A Long-Term Perspective of Seasonal Shifts in Nutrient Dynamics and Eutrophication in the Romanian Black Sea Coast

Abstract

This study investigates the long-term seasonal shifts in nutrient dynamics and eutrophication processes in the Romanian Black Sea coastal waters using multi-decadal data (1960/1976/1980–2023). The findings highlight significant seasonal and interannual changes, revealing a progressive rise in seawater temperature, declining oxygen concentrations, and notable shifts in nutrient stoichiometry, particularly an increasing nitrogen–phosphorus (N:P) ratio. These changes are closely associated with increased occurrences of harmful algal blooms (Noctiluca scintillans), emphasizing the complex relationship between warming, nutrient and dissolved oxygen cycles, and biological activity. Seasonal patterns show that prolonged warmer periods, especially during autumn, exacerbate oxygen depletion and nutrient imbalances, with implications for marine life and food webs. The study underscores the importance of targeted nitrogen reduction strategies, including optimized fertilizer use, improved wastewater treatment, and the establishment of buffer zones to minimize land-based nutrient inputs. Regional cooperation and integrated coastal management aligned with the Marine Strategy Framework Directive are essential for mitigating eutrophication. The results provide critical insights into the impacts of climate change on the Black Sea ecosystems. This research contributes to global efforts under SDG 13 (Climate Action), SDG 14 (Life Below Water), and SDG 3 (Good Health and Well-being), addressing the key challenges to marine biodiversity, water quality, and ecosystem sustainability.

1. Introduction

Climate change is one of the most pressing environmental challenges of our time, reshaping ecosystems across the globe through rising temperatures, altered precipitation patterns, and increasing atmospheric CO₂ levels [1]. Among the most visible impacts in marine environments is the tendency of the rise in sea surface temperatures in recent decades, which significantly affects the physical, chemical, and biological processes of aquatic systems [2,3,4,5].

In semi-enclosed seas like the Black Sea, where water exchange with other seas is limited [6,7], the effects of warming are particularly pronounced, driving profound changes in oxygen dynamics, nutrient cycling, and ecosystem health [8]. Over the past several decades, the Black Sea region has experienced consistent warming of its air and seawater temperatures, a trend attributed to regional and global climate dynamics [9]. This warming intensifies the stratification of water columns, reducing vertical mixing and limiting the oxygenation of seawater. These combined factors increase the vulnerability of the Romanian Black Sea Coast (RBSC) to hypoxia, creating a feedback loop that fuels eutrophication [10]. While rising temperatures are a central driver of these changes, they interact synergistically with additional stressors, particularly freshwater and nutrient inputs. The Danube River, Europe’s second-largest river, significantly contributes to the RBSC’s nutrient loads, delivering large quantities of nitrogen, phosphorus, and organic matter into the coastal waters [11,12]. Climate change exacerbates this pressure by altering the river’s flow patterns; periods of heavy rainfall and snowmelt increase freshwater discharge, while prolonged droughts reduce flow, in both cases the salinity and nutrient concentrations being influenced. These fluctuations fuel eutrophication intensification and threaten marine biodiversity [13,14]. To these, additional coastal source discharges significantly contribute to the nutrient input [15].

The RBSC is an important case study for understanding the intersection of climate change and anthropogenic pressures on marine systems. With its history of nutrient enrichment driven by agriculture, industrial activity, and urban runoff [16,17], the region provides valuable insights into how changing temperatures and hydrological variability interact to alter marine ecosystems’ time scales. The temperature’s longer time scale establishes the environment for the shorter time scale processes, like seasonal variations in nutrients and dissolved oxygen concentrations, or diurnal cycles of the latter. Previous efforts to mitigate nutrient pollution, such as improved agricultural practices and wastewater treatment, have had some success in curbing eutrophication. The Black Sea has been a focal point for studying the combined effects of anthropogenic nutrient pollution and climate change. Mee (2006) [18] provided a comprehensive review of eutrophication in the Black Sea, highlighting how the collapse of the Soviet Union in the early 1990s led to a temporary reduction in nutrient inputs due to decreased agricultural activity. However, recent research indicates that climate change is reversing these gains, with seasonally rising freshwater input from the Danube and increased nutrient runoff from agricultural activities in the basin exacerbating eutrophication. Studies by Capet et al. (2013) [19] have reinforced this by showing how changes in the Black Sea’s circulation patterns, driven by climate variability, are influencing nutrient distribution and intensifying oxygen depletion [15,20].

The current state of research on marine eutrophication and its connection to climate change has expanded significantly in recent decades, with growing recognition of how temperature cycles, altered precipitation patterns, and increased nutrient inputs are reshaping marine ecosystems. Key studies have advanced our understanding of how these processes interact to exacerbate eutrophication, particularly in semi-enclosed seas like the Baltic, North, Adriatic, and Black Seas [2,15,19,21,22,23,24,25]. While significant research has been conducted on eutrophication in semi-enclosed seas like the Baltic [26], North [27,28,29], and Adriatic Seas [30], relatively few studies have integrated long-term, multi-decadal datasets to explore the combined effects of climate change and nutrient dynamics in the Black Sea. Understanding these interactions is essential for developing targeted, science-based strategies to mitigate the impacts of eutrophication and build resilience in marine ecosystems.

In broader global contexts, the Intergovernmental Panel on Climate Change (IPCC) reports [31] have acknowledged that climate change is likely to exacerbate eutrophication, particularly in vulnerable coastal and semi-enclosed seas. Their findings emphasize the critical need for integrated approaches to managing nutrient pollution that account for climate-related changes in hydrological cycles, temperature, and marine circulation.

Diverging views exist within the scientific community regarding the relative importance of human-induced nutrient inputs versus climate change in driving eutrophication. Some researchers argue that human activities remain the dominant force behind nutrient over-enrichment and must be the primary focus of mitigation efforts, while others highlight the role of climate change in exacerbating these impacts, particularly through increased stratification and altered riverine flows [32,33,34]. This ongoing debate highlights the need for comprehensive studies that account for both anthropogenic and climate-driven factors.

Despite significant research on eutrophication and nutrient dynamics in the Black Sea, several critical gaps remain unaddressed, and this research tries to address them, focusing on RSBC. First, long-term analyses spanning multiple decades are scarce [35,36], limiting the ability to detect gradual but significant changes in nutrient time scale cycles and their links to climate change time scale.

This study aims to investigate the long-term seasonal shifts in nutrient dynamics and eutrophication processes along the Romanian Black Sea coast, considering the relationship between climate-driven changes and anthropogenic pressures. By analyzing long-term data series (1960–2023), the research seeks to identify temporal patterns in nutrient stoichiometry, dissolved oxygen variations, and eutrophication intensity, providing insights into the ecological consequences of shifting seasonal cycles and informing adaptive management strategies aligned with the United Nations Sustainable Development Goals (SDGs), particularly SDG 13 (Climate Action) and SDG 14 (Life Below Water).

2. Materials and Methods

2.1. Study Area and Data Acquisition

The study utilizes a long-term dataset (1960/1976/1980–2023) from the Black Sea coastal region in Mamaia Bay (Figure 1). All the parameters were measured and analyzed in the National Institute for Marine Research and Development “Grigore Antipa”, Constanta, Romania.

The key environmental parameters analyzed in this study (Table S1) include sea surface temperature (SST), salinity, and dissolved oxygen, which have been consistently measured since 1960. Additionally, nutrient concentrations—such as phosphate, silicate, nitrite, and ammonium—were incorporated into the analysis as laboratory methods for their detection were introduced in subsequent years, reflecting the evolving capabilities of environmental monitoring over time. Temperature and salinity were measured using various techniques, including a reversible thermometer, titration, and the CastAway CTD multiparameter probe (SonTek Cast-Away-CTD, San Diego, CA, USA). Dissolved oxygen was measured by the Winkler method, a titrimetric technique for measuring dissolved oxygen in water. Manganese (II) sulfate and an alkaline iodide solution react with oxygen to produce iodine, which is titrated with sodium thiosulfate [37]. The amount of thiosulfate used is proportional to the dissolved oxygen concentration, making this method accurate for environmental monitoring. Dissolved nutrient concentrations were determined following the established seawater analysis protocols [37]. Nutrients were quantified using spectrophotometric methods and validated in the laboratory according to the guidelines in the “Methods of Seawater Analysis” [37]. Specifically, nitrate concentrations were determined since 1976 using the method outlined by Mullin and Riley (1955) [38], in which nitrate was reduced to nitrite using hydrazine sulfate. The resulting nitrite then reacted with sulphanilamide in an acidic solution to form a diazonium compound, which coupled with N-(1-Naphthyl)-ethylenediamine dihydrochloride to produce a colored azo dye, measurable via spectrophotometry. Ammonia levels have been measured since 1980 using the indophenol blue method [37], where ammonia reacts in a mild alkaline solution with hypochlorite to form monochloramine. This, in turn, reacts with phenol, nitroprusside ions, and excess hypochlorite, resulting in the formation of indophenol blue. Inorganic phosphate concentrations have been determined since 1960 through their reaction with ammonium molybdate in an acidic medium to form a yellowish phosphomolybdenum complex. This complex is then reduced to a blue-colored compound, with the intensity of the color correlating to the phosphate concentration. The intensity is measured at 885 nm using a spectrophotometer [37]. Similarly, silicon has been measured since 1960. Silicon ions react with ammonium molybdate in an acidic solution to form a yellowish silicomolybdenum complex. In the presence of ascorbic acid as a reducing agent, this complex is reduced to a deep blue compound, and the intensity of the color is proportional to the silicon concentration. The measurement is performed at 810 nm using a spectrophotometer [37].

Between 1981 and 2023, data on Noctiluca scintillans were gathered using a combination of historical records and in situ field surveys from adjacent areas (Figure 1). The historical data used in this study covered the period from 1981 to 2000, during which a standardized data collection procedure was implemented to ensure accuracy, consistency, and reliability. The sampling involved deploying a Juday net in the water column, with the collected samples preserved in 4% buffered formaldehyde for subsequent microscopic identification [39]. The density and biomass of N. scintillans were determined using cell counts, with biomass calculated by applying a standard weight per individual [40]. Maintaining the same methodology across historical and recent data collection periods ensured that the datasets were comparable.

2.2. Statistical and Time-Series Analysis

2.2.1. Trend Analysis

A linear regression model and a 10-year rolling average were applied to detect long-term trends in the data. Linear regression was used to capture long-term trends in temperature and salinity. The significance of the fit was evaluated using the coefficient of determination (R²) and p-values. Trends were visualized using time-series plots.

The 10-year rolling average provides a clearer view of long-term changes by smoothing out short-term variability. This approach helps reveal persistent trends that individual years may obscure and mitigate the effects of seasonal cycles and extreme events, making the overall trend more stable and reliable for analysis.

2.2.2. Correlation Analysis

The relationships between the key environmental variables (e.g., temperature, salinity, oxygen levels, and nutrients) were explored using Pearson correlation coefficients to identify how changes in one variable (e.g., temperature) relate to others (e.g., salinity or oxygen levels). Correlation matrices were computed to evaluate the strength of these relationships and significant correlations were highlighted. Multicollinearity among the independent variables in the regression analysis was assessed using correlation matrices across the full dataset and seasonal subsets. The analysis revealed some significant correlations between variables; however, most correlation coefficients remained below the threshold of concern (r > 0.7). A strong negative correlation was observed between sea surface temperature and dissolved oxygen (r = −0.843), indicating a high degree of redundancy. To enhance the model’s reliability, dissolved oxygen was excluded from the regression analysis, as its variation was largely explained by temperature. The refined model minimized multicollinearity while preserving the explanatory power of the remaining variables.

Principal component analysis (PCA) was utilized to explore the relationships between the N. scintillans abundance and key environmental variables. Data preparation steps were undertaken to address the heterogeneity in data collection methods and temporal coverage to ensure consistency and comparability. Measurements were obtained using different instruments and analytical methods over various periods, potentially introducing variability. Data standardization was applied to minimize discrepancies by transforming variables to a common scale, ensuring equal influence on the principal component analysis (PCA). This multivariate statistical method helped to identify patterns and reduce the dimensionality of the dataset while retaining the most significant variance, enabling the identification of dominant environmental factors influencing N. scintillans dynamics.

2.2.3. Visualization and Interpretation

All the analyses were visualized using Statistica [41], while for the analysis of N. scintillans, the PRIMER v. 7.0 [42] software was employed for multivariate analysis, which provided valuable insights into how environmental factors and seasonal changes influenced the dynamics of N. scintillans, emphasized by graphical tools, such as box plots. This approach allowed for a detailed visualization of the distribution and variability of N. scintillans across different seasons, providing insights into patterns and trends in the data.

3. Results

3.1. Temperature

The analysis of seawater temperature reveals a clear upward trend, suggesting a gradual warming of the Black Sea. The mean has steadily increased over time, with the fitted linear regression model yielding a positive slope which indicates an average increase of approximately 0.037 °C annually (Figure 1). This model demonstrates a statistically significant warming trend, as indicated by the slope’s p-value (p < 0.001,

Table 1. Linear regression summary for the relationship between year and mean annual temperature—Black Sea, 1960–2023 (red-statistically significant, b*—Standardized regression coefficient).

N = 64	Regression Summary for Dependent Variable: T[oC]
	R = 0.59483613 R2 = 0.35383003 Adjusted R2 = 0.34340793 F(1,62) = 33.950 p < 0.00000 Std.Error of Estimate: 94304
	b*	Std.Err. of b*	b	Std.Err. of b	t(62)	p-Value
Intercept			−61.504	12.709	−4.839	0.000009
Year	0.595	0.102	0.037	0.006	5.827	0

). The trend line highlights the progressive increase in temperature over the decades despite some interannual variability and fluctuations (Figure 2).

The linear regression model (Figure 2, red line and Table 1) confirms the annual increase (b) with high confidence (p < 0.0001), but the rather scattered data gave a coefficient of determination (R²) of 35.4% (R²), which means that temporal variability in temperature is a much more complex phenomenon. Moreover, the analysis of variance (ANOVA) F(1,62) = 33.950, p < 0.00000, shows that the model is highly significant, meaning the temperature changes in time are important.

The 10-year rolling average suggests a more pronounced warming trend around the 1990s. This period marks a shift from relatively stable or fluctuating temperatures to an accelerated warming trend in recent decades, a pattern emphasized by the polynomial fit marking the starting point for significant rising seawater temperatures (Figure 3).

The seasonal temperature trends for winter, spring, summer, and autumn with each season’s 10-year rolling average illustrate long-term changes, capturing the upward trends in seasonal temperatures. The polynomial trends’ R² values indicate a steady upward trend in summer and autumn temperatures over time, highlighting accelerated warming in these seasons. Winter and spring also show warming trends, though they are less pronounced indicating that temperature increases are not uniform across seasons, with a more pronounced rise during the warmer months of the year, which may have significant ecological impacts on seasonal biological cycles and regional climate patterns (Figure 4).

3.2. Salinity

Salinity showed high variability over time, with only a slight, statistically insignificant downward trend (F(1,62) = 0.28155, p < 0.59758). This suggests that salinity in this dataset does not respond to long-term, yearly changes, its time scale being more likely influenced by short-term environmental or hydrological factors (freshwater inflow from the Danube, precipitation, evaporation, and currents) (Figure 5).

Salinity varies seasonally, with the lowest levels occurring during spring and summer, reaching a minimum in summer, increasing during autumn, and peaking in winter. This pattern corresponds to the relationship between temperature and salinity, as salinity, given by the concentration of the dissolved species, tends to be higher during colder seasons since the saturation concentration increases with temperature decrease. The reduction in salinity during warmer months could be attributed not only to the decrease in the saturation concentration, but also to the increased freshwater input from rivers, as well, while during colder months, lower riverine influence and upwelling events allow salinity to rise (Figure 6).

The regression equation (Figure 7) illustrates that for every 1 °C increase in temperature during summer, salinity decreases by approximately 0.24‰. The coefficient of determination (R² = 0.1556) shows that around 15.6% of the variation in salinity can be explained by the changes in temperature. This suggests that other factors besides temperature are also influencing salinity. The negative correlation could reflect various oceanographic processes beyond just freshwater input. One possible explanation is upwelling, during which colder, highly saline, low-oxygen, and nutrient-rich waters from deeper layers rise to the surface, resulting in lower surface temperatures and increased salinity. Alternatively, evaporation, which typically increases salinity, could also play a role.

3.3. Dissolved Oxygen

The dissolved oxygen concentrations demonstrate a consistent downward trend of annual means, from above 11 mg/L in the early 1960s to less than 9 mg/L in recent years and an average decrease in oxygen concentration of about 0.024 mg/L per year (Figure 8).

The regression summary (

Table 2. Linear regression summary for the relationship between year and mean annual dissolved oxygen—RBSC, 1960–2023 (red-statistically significant, b*—Standardized regression coefficient).

N = 64	Regression Summary for Dependent Variable: O2 [mg/L] R= 0.58251671 R2= 0.33932572 Adjusted R2= 0.32866968 F(1,62) = 31.844 p < 0.00000 Std.Error of Estimate: 0.63715
	b*	Std.Err. of b*	b	Std.Err. of b	t(62)	p-Value
Intercept			58.752	8.586	6.842	0
Year	−0.583	0.103	−0.024	0.004	−5.643	0

) for dissolved oxygen concentration (mg/L) provides a clear picture of the trend in oxygen levels over time, indicating that approximately 33.9% (R²) of the variability in oxygen concentration can be explained by the passage of time. The Adjusted R² value of 0.329 further supports the model’s robustness, indicating that this trend remains statistically significant even after adjusting for potential overfitting. The F-statistic (F(1,62) = 31.844, p < 0.00000) confirms the statistical significance of this trend underscoring that the decline in oxygen levels reflects a consistent pattern over time.

Oxygen levels show a strong negative correlation with seawater temperatures explaining 30.16% of the observed decline pattern (Figure 9).

The seasonal relationships between temperature and dissolved oxygen in the RBSC show a consistently significant negative correlation across all the seasons (p < 0.0001). Winter shows a moderate correlation (−0.64), where lower temperatures generally imply greater oxygen availability. The negative correlation strengthens in spring (−0.67) and autumn (−0.70), reflecting the increased biological activity and prolonged periods of reduced oxygen solubility, especially during longer, warmer autumns. In summer, stratification and high temperatures also lead to a significant dissolved oxygen concentration decrease (−0.54). These patterns suggest that not only the rising temperatures drive lower dissolved oxygen levels, but also the biological activity, as well, with implications for the marine ecosystems, particularly during warmer and extended autumns (Figure 10).

3.4. Nutrients

3.4.1. Long-Term Variability

Phosphate concentrations have shown a steady decline, with a pronounced peak during the 1970s and 1980s associated with intensive agricultural practices and untreated wastewater discharges [15,16]. The annual decrease of ~0.05 µM post-1990 reflects the impact of reduced phosphorus inputs through the banning of phosphate detergents and improved wastewater treatment [20,43] (Figure 11).

Silicate concentrations exhibit the steepest decline among nutrients, with an average reduction of ~0.27 µM annually. This decline can be attributed to reduced riverine input due to damming (e.g., Porțile de Fier [43]). Unlike phosphate, silicate reduction is more closely tied to physical changes in river discharge rather than direct anthropogenic control measures.

The inorganic nitrogen species have gradually decreased with different levels showing annual reductions of ~0.11 µM (NO₃), 0.02 µM (NO₂), and 0.04 µM (NH₄). Generally, high concentrations in the late 1970s and early 1980s were linked to excessive nitrogen fertilizer use and poor wastewater management. Improvements in agricultural practices and wastewater treatment from the 1990s onwards led to lowering the inputs, although sporadic peaks suggest periodic runoff events or localized discharges. In the case of ammonium, periodic increases suggest that ammonium remains sensitive to episodic pollution and riverine inputs.

The N:P ratio has shown a statistically significant upward trend, increasing by ~1.06 annually. This rise is primarily driven by the greater reduction in phosphorus relative to nitrogen, highlighting an imbalance in nutrient management efforts. Elevated N:P ratios suggest a shift in nutrient limitation dynamics, favoring nitrogen-tolerant phytoplankton such as cyanobacteria [44,45]. The increasing N:P ratio poses the risks of altered ecosystem functioning, reduced biodiversity, and enhanced harmful algal blooms (HABs) [46,47].

3.4.2. Seasonal Variation

The seasonal trends of nutrients in relation to temperature (T) in the Black Sea reveal distinct patterns shaped by physical, biological, and chemical processes (Figure 12). The analysis highlights the seasonal coupling of nutrient dynamics and temperature, influencing nutrient availability, primary productivity, and overall ecosystem functioning.

Phosphate concentrations exhibit a pronounced seasonal decline from winter to summer, corresponding to rising temperatures and suggesting a strong seasonal coupling driven by biological uptake during warmer months when phytoplankton growth peaks. In winter, the lower temperatures and reduced biological activity result in elevated phosphate levels due to limited uptake and replenishment through vertical mixing. During spring, moderate temperatures coincide with increasing biological activity, initiating phosphate depletion. In autumn, cooling temperatures and resumed mixing replenish phosphate concentrations.

As with any dissolved species, silicate concentration displays an inverse relationship with temperature, with the lowest values in summer and autumn due to the decrease in its saturation concentration. Owing to this, together with the biological activity, the increased riverine input, and the mixing between the latter and the seawater during colder months, the silicate concentration in surface waters can fluctuate. As temperatures rise in spring, lowering its saturation concentration, silicate is rapidly consumed by diatom blooms, leading to significant depletion by summer. The depletion is exacerbated by stratification, which limits nutrient replenishment. In autumn, the silicate levels remain at lower levels due to the ongoing biological uptake.

Nitrate concentrations follow a similar seasonal trend to phosphate, declining as temperatures increase. Elevated nitrate levels in winter result from reduced biological uptake and enhanced vertical mixing. During spring, nitrate concentrations are at their highest due to the riverine input. By summer and autumn, stratification and continued biological uptake deplete nitrate levels.

Nitrite concentrations remain relatively low throughout the year but show a slight seasonal decline with rising temperatures. This is attributed to the rapid conversion of nitrite into other nitrogen species during active biological cycling in warmer months. Winter concentrations are higher due to reduced microbial activity and slower cycling rates. Spring and summer show consistent declines as biological processes intensify.

Ammonium concentrations exhibit a gradual seasonal increase from spring to autumn, peaking in late summer. This trend is linked to intensified microbial activity and organic matter decomposition under warmer conditions. While the biological uptake of ammonium for primary production increases during spring, in summer decomposition outpaces consumption, leading to higher concentrations. In winter, ammonium levels start to decline as temperatures decrease and microbial activity slows.

The N:P ratio shows less significant seasonal variation, indicating a decoupling of nitrogen and phosphorus dynamics in relation to temperature. The highest values and variability with extremely high potential for eutrophication occur during spring and summer, likely tied to nutrient inputs and biological activity. However, the seasonal fluctuations indicate increased nitrogen availability relative to phosphorus during spring and summer, potentially favoring nitrogen-tolerant phytoplankton species.

3.5. Harmful Algal Blooms (Noctiluca scintillans)

The proliferation of N. scintillans in the RBSC, contributing to harmful algal blooms (HABs), has been increasingly documented since the 1980s [35], coinciding with the region’s growing eutrophication due to nutrient runoff from agriculture, industry, and urban areas. Significant blooms were noted during the 1990s as nutrient levels peaked, leading to widespread ecological issues such as hypoxia and fish die-offs. In the 2000s and beyond, the frequency and intensity of N. scintillans blooms have continued to rise, driven by seasonal nutrient enrichment, water stratification, and potentially climate-related factors, making these blooms a critical indicator of the ongoing environmental challenges in the Black Sea [48,49]. Additionally, N. scintillans blooms serve as an indicator of eutrophication, a condition often assessed under Descriptor 5—Eutrophication, Criterion 3 (D5C3) of the Marine Strategy Framework Directive (MSFD) [50].

During the summer months, there was a noticeable spike in both density and biomass, as evidenced by the significantly higher medians and the presence of outliers indicating extreme bloom events (Figure 13). The broader range seen in summer suggests considerable variability, likely influenced by environmental factors such as temperature, nutrient availability, and light intensity, all of which are optimal during this period for N. scintillans growth. In contrast, the autumn, winter, and spring seasons displayed much lower and relatively stable values for both abundance and biomass, with minimal variation.

Among the predictors, the N:P ratio stands out as highly significant, with a standardized coefficient (b*) of 0.78 (

Table 3. Regression summary for the relationship between environmental parameters, nutrients, and N. scintillans abundance—RBSC, 1981–2023 (red-statistically significant, b*—Standardized regression coefficient).

N = 29	Regression Summary for Dependent Variable: N. scintillans [ind/m3] R = 0.54865601 R2 = 0.30102342 Adjusted R2 = 0.02143279 F(8,20) = 1.0767
	b*	Std.Err. of b*	b	Std.Err. of b	t(20)	p-Value
Intercept			99,259.44	104,695	0.94808	0.354403
T [°C]	0.482	0.469	1862.24	1813	1.027	0.31661
S [‰]	−0.401	0.244	−9065.06	5513.4	−1.644	0.115765
PO4 [µM]	−0.133	0.296	−2817.5	6266.7	−0.45	0.657837
SiO4 [µM]	0.598	0.474	2163.67	1714.7	1.262	0.221528
NO3 [µM]	−0.796	0.437	−3868.4	2121.6	−1.823	0.083233
NO2 [µM]	0.269	0.312	22,904.15	26,642.3	0.86	0.400148
N/P	0.78	0.341	394.42	172.4	2.288	0.033147
NH4 [µM]	0.004	0.22	30.02	2100.9	0.014	0.988739

). This indicates that changes in the N:P ratio may have a substantial impact on the abundance of N. scintillans, highlighting its ecological importance in the studied system.

The principal component analysis (PCA) biplot illustrates the principal components explaining the ecosystem variability (Figure 14). The high loadings of both N/P and N. scintillans density (

Table 4. PCA factor-variable correlations (factor loadings), Black Sea, 1981–2023.

Variable	Factor-Variable Correlations (Factor Loadings), Based on Correlations Active and Supplementary Variables *Supplementary Variable
	Factor 1	Factor 2
N/P	0.778283	0.627914
N. scintillans [ind/m3]	0.778283	−0.627914
*T [°C]	0.124689	0.010839
*S [‰]	−0.18465	−0.115939
*PO4 [µM]	−0.408249	0.246615
*NO3 [µM]	0.435026	−0.305593
*NO2 [µM]	−0.351395	0.254426
*NH4 [µM]	−0.181113	0.124353

) establish these as the dominant variables driving the principal components in the PCA. The PCA revealed that Factor 1 and Factor 2 together accounted for 100% of the total variance, with Factor 1 explaining 39.43% and Factor 2 contributing 60.57%. Factor 1 likely represents the nutrient enrichment gradient, where nitrate and N/P ratios significantly contribute to Noctiluca proliferation. Factor 2 may be associated with hydrodynamic or seasonal effects, influencing nutrient distributions and salinity levels, with a lesser influence from temperature. N. scintillans thrive in conditions characterized by high nitrate and N/P ratios, whereas elevated phosphate and ammonium levels tend to show an inverse relationship.

To analyze the temporal changes in the eutrophication processes and nutrient dynamics, particularly in response to the increasing sea surface temperature observed after the 1990s, the dataset was divided into two periods: before 1995 and after 1995. A principal component analysis (PCA) was performed separately for each period to identify the key drivers of eutrophication and assess shifts in their influence over time (Tables S7 and S8, Figures S7 and S8). Before 1995, the eutrophication processes in the Romanian Black Sea coastal waters were primarily driven by nutrient dynamics, with Factor 1 explaining 76.68% of the total variance. During this period, N. scintillans abundance and the nitrogen–phosphorus (N/P) ratio were strongly correlated, indicating that nutrient imbalances, particularly nitrate levels, played a dominant role in bloom occurrences. Temperature had a moderate influence, while salinity and phosphate showed negative correlations, suggesting their limiting effects on algal growth. However, after 1995, the system’s complexity increased, as indicated by the reduced variance explained by Factor 1 (58.60%) and the increased contribution of Factor 2 (41.40%), reflecting the growing influence of hydrographic and biological interactions. The projections suggest that the role of nitrate in driving blooms has weakened, while salinity, phosphate, and ammonium have become more influential, pointing to evolving nutrient dynamics and possible shifts in limitation patterns.

4. Discussion

The results point to a trend in the temperature increase in the region, which could have implications for the ecological processes and marine life in the area. This progressive rise highlights the impact of climatic shifts on regional seas. The significant trend identified through regression analysis suggests an intensifying influence of global climate drivers on local marine systems. Notably, the warming is more pronounced from the 1990s onward, marking a transition to accelerated temperature increases. Seasonally (Tables S2–S6), warming is most significant during summer and autumn, which could intensify stratification, reduce vertical mixing, and disrupt thermal conditions essential for marine ecosystems.

The main consequence is the dissolved oxygen decline. Although the decline in oxygen levels is small on an annual basis, it accumulates to a substantial reduction over the decades, posing a significant future threat to marine ecosystems, with implications for ecological health and biodiversity. Seasonal analysis reveals a strong negative correlation between the temperature and oxygen levels, particularly during spring and autumn when biological activity and thermal stratification intensify oxygen depletion. Moreover, as winters become warmer and shorter due to climate change [2,13,14], less oxygen is dissolved. The effects are even more pronounced during autumn. Climate change has extended and warmed the autumn season, thereby prolonging the periods of low oxygen solubility and delaying the breakdown of stratification. This may lead to extended hypoxic conditions. The extended warming of autumn and winter is particularly disturbing, as these conditions lead to reduced oxygen availability, further stressing marine life, especially in regions already prone to hypoxia [22].

Salinity reveals high inter-annual variability, with no significant long-term trend, indicating that changes in salinity are episodic rather than gradual. The observed seasonal cycle, characterized by lower salinity in summer and higher values in winter, is associated with freshwater discharge patterns and seasonal hydrological processes. The weak relationship between salinity and temperature in summer further points to the complex interaction of factors, including riverine inputs, evaporation, and upwelling. The results suggest that salinity is influenced by short-term environmental conditions rather than sustained climatic trends. During summer, salinity negatively correlates with phosphate (−0.306) and silicate (−0.306), reflecting enhanced evaporation processes and biological uptake. Conversely, in winter, salinity exhibits a stronger negative relationship with dissolved oxygen (−0.456), indicating that colder, saline waters may originate from deeper, oxygen-poor layers through upwelling. Thus, salinity indirectly reflects the effects of climate change through mechanisms such as changes in precipitation patterns, riverine inputs, or evaporation rates [51,52] and upwelling; although it does not exhibit a direct and consistent response over time, it shows a consistent seasonal pattern.

In addition, seasonal nutrient dynamics (Figures S1–S6 and Tables S1–S5) experience significant climate change-driven changes. During winter, low temperatures and reduced biological activity limit nutrient uptake, allowing silicate (SiO₄) to replenish (−0.427). However, low salinities reflecting precipitations and riverine input correspond to higher silicate (−0.420), nitrate (−0.449), and ammonium (−0.396) levels. The main contribution of nitrogen in N/P is from nitrate (0.561).

In spring, increased temperatures and light availability drive strong phytoplankton blooms that use PO₄ and SiO₄, reflected in a more pronounced negative correlation between temperature and PO₄ (−0.310) and SiO₄ (−0.710). Although oxygen values decrease (−0.655) due to higher temperatures, silicate contributes to oxygen production (0.514) by diatoms blooms. The main contributor to N/P is nitrite (0.97). Summer represents the peak of biological productivity, driven by high temperatures and extended daylight. Ammonium (NH₄⁺) serves as a key intermediate in nitrogen cycling, produced during organic matter decomposition by microbial activity. Its strong correlation with phosphate (0.527) and silicate (0.526) suggests that enhanced remineralization processes in warmer summer conditions contribute to the simultaneous regeneration of these nutrients. This linkage is typical in systems where biological productivity and the subsequent organic matter degradation dominate nutrient availability. The strong relationship between phosphate and nitrite (0.762) reflects the active role of nitrogen transformations, particularly nitrification and denitrification. Nitrite acts as a transient species in the nitrogen cycle, and its association with phosphate indicates that nutrient recycling is tightly coupled during summer when microbial processes are accelerated by elevated temperatures. The correlations between nitrogen species, such as the positive link between nitrite and ammonium (r = 0.586), indicate active nitrogen cycling under hypoxic or low-oxygen conditions. The main contributor to N/P is ammonium (0.325). By autumn, reduced biological activity and partial mixing lead to the remineralization of organic matter, releasing nutrients back into the water column. The strong correlation between temperature and oxygen (−0.727) likely influences nutrient dynamics by favoring anaerobic processes such as denitrification, which reduces nitrate to nitrogen gas (N₂) and may limit nitrate availability (correlation between NO₃⁻ and temperature: −0.235).

The decline in silicate has important ecological implications. Silicate is a key nutrient for diatom growth, and its reduction may have led to changes in phytoplankton community structure, favoring non-diatom species [53,54,55]. The long-term reduction in silicate may, thus, have consequences for the food web, as diatoms play a crucial role in marine ecosystems by supporting higher trophic levels [53,54,56]. As climate change leads to warmer winters, the intensity of water column mixing may diminish. Winter mixing is crucial for replenishing surface nutrients, including silicate, from deeper layers. Warmer winters may limit this mixing, potentially reducing the availability of silicate in surface waters during the subsequent seasons. Increasing temperatures may shift the composition of phytoplankton communities, favoring species that are less reliant on silicate, such as non-diatom phytoplankton [57,58,59]. This could lead to reduced silicate uptake overall, altering its seasonal dynamics [60,61,62]. Since rivers are a major source of silicate to coastal waters, alterations in their flow could directly impact silicate concentrations [61,63]. This extended nutrient limitation may influence the overall nutrient balance and phytoplankton productivity in surface waters [64,65,66,67]. Climate change could alter the natural seasonal dynamics of silicate in the Black Sea, potentially leading to changes in nutrient availability, phytoplankton community composition, and the overall health of the marine ecosystem [68,69,70,71].

The nitrogen–phosphorus (N:P) ratio is an important metric in marine ecosystems, with deviations from the Redfield ratio (16:1) serving as the indicators of nutrient imbalances that influence ecosystem dynamics [72]. Discussions surrounding the N:P ratio often highlight the relationship between natural processes and anthropogenic impacts, such as agricultural runoff and wastewater discharges, which disproportionately increase nitrogen inputs [73], particularly in coastal and semi-enclosed seas like the Black Sea [74,75]. These changes can drive eutrophication, harmful algal blooms (HABs) [76], and hypoxic conditions, fundamentally altering food web structures and biogeochemical cycles [77,78]. Additionally, shifts in the N:P ratio affect phytoplankton community composition, favoring nitrogen-fixing cyanobacteria in nitrogen-limited environments and potentially impacting zooplankton and higher trophic levels [76,79,80]. While the Redfield ratio remains a benchmark, regional variability underscores the need for context-specific thresholds to guide management strategies [72]. The rising N:P ratio and silicate depletion highlight potential risks for phytoplankton dynamics, with a shift from diatom-dominated systems to cyanobacteria or flagellate-dominated communities.

This study’s principal component analysis (PCA) provides significant insights into the interactions between nutrient dynamics, climate change, and the proliferation of Noctiluca scintillans in the Black Sea. The PCA revealed that the nitrogen–phosphorus (N:P) ratio is a dominant factor influencing N. scintillans abundance. The PCA biplot underscores that nutrient stoichiometry, particularly the N:P ratio, strongly correlates with bloom dynamics, while temperature, although influential, primarily affects blooms indirectly. Elevated temperatures exacerbate water column stratification, reducing vertical mixing and enhancing nutrient trapping in surface layers. This phenomenon creates favorable conditions for N. scintillans growth during the summer months. The pronounced increase in summer points to warmer water temperatures and enhanced stratification as the key drivers of these blooms [81]. Stratification during summer can concentrate nutrients in the surface layers, promoting the growth of N. scintillans. Moreover, the summer season often brings higher levels of organic matter, resulting from increased biological activity and nutrient input from river runoff, which provides a rich food source for this heterotrophic dinoflagellate [48,49,58].

Furthermore, the PCA highlights the complexity of nutrient interactions in the Black Sea ecosystem. While phosphorus concentrations have declined over decades, nitrogen concentrations remain high, maintaining an elevated N:P ratio. This imbalance drives ecosystem changes, shifting phytoplankton communities from diatom-dominated systems to those dominated by non-diatom species, including HAB-forming taxa like N. scintillans. Such shifts can disrupt trophic dynamics, reduce biodiversity, and intensify hypoxia in already vulnerable regions of the Black Sea.

N. scintillans dynamics also fit within this seasonal framework. The broader range of N. scintillans abundance and biomass observed during summer suggests significant variability driven by several environmental factors, such as nutrient availability [82]. In contrast, the lower values from autumn, winter, and spring indicate less favorable conditions for the organism, likely due to lower temperatures and potentially limited nutrient availability.

Overall, this seasonal pattern suggests that N. scintillans blooms are primarily a summer phenomenon in the RBSC. This has significant ecological implications, as the blooms can lead to oxygen depletion in the water column, additionally to the increasing seawater temperature, potentially causing hypoxia and disrupting the marine food web during peak bloom periods [83]. The peak of N. scintillans blooms, particularly during summer, is often associated with increased ammonium levels. As N. scintillans consumes large quantities of organic material, it releases ammonium₄ as a byproduct of its metabolism. This process can lead to elevated concentrations of ammonium in the surrounding water, particularly in areas where blooms are extensive. The correlation between ammonium and N. scintillans abundance is often positive during bloom periods, reflecting this contribution to nitrogen cycling [83,84].

Furthermore, during and after bloom events, the decomposition of N. scintillans biomass by bacterial activity can further elevate ammonium concentrations. This release of ammonium can enhance nutrient availability, potentially leading to a feedback loop that supports further primary production, especially under the stratified, low-oxygen conditions typical of summer [48,49,58,81].

Thus, N. scintillans blooms can play a significant role in the nitrogen cycle by directly releasing ammonium through excretion and indirectly through decomposition processes [85], contributing to the nutrient dynamics in coastal ecosystems like the Black Sea [15].

The nutrient dynamics along the Romanian Black Sea coastal waters are heavily influenced by riverine nutrient inputs, agricultural runoff, and industrial discharges. The Danube River, as one of Europe’s largest rivers, serves as the primary source of nutrient loading in the region [86,87,88]. Its discharge introduces substantial quantities of nitrogen, phosphorus, and organic matter into the Black Sea, driving eutrophication and contributing to harmful algal blooms (HABs) [76,81,86,89,90]. Seasonal fluctuations in river flow, driven by snowmelt, rainfall, and droughts, modulate nutrient delivery, amplifying nutrient loads during high-flow periods and concentrating pollutants during low-flow conditions.

Agricultural runoff significantly exacerbates nutrient imbalances. Fertilizer application across the Danube Basin has historically introduced excessive nitrogen and phosphorus into the water system [86,91,92]. This agricultural input not only sustains primary production but also modifies the nitrogen–phosphorus (N:P) ratio, favoring the proliferation of nitrogen-tolerant plankton species like N. scintillans. Despite improvements in agricultural practices, episodic peaks in runoff continue to reflect inadequate nutrient management [86,93].

Industrial discharges, particularly from urban wastewater and industrial effluents, remain critical contributors to nutrient dynamics [86,94]. Point sources of pollution from coastal areas, combined with untreated or partially treated wastewater, further elevate nutrient concentrations near urbanized regions. These localized discharges exacerbate eutrophication and lead to oxygen depletion, with detrimental effects on marine biodiversity and ecosystem services [86,93,95,96,97,98].

Eutrophication and associated ecosystem changes in the Romanian Black Sea coastal waters have profound societal and economic implications. As nutrient-driven algal blooms, particularly those involving harmful species like N. scintillans, continue to increase, they threaten marine biodiversity, disrupt food webs, and compromise the resilience of coastal ecosystems. This degradation directly affects local fisheries by reducing fish stocks due to hypoxia and habitat loss, leading to diminished livelihoods for communities dependent on fishing. Additionally, the decline in water quality and biodiversity undermines coastal tourism, which is a key economic driver in the region. These ecosystem changes highlight the urgency for integrated nutrient management and sustainable development strategies, as they directly impact human well-being, regional economies, and the achievement of global environmental goals such as SDG 14 (Life Below Water).

The study stands for targeted nitrogen reduction strategies to mitigate eutrophication and its ecological impacts. Policies should prioritize reducing agricultural runoff by optimizing fertilizer use and promoting sustainable farming practices. Enhanced wastewater treatment systems are critical to minimizing nitrogen discharges into the Black Sea. Establishing buffer zones along riverbanks can also play a vital role in curbing nitrogen inputs from land-based sources. These measures, when combined, can significantly reduce the nutrient loads entering the Black Sea and help restore ecological balance.

Integrated coastal management is essential for addressing the transboundary nature of Black Sea nutrient dynamics. Region-specific nutrient reduction strategies aligned with the Marine Strategy Framework Directive should be developed and implemented. Additionally, regional cooperation among Black Sea countries is imperative to harmonize actions for reducing nitrogen and phosphorus inputs, ensuring the effectiveness of mitigation efforts across the basin.

Improved monitoring and modeling efforts are crucial for tracking changes and designing adaptive strategies. High-resolution, real-time monitoring systems should be established to measure nutrient fluxes, oxygen levels, and biological responses. Scenario-based modeling can assess the effectiveness of proposed nitrogen reduction measures under various climate change scenarios, providing valuable insights for policymakers and stakeholders.

Finally, enhancing biodiversity and ecosystem resilience should be a central component of mitigation strategies. Restoring natural habitats such as wetlands can improve nutrient filtration and strengthen ecosystem resilience. Conservation programs targeting species affected by eutrophication and deoxygenation can also play a vital role in maintaining marine biodiversity and ecosystem health.

By implementing these recommendations, the RBSC region can mitigate eutrophication, reduce the frequency and severity of harmful algal blooms, and enhance the sustainability of its marine ecosystems. These actions align with global efforts to achieve SDG 13 (Climate Action), SDG 14 (Life Below Water), and SDG 3 (Good Health and Well-being). An integrated approach is essential for ensuring the long-term health of the Black Sea and the well-being of the communities that depend on its resources.

5. Conclusions

In conclusion, this study highlights the significant long-term impacts of rising sea surface temperatures (SST) on nutrient dynamics, eutrophication, and marine biodiversity along the Romanian Black Sea coast. The analysis revealed key ecological trends, including a consistent increase in SST, shifts in nutrient stoichiometry, and oxygen depletion, which exacerbate eutrophication and promote harmful algal blooms such as Noctiluca scintillans. These changes, driven by warming and nutrient imbalances, pose substantial risks to marine biodiversity and ecosystem resilience, particularly through alterations in phytoplankton community composition and trophic dynamics. To address these challenges, the findings emphasize the need for targeted nitrogen reduction strategies, such as optimizing agricultural practices, enhancing wastewater treatment, and implementing buffer zones to minimize land-based nutrient inputs. By providing actionable insights into climate-driven nutrient dynamics and aligning with global frameworks like the United Nations Sustainable Development Goals (SDGs), specifically SDG 13 (Climate Action) and SDG 14 (Life Below Water), this research contributes to sustainable coastal management and the preservation of marine ecosystems. Future efforts should integrate improved monitoring, scenario-based modeling, and regional cooperation to enhance adaptive strategies for mitigating the impacts of climate change on the Black Sea.

Despite these findings, the study has several limitations. The observational data used covers multiple decades but is not exhaustive, particularly regarding spatial coverage across different Black Sea regions. Additionally, the complexity of marine ecosystems means that confounding factors, such as riverine nutrient input variability, anthropogenic influences, and unaccounted biological interactions, could impact the results. Modeling uncertainties and limitations in the availability of in situ measurements also constrain the ability to make definitive conclusions about the ecosystem’s response to long-term climate stressors. Furthermore, data heterogeneity, stemming from differences in sampling methods, temporal frequency, and spatial resolution across datasets, introduces methodological limitations that could affect the comparability and consistency of results. This remains a critical factor to address in order to improve the accuracy and robustness of future research. Future work should aim to incorporate more extensive datasets, improve spatial coverage, and use higher-resolution models to further understand the complexities of climate change impacts on Black Sea ecosystems.