterns across stations and seasons. Our
analyses demonstrated a significant increase in the temperature of the
surface waters especially in the Arkona and Bornholm Seas in summer
and autumn. The warming of the upper water layers intensifies the
N1
M2
M1
K8
K7 K4
K5
K2
K1
J1
I1
N3
10 11 12 13 14 15 16 17 18 19 20 21 22
Longitude [deg E]
54
55
56
57
58
Latitude [deg N]
SBP
EGS
AS
Darss Sill
KB
MB
Fig. 1. The study area and sampling stations. The sea areas are abbreviated as follows: EGS—Eastern Gotland Sea, SBP—Southern Baltic Proper, AS—Arkona Sea, MB—Mecklenburg
Bight, and KB—Kiel Bight.
Table 1
Investigated stations and data frequency at the stations.
Station name Sea area Latitude (°N, in decimals) Longitude (°E, in decimals) Length of data series Number of phytoplankton samplings (n)
in spring/summer/autumn
Baltic Proper
BMP I1 Western Gotland Sea 57.12 17.67 1979–1996 23/27/32
BMP J1 Eastern Gotland Sea 57.32 20.05 1979–2005 116/126/78
BMP K1 Southern Gotland Sea 55.56 18.40 1979–2005 73/63/50
BMP K2 Bornholm Sea 55.25 15.98 1979–2006 137/141/88
BMP K4 Eastern Arkona Sea 55.00 14.08 1979–2005
BMP K5 Central Arkona Sea (GE) 54.93 13.50 1981–2005 224/191/152
a
BMP K7 Central Arkona Sea (DK) 55.00 13.30 1979–2002
BMP K8 Darss Sill 54.72 12.78 1989–2005 52/34/26
Belt Sea
BMP M1 Kadet Channel 54.47 12.22 1980–2005
BMP M2 Mecklenburg Bight 54.32 11.55 1980–2005 138/154/106
b
BMP N1 Fehmarn Belt 54.57 11.33 1979–1997 46/85/72
BMP N3 Kiel Bight 54.60 10.45 1986–2000 24/37/25
a
Stations K4, K5 and K7 combined.
b
Stations M1 and M2 combined.
147N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
stratification and reduces the upward transport of deep water of
higher salinity and nutrient contents. This may be one reason for the
strongly decreased salinity in the surface water in the Baltic Proper
(Fig. 2). The reduction in major inflows of marine water into the deep
Baltic basins during recent decades is another reason for the reduced
salinity (Matthäus et al., 2008).
A decrease in the concentrations of dissolved inorganic nitrogen
(DIN) has been demonstrated for the southern Baltic Proper in all
seasons. Phosphate concentrations changed only at two stations with a
shorter time series andare therefore less relevant. Silicate concentration
was slightly increasing in the Eastern Gotland Sea during spring and
summer and in the Mecklenburg Bight in autumn (Fig. 2).
Nausch et al. (2008) stated that the main eutrophication pulse
occurred until the 1980s. Afterwards, the nitrate plus nitrite concen-
trations decreased in the winter surface layer (0–10 m) of the Eastern
Gotland Sea and the Bornholm Sea, whereas phosphate concentrations
were strongly fluctuating at a high level. It is expected that also the
silicate concentrations decrease because of reduced river loadings (as
result of river damming) and eutrophication (Conley et al., 2008;
Humborget al., 2008). Papush andDanielsson (2006)found downwards
silicate trends from 1970 to 2001 at most stations of the Baltic proper,
but not if only the period from 1991 to 2001 was considered. Obviously,
the termination of the eutrophication trend stopped the silicate
decrease. The eutrophication ceased also in the northern Baltic Proper
in the 1980s, and the DIN:silicate ratio of winter data declined since the
1990s (Fleming-Lehtinen et al., 2008). As stated above, our results
revealed even increasing silicate trends in different seasons over the
whole investigation period at stations J1 and M12.
4.2. Trends in phytoplankton biomass
The trends in different phytoplankton taxa, separated for stations
and seasons, are summarized in Fig. 2. A selection of the most important
data series is exemplified in Figs. 3–7as a basis for discussion.
4.2.1. Total biomass
The plausibility of the trends of microscopically determined total
A +M biomass, which represents the phytoplankton, can be checked by
comparison with chlorophyll adata.We found interesting differences in
chlorophyll trends (1979–2005) if different seasons were analysed
separately (Fig. 2): The chlorophyll aconcentrations increased in the
Baltic Proper but decreased in Mecklenburg Bight in spring. Summer
data revealed no trend and autumn data only increasing trends at
stations J1 and K457. Similar results were described by Wasmund and
Siegel (2008). In the northern Baltic Proper, even increasing trends in
chlorophyll aconcentrations were found in summer data (Fleming-
Lehtinen et al., 2008; Suikkanen et al., 2007).
Our chlorophyll trends support our phytoplankton trends from
spring (Fig. 3a, b) and from autumn but not those of summer (Fig. 3c).
Therefore, and because the declining summer biomass trends are
caused by only a few extreme values which occurred just at the
beginning of the time series, we regard them with suspicion. The
autumn data of phytoplankton biomass showed peak values at the
end of the 1980s in the Baltic Proper (Fig. 3d, cf. HELCOM, 1996).
Hence, stations with shorter time-series disclose increasing trends if
covering mainly the 1980s (Station I1) and decreasing trends if
covering mainly the 1990s (Station K8). In the Belt Sea, the autumn
values tend to increase.
A recent decrease in chlorophyll concentration seems to be a world-
wide phenomenon, caused mainly by increased water temperature,
stronger water column stratification and reduced recycling of nutrients
from deeper water layers (Behrenfeld et al., 2006; Boyce et al., 2010;see
also Nixon et al., 200 9).Eutrophicationover-compensates this effect and
may lead to a sustained increase in phytoplankton biomass, asfound in
the BalticProper. Especiallythe spring chlorophyll aconcentrations may
be a good indicator of eutrophication because they best reflect the
nutrient concentrations accumulated during the winter.
4.2.2. Nostocophyceae (cyanobacteria)
The pronounced biomass decrease of Nostocophyceae in summer is of
great interest because representatives of the Nostocales, mainly the
potentially toxic genera Aphanizomenon and Nodularia,formlargeblooms
in the Baltic Proper. There has been much debate about whether these
blooms are increasing or not (Finni et al., 2001; Wasmund and Siegel,
2008). As the biomass of Nostocophyceae is related to high temperature
(Fig. 8b), warming should enhance the cyanobacteria in summer. Trend
analyses by Suikkanen et al. (2007) showed an increase in Nostocophy-
ceae in the Gulf of Finland and isolated peaks in 1985, 1995, and 1996 in
the northern Baltic Proper. In contrast, our data revealed strongly
decreasing trends in the southern Baltic (K2, K457) for Nostocophyceae
and its main representatives, Aphanizomenon sp. and Nodularia spumigena
(Fig. 3e), following pronounced peaks in 1979/1980. Obviously, nitrogen-
fixing cyanobacteria blooms are governed not only by temperature but
also some other factors like N:P ratios (Pliński and Jozwiak, 1999).
Also Kononen and Niemi (1984) reported on very high biomass of
Nodularia and Aphanizomenon in the summers 1979–1981 in comparison
to the previous decade. Using satellite images, Kahru (1997) found large
areas covered by cyanobacteria blooms in 1982–1984 and 1991–1994
and Kahru et al. (2007) observed the highest frequency of cyanobacterial
accumulations in 1984, 1999 and 2005. HELCOM presents annual
updates of satellite images on cyanobacteria blooms (Hansson and
Öberg, 2009) and on composition of these blooms (Kaitala and Hällfors,
2008). Representative sampling of the buoyant and highly patchy surface
blooms is difficult and would require many replicates (Rolff et al., 2007).
As an extremely high biomass in Nostocophyceae was coincidentally
recorded at the beginning of the time-series, a long-lasting decreasing
trend is expected, which might represent an artefact resulting from
methodological difficulties and has to be regarded with caution.
4.2.3. Dinophyceae (dinoflagellates)
Only the A+ M Dinophyceae are considered in this section. They
form late spring blooms in the Baltic Proper, sometimes after a diatom
bloom, and are therefore associated with higher temperature and
lower nutrient concentrations than the diatoms (Fig. 8a). Our data
indicate increasing trends in the spring blooms of Dinophyceae at all
stations of the Baltic Proper, except for the shorter time series at
station K8 (Fig. 2). This is most impressive at station I1, whose data
series ends in 1996 (Fig. 3f). Longer data series of stations K1 (Fig. 3g),
K2 and K457, revealed trend breaks in the mid 1990s both for the total
Dinophyceae and their main taxonomic components.
The dinoflagellate spring blooms of the mid 1990s were dominat-
ed by Peridiniella catenata (Fig. 3h, no general trend but break in 1994,
compare with Fig. 3g). This contrasts with the previous decade, when
in 1979–1983 HELCOM (1996) did not report P. catenata in the list of
the 5 most important spring species in the Eastern Gotland Sea. It
seems to move from the Northern Baltic Proper to the south; it was
recorded in the Gulf of Gdansk only since the mid-1980s (Witek et al.,
1997). It generally decreases since the middle of the 1990s which
makes its total trends decreasing in contrast to that of the total A+ M
Dinophyceae. Contrary to the Gotland Sea, P. catenata has rarely
been abundant in the Bornholm and Arkona Seas (Fig. 4a). The same
holds true for the genus Gymnodinium. Dinophysiales with its main
representatives Dinophysis acuminata and D. norvegica is never
forming spring blooms in the upper water layer (Fig. 4b). Their
Table 2
Definition of seasons according to the HELCOM strategy (e.g. HELCOM, 1996).
Season Belt Sea Baltic Proper
Spring February–April March–May
Summer May–August June–September
Autumn September–November October–December
148 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
Fig. 2. Compiled results of trend analyses for different abiotic parameters and biomass of different taxa, separated for stations and seasons. Significant trends are indicated by colours. If trend breaks occurred, the significant trends before and
after the break point are marked by “+”or “−“. These average break points (in contrast to those depicted in Figs. 3–7) have been calculated by averaging the break points of all species for a specific station.
149N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
highest spring biomass is found at the beginning of the 1990 s. The
increase in total A+M Dinophyceae despite a decrease in single
components may be explained by an increase in unidentified small
Dinophyceae in spring. One of these hardly identifiable groups is that
of Scrippsiella/Biecheleria/Gymnodinium (Kremp et al., 2005; Moestrup
et al., 2009; Sundström et al., 2009).
Fig. 3. Biomass of selected taxa over the investigation period at stations and in seasons specified in the headlines. A curve estimated with a locally weighted scatterplot smoother
(LOWESS) is plotted with the solid line, and its 95% confidence interval with a dashed line. The p-value from the Mann–Kendall trend test is also indicated. The vertical dashed lines
indicate break point locations for the taxa or ratios that had significant trends before and/or after the trends breaks.
150 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
Summer A+M Dinophyceae and their representatives show a
decreasing trend. These were detectable in the Gymnodiniales (Gymno-
dinium spp.) and Peridiniales (Fig. 4c) at stations K2, K457, M12, and N1.
They are less important as the biomass of these taxa did in general not
exceed 500 mg m
−3
apart from one peak in the Gymnodiniales
(1400 mg m
−3
) at station M12 in 1994. Peaks of Prorocentrum micans
Fig. 4. Biomass of selected taxa over the investigation period at stations and in seasons specified in the headlines. A curve estimated with a locally weighted scatterplot smoother
(LOWESS) is plotted with the solid line, and it's 95% confidence interval with a dashed line. The p-value from the Mann-Kendall trend test is also indicated. The vertical dashed lines
indicate break point locations for the taxa or ratios that had significant trends before and/or after the trends breaks.
151N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
(900 mg m
−3
)andP. minimum (1500 mg m
−3
) were found at stations
M1 and K2, respectively, in August 1990 and in August 1992. Summer
A+M Dinophyceae, including Gonyaulacales and Prorocentrales, were
positively related to temperature, whereas Dinophysiales preferred
colder water (Fig. 8b). The latter sometimes concentrate at greater
water depths (Carpenter et al., 1995; Gisselson et al., 2002).
Autumn biomass of A+M Dinophyceae was increasing at stations
K1 and K2 (Fig. 2) but it did not exceed 260 mg m
−3
. The decreasing
Fig. 5. Biomass of selected taxa over the investigation period at stations and in seasons specified in the headlines. A curve estimated with a locally weighted scatterplot smoother
(LOWESS) is plotted with the solid line, and it's 95% confidence interval with a dashed line. The p-value from the Mann-Kendall trend test is also indicated. The vertical dashed lines
indicate break point locations for the taxa or ratios that had significant trends before and/or after the trends breaks.
152 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
trends of Prorocentrales and Gymnodiniales in the southern Baltic
Proper were of lower importance because biomass stayed below
250 mg m
−3
in autumn, except for the year 1997, when Gymnodiniales
exceptionally reached 3500 mg m
−3
. Prorocentrales were of some
importance in Mecklenburg Bight, with biomass regularly up to
1000 mg m
−3
.P. micans was continuously decreasing (Fig. 4d) whereas
Fig. 6. Biomass of selected taxa over the investigation period at stations and in seasons specified in the headlines. A curve estimated with a locally weighted scatterplot smoother
(LOWESS) is plotted with the solid line, and it's 95% confidence interval with a dashed line. The p-value from the Mann-Kendall trend test is also indicated. The vertical dashed lines
indicate break point locations for the taxa or ratios that had significant trends before and/or after the trends breaks.
153N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
P. minimum formed the maximum in 1997 in autumn. P. minimum
invaded into the Baltic Sea, expanded until the mid of the 1990s and
decreased afterwards (Olenina et al., 2010).
Dinophyceae were most important in autumn in the Belt Sea
because of the blooms of Ceratium spp. These Ceratium blooms are the
most stable feature in the succession during the last 100 years
Fig. 7. Biomass of selected taxa over the investigation period at stations and in seasons specified in the headlines. A curve estimated with a locally weighted scatterplot smoother
(LOWESS) is plotted with the solid line, and it's 95% confidence interval with a dashed line. The p-value from the Mann-Kendall trend test is also indicated. The vertical dashed lines
indicate break point locations for the taxa or ratios that had significant trends before and/or after the trends breaks.
154 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
(Wasmund et al., 2008). Accordingly, no trends in biomass of
Ceratium tripos were detected (Fig. 4e), except for station N3 with
its shorter time-series. The A + M Dinophyceae of the autumn,
including the Peridiniales, Gonyaulacales, Dinophysiales and Pro-
rocentrales, prefer high temperature and salinity (Fig. 8c). The
hydroclimatic changes that occurred since the late 1980s favour the
growth and earlier succession of dinoflagellates, which are positively
correlated with the NAO (Edwards et al., 2006).
4.2.4. Diatomophyceae (diatoms)
The Diatomophyceae appear early in the year and are related to low
temperature and high nutrient concentrations, as shown in Fig. 8a. A
very interesting feature is the sudden decrease in spring diatoms in the
Baltic Proper at the end of the 1980s, caused by decreasing water
temperature (Alheit et al., 2005; Wasmund et al., 1998; Wasmund and
Uhlig, 2003). Suikkanen et al. (2007) reported a strong decrease in
diatoms atthe end of the 1980s evenin the summer data of thenorthern
Baltic Proper and Henriksen, 2009) in the annual biomass means of the
Kattegat. Because of limited sampling frequency, the peak of the spring
bloom could only be sampled on few occasions. However, (Wasmund
et al., 1998) proved by calculation of the silicate consumption that the
spring diatom communities grew very well in the 1980s but suddenly
failed to grow in the 1990s. Shorter data series, as available from Station
I1 until 1996, revealed a decreasing trend (Fig. 4f). However, the longer
data series of stations J1 and K2 do not confirm a monotonous trend
(Fig. 4g, h), showing high values in the 1980s, followed by a decrease at
the end of the 1980s and a recovery since the year 2000.
The increase after 2000 was most pronounced at Station K2 with a
strong increase in Skeletonema costatum sensu lato in 2000–2003
(Fig. 5a) and of Chaetoceros spp. in 2004–2005. Henriksen (2009) also
reported these taxa especially in the 1980s and after the year 2000.
These important representatives of the spring bloom were also the
dominant Diatomophyceae at the beginning of the 20th century
(Henriksen, 2009; Lohmann, 1908; Wasmund et al., 2008). Achnanthes
taeniata, which is the main representative of the Bacillariales in the
Baltic Proper, was dominant in spring blooms at the beginning of the
1980s (HELCOM, 1996; Kononen and Niemi, 1984) but decreased and
did not recover. At station M12, theDiatomophyceae decreased more or
less continuously, mainly caused by decreases of Thalassiosira (Fig. 5b),
A. taeniata (Fig. 5c), Chaetoceros spp., S. costatum and Leptocylindrus
danicus.Henriksen (2009) also described a strong decline of the cold
water species Detonula confervacea.
The recovery of spring Diatomophyceae, namely only the Centrales,
is most apparent when compared with the competing Dinophyceae. The
ratio of these two dominating classes was used as an indicator of the
ecosystem state by McQuatters-Gollop et al. (2009). They found
increasing ratios of Diatomophyceae:Dinophyceae in the coastal North
Sea, the northern Adriatic and in the north-western Black Sea, but did
not try it for the Baltic Sea. We show this Diatomophyceae:Dinophyceae
ratio for a representative stationof the southern Baltic Proper in Fig. 4h.
Diatomophyceae were strongly dominating over the Dinophyceae
during the 1980s and since the year 2000. A decrease in the
Diatomophyceae:Dinophyceae ratio in the North Seawas also described
by Edwardset al. (2006) and related tochanges in the NAO andresulting
progressive warming and stronger stratification of the water.
The summer data of total Diatomophyceae showed no clear trend.
At station K2, both the Bacillariales and the Eupodiscales were
decreasing because of single high biomass data in 1980 and 1981. At
station M12, S. costatum was decreasing but Pseudo-nitzschia spp. was
increasing.
The autumn biomass of total Diatomophyceae was increasing in
most areas (Figs. 2and 5d, e). It is interesting that exceptionally high
biomass occurred in 1988–1990, the years of the strongest decline of the
spring diatoms (cf. Fig. 4f, g). The relatively low diatom biomass in
spring of theyears 1989–1999 is obviously compensated by highdiatom
biomass in autumn in the same period. The strong autumn bloom in
a
b
c
Axis 1 (λ1= 0.087)
Axis 2 (λ2= 0.030)
Axis 1 (λ1= 0.085)
Axis 2 (λ2= 0.048)
Axis 1 (λ1= 0.187)
Axis 2 (λ2= 0.035)
Fig. 8. Correlation plots of the redundancy analysis (RDA) for (a) spring, (b) summer
and (c) autumn data, on the relationship between the biomass of phytoplankton taxa
(solid vectors) and environmental variables (dashed vectors). The plots display 11.7,
13.3 and 22.2% of the variance in the phytoplankton biomass in spring, summer and
autumn, respectively, and eigenvalues of the first two axes are indicated by λ1 and λ2.
Codes of the phytoplankton taxa are explained by bold letters in Table 3. Asterisks
indicate statistical significance (pb0.05) of environmental variables.
155N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
1988–1990 was found at stations I1,J1, K1, K457 and M12 and is already
known fromliterature (HELCOM, 1996). Thediatom autumn blooms are
regularly formed by Coscinodiscus granii (cf. Fig. 5f). Cerataulina pelagica
and Actinocyclus sp. show pronounced peaks in autumn 1993/1994 at
station M12. Other species of Diatomophyceae decrease more or less
continuously. However, Chaetoceros increased in comparison to
Skeletonema in areas of higher salinity (Arkona Sea and Belt Sea;
Fig. 5g). The potentially toxic Pseudo-nitzschia spp. occurred in summer
and autumn mainly in the Mecklenburg Bight and is strongly related
with salinity (Fig. 8c). Peaks were noted in summer 2002 and inautumn
2003 (Fig. 5h). An increase in this genus was also shown by Henriksen
(2009).
4.2.5. Smaller phytoplankton groups
Dictyochophyceae are sometimes bloom-forming in Mecklenburg
Bight in late spring or early summer and are associated with high
salinity and high nutrient concentrations but low temperature
(Fig. 8a, b). They are strongly increasing (Fig. 6a). In autumn, they
occur in lower biomass (up to 250 mg m
−3
) with a peak in the year
2000. The main representative is Dictyocha speculum, which occurs
mostly in its naked form (Jochem and Babenerd, 1989). Therefore it is
possible that it was not recognized in earlier years of investigation.
Differentiation between this naked D. speculum and Pseudochattonella
verruculosa is difficult but both belong to the same class.
Cryptophyceae are not bloom-forming. They increased in the southern
Baltic Proper in spring and autumn (e.g. Fig. 6b) but decreased at station
K457 in summer. At station N1, the trends were decreasing in summer
and autumn (Fig. 6c), but because of the shorter data series we cannot
exclude a strong increase after 1997, as it was found in the neighbouring
Mecklenburg Bight and the Baltic Proper. The spring Cryptophyceae
preferred low temperature and high salinity but the summer populations
preferred high temperature and low salinity (Fig. 8a, b).
Chrysophyceae decreased at station K457 in spring. This decrease
was mainly caused by a strong occurrence of Dinobryon sp. in 1987
(up to 450 mg m
−3
). At station M12, however, Chrysophyceae
increased strongly (Fig. 6d), together with Dinobryon sp. This species
contributed 5–50% to the increase in Chrysophyceae. Suikkanen et al.
(2007) identified increasing trends for Chrysophyceae in the northern
Baltic Proper and the Gulf of Finland. Chrysophyceae preferred high
temperature and low salinity in spring (Fig. 8a).
Chlorophyceae are generally less important in the open sea, especially
in the more saline areas, because they are related to freshwater (cf.
Fig. 8a–c). Our study showed biomass decreases in different sea areas
and seasons. One example is displayed in Fig. 6e. However, Suikkanen
et al. (2007) reported increasing trends in the northern Baltic Proper
and the Gulf of Finland, in line with decreasing salinities, in summer
1979–2003.
Euglenophyceae revealed decreasing trends. In spring, a strong
peak (up to 1100 mg m
−3
) occurred in the southern Baltic Proper and
Mecklenburg Bight in 1994 and 1995, respectively. Also the summer
data have maxima (up to 800 mg m
−3
) between 1991 and 1995
(Fig. 6f). A negative trend for Euglenophyceae was also found by
Suikkanen et al. (2007) in the northern Baltic Proper. This class is
associated with high temperature and low salinity (Fig. 8a, b).
Prasinophyceae are generally decreasing. The peaks are frequently
found in 1990–1993 (Fig. 6g). The apparent slight increase at station
I1 in autumn is again more likely to be due to the shorter time series in
comparison with the other stations. This class is associated with high
temperature and low salinity (Fig. 8a–c).
Ciliophora A is identical with the only auto- or mixotrophic species
of Ciliophora in the Baltic Sea, Mesodinium rubrum. It is related to low
salinity and partly to high temperature (Fig. 8a–c) and is increasing in
all seasons and all regions in the Baltic proper (Fig. 6h), as expected
from the trends in salinity and temperature. M. rubrum formed spring
blooms mainly in the years 1999 and 2000. Its biomass was much
lower in autumn (Fig. 7a). It was strongly increasing during the 20th
century (Henriksen, 2009; Wasmund et al., 2008).
4.2.6. Heterotrophs
In the microscopical analysis of phytoplankton, heterotrophic
flagellates are also counted which are of the same size range as
phytoplanktonic nano- and microalgae. This heterotrophic biomass,
called Nanoflagellates H in Figs. 2, 7 and 8, is obviously decreasing in all
sea areas and more or less in all seasons (mainly spring/summer;
Fig. 7b). Theslight increase at station N3 inautumn is of little explanatory
power because it is based on only 10 records and the data series is short.
At most stations, a peak occurred in 1993/1994.
Ebriidea (with Ebria tripartita) show a decreasing trend, as well
(Fig. 7c). They prefer high temperature and low salinity (Fig. 8a, b).
Choanoflagellidea are a small group with biomass smaller than
35 mg m
−3
, except for outliers in 1984 at station N1 (autumn), and in
1990 at stations M12 (summer) and K8 (spring). They may sometimes
be overlooked and underestimated. They were strongly decreasing in
the southern Baltic Proper (Fig. 7d) and slightly increasing in
Mecklenburg Bight in different seasons. This classpreferred high salinity
and low temperature (Fig. 8a, b). Although usually considered a minor
group, they are nevertheless significant predators of bacteria and as
such can be an important component in the carbon cycle.
Many Dinophyceae (e.g. Protoperidinium spp.) are heterotrophic. Just
like the A+M Dinophyceae, they show increasing trends in the spring
and autumn data. At more or less all stations heterotrophic dinoflagellates
peaked in the 1990s with a subsequent decrease (Fig. 7e), an exception
being station K8 in autumn as this time series started only in 1989 and
therefore only represents the late phase of biomass decrease. The only
species of the heterotrophic Dinophyceae that showed trends was
Protoperidinium bipes. In contrast to the total heterotrophic Dinophyceae,
itsbiomasswasdecreasingatsomestations(Fig. 7f). As its biomass was
always very low, trends in P. bipes are of low relevance.
One interesting indicator for the state of the ecosystem, particularly of
its food web, might be the total biomass of the phytoplankton (total A + M
biomass) in comparison with the heterotrophic biomass of the same size
range. Whereas the phytoplankton biomass is increasing in most cases
(e.g. Fig. 3a), that of the heterotrophic flagellates is almost exclusively
decreasing. Therefore, the ratio of total A +M biomass and total H biomass
is strongly increasing (Fig. 7g, h). We have to admit that the counting of
heterotrophic flagellatesisaby-productofthe phytoplankton analysis
and may therefore be of lower precision. But if the decrease in
heterotrophic flagellates will be confirmed with the continuation of the
data series, it may indicate an increasing mismatch between phytoplank-
ton growth and the consumers of the phytoplankton's exudates. The
consequences for the food web and the ecosystem have to be investigated.
4.3. Relationships between environmental factors and phytoplankton
groups
In all seasons, the RDA yielded four significant environmental
variables explaining the variability in the phytoplankton biomass:
salinity, temperature, PO
4
and SiO
4
concentration. The effect of DIN
concentration was only significant in spring and autumn, but not in
summer (pN0.05). Together, all environmental variables considered
(and all canonical axes) accounted for 14.6, 15.3 and 25.0% of the
variation in the phytoplankton biomass data during spring, summer
and autumn, respectively. In the RDA ordination (Fig. 8), the first two
axes explained 11.7% (spring), 13.3% (summer) and 22.2% (autumn)
of the total variance in the phytoplankton biomasses.
The phytoplankton taxa in the different seasons could be classified
according to their association with these environmental variables. In
spring, the phytoplankton formed two main groups (Fig. 8a). All
Diatomophyceae including Bacillariales and Pseudo-nitzschia spp., as
well as the Dictyochophyceae and Cryptophyceae were associated
with high salinity and high inorganic nutrient concentrations, but
156 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
with low temperature. Most of the other taxa were associated with
high temperature, low salinity and low nutrient concentrations. The
opposed requirements of Diatomophyceae and Dinophyceae con-
cerning temperature, salinity and nutrient concentrations can explain
their interplay in the spring season.
In summer, the largest cluster comprised taxa that occurred mainly
at high temperature and silicate concentrations, but low salinity and
phosphate concentrations: Nostocophyceae, Cryptophyceae, Chloro-
phyceae, Euglenophyceae, Prasinophyceae, Ciliophora A and Ebriidea
(Fig. 8b). Another group was formed by the A+ M Dinophyceae,
Gonyaulacales and the Diatomophyceae, which were positively related
to temperature, salinity and phosphate concentration, but negatively to
silicate concentration. Some taxa, such as the Dinophysiales, Dictyocho-
phyceae, and Choanoflagellidea, were associated with low temperature
and silicate concentration, but high salinity and phosphate concentra-
tion, whereas the total summer phytoplankton biomass and that of the
Prorocentrales was associated with high temperature, salinity and
silicate concentration, but low phosphate concentration.
During autumn, high temperature, salinity, phosphate and silicate
concentrations, but low DIN concentrations, were related to the total
A+M phytoplankton, all dinoflagellates except Peridiniales, Dictyo-
chophyceae, Chrysophyceae, Flagellates H and Ebriidea (Fig. 8c). The
diatoms were associated with high temperature, salinity and nutrient
concentrations, and the Prasinophyceae and Ciliophora A with high
temperature and silicate concentrations, but with low salinity,
phosphate and DIN concentrations. The biomass of Nostocophyceae
and Choanoflagellidea was highest in low salinity, temperature and
nutrient, especially PO
4
concentrations.
The taxa that were associated with a high temperature in all seasons
included the Dinophyceae A+M, Prorocentrales, Prasinophyceae,
Ciliophora A and Ebriidea. Nostocophyceae, Chlorophyceae, Eugleno-
phyceae, Prasinophyceae and Ciliophora A were mainly found at low
salinity and low PO
4
concentration, and the three diatom taxa and
Dictyochophyceae at high salinity and PO
4
concentration. The total
phytoplankton biomass, Nostocophyceae, DinophyceaeA + M, Gonyau-
lacales, Prorocentrales and Choanoflagellidea were strongest at low DIN
concentrations throughout the year, whereas Cryptophyceae were
related to high and Choanoflagellidea to low SiO
4
concentrations.
The relationships between nutrient concentrations and phytoplank-
ton biomassare complex, as nutrient concentrations are a precondition
but also the result of phytoplankton growth. If, for example, phosphorus
concentrations are low during cyanobacteria blooms, this does not
mean that cyanobacteria grow best without phosphorus but that they
have already consumed it. We aimed at accounting for the growth
conditions by introducing a time lag that relates the actual biomass with
the nutrient concentration measured at the preceding sampling event.
Nevertheless, the relationships between nutrient concentrations and
biomass extracted from the RDA may not reflect the real causal
connection. In some cases, the RDA could not support the relationships
between abiotic and biotic trends because we have considered only a
small part of the environmental variables that may affect individual
phytoplankton species in nature and had to neglect other important
factors like micronutrients, pH, interspecific competition, allelopathy,
grazing, turbulence etc. The explained variability is therefore rather low
in complex natural systems.
5. Summary and conclusions
The trends discussed above are summarized in Table 3.Itismost
interesting that the interactions among the main components of the
phytoplankton themselves (and with relevant environmental drivers)
actually appears to be oscillating rather than linear. Hickel (1998)
analysed the phytoplankton data of Helgoland Roads (North Sea) and
found diatom and dinoflagellate cycles which mostly alternated.
Table 3
Summary of the most relevant trends. In the column “Trend”,“+”means increase and “−“means decrease. The bold and underlined letters in the taxa names explain the abbreviations
in Fig. 8. The sea areas are abbreviated as follows: BP—Baltic Proper, SBP—Southern Baltic Proper, EGS—Eastern Gotland Sea, AS—Arkona Sea, BS—Belt Sea,MB—Me cklenburg Bight,
and KB—Kiel Bight.
Abiotic factor or taxonomical group Sea area Season Trend Example
Temperature SBP Summer+autumn +
Salinity BP Spring–autumn −
DIN SBP Spring–autumn −
Silicate (SiO
4
) EGS Spring–summer +
PHYTOPLANKTON and chlorophyll aBP Spring + Fig. 3a
PHYTOPLANKTON and chlorophyll aMB Spring −Fig. 3b
Nostocophyceae (with Nostocales and Chroococcales) SBP Summer −Fig. 3e
Dinophyceae (A+M) BP+MB Spring +, trend break in mid 1990s Fig. 3f, g
Gonyaulacales (with Peridiniella catenata) BP Spring −, peak in mid 1990s Figs. 3h and 4a
Dinophysiales (with Dinophysis acuminata) BP Spring −, peak at beginning of 1990 s Fig. 4b
Peridiniales (A+M) BP+BS Summer −Fig. 4c
Prorocentrales (with Prorocentrum micans) AS+ MB Autumn −Fig. 4d
Diatomophyceae; ratio Diatomophyceae/Dinophyceae BP Spring Tendency to decrease up to the 1990 s
and to increase since 2000
Figs. 4f–h and 5a
Diatomophyceae (with Chaetoceros spp.,
Thalassiosira spp., etc.)
MB Spring −Fig. 5b
Bacillariales (with Achnanthes taeniata) BP+MB Spring −Fig. 5c
Diatomophyceae (with Coscinodiscus spp.) BP+BS Autumn +, peak around 1990 Fig. 5d–f
Ratio Chaetoceros/Skeletonema AS - KB Autumn + Fig. 5g
Pseudo-nitzschia spp. MB Summer + autumn Peak in 2002/2003 Fig. 5h
Dictyochophyceae MB Spring + summer + Fig. 6a
Cryptophyceae BP Spring + Fig. 6b
Chrysophyceae (with Dinobryon spp.) MB Spring + Fig. 6d
Chlorophyceae BP+ BS Spring–autumn −Fig. 6e
Euglenophyceae BP+BS Spring–autumn −Fig. 6f
Prasinophyceae BP Spring–autumn −Fig. 6g
Ciliophora A(with Mesodinium rubrum) BP Spring–autumn + Figs. 6h and 7a
NANOFLAGELLATES HBP +BS Spring–autumn −Fig. 7b
Ebriidea BP+BS Spring+summer −Fig. 7c
Choanoflagellidea SBP Summer+ autumn −Fig. 7d
Dinophyceae HBP+BS Spring+autumn +, trend break in mid 1990s Fig. 7e
Ratio phytoplankton (A+M)/Flagellates H BP+BS Spring–autumn + Fig. 7g, h
157N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
Henriksen (2009) reported on unusually high dinoflagellate biomass in
the Kattegat/Belt Sea area in 1987–1989, which was related to
exceptionally low diatom biomass. If data of Kononen and Niemi
(1984) from the entrance to the Gulf of Finland from the period before
our data series are added, an interesting pattern of diatom versus
dinoflagellate dominance in spring emerges in the Baltic Proper:
dinoflagellate (Peridiniella) dominance in 1968–1975 and 1989–1999,
diatom dominance in 1978–1988 and since 2000. The phase of these
oscillations is approximately 10 years. This would agree with the
succession of regime shifts described in literature, as shown above. Our
data confirmed the known shift from 1988/89, which was characterised
by a quick decline of spring Diatomophyceae (Fig. 4f, g) and an increase
of spring Dinophyceae to a higher level, which lasted for almost 10 years
(Figs. 3f–hand4a, b). Our data series covered a period that was mainly
characterised by an increase in Dinophyceae in spring, which took
advantage of the decreasing salinity and nutrient concentrations, and
probably of the generally increasing temperature. A possible regime
shift in the years 1998/99, as reported by Weijerman et al. (2005),
Behrenfeld et al. (2006) and Overland et al. (2008), might be indicated
by a recovery of the spring Diatomophyceae (Figs. 4hand5a) and a
sudden but sustained increase in M. rubrum (Figs. 6hand7a) at that
time.
In years of low diatom growth in spring, their growth is strong in
autumn (compare Figs. 4f with 5d and 4g with 5e).
Our data show that the length of the data series is decisive for the
result. Short data series (e.g. station I1) show trends whichmust not be
extrapolated because they may even invert with a continuation of the
series, as shown by trend breaks in the middle of the 1990s at many
stations. For our data series of 27 years, the analyses for monotonic
trends with one break point were sufficient but a second trend break
may have occurred in somecases. A continuation of the data series may
reveal oscillations, whose identification will require other statistical
methods (e.g. wavelet analysis).
Our results also re-emphasize that the investigated area of the Baltic
Sea is not a uniform water body. Our study area alone can be divided in
two areas, the Belt Sea area and the Baltic Proper which sometimes even
show contrary trends, e.g. in spring phytoplankton biomass and
chlorophyll a. The contrast in chlorophyll trends is supported by literature
data, which show a decrease in the Kattegat/Belt Sea area (Henriksen,
2009) and an increase in the Baltic Proper (Fleming-Lehtinen et al., 2008;
Nakonieczny et al., 1991). Darss Sill is obviously a strong biological
border, as already suggested by Kell (1973) and Witkowski et al. (2005).
This paper shows that systematic monitoring programmes, con-
ducted with comparable methods are valuable tools for the identifica-
tion of long-term changes in biota if the sampling covers all seasonal
stages of the biocoenosis and all relevant sub-regions of the water body
in question. The retrospective causal analysis of ecosystem changes on
the basis of few routinely taken abiotic parameters is, however, difficult
and not always satisfying, especially if undersampling and high natural
variability mask causal relationships. The large number of samples
required for representing short-lived populations and large stretches of
water can be managed by combining the efforts of different institutions
in joint programmes. The integration of different datasets is a dedicated
strategy to improve the data coverage, provided that various pre-
conditions concerning the data quality are fulfilled (Kraberg et al., 2011;
Vandepitte et al., 2010). The methodological frameworks and in-
frastructures put in place by HELCOM provide a unique opportunity for
such large-scale analyses.
Acknowledgements
The authors wish to thank the community of phytoplankton experts
who provided data to the HELCOM database during the last three
decades. The Professional Secretary of the Helsinki Commission, Juha-
Markku Leppänen, put the HELCOM data at our disposal in April 2007.
We are grateful for the personal supply of more recent data by Susanna
Hajdu and Svante Nyberg (Stockholm University), Slawomira Gromisz
and Janina M. Kownacka (MIR Gdynia, PL), Henrik Jespersen (Born-
holms Regionskommune, DK) and Bente Brix Madsen (Orbicon, DK) as
well as Irina Olenina (Centre of Marine Research Klaipeda, LT).
References
Alheit, J., Möllmann, C., Dutz, J., Kornilovs, G., Loewe, P., Mohrholz, V., Wasmund, N.,
2005. Synchronous ecological regime shifts in the central Baltic and the North Sea
in the late 1980s. ICES Journal of Marine Science 62, 1205–1215.
BACC, 2008. Assessment of climate change for the Baltic Sea basin. Regional Climate
Studies. Springer-Verlag, Berlin, Heidelberg. 473 pp.
Behrenfeld, M.J., O'Malley, R.T., Siegel, D.A., McClain, C.R., Sarmiento, J.L., Feldman, G.C.,
Milligan, A.J., et al., 2006. Climate-driven trends in contemporary ocean productivity.
Nature 444, 752–755.
Boyce, D.G., Lewis, M.R., Worm, B., 2010. Global phytoplankton decline over the past
century. Nature 466, 591–596.
Carpenter, E.J., Janson, S., Boje, R., Pollehne, F., Chang, J., 1995. The dinoflagellate
Dinophysis norvegica: biological and ecological observations in the Baltic Sea.
European Journal of Phycology 30, 1–9.
Chiba, S., Aita, M.N., Tadokoro, K., Saino, T., Sugisaki, H., Nakata, K., 2008. From climatereg ime
shifts to lower-trophic level phenology: synthesis of recent progress in retrospective
studies of the western North Pacific. Progress in Oceanography 77, 112–126.
Conley,D.J., Humborg, C., Smedberg, E.,Rahm, L., Papush, L.,Danielsson, A., Clarke, A., et al.,
2008. Past, present and future state of the biogeochemical Si cycle in the Baltic Sea.
Journal of Marine Systems 73, 338–346.
Edwards, M., Johns, D.G., Leterme, S.C., Sverdsen, E., Richardson, A.J., 2006. Regional
climate change and harmful algal blooms in the northeast Atlantic. Limnology and
Oceanography 51, 820–829.
Elmgren, R., 2001. Understanding human impact on the Baltic ecosystem; changing
view in recent decades. Ambio 30, 222–231.
Finni, T., Kononen, K., Olsonen, R., Wallström, K., 2001. The history of cyanobacterial
blooms in the Baltic Sea. Ambio 30, 172–178.
Fleming-Lehtinen, V., Laamanen, M., Kuosa, H., Haahti, H., Olsonen, R., 2008. Long-term
development of inorganic nutrients and chlorophyll ain the open Northern Baltic
Sea. Ambio 37, 86–92.
Gisselson, L.-Å., Carlsson, P., Granéli, E., Pallon, J., 2002. Dinophysis blooms in the deep
euphotic zone of the BalticSea: do they grow in the dark? Harmful Algae 1, 401–418.
Goffart,A., Hecq, J.-H.,Legendre, L., 2002.Changes in the developmentof the winter–spring
phytoplankton bloom in the Bay of Calvi (NW Mediterranean) over the last two
decades: a response to changingclimate? Marine EcologyProgress Series 236, 45–60.
Hansson, M., Öberg, J., 2009. Cyanobacterial blooms in the Baltic Sea. HELCOM Indicator
Fact Sheet. Helsinki Commission http://www.helcom.fi/BSAP_assessment/ifs/
ifs2009/en_GB/cyanobacterial_blooms/.
Hare, S.R., Mantua, N.J., 2000. Empirical evidence for North Pacific regime shifts in 1977
and 1989. Progress in Oceanography 47, 103–145.
HELCOM, 1988. Guidelines for the Baltic Monitoring Programme for the third stage. Part
D. Biological determinands. Baltic Sea Environment Proceedings 27 D. 161 pp.
HELCOM, 1996. Third periodic assessment of the state of the marine environment of the
Baltic Sea, 1989–1993; background document. Baltic Sea Environment Proceedings
64 B. 252 pp.
HELCOM, 2002. Environment of the Baltic Sea area 1994–1998. Baltic Sea Environment
Proceedings 82 B. 215 pp.
HELCOM, 2007. Climate change in the Baltic Sea area. HELCOM Thematic Assessment in
2007. Baltic Sea Environment Proceedings 111. 49 pp.
HELCOM, 2009. Eutrophication in the Baltic Sea—an integrated thematic assessment of
the effects of nutrient enrichment and eutrophication in the Baltic Sea region. Baltic
Sea Environment Proceedings 115B. 148 pp.
Henriksen, P., 2009. Long-term changes in phytoplankton in the Kattegat, the Belt Sea,
the Sound and the western Baltic Sea. Journal of Sea Research 61, 114–123.
Hickel, W., 1998. Temporal variability of micro- and nanoplankton in the German Bight
in relation to hydrographic structure and nutrient changes. ICES Journal of Marine
Science 55, 600–609.
Humborg, C., Smedberg, E., Medina, M.R., Mörth, C.-M., 2008. Changes in dissolved
silicate loads in the Baltic Sea—the effects of lakes and reservoirs. Journal of Marine
Systems 73, 223–235.
Jaanus, A., Andersson, A., Hajdu, S., Huseby, S., Jurgensone, I., Olenina, I., Wasmund, N.,
et al., 2007. Shifts in the Baltic Sea summer phytoplankton communities in 1992–
2006. HELCOM Indicator Fact Sheet. Helsinki Commission http://www.helcom.fi/
environment2/ifs/ifs2007/en_GB/Phytoplankton/.
Jochem, F., Babenerd, B., 1989. Naked Dictyocha speculum—a new type of phytoplankton
bloom in the western Baltic. Marine Biology 103, 373–379.
Kahru, M., 1997. Using satellites to monitor large-scale environmental change: a case
study of cyanobacteria blooms in the Baltic. In: Kahru, M., Brown, C.W. (Eds.),
Monitoring algal blooms. Springer, Berlin, pp. 43–61.
Kahru, M., Savchuk, O.P., Elmgren, R., 2007. Satellite measurements of cyanobacterial
bloom frequency in the Baltic Sea: Interannual and spatial variability. Marine
Ecology Progress Series 343, 15–23.
Kaitala, S., Hällfors, S., 2008. Cyanobacteria bloom index. HELCOM Indicator Fact Sheet.
Helsinki Commission http://www.helcom.fi/BSAP_assessment/ifs/ifs2008/en_GB/
CyanobacteriaBloomIndex/.
Kell, V., 1973. Die Darßer Schwelle - eine biologische Grenze? Wissenschaftliche
Zeitschrift der Universität Rostock, Mathematisch-naturwissenschaftliche Reihe 22
(6/7), 617–623.
158 N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159
Author's personal copy
Kononen, K., Niemi, Å., 1984. Long-term variation of the phytoplankton composition at
the entrance to the Gulf of Finland. Ophelia (Suppl. 3), 101–110.
Kraberg, A.C., Wasmund, N., Vanaverbeke, J., Schiedek, D., Wiltshire, K.H., Mieszkowska,
N., 2011. Regime shifts in the marine environment: the scientific basis and political
context. Marine Pollution Bulletin 62, 7–20.
Kremp,A.,Elbrächter,M.,Schweikert,M.,Wolny,J.F.,Gottschling,M.,2005.Woloszynskia
halophila (Biecheler) comb. nov.: a bloom-forming cold-water dinoflagellate co-
occurring with Scrippsiella hangoei (Dinophyceae) in the Baltic Sea. Journal of Phycology
41, 629–642.
Lohmann, H., 1908. Untersuchungen zur Feststellung des vollständigen Gehaltes des
Meeres an Plankton. Wissenschaftliche Meeresuntersuchungen Kiel, Neue Folge,
10, 130–370.
Matthäus, W., Nehring, D., Feistel, R., Nausch, G., Mohrholz, V., Lass, H.-U., 2008. The
inflow of highly saline water into the Baltic Sea. In: Feistel, R., Nausch, G.,
Wasmund, N. (Eds.), State and evolution of the Baltic Sea, 1952–2005. John Wiley
and Sons, Hoboken, New Jersey, pp. 265–309.
McLeod, A.I., 2005. Kendall rank correlation and Mann–Kendall trend test, R package
version 2.0.
McQuatters-Gollop, A., Gilbert, A.J., Mee, L.D., Vermaat, J.E., Artioli, Y., Humborg, C.,
Wulff, F., 2009. How well do ecosystem indicators communicate the effects of
anthropogenic eutrophication? Estuarine, Coastal and Shelf Science 82, 583–596.
Moestrup, Ø., Lindberg, K., Daugbjerg, N., 2009. Studies on woloszynskioid dinoflagel-
lates IV: the genus Biecheleria gen. nov. Phycological Research 57, 203–220.
Möllmann, C., Müller-Karulis, B., Diekmann,R., Flinkman, J., Kornilovs, G., Łysiak-Pastuszak, E.,
Modin, J., Plikshs, M., et al., 2006. An integrated ecosystem assessment of the Central
Baltic Sea and the Gulf of Riga. ICES Document CM 2006/P: 03. 24 pp.
Nakonieczny, J., Ochocki, S., Renk,H., 1991. Long-term changes in primary production and
chlorophyll concentrations in the southern Baltic. Acta Ichthyologica et Piscatoria
(Suppl. 21), 145–152.
Nausch, G., Nehring, D.,Nagel, K., 2008. Nutrient concentrations, trends and their relation
to eutrophication.In: Feistel, R., Nausch, G., Wasmund, N. (Eds.),State and evolutionof
the Baltic Sea, 1952–2005. John Wiley and Sons, Hoboken, New Jersey, pp. 337–366.
Nixon, S.W., Fulweiler, R.W., Buckley, B.A., Granger, S.L., Nowicki, B.L., Henry, K.M.,
2009. The impact of changing climate on phenology, productivity, and benthic–
pelagic coupling in Narragansett Bay. Estuarine, Coastal and Shelf Science 82, 1–18.
Olenina, I., Hajdu, S., Andersson, A., Edler, L., Wasmund, N., Busch, S., Göbel, J., et al.,
2006. Biovolumes and size-classes of phytoplankton in the Baltic Sea. Baltic Sea
Environment Proceedings 106. 144 pp.
Olenina, I., Wasmund, N., Hajdu, S., Jurgensone, I., Gromisz, S., Kownacka, J., Toming, K.,
Vaiciūtė, D., Olenin, S., 2010. Assessing impacts of invasive phytoplankton: the
Baltic Sea case. Marine Pollution Bulletin 60, 1691–1700.
Overland, J., Rodionov, S., Minobe, S., Bond, N., 2008. North Pacific regime shifts:
definitions, issues and recent transitions. Progress in Oceanography 77, 92–102.
Papush, L., Danielsson, A., 2006. Silicon in the marine environment: dissolved silica
trends in the Baltic Sea. Estuarine, Coastal and Shelf Science 67, 53–66.
Pliński, M., Jozwiak, T., 1999. Temperature and N:P ratio as factors causing blooms of
blue-green algae in the Gulf of Gdansk (Poland). Oceanologia 41, 73–80.
R Development Core Team, 2009. R: a language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria3-900051-07-0.
Reid, P.C., de Fatima Borges, M., Svendsen, E., 2001. A regime shift in the North Sea circa
1988 linked to changes in the North Sea horse mackerel fishery. Fisheries Research
50, 163–171.
Rolff, C., Almesjö, L., Elmgren, R., 2007. Nitrogen fixation and the abundance of the
diazotrophic cyanobacterium Aphanizomenon sp.in the Baltic Proper. Marine Ecology
Progress Series 332, 107–118.
Rönnberg, C., Bonsdorff, E., 2004. Baltic Sea eutrophication: area-specific ecological
consequences. Hydrobiologia 514, 227–241.
Sonderekker, D., 2008. SiZer: significant zero crossings, R package version 0.1-0.
Suikkanen, S., Laamanen, M., Huttunen, M., 2007. Long-term changes in summer
phytoplankton communities of the open northern Baltic Sea. Estuarine, Coastal and
Shelf Science 71, 580–592.
Sundström, A.M., Kremp, A., Daugbjerg, N., Moestrup, Ø., Ellegaard, M., Hansen, R.,
Hajdu, S., 2009. Gymnodinium corollarium sp. nov. (dinophyceae)—a new cold-
water dinoflagellate responsible for cyst sedimentation events in the Baltic Sea.
Journal of Phycology 45, 938–952.
ter Braak, C.J.F., 1994. Canonical community ordination. Part I: basic theory and linear
methods. Ecoscience 1, 127–140.
ter Braak, C.J.F., Šmilauer, P., 1998. CANOCO reference manual and user's guide to
CANOCO for Windows: software for canonical community ordination (version 4).
Microcomputer Power, Ithaca, NY. 351 pp.
Tian, Y., Kidokoro, H., Watanabe, T., Iguchi, N., 2008. The late 1980s regime shift in the
ecosystem of Tsushima warm current in the Japan/ East Sea: evidence from
historical data and possible mechanisms. Progress in Oceanography 77, 127–145.
Toms, J.D., Lesperance, M.L., 2003. Piecewise regression: a tool for identifying ecological
thresholds. Ecology 84, 2034–2041.
Vandepitte, L., Vanhoorne, B., Kraberg, A., Anisimova, N., Antoniadou, C., Araújo, R.,
Bartsch, I., et al., 2010. Data integration for European marine biodiversity research:
creating a database on benthos and plankton to study large-scale patterns and
long-term changes. Hydrobiologia 644, 1–13.
Wasmund, N., Siegel, H., 2008. Phytoplankton. In: Feistel, R., Nausch, G., Wasmund, N.
(Eds.), State and evolution of the Baltic Sea, 1952–2005. John Wiley and Sons,
Hoboken, New Jersey, pp. 441–481.
Wasmund, N., Uhlig, S., 2003. Phytoplankton trends in the Baltic Sea. ICES Journal of
Marine Science 60, 177–186.
Wasmund, N., Nausch, G., Matthäus, W., 1998. Phytoplankton spring blooms in the
southern Baltic Sea—spatio-temporal development and long-term trends. Journal
of Plankton Research 20, 1099–1117.
Wasmund, N., Göbel, J., v.Bodungen, B., 2008. 100-years-changes in the phytoplankton
community of Kiel Bight (Baltic Sea). Journal of Marine Systems 73, 300–322.
Weijerman, M., Lindeboom, H., Zuur, A.F., 2005. Regime shifts in marine ecosystems of
the North Sea and Wadden Sea. Marine Ecology Progress Series 298, 21–39.
Witek, Z., Ochocki, S., Maciejowska, M., Pastuszak, M., Nakonieczny, J., Podgórska, B.,
Kownacka, J.M., Mackiewicz, T., Wrzesińska-Kwiecień, M., 1997. Phytoplankton
primary production and its utilization by the pelagic community in the coastal zone
of the Gulf of Gdańsk (Southern Baltic). Marine Ecology Progress Series 148,
169–186.
Witkowski, A., Broszinski, A., Bennike, O., Janczak-Kostecka, B., Jensen, J.B., Lemke, W.,
Endler, R., Kuijpers, A., 2005. Darss Sill as a biological border in the fossil record of
the Baltic Sea: evidence from diatoms. Quaternary International 130, 97–109.
Wood, S.N., 2000. Modelling and smoothing parameter estimation with multiple
quadratic penalties. Journal of the Royal Statistical Society (B) 62 (2), 413–428.
159N. Wasmund et al. / Journal of Marine Systems 87 (2011) 145–159