The Europaen Commission The Commission on the Protection of the Black Sea Against Pollution
BSC LogoEC Logo

Report Contents

Preface Chapter 1A Chapter 1B Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11 Chapter 12
List of Tables List of Figures

State of Environment Report 2001 - 2006/7

Chief Editor, Prof. Dr. Temel Oguz, Institute of Marine Sciences, Middle East Technical University, Erdemli, Turkey

Chapter 2 - State of Chemical Pollution

State of the Environment of the Black Sea - 2009

CHAPTER 2 THE STATE OF EUTROPHICATION(T. Oguzet al.)

T. Oguz

Institute of Marine Sciences, Middle East Technical University, Erdemli, Turkey

V. Velikova

Inebolu Sokak 29, Kabatas, Istanbul, Turkey

A. Cociasu

National Institute for Marine Research and Development (NIMRD), Constanta, Romania

A. Korchenko

GOIN, Moscow, Russian Federation

2.1. Introduction

Marine eutrophication of coastal waters is considered to be a significant problem worldwide. The Black Sea is no exception, where considerable increase in the nutrient load has led to marked changes in the ecosystem structure and functioning. Eutrophication is defined here as excessive supply of nutrients (silicate, nitrogen and phosphorus) into water that subsequently leads to accelerated growth and over-production of algae and species of higher trophic levels, high rate of oxygen depletion, development of hypoxia or anoxia near the bottom of productive areas and subsequent degradation of benthic community structure. A key to successful management of coastal waters is reliable scientific assessment of eutrophication and of its governing processes. This chapter evaluates the current state and the long-term trend of eutrophication in the Black Sea using the available data for river nutrient (both organic and inorganic) loads, dissolved inorganic and organic nutrient, chlorophyll-a, surface and subsurface oxygen concentrations along the coast with respect to the reference conditions of the interior basin. The coastal stations are chosen from the sites that either receive river loads (like Sulina station) or are in close proximity of industrial complexes (like Constanta) or urban settlements (like Bay of Bourgas). Unless otherwise specified, the data described below are from the data-base of the Commission on the Protection of the Black Sea Against Pollution.

2.2. Long-term changes in river nutrient loads

The excessive nutrient enrichment was originated predominantly by enhanced river-based nutrient supply into the northwestern shelf starting by the early 1970s. The River Danube was used to be the major supplier of nutrients during the 1970s and 1980s that amounted to almost 80% of the total load into the sea. They were derived by agriculture, industry and urban settlements and supplied mainly through diffuse sources (Table 2.2.1). Based on more recent measurements at the discharge point of the Sulina branch, the River Danube contribution was reduced to nearly 50% of the overall river-based DIN and P-PO4 loads (Table 2.2.2). The remaining half was contributed almost equally by the Ukrainian rivers along the northwestern coast (Dniepr, Dniester, Bug) and the Turkish Rivers along the southern coast (Table 2.2.2).

Table 2.2.1. The sources and causes of nutrient enrichment in the western coastal waters of the Black Sea.

Drivers of nutrient enrichment

Causes of nutrient enrichment
Agriculture/farmingLack of fertiliser storage facilities

Unsustainable/inefficient farming practices

Intensive livestock production

Intensive fertiliser utilization and detergents

Lack of proper effluent treatments of discharges from livestock and agricultural farms

Point and diffuse sources from agriculture/farming, industry and settlements

Deposition from atmospheric emissions originated from land-based sources

Background emissions

IndustryUntreated or improperly treated industrial effluents due to outdated or absence of treatment technology?

Insufficient treatment plants and their poor management

Lack of control for waste water treatment plants

Table 2.2.2 Annual-mean river-borne nutrient loads (in kilotonnes y-1) into the Black Sea from each country during 2003-2005 (from TDA, 2007). The superscripts (a) and (b) denote the estimates based on the TNMN measurements at Reni and the NMRD measurements at Sulina discharge point, respectively.

Nutrient LoadUkraineRomaniaBulgariaTurkeyGeorgiaRussiaTotal
DIN29.85304.10(a)

?? 68.86(b)

2.3524.870.540.84362.55(a)

127.31(b)

P-PO42.308.80(a)

8.52(b)

0.246.130.020.3217.81(a)

17.53(b)

The total nitrogen emission from the River Danube catchments increased from about 400 kilotonnes (kt) y-1 in the 1950s to 900 kt y-1 in 1985-1990 and then reduced to 760 kt y-1 in 2000-2005 (Fig. 2.2.1a). Phosphorus emission was an order of magnitude smaller and changed from 40 kt y-1 in 1950s to its peak value of 115 kt y-1 during? the first half of the 1990s and then to 70 kt y-1 in 2000-2005 (Fig. 2.2.1b). Both emissions are thus still roughly 1.5 times higher than the 1950s. The construction of Iron Gate 1 in 1975 was estimated to impose a minor influence on the total-N load but was more critical on controlling the total-P load (daNUbs Project Final Report, 2005). The nitrogen?based agricultural run-off contributed around half of the total nitrogen emission since 1955 and contribution from urban settlements was around 20-30% (Fig. 2.2.1a). The increase in nitrogen emission is evidently well-correlated with fertilizer consumption prior to 1990 that increased from 0.5 million tonnes (mt) y-1 in the 1950s to 3.0 mt y-1 during the green revolution era in the eastern block countries during the 1980s (Fig. 2.2.1). According to this data set, phosphorus emission was mainly supplied by urban settlements up to 80% in 1990; the contribution of agriculture remained less than 20% (Fig. 2.2.1b). The effect of increase in phosphorus fertilizer consumption to 1.5 mt y-1 did not appear to increase agricultural-based phosphorus emission.

Following the collapse of their centrally-planned economy and economic recession, nitrogen and phosphorus fertilizer consumptions reduced to 1.5 mt y-1 and 0.5 mt y-1, respectively, during 1990-1991. Fig. 2.2.1 however suggests surprisingly minor change in agricultural nitrogen emission and no change in agricultural phosphorus emission. The phosphorus emission was reduced more predominantly by the improvement of regional environmental management, the introduction of phosphorus-free detergents, the improved nutrient removals at treatment plants (daNUbs Project Final Report, 2005).

Fig. 2.2.2 Annual DIN and P-PO4 loads measured at Sulina discharge point of the River Danube to the sea and at Reni located at the upstream end of the Sulina branch.

The lack of major reduction in nitrogen emissions from the Danube catchments basin is reflected in relatively high DIN load into the sea. At Reni (located at the upstream end of the Sulina branch), it fluctuated within the range of 300-500 kt y-1 since the early 1990s (Fig. 2.2.2a) in consistent with the N-emission data (Fig. 2.2.1a). Such relatively high DIN load was explained by continuing emissions from large nitrogen stocks deposited in soils and groundwater in the catchments areas (daNUbs Project Final Report, 2005). On the basis of Reni data, a factor of three reductions is necessary to accomplish its pristine level prior to 1960 (~150 kt y-1). According to the NIMRD (National Institute of Marine Research and Development, Romania) measurements at the discharge point of Sulina branch of the River Danube, the DIN flux entering the sea tended to decrease gradually during the 1990s to ~100 kt y-1 at 2000 and increased afterwards to 250 kt y-1 in 2005 that was twice higher than its pristine level. It is hard to justify the considerable difference between two ends of the Sulina branch (~40 km), but the concomitant reduction in the interior basin subsurface nitrate peak (section 2.3) supports the Sulina data. On the other hand, both measurements reveal similar P-PO4 load between 10 and 20 kiloton yr-1 after the mid-1990s (Fig. 2.2.2b).

No systematic data sets are available to assess the current state of organic nutrient loading from the northwestern rivers. But, the BOD5 data may be used to deduce indirectly their DON loads. As inferred from Fig. 2.2.3, the total BOD5 data from the Rivers Danube and Dniester, Dniepr and Bug display a gradual decreasing trend from the peak values close to 1000 kt y-1 during 1997-1999 to ~500 kt y-1 after 2000, which implies approximately 50% reduction. The Danube contribution to the total BOD load amounts to 80%. Assuming the DON (dissolve organic nitrogen)/ BOD5 ratio ~0.3 (San Diego-Mcglone et al., 2000), the DON load into the northwestern Black Sea is estimated as 150 kt yr-1 in the present decade. The Danube contribution turns out to be approximately 130 kt yr-1, which roughly corresponds to one-third of the Reni DIN load. This is a rather conservative estimate and represents an average condition over the NWS. It may in fact be locally much higher.

Fig. 2.2.3. BOD5 load from the Danube River and other rivers (Dniester, Dniepr and Bug) discharging into the northwestern shelf during 1995-2005.

2.3. Long-term changes in nutrient concentrations

Nutrients along the western coast acquire maximum concentrations in the Danube delta region that is characterized by relatively colder and less saline water mass and separated from offshore waters through a sharp frontal zone (Fig. 2.3.1). A predominant feature of the delta region is high rates of sediment deposition and resuspension mechanisms and rapidly changing physical conditions (in daily-to-weekly time scales) that introduce great deal of patchiness in nutrient concentrations. Almost 30-50% of nutrients supplied by the Danube is estimated to deposit near its mouth (Velikova and Cociasu, 2004) and thus does not contribute directly to the local biological production. At longer time scale (i.e. ~the last 50 years), sediment cores suggested 20% of total riverine nitrogen input accumulated in sediments of the western coastal waters, whereas denitrification eliminated 55% of the input (Teodoru et al., 2007).? When the Danube plume extends southward, as shown in Fig. 2.3.1, low salinity and high nutrient water mass looses its character immediately to the south of the delta region and N-NO3 concentration generally decreases by more than 50% (Fig.2.3.1).

Changes in nitrogen concentration: Consistent with the DIN load data shown in Fig. 2.2.2a, the annual-mean DIN concentration at Sulina discharge point shows a gradual reduction from ~300 μM at 1990 to ~40 μM at 2000 and then exhibits a slight rise to ~70 μM at 2005 (Fig. 2.3.2). Albeit its large drop, recent values of DIN are still too high. The contribution of N-NH4 to DIN remained negligibly small during the 1990s, but was around 10% after 2000. On the contrary, the annual-mean DIN concentration persistently varied around 150 μM at Reni during 1997-2005 (Fig. 2.3.2).

Fig. 2.3.2 Annual DIN concentration measured at Sulina discharge point of the River Danube to the sea and at Reni located at the upstream end of the Sulina branch.

As inferred by the amalgamated annual-mean surface NO3 data (Fig. 2.3.3), following an increase in the first half of the 1990s, N-NO3 concentration decreased during the second half and then started rising in the present decade. Its mean values for the previous and present decades are approximately 5 μM and 7 μM, respectively. A notably similar long-term DIN structure also took place to the north of the Danube delta. Along the western coastal waters of Ukraine, DIN concentration increased from 1 μM in the 1960s to 9 μM in the 1980s and stabilized thereafter around 7.5 μM (Fig. 2.3.4) that was similar to that in the Romanian shelf. Although the eastern coastal waters of Ukraine had relatively better conditions with lower concentrations (~4 μM in the present decade) an increasing tendency over the last two decades is evident (Fig. 2.3.4).?

Surface nutrient concentrations measured monthly at the coastal sampling station near Constanta constitute the most comprehensive long-term data set to monitor their multi-decadal changes along the western coastal waters. Following a major reduction during the second half of the 1970s, N-NO3 and N-NH4 concentrations fluctuated within 5-10 μM during the 1980s and 1990s except a rising trend in 2004-2005 (Fig. 2.3.5). The decadal-mean N-NO3 concentration reduced from 6.90 μM in the 1980s to 5.90 μM in the 1990s and then elevated to ~8 μM during 2000-2005 (Table 2.3.1). N-NH4 concentration experienced an opposite trend and increased roughly 2.0 μM from the 1980s to the 1990s and then reduced by 1.0 μM during the present decade (Table 2.3.1). N-NO2 concentration always remained around 1 μM and constituted only 6% of the total DIN. In general, the DIN possessed a slight rising trend from 12 μM in the 1980s to 13 μM in the 1990s and 14 μM in the 2000s (Table 2.3.1). This trend is consistent with the Sulina data (Fig. 2.2.2a) although the values are four-fold lower.

Table 2.3.1. Multi-annual mean surface nutrient concentrations (μM) in Constanta (after Velikova and Cociasu, 2004)

Period1959-651983-901991-002001-05
N-NO31.606.905.907.98
N-NH4-5.117.066.12
P-PO40.266.541.860.49
SiO440.511.012.613.7

According to the monthly surface NO3 and N-NH4 concentration variations at Constanta during 2000-2006 (Fig. 2.3.5), the period 2000-2004 was characterized by a major N-NO3 peak during April or April-May that coincides with the highest discharge period of the Danube. N-NO3 concentration then decreased linearly up to September and then started increasing linearly up to next April. As an exception, a second and stronger peak appeared during the 2003 severe winter (January-February). This annual structure changed appreciably during 2005-2006 in terms of an extended spring peak to July and reduction in N-NO3 concentration by September up to a minimum in November; thus the minimum N-NO3 concentration phase shifted from August to November.? Moreover, NO3 concentrations from 2000 to 2006 indicated approximately 5 μM linear trend of increase. NH4 concentration attained higher values predominantly in summer months.

Considerable interannual and local variabilities are also noted along the northwestern coastal waters (Fig. 2.3.7). N-NO3 and N-NH4 concentrations were exceptionally high (60-90 μM) near the waste water treatment plant sites Pivnichi and Pivdenna during 2003-2006 contrary to their much lower values during 2000-2002.? N-NO3 concentrations in the Odessa, Uzhnyi, and Ilichevsk ports were typically around 10-20 μM.

Measurements along the Romanian and Bulgarian coastal waters to the south of the Danube delta revealed comparable N-NO3 and N-NH4 concentrations with the Ukrainian coastal waters. N-NO3 concentrations at Mamaia beach and two adjacent offshore stations along 5m and 20 m isobaths indicate uniformity of the Danube plume within 20 m isobath zone (Fig.2.3.8a).? The features such as the April-May N-NO3 peak within 10-20 μM range and relatively high spring concentrations during 2002 and 2005 resemble those observed at Constanta station (Fig.2.3.6).

Further south along the southern Bulgarian coast, N-NO3 and N-NH4 at stations near Ahtopol and Bourgas Bay reveal concentrations as high as 20 μM in summer months that are comparable with their winter values (Fig. 2.3.8b,c). Their annual structure was therefore somewhat different than Constanta station, and likely signifies contribution from local sources in addition to the Danube upstream influence.

More dense and regular measurements in the Bosporus exit section during 1996-2003 possessed an order of magnitude lower N-NO3 concentration (Fig. 2.3.8d). More importantly, they revealed a typical open-sea type annual structure with an N-NO3 peak during winter due to enhanced vertical mixing, and N-NO3 depletion in summer due to its uptake in phytoplankton production. An additional spring (April-May) peak exists during high Danube discharge years, reflecting efficiency of southward coastal transport. Opposite to the Bourgas Bay and Ahtopol, in the Bosporus section the summer N-NO3 concentration was low due to its intense uptake in phytoplankton production.?

As inferred from the BOD5 data (Fig. 2.1.3), the western Black Sea coastal waters seem to continue receiving considerable anthropogenic-based organic nitrogen. An estimate of 130 kt y-1 DON load inferred earlier from the Danube BOD5 data implies ~30 μM DON concentration supplied by the River Danube during the present decade that is comparable to ~40 μM in the northwestern coastal waters (Fig. 2.2.9) due to continuous supply of high DON load from Dniepr, Dniestr and Bug Rivers as evident by a steady increase of its concentration since the 1970s. Similarly, the annual- and regional-average dissolved organic nitrogen concentration of ~10 μM was observed along the Bulgarian coastal waters during 2001-2003 (Table 2.2.2).? In addition, more than 10 μM difference between total nitrogen and N-NO3 concentrations in the Bosporus exit station during the peak Danube discharge period (March-May) throughout 1996-2001 (Fig. 2.2.8d) provides a further support for organic nitrogen enrichment all along the western coastal waters.

On the basis of available data for 2000-2005, Fig. 2.3.10 summarizes the current status of the nitrogen enrichment along the periphery of the basin using different color codes for different ranges of annual-mean N-NO3 concentration. According to this classification, the threshold for the highest N-NO3 concentration in western coastal waters were set to 2.8 ?M that in reality was far more higher and easily exceeded 10 ?M as documented above. The entire southern coast is identified by the second highest N-NO3 concentration that covers the range between 1.4 ?M and 2.8 ?M. A small number of sites offshore of the Turkish/Georgian border show high levels of enrichment, which are likely to be the result of local discharges from mining industry. Northeastern and Crimean coastal waters are less eutrophic and characterized by N-NO3 concentration less than 1.4 ?M. It is however important to note that, for southern and northeastern coastal waters, this classification scheme was based on rather limited measurements, such as twice a year (April and October) sampling during only 2004 and 2005 along the southern coast. Fig. 2.3.10 therefore should be interpreted with some caution. More frequent sampling strategy is highly desirable for more reliable coverage of the annual range of variations.

The surface productive zone of the interior basin has been nitrate limited even during the intense eutrophication phase (Oguz, 2005), and therefore the change of peak N-NO3 concentration in the nitracline layer (see Chapter 1) is an important indicator for monitoring basin-scale response of eutrophication. The change in subsurface nitrate maximum concentration (Fig. 2.3.11) from ~2-3 ?M during the pristine state to 7-9 during the intense eutrophication period (i.e. the 1980s) confirms devastating effect of eutrophication for the entire sea that persisted up to 1992. Afterwards, the interior basin responded to the decline in the anthropogenic DIN load by reducing the peak subsurface nitrate concentration below 6 ?M after the 1990s. The present value of this peak varies around 4.0-5.0 ?M and thus implies a gradual shift of the system towards low nutrient conditions. As the nitrogen (both organic and inorganic) load supplied by the northwestern rivers was decreasing according to the Sulina measurements, primary production in the interior basin at present appears to consume nutrients which have already accumulated within during the intense eutrophication phase.?

Figure 2.3.10. Ranges of mean nitrate (NO3-N) concentrations during 2000-2005 in surface waters (0-10 m) along the coast of the Black Sea (modified from TDA Report, 2007).

Changes in phosphorus concentration: P-PO4 concentration in the Sulina discharge point experienced an abrupt reduction from 7.5 μM at 1990 to 2.5 μM at 1993. It fluctuated afterwards between 2 μM and 3 μM, whereas it was changing between 1.0 and 2.0 μM at Reni (Fig. 2.2.12).

Along the northwestern coastal waters, it increased to around 3 ?M in the mid-1970s and retained this level for a decade (Fig. 2.3.13a) reflecting upstream influence of the Danube in addition to local contributions from hot spots and discharges from Dniepr, Dniestr and Bug Rivers. This level was maintained until 2005 even though it locally acquired order of magnitude higher values during 2004-2005 (Fig. 2.3.13b). Long-term P-PO4 data at Constanta monitoring station (Fig. 2.3.14) indicate two major peaks (~10-12) ?M at 1975 and 1987 and minima (~2 ?M) at 1981 and 1992 following its background values of <1 ?M during the 1960s (the mean value = 0.26 ?M). The decadal-mean P-PO4 concentration was highest (~6.0 μM) during the 1980s and decreased to values less than 1.0 ?M during 1990-1992 with a mean value of 0.49 ?M for 1993-2005 (Table 2.3.1). The Constanta data is however biased due to the strong impact of the land-based source from the Navodari Factory and thus may not be representative for all the Romanian coastal waters. Further reduction to the mean value of ~0.2 ?M is noted along the Bulgarian coast (Table 2.3.2) and northeastern coastal waters (Fig. 2.3.15).

Fig. 2.3.12. Annual-mean P-PO4 concent-ration measured at Sulina discharge point of the River Danube to the sea and at Reni located at the upstream end of the Sulina branch.

Fig. 2.3.13a. Decadal variability of surface phosphate (P-PO4) concentration along the western and eastern coastal waters of the NWS based on the averaging of available data from several stations (redrawn from Loveya et al., 2006).

According to the colour-coded classification scheme (Fig. 2.3.16), the western coastal waters are classified as highest phosphate levels (>0.39 ?M). A comparable level of elevated phosphate concentrations was assigned in Fig.2.3.16 for the Russian coast along the opposite side of the sea, but it disagrees with lower values shown in Fig. 2.3.15. The Turkish coast is characterized by relatively low P-PO4 concentration albeit isolated sites of higher contamination exist presumably due to local industrial discharges.

The annual-mean surface silicate concentration (SiO4) measured at Constanta monitoring station prior to 1970s was quite high with an average value of 40 μM and reached at 65 μM at 1975 (Fig. 2.3.14). Following the construction of Iron Gate I during the early-1970s, it started reducing to around 10-20 μM within the second half of 1970s due to the retention mechanism in the upper reaches of the River Danube. The declining trend since the 1980s was disrupted by an abrupt increase from 10 μM in 1996 to 22 μM in 1998 and then declining trend continued up to 11.0 μM in 2004 except a minor increase to 13.7 μM in 2005.

Changes in nutrient ratios: Owing to relatively high DIN emissions, the River Danube has always been a phosphorus limited system and supplied DIN at much higher rate with respect to P-PO4. The P-limitation is evident at the Sulina discharge point where N/P ratio remained above its Redfield ratio even though both N-NO3 and P-PO4 concentrations tended to decrease since 1990 and P-limitation becomes weaker than the previous decades (Fig. 2.3.17a). The increasing trend during 1990-1993 reflects much faster decline of P-PO4 concentration (Fig. 2.3.12), whereas the gradual decline during 1993-1999 and increase afterwards are largely controlled by the variations of N-NO3 concentration (Fig. 2.3.2) as P-PO4 concentrations remained rather steady after 1993. The data at Constanta monitoring station also support the P-limitation at an increasing rate after 1997 as compared to the N-limitation before (Fig. 2.3.18a). On the contrary, the amalgamated data (Fig. 2.3.17a) indicate preferentially N-limitation for the Romanian shelf waters. The data shown in Table 2.3.3 also support N-limitation for the 1980s and 1990s but not for the 2000-2003 period for which the N/P ratio value of 32 contradicts with a value of ~12 suggested by the amalgamated data. The difference seems to arise due to the bias of the data given in Table 2.2.3 by the Constanta N/P data (Fig. 2.3.18a) that showed P-limitation after 1999.

The N-limitation also applies for both the northwestern and northeastern Ukrainian coastal waters (Fig. 2.3.17b and Table 2.3.3) as well as the Bosporus exit section (Fig. 2.3.17a), except for 2003 that was based on an average of only two measurements; one below the Redfield ratio and the other above and therefore the 2003 value of N/P ratio is not statistically significant. It is hard to identify precisely the type of nutrient limitation for Bulgarian waters because of the absence of systematic and sufficiently long time series for all major nutrients. On the basis of available data summarized in Table 2.3.3, Bulgarian coastal waters appear to be P-limited during winter months due to their excessive nitrogen enrichment (Fig. 2.3.8b,c, 2.3.18b), while they attained either N-limitation or Redfield value the rest of the year (Fig. 2.3.18b). Bulgarian waters therefore switch seasonally between P- and N-limitation (Velikova and Cociasu, 2004). The seasonal alternation does not show a regular seasonal pattern, but varies interannually. On the basis of all these evidences, it appears that inner shelf along the western coast is currently P-limited (Fig. 2.3.18a) whereas the outer shelf is either weakly N- or P-limited (Fig. 2.3.17), and the interior basin is predominantly N-limited (Fig. 2.3.19a).

The long-term variation of Si/N ratio at Constanta (Fig. 2.3.18a) since 1980 falls into three phases. The first phase (1980-1988) was characterized by the monthly Si/N values varying between 0.5-1.5.? In the second phase (1989-1997), Si/N values reduced to the range 0.4-1.2. The third phase (1998-2005) resembled the first phase with the values changing between 0.6 and 2.0.? In general, Si/N ratio remained below 2.0 and favored Si limitation in the Constanta region. Because of the lack of comprehensive data coverage, it is not clear how well the Si limitation applies to other regions along the western coast. Because the deep interior basin is more severely limited by nitrogen,? Si/N values are much higher than its threshold for Si limitation (Fig. 2.3.19b).

Fig. 2.3.18. Long-term changes of nutrient concentration ratios at (a) Constanta monitoring station, and (b) monthly changes in Bulgarian coastal waters during 2001-2003.

Atmospheric input of nutrients: In addition to nutrient inputs from rivers, coastal sources and sediments, atmospheric wet and dry depositions may likely constitute an important component of the long-term nutrient enrichment of the Black Sea that however remains to be poorly studied. According to the State of Environment Report (2000), the atmospheric nutrient input onto the Black Sea surface was about 400 kt y-1, and therefore higher than the present level of riverine supply.

Within the framework of BSERP/GEF monitoring program, atmospheric precipitation was sampled at a remote site Katsively near Sevastopol-Crimea during 2004-2005 (Chaykina et al., 2005). They measured the total DIN wet deposition as 550 mg m-2 during March-December 2004 and 910 mg m-2 during 2005 that amounts to atmospheric fluxes of 240 kt y-1 and ?400 kt y-1, respectively, if extrapolated over the surface area of the Black Sea. Similar measurement at Zmeiny Island at 40 km offshore from Odessa during 2003-2007 indicated approximately 240 ktons y-1 nitrogen and 16 ktons y-1 of phosphorus fluxes from the atmosphere (Medinets and Medinets, 2008). Atmospheric deposition of nutrients is therefore comparable to the current loads from the River Danube and therefore appears to constitute an important element of the overall nitrogen budget of the sea.

Table 2.3.3. Multi-annual mean nutrient ratios in Romanian surface coastal waters, the Bay of Varna and Bulgarian coastal waters (BW) as an average of all measurements within the 50 km zone (after Velikova and Cociasu, 2004)

Romania

Bulgaria

Ukraine

Period

1960-1970

1980-1991

1992-1999

2001-2005

2001-03

2000-05

BWVarna BayNW

coast

NE

Coast

Si / P142.92.611.230.634.488.4--
Si / N-0.90.91.29.07.0--
N / P-3.110.232.055.231.010.715.7

2.4. Surface chlorophyll concentration

Few long-term data exist on in situ chlorophyll-a measurements in the Black Sea and therefore remote sensing ocean color data are used to infer its seasonal-to-interannual variations, even though chlorophyll signal is likely deteriorated by rich particulate material and dissolved organic substance along the western coastal waters. The average chlorophyll concentrations are computed from 9 km gridded, 8-daily SeaWiFS products for the northwestern (Ukrainian) coastal waters (the region 1), the Romanian and Bulgarian coastal waters (the regions 2 and 3), the Bosporus-Black Sea junction region (the region 4), and the western interior basin (the region 5) (Fig. 2.4.1). The Danube delta region between the regions 1 and 2 is not included into the analysis because of its unrealistically high chlorophyll concentrations (≥ 20 mg m-3), most of which possibly reflect contributions from yellow substance and particulate matter discharged by the Danube.? The analysis of the data is depicted in Fig. 2.4.2.

An immediate feature noted in Fig. 2.4.2 is approximately four-fold decrease in chlorophyll concentration from the northwestern shelf (region 1) to the Bosporus-Black Sea junction zone (region 4). In the region 1, highest concentrations extend from early spring to late autumn in the range of 6-8 mg m-3. Lowest concentrations are observed in winter months and often exceed those in further south. The region 2 possesses three peaks; the first one comprises the late winter-early spring (February-March) and arises due to typical spring bloom dynamics. This is the weakest peak (~3 mg m-3). It is followed by a stronger late-spring peak (in May-June with ~5?1 mg m-3) that evidently coincides with the period of high Danube discharge. For some years (i.e. 2006), these two peaks are combined to form one extended peak. The third one occurs in autumn (~4?1 mg ?m-3) and is often separated from the former one by a period of relatively low summer concentrations of ~2 mg m-3. The measurements in Constanta monitoring station during 2001-2005 (Fig. 2.4.3) generally support these features with some differences such as a strong early autumn (September) bloom in 2002, a late winter (February-March) bloom in 2003, and a late autumn bloom (November) in 2004. The Bulgarian and the Bosporus regions also possess three-peaks between 2 mg m-3 and 4 mg m-3 with considerable year-to-year changes whereas the lowest concentrations in summer months are about 0.5 mg m-3.

The long-term monthly measurements at the Bosporus exit section during 1996-2001 (Fig. 2.4.4) generally support the ocean colour data both in terms of timing and magnitude. The most pronounced features are the late winter and autumn peaks that are connected with low summer chlorophyll concentrations. The measurements in Bourgas Bay during 1987-1997 (Fig. 2.4.4) that comprises the period before the availability of satellite ocean color data suggest somewhat different seasonal chlorophyll concentration structure; the late winter-spring peak emerges the most dominant feature at either moderate (~4-5 mg m-3 ) or high (>8 mg m-3) concentrations and is followed by a secondary peak during summer months. No appreciable chlorophyll peak however existed during autumn months.? This annual structure persisted during 1987-1995 period, after which the annual structure seems to shift to the structure described for the region 4 with higher intensity of autumn concentrations.

Fig. 2.4.4. Monthly surface chlorophyll concentration during 1987-2001 measured in the Burgaz Bay (red dots) and the Bosphorus northern exit (green squares), and the SeaWiFS ocean color data for the region 4 (bold lines). The dashed line shows decreasing trend of peak Chl concentration since the 1980s. The field data are provided by Hibaum (2005); Moncheva (2005) and Okus (2005) and satellite data by daily-8, 9 km resolution SeaWiFS ocean colour product.

Fig. 2.4.5. Surface chlorophyll-a concentration measured along the Crimean topographic slope zone (dots; after Churiliova et al., 2005), near the Cape Sinop at the central part of the southern coast (triangles; after Feyzioglu, 2006; Bat et al., 2007), and Surmene Bay near the southeastern corner of the sea (squares; after Feyzioglu, 2006).

On the other hand, measurements at two coastal stations along the southern coast and along the topographic slope zone to the south of Crimea (Fig. 2.4.5) show peak concentrations in the range of 1.0-2.0 mg m-3 during February-March and November-December that are comparable to minimum concentrations measured at Bourgas and Bosporus stations. The lowest concentrations which are generally observed during the summer were typically less than 0.5 mg m-3.

The surface chlorophyll concentration variations within the central part of the western basin (region 5) vary within 0.5 - 2.0 mg m-3 (Fig. 2.4.2). The most notable feature that repeats almost every year is a linear decreasing trend from a peak in November (1.5 - 2.0 mg m-3) to a minimum (~0.5 mg m-3) in July followed by a sharp increase from August to November again. This annual structure differs from the western coastal waters (the regions 1-to-4) by the absence of additional spring peak due to the Danube nutrient supply. The only exception is the anomalously high concentrations in summer 2001 as also supported by direct measurements (Oguz and Ediger, 2007).

The long-term data using all available direct measurements within the interior basin (> 1500 m) for the summer-autumn period (May-November) reveal surface chlorophyll values less than 0.6 mg m-3 during 1978-1987 followed by the phase of high concentrations up to about 2.0 mg m-3 during 1989-1992 during the anchovy collapse and Mnemiopsis population outburst period (i.e. the changes in top-down control) (Fig. 2.4.6). There is no sufficient data coverage to assess chlorophyll variations during the mid-1990s even though values up to 1.0 mg m-3 were measured during 1998-1999. The SeaWiFS summer (May-September) data for more recent period (averaged over the rectangular region defined by 31-41oE and 41.5-44oN) reveal two distinct phases: high summer chlorophyll concentrations (~1.4 mg m-3) during 1998-2001 and the subsequent moderate chlorophyll concentrations (~0.8 mg m-3) afterwards.??

Fig. 2.4.6. Long-term changes of surface chl-a concentrations in the interior Black Sea (depth >1500m) for May-November period, 1978-1999. Each data point represents a single measurement taken from Yunev et al. (2002) (blue symbols) and Churilova et al (2004) (red symbols). This data set is complemented by the monthly-average 9 km gridded May to September-mean SeaWiFS data for the interior basin defined by 31-41oE longitudes and 41.5-44oN latitutes (green symbols).

The annual-mean chlorophyll concentrations deduced from the satellite ocean color data do not indicate appreciable interannual variability for the regions 3, 4 and 5 (Fig. 2.4.7), but attained relatively low values during 2003 in regions 1 and 2 due possibly to the anomalous cold winter (Chapter 1). The decrease in annual-mean concentrations from north-to-south and from 1998 to 2007 is also noted in Fig. 2.4.7.

Fig. 2.4.7. Annual-mean surface chlorophyll concentration for five regions of the western Black Sea basin obtained from 8-daily 9 km resolution SeaWiFS ocean colour product. The colour code for the regions are as shown in Fig.2.4.2.

The surface chlorophyll maps presented in Fig. 2.4.8 provide the monthly average distributions for 2001, 2003 and 2005 that were chosen to represent examples for highly, weakly, and moderately productive years, respectively. Throughout these years, the northwestern part of the Black Sea possess highest chlorophyll concentrations (>10 mg m-3) especially in summer months. This over-productive zone weakens gradually southward along the Bulgarian and Turkish coastal waters that are characterized by the values around 2-4 mg m-3. The Danube effect is most often visible up to the Bosporus coastal region, and its signature gradually weakens further eastward along the Anatolian coast. Some localized moderately productive zones are occasionally observed along the south-eastern and eastern coasts. The northern coastal zone was identified by lowest concentrations. The Rim Current system forms a barrier between coastal and interior waters during winter and spring periods when it is relatively strong, but more efficiently distributes chlorophyll between coastal and interior waters during summer and autumn when it is weakly stable and prone to mesoscale instabilities.

Fig. 2.4.8. Monthly composites of surface chlorophyll concentration derived from daily-2km resolution SeaWiFS (NASA) ocean colour products for 2001, 2003 and 2005 (EC-Joint Research Centre, Global Environment Monitoring Unit Ocean Colour Archive, http://oceancolour.jrc.ec.europa.eu/).

Fig. 2.4.8. (continued) Monthly composits of surface chlorophyll concentration derived from daily -2km resolution SeaWiFS (NASA) ocean colour products for 2001, 2003 and 2005 (EC-Joint Research Centre, Global Environment Monitoring Unit Ocean Colour Archive, http://oceancolour.jrc.ec.europa.eu/).

2.5. Surface and near-bottom oxygen concentrations

Northwestern shelf: Following the first sight of hypoxia (oxygen concentrations below 2 mg/l, or oxygen concentration less than 30% saturation value) in the coastal zone between Dnestr and Danube delta in August 1973, zones of seasonally low oxygen have been detected in the bottom waters of the northwestern shelf, most commonly within 30-40 mile zone along its north and west coasts; i.e. in the discharge regions of the northwestern rivers. Hypoxia development typically started in June-July, stretched towards offshore up to the depths of 20-30 m and attained its maximum coverage in August. In October, it largely retreated except in close vicinity of the Danube discharge region and near Odessa. According to the long-term observations (Loyeva et al., 2006), the largest hypoxia coverage (about one-third of the NWS area or more) was observed in the years 1978, 1983, 1989, 1990, 1999 and 2000, among which the 1983 event covered 52% of the NWS total area (Fig. 2.5.1). A similar analysis by Zaitsev (1992) reported almost twice larger spatial coverage for the 1970s and 1980s. It was reported to decrease in Romanian coastal waters and no hypoxia case was observed in Bulgarian waters in 2001-2005, where lower oxygen concentrations tended to occur more predominantly in the northern sector (Parr et al., 2005).

Berlinskii (1989; 2003) proposed a linear empirical relationship between the rate and timing of river discharge and the scale of hypoxic conditions. In years with high river discharge and hence high nutrient input and organic matter production, the oxygen concentration in bottom layers becomes ~20% lower in comparison with years of low river discharge. Frequency of hypoxic conditions in front of the Danube river mouth was found to decrease when roughly half of the fresh water discharge took place before the first half of April whereas the maximum river input in April-May favoured oxygen depletion later in June-October. Loyeva et al. (2006) suggested a close correlation (r=0,86) between spatial coverage of brackish water with salinity less than 17.5 psu and the area of near bottom hypoxia. The data shown in Fig. 2.5.1, however, suggest no clear correlation of the extent of hypoxia with neither Danube fresh water discharge nor NO3 load, but it seems to correlate with the SeaWiFS chlorophyll concentration during 1998-2000 for the northern part of the NWS. In fact, hypoxia development depends not only the intensity of eutrophication but also circulation, stratification and meteorological (wind, heat flux, etc.) conditions and therefore may be subject to large year-to-year variations and a simple correlation to environmental factors may not always hold. Its sensitivity to different number of hydro-meteorological and biogeochemical factors complicates estimation of its spatial coverage by proxy data and therefore limits the use of such indirect approach as a reliable indicator of eutrophication. Fig. 2.5.1 does not indeed show stable tendency of increase or decrease in hypoxia coverage in the northwestern Black Sea.

Vertical distribution of oxygen saturation ratio along the Sulina transect exhibited oxygen depletion in summers 2001 and 2002 following relatively warm winters whereas bottom waters in the summer 2003 after the cold winter did not show any significant depletion (Fig. 2.5.2).

Long-term surface oxygen measurements in Constanta (Fig. 2.5.3) suggest persistently low concentrations (between 200-250 ?M) from the late-May to the end of September during the 1980s and the 1990s. The recovery in oxygen status in surface waters is evident after 2000 which are further supported by benthic data (e.g. mussel bed and community age class distributions) in the following chapters.


Interior basin: Modification in sub-surface oxygen structure of the interior basin in the 1980s due to considerable increase in organic matter production has been monitored by long-term changes in the thickness of the suboxic layer (Orlova et al., 1999; Konovalov et al., 2001). This layer is defined in Fig. 2.5.4 by the sigma-t surfaces of 20 μM oxygen and 5 μM hydrogen sulphide concentrations and has been constructed using all available data from the interior basin since the 1960s. It is located at depths away from the direct impact of atmospheric ventilation and its temporal changes mainly reflect the changes due to biochemical processes. Its thickness had a relatively stable structure with a mean value of 20 m or equivalently Δσt~0.4 kg m-3 up to the late 1970s. It was then broadened to ~40 m (Δσt ≥ 0.55 kg m-3) during the 1980s and preserved this structure until the early-1990s due to higher rate of oxygen consumption associated with intensified biological production. This period was followed by a decline of the thickness to Δσt~0.35 kg m-3 during the mid-190s after which the data from 2001 and 2003 indicated further broadening due to intense phytoplankton bloom events.

Fig. 2.5.4 further includes the long-term oxygen concentration variation of the layer between σt ~14.45 and 14.6 kg m-3 surfaces (roughly the base of euphotic zone) along the eastern coast formed by annual averages of all deep stations irrespective of their seasons and locations. In consistent with the period of maximum SOL thickness, oxygen concentration of this layer attained low values below 200 μM and then was followed by an increasing trend up to ~300 μM during the 1990s at the time of narrower SOL thickness. The mid-1990s represented best conditions of the subsurface oxygen concentration, after which it started to deteriorate again during the period of SOL broadening. The interior basin and the eastern part of the sea presently reflect an intermediate state of the subsurface oxygen structure between worst conditions of the 1980s and the best conditions of the mid-1990s.

The surface DO concentrations in the northwestern coastal waters (Fig. 2.5.4), mostly in the vicinity of Constanta, respond less markedly to adverse effects of eutrophication because of atmospheric ventilation of surface layers. Nevertheless, slight changes are observed between relatively higher concentrations during cold periods (e.g. 1985-1993, after 2003, and prior to 1975) and warm periods (1975-1985 and 1994-2002). Instead it responds more strongly to climatic effects as indicated by its elevated values during short term severe winter episodes of 1985-1987, 1992-1993, 2003-2004.?? Fig. 2.5.4 further indicates that subsurface concentration within the eastern basin and surface concentration in the northwestern coastal waters tend to have opposite interannual variations after 1998 due to different environmental factors governing oxygen concentration. The eastern coastal waters appear to be controlled primarily by physical mechanisms that are regulated directly by climatic cooling-warming cycles, whereas northwestern waters by the changes in Danube organic and inorganic nitrogen loading, in addition to the direct climatic forcing.

2.6. Conclusions and key assessments

Nitrogen and phosphorus emissions continue to reduce, but their 2000-2005 average values are still 1.5 times higher than their pristine levels at 1955-1965. P-PO4 flux supplied by the River Danube declined considerably at 1990-1993, after which it maintained a steady level between 10-20 kt y-1 at the Sulina discharge point to the sea. The River Danube DIN flux at the Sulina discharge point decreased gradually during the 1990s to about 100 kt y-1 at 2000, but possessed a weak increasing trend to 250 kt y-1 during 2005. The DON flux supplied by the northwestern rivers (Danube, Dniepr, Dniestr, and Bug) is currently even higher than the DIN load. Consequently, an increasing trend is observed in DIN concentration along the Ukrainian, Romanian and Bulgarian coastal waters in the present decade. The western coastal waters are presently subject to both dissolved inorganic and organic nitrogen enrichment, and they do not present an apparent status of improvement in terms of nitrogen during the last 10 years. DIN and PO4 concentration levels along other coasts (southern, northeastern and northern) are, on the other hand, about 3-4 folds lower than the western coast. In spite of ongoing nitrogen enrichment, the western coastal waters tend to be weakly nitrogen limited as for the case of the interior basin, except the predominantly P-limited shallow inner shelf zone. The data suggest a weak Si limitation in western coastal waters as well. This is partially due to decreasing SiO4 concentration and partially due to increasing DIN concentration. Limited measurements at Sevastopol-Crimea quantitatively supported the importance of atmospheric deposition of nitrogen species.

Monitoring winter nutrient concentrations as an indicator of eutrophication set by EEA does not suit well for the Black Sea western coastal waters, because they possess peaks at April-May at the time of highest Danube discharge. More frequent (e.g. monthly) monitoring at less number of stations around the basin can be considered as an option of monitoring.? Atmospheric deposition need to be monitored regularly around the basin in order to arrive at a more reliable nutrient budget of the sea. Moreover, nitrogen losses due to sedimentation and denitrification along the western coastal waters remain to be quantified in order to assess the fate of nitrogen enrichment of the system.

Surface chlorophyll concentration varies five-folds from its highest annual-mean values (~5 mg m-3) in the northwestern region to lowest values (~1 mg m-3) within the interior basin. The annual structure in the northwestern shelf is characterized by relatively high concentrations from early spring to late autumn. Lowest concentrations are observed in winter months. Romanian, Bulgarian and southwestern coastal waters generally reveal three peaks over the year; the first one comprises a weak late winter-early spring (February-March) algae bloom period in relation to typical spring bloom dynamics. This is the weakest one and does not exist for some years. It is followed by a stronger late-spring peak (in May-June) that evidently coincides with the period of high Danube discharge. The third one occurs in autumn and is separated from the former one by a period of relatively low summer concentrations. Interior basin depicts a different annual structure. Surface chlorophyll concentration decreases linearly from a peak in November to a minimum in July, followed by a sharp increase from August to November again. The autumn peak is the strongest one with an occasional weak chlorophyll peak in spring.? Monitoring surface chlorophyll concentration only in summer months, as set by EEA monitoring strategy, therefore may not be well-suited for the Black Sea, and may even lead to a wrong assessment. While summer months possess lowest chlorophyll concentration at the surface, they tend to show high values below the seasonal thermocline (i.e. the subsurface chlorophyll maximum layer), which therefore need to be monitored systematically. Comparison of satellite ocean colour and in situ chlorophyll concentrations is encouraging, and indicate that satellite data provide a potential tool for chlorophyll monitoring.??

According to recent studies, hypoxia decreased in Romanian coastal waters as compared to the previous decade, and no hypoxia case was observed in Bulgarian waters in 2001-2005. Relatively low oxygen concentrations tend to occur more predominantly in the northern sector upstream of the Danube delta region. Long-term data at Constanta monitoring station suggest surface oxygen concentration in 200-250 μM range during May-September period, including the intense eutrophication period of the 1980s.? This information therefore is not particularly useful for assessment purposes. It is indeed essential to monitor the near-bottom summer oxygen concentration in addition to its surface value. On the other hand, the annual-mean surface oxygen concentration is a good indicator for monitoring long-term climate-induced modulations of eutrophication phenomenon in the western coastal waters. Subsurface oxygen concentration (below the euthopic zone) data from the interior basin and the eastern part of the sea presently reflect an intermediate state between worst conditions of the 1980s and the best conditions of the mid-1990s.

References

Bat, L., Şahin, F., Satılmış, H.H., ?st?n, F., Birinci ?zdemir, Z., Kıdeys A.E., ve Shulman, G.E. (2007). Karadeniz'in değişen ekosistemi ve hamsi balık?ılığına etkisi. Journal of Fisheries Sciences, 1 (4), 191-227.

Berlinskii, N. (1989) Mechanisms of bottom hypoxy formation in shelf ecosystems. Water Resources, 4, 112-121.

Berlinskii, N, Garkavaia, G., Bogatova, Y. (2003) Problems of anthropogenic eutrophication and evolution of hypoxic situations in the North-Western part of the Balcak Sea. Ecology of Sea, 17-22.

Chaykina, A.V., V.V., Dolotov, Y.P., Ilyin, S.K., Konovalov, A.S., Kuznetsov, L.N., Repetin, O.V. Voitsekhovich (2005) Observational studies of nutrient loads on the Black Sea with atmospheric precipitations. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Churilova T., G. Berseneva, L.Georgieva. (2004) Variability in bio-optical characteristics of phytoplankton in the Black Sea. Oceanology (English translation), 44(2), 192-204.

Churilova, T., Z. Finenko, S. Tugrul, (2006)? Light absorption and quantum yield of? photosynthesis during autumn phytoplankton? bloom? in the western? Black Sea. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Cociasu, A., L. Popa, S. Dineva, P. Ivanova, V. Velikova (2004) Summary report on field and laboratory work in 2001-2003 in comparison with previous observations in the Western Black Sea. Nutrient management in the Danube Basin and its impact on the Black Sea. daNUbs Projects (EVK1-CT-2000-00051), Deliverable 7.6, Nutrient Management in the Danube basin and its impact on the Black Sea. EVK1-CT-2000-00051.

daNUbs Project Final Report (2005) Nutrient management in the Danube basin and its impact on the Black Sea. Final Report, December 2004.

Feyzioglu, M., A. Guneroglu, I. Yildiz (2006) Weekly changes of? Synechococcus spp. and pigment concentration in the Southeastern Black Sea coast. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Hibaum, G, V. Karamfilov (2005) Regime shifts in the annual dynamics of primary production and chlorophyll_a concentrations in the costal zone of the Bourgas Bay (Western Black Sea). Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Horstmann, U., Davidov, A. (2002) IFM Annual Report. daNUbs project.

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.

Konovalov, S., J.W. Murray, and G.W. Luther III (205) Basic Processes of Black Sea Biogeochemistry. Oceanography, 18, 25-35.

Kucheruk, N. (2005) Integrated study of eutrophication and biodiversity in the Russian coastal waters. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Loyeva, I., I. Orlova, N. Pavlenko, Yu. Popov, V. Ukrainskiy, Yu. Denga, V. Komorin,? V. Lepeskin (2006) Estimation of the ecological state of the north-western part of the Black Sea. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

Medinets, S. and Medinets, V. (2008) Results of investigations of atmospheric pollutants fluxes to the surface of Zmeiny Island in western part of the Black Sea in 2003-2007. In: 2nd Biannual and Black Sea Scene EC project joint conference on climate change in the Black Sea ? hypothesis, observations, trends, scenarios and mitigation strategy for the ecosystem. 6-9 October 2008, Sofia, Bulgaria.

Moncheva, S. (2005) Phytoplankton shifts in the Black Sea ? Driving forces and implication for reference conditions. Proceedings of the JRC Workshop, 26-28 October, 2005-Istanbul

Okus, E.? (2005)? Time series analysis of nutrients in southwestern Black Sea and the Sea of Marmara. Proceedings of the JRC Workshop, 26-28 October, 2005-Istanbul

Oguz, T. (2005) Long-Term Impacts of Anthropogenic Forcing on the Black Sea Ecosystem. Oceanography, 18, 112-121.

Oguz, T. and D. Ediger (2007) Comparision of in situ and satellite-derived chlorophyll pigment concentrations, and impact of phytoplankton bloom on the suboxic layer structure in the western Black Sea during May?June 2001. Deep Sea Res. II, 53, 1923-1933.

Parr et al. (2006) Impact of the Danube River on the Black Sea. Estimation of the ecological state of the north-western part of the Black Sea. Proceedings of The First Biannual Scientific Conference: Black Sea Ecosystem 2005 and Beyond, organized by Commission on the Protection of the Black Sea Against Pollution, May 2006, Istanbul.

PMA AG and AC Activities and Reporting (2007) Final Report on Policy Measures, Development and Data Analysis for 2006/2007. Black Sea Commission, 20 November, 2007 San Diego-Mcglone et al. (2000)

State of the Environment of the Black Sea, 2002. Pressures and trends: 1966-2000. Commission on the Protection of the Black Sea Against Pollution. Istanbul. 101 pp.

Teodoru, C. R., G. Friedl, J. Friedrich, U. Roehl, M. Sturm, B. Wehrli (2007) Spatial distribution and recent changes in carbon, nitrogen and phosphorus accumulation in sediments of the Black Sea. Marine Chemistry, 105, 52?69. TDA (2007) The Black Sea Transboundary Diagnostic Analysis.

Yakushev, E.V., O.I. Podymov, and V.K. Chasovnikov (2005) Seasonal Changes in the Hydrochemical Structure of the Black Sea Redox Zone, Oceanography, 18, 49-55.

Yunev, O. A., V. I. Vedernikov, O. Basturk, A. Yilmaz, A. E. Kideys, S. Moncheva, S. K. Konovalov (2002) Long-term variations of surface chlorophyll a and primary production in the open Black Sea. MEPS, 230, 11?28.

Velikova, V. and Cociasu, A. (2004) Summary report on field and laboratory work in 2001-2003 in comparison with previous observations in the Western Black Sea Part. III. Comparison between Roumanian and Bulgarian waters. Nutrient management in the Danube Basin and its impact on the Black Sea. daNUbs Projects (EVK1-CT-2000-00051), Deliverable 7.6, Nutrient Management in the Danube basin and its impact on the Black Sea. EVK1-CT-2000-00051.

Zaitsev Yu.P. (1992) Recent changes in the trophic structure of the Black Sea. Fisheries Oceanography, 1, 180-189.