1
Submited to DSR-II, Black Sea Special Issue
Regulation of the Black Sea physical and ecosystem structure
by climate variability and anthropogenic forcing
Temel Oguz1 and Joachim W. Dippner2
(1) Institute of Marine Sciences, Middle East Technical University, Erdemli, Turkey
(2) Baltic Research Institute, Warnemünde, Seestr. 15, D-18119 Rostock, Germany.
Abstract
An examination of a wide spectrum of biogeochemical records from the Black Sea indicates
robust
interannual-to-interdecadal
climate-induced
fluctuations
superimposed
on
the
anthropogenic signal at all components of the ecosystem structure from nutrient concentrations
to plankton biomass and pelagic fish stocks. They generally follow temporal variations of the
physical properties, such as the upper layer temperature, salinity and sea level anomaly, in terms
of their phases and durations. Synchronization of these physical-ecosystem variations with the
North Atlantic Oscillation (NAO) index and its signature on the regional atmospheric properties
(air temperature, wind stress magnitude, surface air pressure, evaporation minus precipitation)
indicates a dominant role of the low frequency, large-scale climatic oscillations in the Black Sea.
The analysis suggests an efficient positive feedback mechanism between the climatic and
anthropogenic forcing, which brings a dynamical perspective to the interpretation of
eutrophication-induced ecosystem transformations encountered during the 1980s and 1990s.
1. Introduction
A comprehensive understanding of the impacts of natural climatic variability and anthropogenic
forcing on marine ecosystems is essential in order to effectively predict and manage marine
living resources. Climate-ocean interactions and their response on marine ecosystems have
recently been focus of considerable world-wide attention. The decadal-scale climatic changes on
stocks of plankton and fish have been reported for open ocean ecosystems (e.g., Drinkwater et
al., 2003 for the North Atlantic, Chavez et al., 2003 for the Pacific Ocean), as well as for coastal
2
and marginal seas such as southeastern Bering Sea (Dippner and Ottersen, 2001; Hunt et al.,
2002), North Sea (Reid et al., 2001), East Sea/Japan Sea (Kang et al., 2002), northeastern US
continental shelf (Conversi, et al., 2001). In identifying ecological impacts of climate variability,
coastal and marginal sea ecosystems offer a particular challenge since they are most heavily
affected by human interventions, and natural climatic variations are often masked by local
transients.
Among marginal sea ecosystems, the Black Sea is of special interest due to its intense three
decade-long eutrophication (see Zaitsev and Mamaev,1997 and Kideys, 2002 for a review), as
well as its remote location to major hemispherical atmospheric systems. Moreover,
teleconnection between the large scale atmospheric systems and physical climate and ecosystem
structure of the Black Sea has been poorly investigated, disperse, and focused only on some
specific features. Niermann et al. (1999) has drawn attention to long-term climate-plankton
connection in different regions of the world oceans, and then suggested a possible role of the
NAO-driven large scale atmospheric changes for the outburst of Mnemiopsis and decline of the
anchovy stocks in the Black Sea during the late-1980s. Mikaelyan (1997) pointed out a close
correlation among the summer phytoplankton biomass and the winter average air temperature
variations, suggesting a link to climatic variations.
Kovalev and Piontkovski (1998) and
Kovalev et al. (1998) noted some dominant interannual fluctuations in the Black Sea
zooplankton data. Similarly, Prodanov et al. (1997) emphasized the role of natural climatic
factors, in addition to human-induced effects, regulating pelagic fish stocks. Konovalov and
Murray (2001) stated the interannual-to-interdecadal variability of nutrient, oxygen, hydrogen
sulfide concentrations in the suboxic and anoxic layers. All these studies, however, did not
either mention specifically the NAO link or examine the details of the link. Oguz et al. (2003)
documented an intimate relationship between a decade-long winter warming during the 1990s
and unfavorable plankton growth, and reduced stocks at higher trophic levels. The footprints of
the climatic fluctuations in various hydro-meteorological properties of the Black Sea have been
pointed out by Polonsky et al. (1997), Reva (1997), Belokopytov (1998), Ozsoy (1999), Stanev
and Peneva (2002), Tsimplis and Rixen (2003).
The present study brings together the available published data relevant to this work, and offers a
synthesis to provide a unified and comprehensive description of the long-term variations of
physical structure and ecosystem at interannual-to-interdecadal scales in the Black Sea.
Synchronious changes in the physical properties of the upper layer water column are first
3
documented in section 2. Their correlations with the North Atlantic Oscillation (NAO) index,
and physical mechanisms driving the connection between the NAO and Black Sea climatic
fluctuations are provided in section 3. Section 4 relates these subtle changes in the Black Sea
physical climate to those of nutrients, primary and secondary productions, and ultimately
pelagic fish populations. The conclusions are given in section 5.
2. Long-term variations in the Black Sea physical climate
On the basis of available data, the interannual-to-interdecadal variations observed in the Black
Sea physical climate are best documented through the examination of temperature and salinity of
the upper layer water column, and of sea level anomaly. The sea surface temperature (SST) data
(Rayner et al., 2003) span from 1875 to 2000 and are based on the average of all the available
measurements performed within the interior part of the basin deeper than 1500 m during the
winter (December-March) period. At multi-decadal time scales, the winter SST variations fall
into four different amplitude ranges with a gradual warming trend of 0.25 oC during the last
century (Fig.1a). After the 1950s, the warming trend is about 0.12 oC, some of which might be a
contribution from the green house–related global warming, whose indication in global oceans
was recently provided by Levitus et al. (2000). The period from 1875 to 1915 (Phase 1) is
characterized by the range of SST variations between 7.6 to 8.7 oC, and comprises a sequence of
three cold cycles (denoted by C1, C2, C3 in Fig. 1a) and three warm cycles (W1, W2, W3); each
of them lasts approximately for ~5 years. Thereafter, the warm and cold cycles occur at
gradually higher temperature ranges of 7.9 to 9.0 oC up to 1960 (Phase 2), and of 8.3 to 9.3 oC
within the next two decades (Phase 3). The period from the early 1960s to the early 1980s thus
indicates the most significant warming phase of the last century dominated by the persistently
positive winter SST anomalies above the long-term trend. It is then followed by a strong decadal
cooling phase from 1980 to 1995 (Phase 4). This phase is identified by the SST variations
between the minimum of 7.2 oC (C9, C10) and the maximum of 8.5 oC, and a sharp temperature
drop of about 1.7 oC between the years 1980 and 1985. The highest temperature of 8 oC within
the warm cycle (W10) of this phase is approximately 0.4 oC lower than the minimum
temperature of the cold cycle (C8) of the previous phase. From 1995 onwards, a new warming
cycle (W11) dominates the SST.
4
The winter-mean SST data is complemented by the mean temperature of the Cold Intermediate
Layer (CIL) averaged over the May-November period since the 1950s (Belokopytov, 1998). The
CIL refers to as the cold core layer just below the seasonal thermocline, and is remnant of the
winter surface mixed layer water with temperatures less than 8oC. Approximately the 50 yearslong time series of the summer-autumn (May-November) mean temperature of the CIL are also
shown to possess very similar variations in terms of both timing and duration of the warm cycles
W7-W11, and the cold cycles C6-C10 (Fig. 1b). It therefore implies extension and persistence of
the cooling/warming cyclical events below the surface mixed layer and during the rest of the
year. The decadal climatic signal has also been noted with some phase shift at temperatures of
deeper levels around the permanent pycnocline (Konovalov and Murray, 2001).
The temporal variations of the upper 200 m layer-average salinity anomaly (Fig. 1c) taken from
Tsimplis and Rixen (2003) also reflect similar climatic signature in terms of phases and duration,
and are inversely related to the temperature variations observed in Figs. 1a,b. As expected,
higher positive salinity anomalies correspond to the cold cycles since cold years associated with
stronger surface cooling and intensified wind stress are prone to stronger convective mixing of
surface waters with more saline subsurface waters, and involve more pronounced evaporative
losses to the atmosphere (see below). Conversely, the negative salinity anomalies are associated
with the warm cycles due to weaker vertical mixing within the upper layer water column and
evaporation minus precipitation. The absence of negative salinity anomaly W10 towards the end
of the 1980s seems to be a result of the decade-long strong cooling phase (see Fig. 1a), giving
rise to a continual vertical mixing and subsequently gradually increasing salinity anomaly within
the upper layer water column. As suggested by the SST data, the warm cycle W10 appears to be
still cold enough to generate sufficiently strong vertical mixing comparable to its previous cold
cycle C9.
The detrended sea level anomaly time series (Fig. 1d) given by Reva (1997) (see also Tsimplis
and Josey, 2001 and Stanev and Peneva, 2002) represent an average of the measurements at 20
coastal stations around the Black Sea since 1890. It oscillates within the amplitude range of
about 10 cm. The higher (lower) values of the sea level anomaly coincide with the cold (warm)
cycles of the water temperature, indicating that a part of the observed sea level change has a
thermal origin, due to the steric effect. Similarly, the linear rise of 0.18 cm yr-1 seen in the
original data (Reva, 1997) is compatible with the long-term warming trend of ~0.0022 oC yr-1 in
the SST. In fact, all the physical properties shown in Fig. 1a-d are significantly correlated with
5
each other (significance level p<0.001) (Table 1). Moreover, the fact that the sea level anomaly
variations in the Black Sea are correlated with those in the North Atlantic (Reva, 1997), and in
the Eastern and Western Mediterranean (Tsimplis and Josey, 2001) suggests a regional
dynamical link to a large scale atmospheric system. Various measurements in the Mediterranean,
Black and Caspian Seas (e.g., Ozsoy, 1999; Tsimplis and Josey, 2001; Stanev and Peneva, 2002)
have already established teleconnection to the North Atlantic Oscillation (NAO). The NAO
signature has even been recorded in streamflow changes of the Tigris and the Euphrates Rivers,
which are two well-known waterways flowing from Anatolia to Middle East (Cullen and
deMenocal, 2000), as well as in the River Danube discharge (Polonsky et al., 1997). The data
presented in the next section provide further details of the NAO-Black Sea link.
3. Possible link between the Black Sea physical climatic variations and the NAO
The NAO is regarded as the major large scale weather system controlling the atmospheric
circulation over the North Atlantic and the Eurasia (Marshall et al., 1997), thus affecting their
marine and terrestrial ecosystems. The positive winter NAO index is associated with the strong
pressure gradient between Azores high pressure and Iceland low pressure systems bringing
strong westerly winds, cold and dry air masses to southern Europe and Black Sea region
(Hurrell, et al., 2003). In these periods, southern Europe and the Black Sea region remains under
the influence of Azore high pressure center and thus are characterized by the relatively higher
surface air pressure values. Conversely, the negative NAO index implies a weaker pressure
gradient, lower surface atmospheric pressure values, milder winters with warmer air
temperatures and less dry/more wet atmospheric conditions over the Black Sea.
The major physical mechanisms which link the Black Sea physical climate to the large scale
atmospheric system over the North Atlantic are changes in the air temperature due to cold/warm
air advection to the region, the wind stress, as well as changes in the sea level atmospheric
pressure (SAP), and evaporation minus precipitation (E-P) over the Black Sea. If they are all
controlled by a common large scale atmospheric system, the temporal variations of these
properties are then expected to be compatible with those of the oceanographic properties
reported in the previous section. Here, we first describe interannual-to-interdecadal variations of
the meteorological characteristics over the Black Sea during the last century. We then show their
6
compatibility with those within the water column, and relate them to the long term variations of
the North Atlantic Oscillation.
The data shown by Titov (2000, 2002) suggest that the winter-mean air temperatures measured
at various coastal stations around the periphery of the basin exhibit very similar temporal
variations during the last century, even though the mean temperatures and their range of
variations may differ from the western to eastern end of the basin. Fig. 2a provides an example
at a station near the Kerch Strait (connecting the Black Sea to the Sea of Azov) along the
northern coast of the Black Sea. Since 1885, it reveals 11 successive warming and cooling
cycles between the extreme values of 1.5 and 6.5 oC at intervals of ~5 years. The entire data set
exhibits a linear warming trend with the overall temperature rise of 0.9 oC. This is consistent
with the warming in the winter temperature over Eurasia, which was explained partly by
temperature advection connected to the NAO and partly by the radiative forcing due to increased
greenhouse gases (Hurrell, 1996). As in the case of the SST variations, the last two decades
have been subject to abrupt variations over 10 year cycles from the maximum value of 5.5 oC to
the minimum value of 1.5 oC. This cooling phase (C10) during the 1980s was so dramatic that
the subsequent decade-long warming was only able to bring the air temperature back to its
previous state maintained at the end of the 1970s. Both the timing and duration of the warm and
cold cycles fit reasonably well with those shown previously in the winter-mean (DecemberMarch) SST and the summer-autumn (May-November) mean CIL temperature time-series (Fig.
1a,b). Its correlation with the SST and CIL, respectively, are -0.52 and -0.37 at the significant
level p<0.001. The only major difference is the lack of the W10 peak in the water temperatures
during the second half of the 1980s. Instead, air temperature continues to decrease steadily up to
the early 1990s. This peak is however captured at some other stations (e.g., see Fig. 1 in Titov,
2000).
To our knowledge, no reliable long-term wind stress time series data are available for the Black
Sea. Here, we present 17 year-long, winter-mean, basin-averaged wind stress time series for the
1980-1997 period which we have generated from the NCEP re-analysis data set (Fig. 2b). Even
though it comprises a relatively short time series, it is sufficient to support consistency between
the wind stress temporal variations and other characteristics of the NAO-driven atmospheric
circulation over the region. The periods with more intense wind stress fields correspond to the
cold year cycles (C9, C10), as compared to weaker ones during the warm winter cycles (W9,
W10, W11).
7
The winter (December-March mean) surface air pressure (SAP) distribution over the Black Sea
(Fig. 2c) follows closely the winter air temperature time series.
The severe winters are
preferentially associated with the higher pressures up to 1022 hPa, whereas the lower surface air
pressures prevail in milder winter seasons. A general tendency of severe winters during the
1980s leads to relatively higher minimum surface air pressures during its occasional milder
years. We recall a similar tendency in the sea surface and air temperature time series shown
before in Figs 1a and 2a.
The Black Sea has always been characterized by the excess of evaporation over precipitation,
but their difference oscillates depending on the intensity of winter conditions (Fig. 2d). The
severe winter cycles coincide with the higher values of (E-P), and therefore support the presence
of more dry conditions of the NAO-driven atmospheric systems. Conversely, (E-P) signal
weakens during the mild winter cycles due to their domination by stronger precipitation.
Moreover, a linearly decreasing trend of (E-P) from about 250 km3 yr-1 to zero during the last 50
years, caused by gradual decrease in evaporation and increase in precipitation, accompanies the
long-term warming trend of the AT seen in Fig. 2a.
As shown in Fig. 3, the general consistency between periods with positive (negative) values of
the NAO index and those of the cold (warm) air temperatures (Fig. 2a), stronger (weaker) wind
stress (Fig. 2b), higher (lower) surface air pressures and evaporation minus precipitation values
(Fig. 2c,d) supports the specific role of the NAO-driven large scale atmospheric motions over
the Black Sea. When gross features of the entire data set are taken into account, the agreements
are fairly good, for which the level of correlations is shown in Table 2. In terms of duration and
intensity of events, the sequence of mild and severe winter cycles with ~5 year duration follows
the temporal pattern of the negative and positive NAO cycles. In general, all warm cycles
except W4, W6, W10, and all cold cycles except C6 and C7 correspond to the strong negative
and positive NAO phases, respectively.
The regular 5-year cyclic patterns were however
interrupted by two specific decadal scale oscillations. The first comprises the 1955-1970 period
preferentially dominated by the negative NAO index, and briefly interrupted by two rather short
and weak positive NAO events. This explains the cause of weak signature of the C6 and C7
cold cycles in the air and sea surface temperature as well as in the other related atmospheric and
oceanographic properties. Similarly, the period of 1980-1995 characterizes an extended strong
positive NAO index phase, and therefore is dominated by the cold AT and SST cycles without
8
an appreciably strong warm cycle W10. Once again, the weak W4 and W6 warm cycles in the
SST and AT time series correspond to the weakly negative NAO cycles during these years.
The sea level change time series are the best data set to examine the link between NAO-driven
atmospheric motion and the Black Sea physical climate, because the link includes (i) changes in
the surface atmospheric pressure through the inverse barometer effect, (ii) water density changes
in response to temperature and salinity variations, (iii) changes in precipitation, evaporation and
river runoff. As shown in Table 2, the NAO-SLA correlation is -0.85 for 111 pair of data points.
Moreover, the SLA-(E-P) correlation of -0.545, the SLA-SAP correlation of -0.600, and the
SLA-SST correlation of 0.375, with significant levels less than 0.001, support presence of a
robust and well-defined link between the NAO and the Black Sea marine climate.
4. Biogeochemical impacts of the NAO-driven climatic variations
The traditional view (see Sorokin, 2002 for a review) describes the long-term changes in the
biogeochemical properties of the Black Sea in terms of eutrophication, which was developed by
enhanced nutrient delivery into the shallow northwestern part of the basin as a result of
widespread use of fertilizers and different types of pollutants in the former Eastern Block
countries during the 1970s and 1980s.
The eutrophication then impacted the entire basin
through the dynamic mesoscale-dominated circulation system. According to this view, the
eutrophication was reflected as an order of magnitude increase in phosphate concentrations of
the northwestern coastal waters from around 1 μM during the late 1960s to ~8 μM during the
mid-1970s (Fig. 4a) (Cociasu et al., 1996). The measures introduced to control the level of
pollution were then suggested to decrease nutrient concentrations abruptly to their preeutrophication level in the 1990s. The response of eutrophication was evidently felt at all levels
of the food web as well from phytoplankton to pelagic fishes. The phytoplankton biomass data
from the interior basin (Fig. 4b) increased from ~2 g m-2 during the late 1960s to ~20 g m-2
during the 1980s (Mikaelyan, 1997). The summer surface chlorophyll concentrations along the
northeastern part of the Black Sea, which was the least affected region by the eutrophication,
changed from 0.15±0.10 mg m-3 during 1973-1982 (early-eutrophication period) to 0.89±1.27
mg m-3 during 1985-1992 (eutrophication period) and to 0.35±0.06 mg m-3 during 1995-2000
(post-eutrophication period) (Vedernikov and Demidov, 2002). They were accompanied by
substantial decrease in the visibility of surface waters in the 1980s (Fig. 4c), followed by
9
improved visibility during the less productive post-eutrophication phase of the ecosystem in the
1990s (Mankovsky et al., 1998). Organic matter, continually sinking and decomposing during
the intense eutrophication phase, led to a linear trend of increase in sulfide concentrations from
about 8 μM during the early 1980s to about 20 μM in the 1990s (Fig. 4d) (Konovalov and
Murray, 2001).
The traditional view also described changes in the fish stocks in terms of human-induced
interventions. Once the stocks of large pelagic predatory fishes (e.g. tuna, swordfish, large
bonito, blue fish) were severely depleted in the early 1970s as a result of heavy and unregulated
fishing (Daskalov, 2002), the planktivorous small pelagic fish stocks (e.g. anchovy, sprat, horse
mackerel) increased considerably in the absence of their predator grazing pressure (Gucu, 2002).
Additionally, enhanced food supply due to eutrophication contributed favorably to small
pelagics, which then became the main commercially exploitable fish stocks during the 1970s and
1980s. The sum of sprat and anchovy parental stocks increased from about 400 ktons in the mid1960s to 700 ktons at the beginning of 1970s, and to the maximum level of 1100 ktons during
the first half of the 1980. Thereafter, the stock started to drop up to 300 ktons at the end of the
1980s, the major cause of which was overfishing and food competition with Mnemiopsis
(Daskalov, 2002).
More in-depth analysis of the same biogeochemical data, on the other hand, identifies some
quasi-periodic oscillations superimposed on the anthropogenically-derived trends, and links
them to changes in the large scale atmospheric circulation system over the North Atlantic. For
example, phosphate concentration time series (Fig. 4a) reveal a remarkable consistency between
the low and high concentrations and the physical properties of the water column. The years with
mild winters characterized by warmer temperatures (Fig. 1a), weaker winds (Fig. 2b) and
negative NAO index (Fig. 3) encountered towards the end of the 1960s, 1970s, and 1980s (W8,
W9, W10) correspond to the periods with lower phosphate concentrations. Conversely, the
periods with increased phosphate concentrations encountered during the mid-1970s and mid1980s (C8, C9) match very well with the severe winter years characterized by colder water
temperatures, stronger winds, and positive cycles of the NAO index. The severe winter period
during the early 1990s (C10) does not however show increased phosphate concentrations. The
correlation of the overall phosphate data (p<0.001) with NAO and SST is -0.53 and 0.35,
respectively (Table 3). The fact that the periods of phosphate concentrations are negatively
correlated with the Danube discharge (Polonsky et al. 1977; Stanev and Peneva, 2002) implies
10
primary role of the vertical biological pump and mixing processes governing the interannual-tointerdecadal variations. It seems that combination of these vertical processes allow efficient and
continuous turn-over of phosphate between deep and surface waters. The River Danube acts as
the continuous nutrient supplier into the deep phosphate reservoir in the region.
The cyclic variations are also noted in the summer phytoplankton biomass and the surface
chlorophyll concentration data from the interior of the basin, both are shown in Fig. 2b. Between
two less productive periods in the early 1970s and 1980s corresponds to a period with higher
phytoplankton biomass during the mid-1970s. These cyclic variations, once again, fits very well
with the sequence of two mild periods and negative NAO cycles (W8, W9) interrupted by the
severe period (C8) with positive NAO cycle in between. The upward rise of both phytoplankton
biomass and chlorophyll concentrations later in the 1980s resembles similar trends in the
physical data as well as the prolonged period of the strong positive NAO cycle. The maximum
biomass and concentrations occured during the early 1990s corresponding to the time of the
coldest years of the last century (C10) followed later by their gradual reduction during the warm
cycle of the 1990s. The correlation of phytoplankton biomass with the NAO and SST are 0.44
and -0.79, respectively (Table 3).
The bottom-up limited unfavorable phytoplankton growth during mild winter years might be
explained by more limited upward nutrient supply from the nutricline. On the contrary, strong
westerly winds and frequent cold air outbreaks into the region during the periods of the positive
NAO index induce strong vertical mixing that leads to elevated nutrient levels in the surface
layer in winter months. A strong new production-based phytoplankton growth develops in
spring, and is followed by regenerated-based production later in summer months. Moreover,
delay in spring primary production in the case of colder water temperatures can promote more
active regenerated production and higher phytoplankton biomass in summer months.
The quasi-periodic, climate induced fluctuations observed in the phytoplankton biomass and
chlorophyll concentration time series are also supported by the Secchi disc data (Fig. 4c), which
yield decreasing Secchi disc depths during more productive and turbid cycles in the mid-1960s,
the mid-1970s as well as during the 1980s up to the early 1990s (C10). Conversely, increasing
Secchi disc depths occur during less productive and thus less turbid, milder climatic cycles (W7,
W8, W9) of the early years of the 1960s, 1970s, and 1990s. The NAO and SST correlations of
the Secchi disc data are -0.51 and 0.65, respectively.
11
Hydrogen sulfide concentration variations along the density surface of σ t=16.4 kg m-3 (Fig. 4d)
also demonstrate contributions of the climate-induced variations to the Black Sea
biogeochemistry. This density surface lies along the base of the permanent pycnocline, and
therefore variations of sulfide concentrations along this surface is a measure for how deep the
surface signature can effectively propagate within the water column. As shown in Fig. 4d for the
mid-1970s and the 1980s, the years with severe winters (C8, C10) represent relatively higher
sulfide concentrations. This was caused by more active biological production as well as higher
rate of advective supply from the deep anoxic pool due to presence of intense cyclonic
circulation, whose signature is provided in Fig. 1d by the relatively low sea level anomalies in
cold years. On the contrary, both upward supply and local production of sulfide concentration
from sinking organic matter source during the mild winter years are limited; thus warm cycles
(W8, W9) correspond to relatively lower sulfide concentrations in Fig. 4d. Once again, NAO and
SST correlations are quite high; 0.85 and -0.76, respectively.
In practice, long-term variations of mesozooplankton biomass are regulated by the combination
of complicated bottom-up and top-down interactions among different trophic levels of the food
web as well as temperature dependence of their growth. A regular pattern of their temporal
biomass variations is thus hardly expected unless any of these signals dominates entirely the
system. The annual-mean mesozooplankton biomass time series (Fig. 5), beyond all
expectations, also possess a cyclic pattern with higher (lower) biomass during the years with
warm (cold) winter cycles; thus they are negatively correlated with the NAO index (Table 3).
The preferential increase of mesozooplankton biomass during the warm years may be explained
by limited role of the bottom up control on mesozooplankton production. On the other hand,
synchronous oscillations between the stocks of mesozooplankton (Fig. 5) and total pelagic fish
(see Fig. 6 in Gucu, 2002) contradict with the concept of top-down trophic cascade. The
mesozooplankton growth thus seems to be primarily controlled by water temperature (Kovalev
et al., 1998); warmer waters increase the rate of metabolic processes and favor more efficient
growth, whereas colder waters limit mesozooplankton production albeit producing stronger
phytoplankton blooms. The observational evidence for temperature dependence of copepod
production has been given by Huntley and Lopez, (1992). The primary role of water temperature
has been pointed out by Planque and Taylor (1998) for regulation of C. finmarchicus stocks in
the North Sea and the northeast Atlantic, and by Hunt et al. (2002) for the southeastern Bering
Sea.
12
Some studies (e.g., Shushkina et al., 1998; Shiganova, 1998; Kideys, et al., 2000) explained the
dramatic reduction in mesozooplankton biomass in the early 1990s in terms of their strong
predation by Mnemiopsis, whose population was exploded towards the end of the 1980s. To our
opinion, however, the Mnemiopsis grazing should have only a secondary effect since
Mnemiopsis biomass decreased substantially from ~1000 gm-2 during 1989-1990 to 200 gm-2
during 1991-1993 as their community was negatively affected by cold waters as well. Only
certain mesozooplankton species (e.g., Acarthia Spp.) permanently residing in the upper mixed
layer have been affected by Mnemiopsis, whereas deep water species (Calanus euxinus, P.
elongates) have been less exposed to Mnemiopsis grazing pressure (Vinogradov et al., 1999).
After 1993, both mesozooplankton and Mnemiopsis biomass started increasing once again, when
the Black Sea switched to a new mild winter cycle for the rest of the 1990s (cf. Oguz et al.
2003).
The climatic contribution to the rise and fall of pelagic stocks also emerges very clearly in the
temporal distributions of sprat and anchovy biomass (Figs. 6 and 7). Sprat is a cold water
species spawning in autumn and winter months. Sprat stocks are therefore expected to be higher
during the cold years. A substantial increase in the parental sprat biomass, from 100 to 600
ktons, occurred during the first half of the 1970s (C8) at the times of cold years of the climatic
oscillations (Fig. 6). It was followed by a decrease to 200 ktons during the subsequent warm
cycle (W9) in the second half of the 1970s, up to the early-1980s. The biomass then increased
first to 450 ktons during the mid-1980s (C9), and finally dropped to its lowest level of ~100
ktons at the end of the 1980s and the early 1990s. Sprat recruitments followed a similar pattern
of oscillations roughly between 20 to 240 billions of juveniles during the 1970s and 1980s and,
once again, dropped to its minimum level at the end of the 1980s. While the sprat time series are
weakly correlated (r2=0.09) with that of the NAO, its correlation with the SST is quite
reasonable, r2= -0.44.
Because sprat is a cold water species, the low level adult and juvenile sprat stocks observed
during the early 1990s do not fit into the climate change scenario, and deserve an alternative
explanation.
As suggested by Gucu (2002), the overall small pelagic stock exceeded a
sustainable level in 1987, Size of the spawning stocks was reduced to such an extent that they
were not able to produce enough offspring anymore to sustain future generations. Thus the small
pelagic fish stock collapsed a year after suggesting that overfishing must have played inevitable
13
role for the sharp declines of the adult and juvenile pelagic stocks. The excessive feeding
pressure introduced by Mnemiopsis population a year later should contribute as well, to some
extent, to depletion of the pelagic stocks (e.g. Shulman et al., 1998; Shiganova, 1998;
Vinogradov et al., 1999; Kideys et al., 2000).
Anchovy is a warm water species, spawns preferentially during summer and autumn months.
Thus, anchovy stocks of both recruitments and adults reveal an opposite temporal variations
(Fig. 7). The stocks were limited during the cold years, since they were ecologically not suitable
for anchovy growth. The parental stock was as low as 300 ktons during the mid-1960s (C7) and
mid-1970s (C8), then decreasing continuously up to 150 ktons during the early 1990s (C10).
Much higher stocks were accomplished during the warm years; stocks raised up to 500 ktons at
the end of the 1960s (W8), and to 700 ktons at the end of the 1970s (W9). Similarly, anchovy
recruitments around 50 billions in the cold period of the early 1970s increased up to 150 billion
at the end of the subsequent warm cycle (W9), decreased thereafter up to about 25 billion in the
early 1990s. The overfishing and adverse effects of Mnemiopsis are also expected to affect the
anchovy stocks towards the end of the 1980s and during the early 1990s. Anchovy eggs and
larvae preferentially live in the surface mixed layer, where Mnemiopsis population generally
inhabits too (Prodanov et al., 1997). The correlation of the anchovy time series with the NAO
and SST is -0.66 and 0.53, respectively.
Our analysis therefore shows that climatic variations govern the alternating periods of high and
low abundance of anchovy and sprat stocks. Large catches of sprat coincided with times when
the NAO index was positive, characterized by severe winters with cold air and water
temperatures in the region. The reverse situation occurs during the negative phases of the NAO
index, when the relatively warm waters favor anchovy spawning stocks. This system resembles
the NAO-driven alternating herring and sardine periods of high and low abundance in the
Northeast Atlantic (Alheit and Hagen, 1997), and alternating anchovy and sardine periods in the
Pacific Ocean controlled by the Pacific Decadal Oscillations (Chavez et al., 2003).
5. Conclusions
The present study demonstrates prominent role of the climate-induced, interannual-tointerdecadal variabilities both in the physical structure of water column and different trophic
14
levels of the Black Sea ecosystem. It then interprets the ecosystem changes as an integrated
response of the natural climatic variability and various anthropogenic and human-induced
perturbations. First, a robust anthropogenic signal was identified in all the available long-term
biogeochemical data in the form of a steady increase in different properties during the 1970s and
1980s (eutrophication phase) from their background levels of the pre-uetrophication phase.
Then, they have been subject to abrupt changes once again at the end of 1980s in response to
population explosuion of gelatinous carnivore Mnemiopsis and complete exploitation of pelagic
fish stocks due to overfishing.
Finally, the 1990s were considered as the post-eutrophication
phase with gradually diminishing anthropogenic signal at all components of the ecosystem.
Second, these abrupt rise and fall type decadal variations in the ecosystem properties were
shown to accompany synchronuous fluctuations in the hydro-meteorological data linked to the
North Atlantic Oscillations (NAO). The teleconnection imply governing role of the remote large
scale atmospheric forcing on the basin-scale dynamical and thermodynamical characteristics of
the Black Sea. The duration and intensity of the oscillations are found to be consistent with
those of the NAO index. Generally speaking, 11 simultaneous the so-called cold and warm
cycles, each with an approximate duration of 5 years, are noted in the long term physical data
sets. They are interrupted by the longer cycles of predominantly negative NAO index during
1955-1970, and the positive NAO index during 1980-1995. These fluctuations are also
accompanied by the long-term trends.
An interesting issue which came out from the analysis of the biogeochemical data is the presence
of a positive feedback mechanism between the anthropogenic and natural climatic forcing.
Anthropogenic-based, eutrophication-induced alterations started building up the ecosystem in
the 1970s and were felt more substantially during the 1980s. The latter era was known to be
characterized by drastic increases in nutrient concentrations, plankton biomass and pelagic fish
stocks.
In the long run, the nutrients primarily supplied by the River Danube into the
northwestern shelf region were distributed within the entire basin by the circulation system and
deposited eventually within the chemocline zone confined between the euphotic layer above and
the suboxic layer below. For example, peak concentrations of the subsurface nitrate maximum
layer increased from about ~1.5-2.5 µM in the 1960s to ~6.0-9.0 µM in the 1980s within the
interior parts of the basin (see Fig. 2 in Oguz et al., 2000). In the absence of any physical
mechanism to efficiently supply these inorganic nutrients into the euphotic zone, the
eutrophication wouldn’t play effective role in bıological production throughout the entire basin,
except coastal regions. The dynamical processes necessary to supply nutrients into the surface
15
layer were generated by the strongest NAO cycle of the 20th century during the 1980s. The large
scale atmospheric motion provided strong wind stress forcing and intense cooling in the region,
which subsequently led to strong convective mixing in the water column to bring the nutrients
from its subsurface pool. Therefore, the climatic forcing introduced an efficient feedback
mechanism, which intensified the anthropogenically-driven biogeochemical processes during the
1980s. The interesting question to be asked is “what would be the impact of eutrophication if the
Black Sea would be controlled by a weakly positive or negative NAO cycle during the 1980s?”
Probably, the Black Sea ecosystem would shift to another state, quite different than what it has
been observed during the last two decades.
Conversely, without appreciable subsurface nutrient accumulation in the absence of the
eutrophication, strong vertical advective and convective processes in the water column wouldn’t
be so effective for generating strong biological production in the basin. The strong NAO-induced
regional atmospheric motion alone would thus be unable to promote an order of magnitude
changes in the ecosystem properties as encountered during the 1980s.
Consequently, the
ecosystem conditions observed in the Black Sea during the 1980s are interpreted as the joint
operation of anthropogenic and natural climatic forcing.
The positive feedback argument can also be used for the interpretation of the ecosystem
conditions during the second half of the 1990s. As described in more detail by Oguz et al.
(2003), the Black Sea has been experiencing milder conditions, less favorable for biological
production from 1995 onwards. At the same time, more effective control measures were
implemented by the riparian countries for reducing the anthropogenic nutrient supply and
eutrophication. These climatic and anthropogenic forcing agents together thus led to a less
efficient biological cycle in the Black Sea.
This exploratory study merely elucidates a link between changes in various components of the
Black Sea ecosystem and large scale climatic variability with special emphasis on the NAO.
The correlations between the biogeochemical properties examined and the NAO index are quite
convincing, but are still needed to be confirmed quantitatively by future modeling studies.
Deterministic models will be particularly useful for mechanistic understanding of the response of
the system as a whole to different forcing agents, testing various hypotheses put forward for
explaining observed features, and forecasting the future state of the ecosystem. Further research
is required to find out possible teleconnection mechanisms with other large scale climatic
16
oscillations, and the ways in which they affect the ecosystem structure and functioning. For
example, Ozsoy (1999) and Ginzburg et al. (2004) pointed out teleconnection between some
cold years (e.g., 1982-1983, 1986-1987, the early 1990s) and negative values of the Southern
Oscillation index. More comprehensive future studies on the climate-driven ecosystem changes
are expected to provide insights for sustainable use of marine living resources in the Black Sea,
and effective management policies.
Acknowledgements
The sea surface temperature data was provided by Met Office HadISST v.1.1 data set in the
British Atmospheric Data Centre. The NAO index values were taken from the URL site
http://www.tao.atmos.washington.edu/data_sets/nao. We thank Erhan Mutlu and Murat Gunduz
for their assistance in the data processing. This work is a contribution to ARENA project of the
5th Framework Program of the European Union.
References
Alheit, J., and E. Hagen, 1997. Long-term climate forcing of European herring and sardine
populations. Fisheries Oceanography, 6, 130-139.
Belokopytov, V., 1998. Long-term variability of Cold Intermediate Layer reneval conditions in
the Black Sea. In Ecosystem modeling as a management tool for the Black Sea, NATO Sci.
Partnership Sub-ser., 2, vol. 47, edited by Ivanov, L and Oguz, T., Vol 2, 47-52pp, Kluwer
Academic Publishers.
Chavez, F.P., J. Ryan, S.E. Lluch-Cota, M.Niquen, 2003. From anchovies to sardines and back:
multidecadal change in the Pacific Ocean. Science, 299, 217-221.
Cociasu, A., L. Dorogan, C. Humborg and L. Popa, 1996. Long-term ecological changes in the
Romanian coastal waters of the Black Sea. Marine Pollution Bulletin, 32, 32–38.
17
Conversi, A., S. Piontkovski, S. Hameed, 2001. Seasonal and interannual dynamics of calanus
finmarchicus in the Gulf of Maine (Northeastern US shelf) with reference to the North Atlantic
Oscillation. Deep-Sea Res. II, 48, 519-530.
Cullen, H.M., and P.B. de Menocal, 2000. North Atlantic Influence on Tigris-Euprates
streamflow. Int. J. Climatol., 20, 853-863.
Daskalov, G.M., 2002. Overfishing drives a trophic cascade in the Black Sea. Mar. Ecol. Prog.
Ser., 225, 53-63.
Dippner, J.W., and G. Ottersen, 2001. Cod and climate variability in the Barents Sea. Climate
Research, 17, 73-82.
Drinkwater, K.F., A. Belgrano, A. Borja, A. Conversi, M. Edwards, C.H. Greene, G. Ottersen,
A.J. Pershing, H. Walker, 2003. The response of marine ecosystems to climate variability
associated with the North Atlantic Oscillation. In The North Atlantic Oscillation: Climatic
significance and environmental impact. Geophysical Monograph 134, Edited by J.W. Hurrell, Y.
Kushnir, G. Ottersen, M. Visbeck, 211-233pp, published by the American Geophysical Union.
Ginzburg, A.I., A.G. Kostianoy, N.A. Sheremet, 2004. Seasonal and interannual variability of
the Black Sea surface temperature as revealed from satellite data (1981-2000). J. Mar. Syst. In
press.
Gucu, A.C., 2002.
Can overfishing be responsible for the successful establishment of
Mnemiopsis leidyi in the Black Sea?, Estuarine, Coastal and Shelf Scienc, 54, 439-451.
Hunt Jr. G. L., P. Stabeno, G. Walters, E. Sinclair, R.D. Brodeur, J.M. Napp, N.A. Bond, 2002.
Climate change and control of the southern Bering Sea pelagic ecosystem. Deep-Sea Res. II, 49,
5821-5853.
Huntley, M.E., and Lopez, M.D.G., 1992. Temperature-dependent production of marine
copepods: a global analysis. American Naturalists, 140, 201-242.
18
Hurrell, J.W., 1996. Influence of variations in extratropical wintertime teleconnections on
Northern Hemisphere temperature. Geophys. Res. Lett., 23, 665-668.
Hurrell, J.W., Y. Kushnir, G. Ottersen, M. Visbeck, 2003. An overview of the North Atlantic
Oscillation. In The North Atlantic Oscillation: Climatic Significance and Environmental Impact,
Geophysical Monograph 134, Edited by J.W. Hurrell, Y. Kushnir, G. Ottersen, M. Visbeck, 136pp, published by American Geophysical Union, Washington DC.
Kang, S., S.Kim, and S-W, Bae, 2000. Changes in ecosystem components induced by climate
variability off the eastern coast of the Korean Peninsula during 1960-1990. Progr. Oceanogr., 47,
205-222.
Kideys, A.E., 2002. Fall and Rise of the Black Sea ecosystem. Science, 297, 1482-1484.
Kideys, A.E., A.V. Kovalev, G. Shulman, A. Gordina, F. Bingel, 2000. A review of zooplankton
investigations of the Black Sea over the last decade. J. Marine Syst., 24, 355-371.
Konovalov, S.K., and J.W. Murray, 2001. Variations in the chemistry of the Black Sea on a time
scale of decades (1960-1995). J. Mar. Syst., 31, 217-243.
Kovalev, A.V. and S.A. Piontkovski, 1998. Interannual changes in the biomass of the Black Sea
gelatinous zooplankton. J. Plankton Res., 20, 1377-1385.
Kovalev, A.V., U., Neirman, V.V., Melnikov, V., Belokopytov, Z., Uysal, A.E., Kideys, M.,
Unsal, and D., Altukov, 1998. Long term changes in the Black Sea zooplankton: the role of
natural and anthropogenic factors. In: Ecosystem Modeling as a Management Tool for the Black
Sea. L. Ivanov and T. Oguz, eds. NATO ASI Series 2, Environment-Vol.47, Kluwer Academic
Publishers, Vol. 1, pp.221-234.
Levitus, S., J.I. Antonov, T.P. Boyer, C. Stephens., 2000. Warming of the world oceans,
Science, 287, 2225-2229.
Mankovsky, V.I., V.L. Vladimirov, M.V. Solovev, and S. Besiktepe, 1998. Optical properties of
the Black Sea: Results of the CoMSBlack and TU-Black Sea Programs. In Ecosystem modeling
19
as a management tool for the Black Sea, NATO Sci. Partnership Sub-ser., 2, vol. 47, edited by
Ivanov, L and Oguz, T., Vol 2, 145-162pp, Kluwer Academic Publishers.
Marshall, J., Y. Kushnir, D. Battisti, P. Chang, J. Hurrell, M. McCartney, and M. Visbeck, 1997.
A North Atlantic Climate Variability: phenomena, impacts and mechanisms.Inter. Jour.
Climatology, 21(15), 1863-1898
Mikaelyan, A.S., 1997. Long-term variability of phytoplankton communities in open Black Sea
in relation to environmental changes, in Sensitivity to change: Black Sea, Baltic Sea and North
Sea, NATO-ASI Series, Environment-vol.27, edited by E. Ozsoy and A. Mikaelyan, pp.1105116, Kluwer Academic Publishers, Dordrecht.
Niermann, U., A.E. Kideys, A.V. Kovalev, V. Melnikov, V. Belokopytov, 1999. Fluctuations of
pelagic species of the open Black Sea during 1980-1995 and possible teleconnections. in
Environmental degradation of the Black Sea: Challenges and Remedies, NATO Sci. Partnership
Sub-ser., 2, vol. 56, edited by Besiktepe, S. et al., 147-174pp, Kluwer Academic Publishers.
Oguz, T., H. W. Ducklow, and P. Malanotte-Rizzoli, 2000. Modeling distinct vertical
biogeochemical structure of the Black Sea: Dynamical coupling of the Oxic, Suboxic and
Anoxic layers. Global Biogeochem. Cycles., 14, 1331-1352.
Oguz, T., T. Cokacar, P. Malanotte-Rizzoli, H. W. Ducklow, 2003. Climatic warming and
accompanying changes in the ecological regime of the Black Sea during 1990s. G. Biogeochem.
Cycles, 17(3), 1088, doi:10.1029/2003GB002031.
Ozsoy, E., 1999. Sensitivity to global change in temperate Euro-Asian Seas (The Mediterranean,
Black Sea and Caspian Sea): A review. In The Eastern Mediterranean as a laboratory basin for
the assessment of contrasting ecosystems, NATO Sci. Partnership Sub-ser., 2, vol. 51, edited by
Malanotte-Rizzoli, P., and Eremeev V., 281-300pp, Kluwer Academic Publishers.
Planque, B. and A.H. Taylor, 1998. Long-term changes in zooplankton and the climate of the
North Atlantic. ICES Journal of Marine Science, 55, 644-654.
20
Polonsky, A., E. Voskresenskaya, V. Belokopytov, 1997. Variability of Northwestern Black Sea
hydrography and river discahrages as part of the global ocean-atmosphere fluctuations. in
Sensitivity to change: Black Sea, Baltic Sea and North Sea, NATO-ASI Series, Environmentvol.27, edited by E. Ozsoy and A. Mikaelyan, pp.11-24, Kluwer Academic Publishers,
Dordrecht.
Popa, L., 1995. Observations sur le phosphore dissous dans les eaux du littoral Rouman de la
Mer Noire. Cercetari Marine, 28, 19-28.
Prodanov, K., K. Mikhailov, G. Daskalov, K. Maxim, A. Chashchin, A. Arkhipov, V.
Shlyakhov, E. Ozdamar, 1997. Environmental impact on fish resources in the Black Sea. in
Sensitivity to change: Black Sea, Baltic Sea and North Sea, NATO-ASI Series, Environmentvol.27, edited by E. Ozsoy and A. Mikaelyan, pp.163-182, Kluwer Academic Publishers,
Dordrecht.
Rayner, N.A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C.
Kent, A. Kaplan, 2003. Global analyses of sea surface temperature, sea ice, and night marine
air temperature since the late nineteenth century. J. Geophys. Res., 108 (D14), 4407,
doi:10.1029/2002JD002670.
Reid, P.C., M.F. Borges, E. Svendsen, 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.
Reva, Yu.A., 1997. Interannual oscillations of the Black Sea level. Oceanology (Eng. Transl.),
37, 193-200.
Shiganova, T.A., 1998. Invasion of the Black Sea by the ctenophore Mnemiopsis leidyi and
recent changes in pelagic community structure, Fish. Oceanogr., 7, 305-310.
Shulman, G.E., T.V. Yuneva, O.A. Yunev, L.S. Svetlichny, B.E. Annisky, Z.A. Romanova, E.G.
Araskevich, A.E. Kideys, F. Bingel, Z. Uysal, A. Yilmaz, 1998. A biological approach for the
estimation of food provision for heterotrophic organisms of the Black Sea. In: Ecosystem
Modeling as a Management Tool for the Black Sea, vol. 1, NATO Sci. Partnership Sub-ser., 2,
vol. 47, L.I. Ivanov, and T. Oguz, eds. Kluwer Acad., Norwell, Mass., pp. 263-280.
21
Shushkina, E.A., M.E. Vinogradov, L.P. Lebedeva, T. Oguz, N.P. Nezlin, V. Yu. Dyakonov, and
L.L. Anokhina, 1998: Studies of structural parameters of planktonic communities of the open
part of the Black Sea relevant to ecosystem modeling. In: Ecosystem Modeling as a
Management Tool for the Black Sea, vol. 1, NATO Sci. Partnership Sub-ser., 2, vol. 47, L.I.
Ivanov, and T. Oguz, eds. Kluwer Acad., Norwell, Mass., pp. 311-326.
Sorokin, Y.I., 2002. The Black Sea ecology and oceanography. Backhuys Publishers, Leiden,
875 pp.
Stanev, E.V., and E. L. Peneva, 2002. Regional sea level response to global climatic change:
Black Sea examples, Global and Planetary Changes, 32, 33-47.
Titov, V.B., 2000. Dependence of the formation of the winter hydrological structure in the Black
Sea on the severity of winter conditions. Oceanology (Engl. Transl.), 40, 777-783.
Titov, V.B., 2002. Seasonal and interannual variability of climatic conditions over the Black
Sea, in the Black Sea, in Multidisciplinary investigations of the northeast part of the Black Sea,
edited by A.G. Zatsepin and M.V. Flint, pp.9-19, Moscow, Nauka.
Tsimplis, M.N. and S.A. Josey, 2001. Forcing of the Mediterranean Sea by atmospheric
oscillations over the North Atlantic. Geophys. Res. Lett., 28, 803-806.
Tsimplis, M.N. and M. Rixen, 2003. Variability of Mediterranean and Black Sea level and its
forcing. In Oceanography of Eastern Mediterranean and Black Sea: Similarities and differences
of two interconnected basins. Ed. by Aysen Yılmaz, 137-144pp, Published by TUBITAK
Publishers, Ankara, Turkey.
Vedernikov, V.I. and A.B. Demidov, 2002. Long-term and seasonal variability of chlorophyll
and primary production in the eastern regions of the Black Sea, in Multidisciplinary
investigations of the northeast part of the Black Sea, edited by A.G. Zatsepin and M.V. Flint,
pp.212-234, Moscow, Nauka.
22
Vinogradov, M.E., E.A. Shushkina, A.S. Mikaelyan, and N.P. Nezlin, 1999. Temporal (seasonal,
interannual) changes in ecosystem of the open waters of the Black Sea. in Environmental
degradation of the Black Sea: Challenges and Remedies, NATO Sci. Partnership Sub-ser., 2, vol.
56, edited by Besiktepe, S. et al., 109-129pp, Kluwer Academic Publishers
Yunev, O.A., V.I. Vedernikov, O. Basturk, A. Yilmaz, A. E. Kideys, S, Moncheva, and S.
Konovalov, 2002. Long-term variations of surface chlorophyll-a and primary production in the
open Black Sea, Mar. Ecol. Prog. Ser., 230, 11-28.
Zaitsev, Yu. and V. Mamaev, 1997. Marine Biological Diversity in the Black Sea: A Study of
Change and Decline, GEF Black Sea Environmental Programme, United Nations Publications,
208pp.
23
Table 1. Correlation matrix among the variations of sea surface temperature (SST), Cold
Intermediate Layer temperature (CIL), salinity anomaly of the upper 200 m layer (SAL), and sea
level anomaly (SLA).
Variable
SST
CIL
SAL
SLA
SST
1.000
0.887
-0.579
0.375
CIL
0.887
1.000
-0.529
0.520
SAL
-0.579
-0.529
1.000
0.373
SLA
0.375
0.520
0.373
1.000
Table 2. Correlations of the NAO with hydro-meteorological properties of the Black Sea
including areal averages of the surface air temperature (AT), evaporation minus precipitation (EP), surface atmospheric pressure (SAP), sea surface temperature (SST), Cold Intermediate layer
temperature (CIL), salinity anomaly of the upper 200 m layer (SAL), and sea level anomaly
(SLA).
Pair of variables
NAO-AT
NAO-(E-P)
NAO-SAP
NAO-SST
NAO-CIL
NAO-SAL
NAO-SLA
Number of
measurements
125
54
43
128
44
47
111
Correlation
Coefficient
- 0.277
- 0.095
0.474
- 0.267
- 0.347
0.229
- 0.198
Table 3. Correlations of the NAO and SST with biogeochemical properties of the Black Sea
including phosphate concentration (PO4), phytoplankton biomass (Phyto), Secchi disk depth
(Sechi), hydrogen sulfide concentration (H2S), mesozooplankton biomass (Mesozoo), parental
sprat and anchovy stocks (Sprat, Anchovy).
Pair of variables
Number of
measurements
NAO-PO4
38
SST- PO4-PO4
NAO-Phyto
21
SST-Phyto
NAO-Sechi
43
SST-Sechi
NAO-H2S
24
SST-H2S
NAO-Mesozoo
38
SST-Mesozoo
NAO-Sprat
26
SST-Sprat
NAO-Anchovy
26
SST-Anchovy
Correlation
Coefficient
- 0.53
0.35
0.44
- 0.79
- 0.51
0.65
0.85
- 0.76
0.17
0.39
0.09
- 0.44
- 0.66
0.53
24
FIGURE CAPTIONS
Figure 1. Long-term variations of (a) the winter (December-March) mean sea surface
temperature (oC) averaged over the interior basin with depths greater than 1500 m (after Rayner
et al., 2003), (b) the mean temperature (oC) of the Cold Intermediate Layer for the MayNovember period (after Belokopytov, 1998), (c) the annual mean salinity anomaly of the upper
200 m layer (after Tsimplis and Rixen, 2003), (d) the detrended sea level anomaly (cm) (after
Reva, 1997). High frequency oscillations in the data have been filtered by the three point
moving average.
Figure 2. Long-term variations of (a) the winter (December-March) mean air temperature (oC)
measured at the meteorological station near the Kerch Strait along the north coast of the Black
Sea (after Titov, 2002), (b) the winter (January-March) mean wind stress magnitude averaged
over the basin obtained from the NCEP data set, (c) the winter (December-March) mean surface
atmospheric pressure (hPa) averaged over the basin obtained by ECMWF Re-analysis data set,
(d) the winter (December-March) mean evaporation minus precipitation (km3 yr-1) averaged over
the basin obtained by ECMWF Re-analysis data set. High frequency oscillations in the data
have been filtered by the three point moving average.
Figure 3. Long-term variations of the winter North Atlantic Oscillation index based on pressure
differences between Azores and Iceland, using the normalization period 1871-1995. High
frequency oscillations in the data have been filtered by the three point moving average.
Figure 4. Long-term variations of (a) the surface phosphate concentration in Romanian coastal
waters, off Constantsa (after Popa, 1995), (b) the phytoplankton biomass (squares) surface
chlorophyll (stars) distributions for the May-September period within the interior part of the sea
(after Mikaelyan, 1997 and Yunev et al., 2002), (c) the annual mean secchi disc depth for the
interior basin (after Mankovsky et al., 1998), (d) the annual mean hydrogen sulfide concentration
(µM) at 16.4 kg m-3 sigma-t surface (after Konovalov and Murray, 2001)
Figure 5. Long-term variations of the annual-mean mesozooplankton biomass (g m-2) within the
interior basin (after Kovalev et al., 1998).
25
Figure 6. Long-term variations of (a) the sprat parental biomass (ktons), and (b) sprat
recruitment (number *109) stock (after Daskalov, 2002).
Figure 7. Long-term variations of (a) the anchovy parental biomass (ktons), and (b) anchovy
recruitment (number *109) stock (after Daskalov, 2002).
26
FIGURE 1
27
FIGURE 2
28
FIGURE 3
FIGURE 4
29
FIGURE 5
FIGURE 6
30
FIGURE 7