The Effects of Reduced Wastewater Load in the Marine Area off Turku in the Archipelago Sea During the Period 1965–2025

Abstract

In Finland, municipal wastewater treatment has significantly improved in recent decades, leading to a substantial reduction in nutrient loads from wastewater discharged into water bodies. For example, in the marine area off Turku in the Archipelago Sea, located in the northern Baltic Sea, the total phosphorus load from wastewater has decreased to approximately one-eighth of its level in the early 1990s. Simultaneously, the total nitrogen load has been reduced to one-fifth, and the ammonium nitrogen load is now less than 5% of its peak in 1994. This study examines in detail how water quality parameters and phytoplankton indicators in wastewater-affected areas have changed during the same period in which wastewater loads have significantly decreased. This reduction has contributed to positive developments in the marine area off Turku, although the goal of achieving good ecological status remains unmet. In Raisio Bay, chlorophyll a concentrations decreased by 68% following the relocation of wastewater discharge. In Rauvola Bay, the reduction was estimated to be 36%. Over the past 15 years, the biomass of nitrogen-fixing cyanobacteria has increased in northern Airisto. This trend appears to be driven by a decrease in external nitrogen loading in combination with increased internal phosphorus loading. Water bodies in the inner archipelago continue to receive excessive nutrient inputs from the surrounding catchment area, while internal loading significantly delays the restoration process.

1. Introduction

Almost all indices related to anthropogenic environmental change and natural resource consumption began accelerating in the 1950s [1,2]. This trend was also evident in the increasing nutrient loads in the Baltic Sea. In Finland, for instance, the volume of water pumped into the sewage system in 1969 was 3.7 times higher than in 1950 [1]. The rise in nutrient discharge was likely even greater, as phosphate-containing detergents became widely used during the same period. At the time, most municipal and industrial wastewater from Baltic Sea countries was discharged into waterways either with minimal filtration or entirely untreated. This era also marked a shift in agricultural practices, with farms increasingly replacing manure with synthetic fertilizers [1]. It was estimated that the nutrient load entering the Baltic Sea increased up to eightfold during the 20th century, with the majority of this increase occurring between 1950 and 1980 [3].

Various statistics suggest that eutrophication related to sewage pollution began around 1955 in the central Baltic Sea [4]. The Eutrophication Ratio (ER) [5] indicates that eutrophication emerged in the mid-1950s, peaked in the 1980s, and subsequently improved significantly. Andersen et al. [5] documented these improvements, which are a direct result of long-term efforts to reduce nutrient inputs.

In Finland, wastewater treatment only began after the Water Act came into effect in 1962. The law prohibited activities that polluted water bodies and required polluters to apply for permits to discharge wastewater. However, the processing of wastewater permits was slow, and obtaining decisions took a long time. The largest wastewater polluter in the Archipelago Sea was the city of Turku. Turku’s first wastewater treatment plant was completed relatively late, only at the end of the 1960s. Until then, the city’s wastewater was discharged into the Aura River, which became increasingly polluted year by year. Over time, the lower reaches of the river began to resemble an open sewer.

In the 1960s, Turku had approximately 120,000 residents, while its neighboring municipalities—Raisio, Naantali, and Kaarina—had a combined population of around 25,000. By the late 1960s, before wastewater treatment plants were built, the wastewater from 145,000 people was discharged directly into the sea off Turku. It is estimated that this wastewater contributed approximately 120 tons of phosphorus and 740 tons of nitrogen per year to the Archipelago Sea. By 2023, the estimated phosphorus load from wastewater discharged into the sea by the Turku region’s central wastewater treatment plant had decreased to 3.4 tons per year, while the nitrogen load had dropped to 212 tons per year [6].

The ecological state of the Archipelago Sea has deteriorated over the decades, like the rest of the Baltic Sea, but there are significant regional differences between the various archipelago zones [7]. The inner archipelago is predominantly sheltered and strongly influenced by riverine inputs and, in the past, by local wastewater loads. In the joint monitoring program conducted in the sea area off Turku, the development of water quality has been comprehensively monitored annually since the 1960s. Despite this, the long-term effect of reduced wastewater loads on the ecological state of the sea area have been surprisingly little studied in the Archipelago Sea region. The first summaries in Finnish [8,9] were made in the 1990s, when, for example, the phosphorus load from wastewater was still more than four times higher than the current level. In his 2011 review on the development of the state of the Archipelago Sea, Suomela [10] examined the overall effects of wastewater load reduction.

Estuaries and inner archipelagos are central in the land–ocean continuum, acting as strong nutrient filters and transforming terrestrial nutrient inputs by complex biogeochemical interactions [11]. The recovery of estuaries and bays can follow different pathways and is further complicated by the overall progression of eutrophication and climate change, emphasizing the need for comparative studies [12]. For example, in the inner archipelago of Stockholm, where phosphorus and biochemical oxygen demand (BOD) loads from wastewater were drastically reduced in the early 1970s, the responses of the marine environment have been examined through phosphorus modeling [13].

This study examines in greater detail how water quality parameters in wastewater-affected areas off Turku, in the eastern Archipelago Sea, have changed during the same period in which wastewater loads have significantly decreased. For this research, various datasets on wastewater loads were compiled, and environmental management information systems—containing comprehensive water quality data—were utilized. The analysis also examines changes in the abundance of cyanobacteria based on phytoplankton sample data. Based on this extensive analysis, the aim is to assess whether achieving good ecological status in the inner archipelago of the Archipelago Sea is feasible through further reducing the critical nutrient load.

2. Materials and Methods

2.1. Study Areas

The Archipelago Sea is located between the Baltic proper and the Bothnian Bay (Figure 1). Descriptions of its characteristics, nutrient loads, and water quality trends are presented in previous articles [7,14]. This study has focused on the eastern inner archipelago zone of the Archipelago Sea, whose area has previously been estimated at 679 km² [7], particularly in the coastal area off Turku. In this context, it is important to emphasize that the total phosphorus and nitrogen load brought to the sea by the largest river flowing into the marine area off Turku, the Aura River, has not significantly changed over the period from 1995 to 2025. According to statistics from the Finnish Environment Institute, the total phosphorus load of the Aura River is approximately 56 tons per year, while the total nitrogen load is around 610 tons per year.

2.2. Wastewater Treatment Plants and Load Data

The Turku central wastewater treatment plant officially began operations in 1968, when wastewater from the western part of Turku was directed to the plant. The treatment method used was biological purification based on the activated sludge process. When the central treatment plant was completed in the late 1960s, wastewater from the eastern part of Turku was still discharged directly into the river mouth. The sewer tunnel under the Aura River was completed in 1972, after which almost all of Turku’s wastewater was treated before being discharged into the sea. In Raisio and Kaarina, wastewater treatment with their own treatment plants began around the late 1960s. The Turku central, Raisio, and Kaarina wastewater treatment plants ceased operations when the current so-called Kakola wastewater treatment plant was commissioned in 2009. In Turku, the wastewater discharge location did not change after centralization (Figure 1) [15,16,17].

The Kakola wastewater treatment plant is a biological–chemical activated sludge facility, enhanced with post-filtration of wastewater using sand filters. Phosphorus is precipitated through simultaneous precipitation with ferrous sulfate. The plant’s enhanced total nitrogen removal is based on the denitrification–nitrification process in aeration [6]. Its designed population equivalent is 315,000, but the current influent load corresponds to a population of just under 300,000. In 2023, the wastewater treatment plant’s influent load was 580 kg/day of total phosphorus, equivalent to 211.7 tons per year, and 4200 kg/day of total nitrogen, amounting to 1533 tons per year, of which 76.2% was ammonium nitrogen. The treatment plant’s environmental permit requires a removal efficiency of 95% for total phosphorus and 75% for total nitrogen. In 2023, the actual removal rates achieved were 98% for total phosphorus, 87% for total nitrogen, and 99% for ammonium nitrogen [6].

Nutrient load data from wastewater treatment plants for different years were obtained from the environmental administration’s environmental protection information system (Vahti). Load data have been available since 1983. Previous assessments of wastewater load have been based on population data from the city of Turku and its neighboring municipalities, along with the load coefficients provided in Government Decree 157/2017. According to the decree, the untreated domestic wastewater load per person per day is 2.2 g of total phosphorus and 14 g of total nitrogen.

2.3. Water Quality and Phytoplankton

2.3.1. Data and Analytics

The water quality of the Archipelago Sea has been monitored for an extended period with standard methods, and the data are available through the open data service VESLA of the Finnish Environment Institute (https://www.syke.fi/en/environmental-data/maps-and-information-services/open-environmental-information-systems, accessed on 10 April 2025). VESLA includes physio-chemical measurement results of national and regional monitoring, carried out by regional environment centers, as well as local statutory monitoring results conducted by private companies and water protection associations.

In general, the basic analytical methods and chemistries that are used to determine concentrations of inorganic nutrients in seawater are well established [18]. Strickland and Parsons outlined the manual methods in their book, A Practical Handbook of Seawater Analysis [19]. The chemical methods have been changed, optimized, and automated over the decades by numerous authors, but the basic chemistries remain the same and are based on colorimetric reactions. The exception to this is the newer methods for ammonium/ammonia determination, which are based on fluorometry [18].

The dataset used in this study is based on the analysis of various water quality parameters conducted in accredited laboratories, in accordance with the accreditation standard [20]. In laboratories, nutrients were usually analyzed from unfiltered samples within 5–8 h after sampling. The effects of turbidity and color on absorbance readings were taken into account in all analyses. Samples for total P were digested with K₂S₂O₈ in acidic conditions and measured spectrophotometrically as ammonium molybdate blue complex [21]. DIP was analyzed as total P, but without the digestion phase [22]. Total N was measured by digesting the sample with K₂S₂O₈ to nitrate, which was further reduced to nitrite and measured in an FIA-ionanalysator application according to the EPA 353.2 method [23]. NO₂ + NO₃ was analyzed as total N, but without the digestion. NH₄ was measured spectrophotometrically with the indophenol blue method [24]. Chlorophyll-a samples were filtered through Whatman GF/F filters (nominal pore size 0.7 mm) and measured spectrophotometrically from ethanol extract [25].

Phytoplankton analyses were conducted on composite samples from the production layer collected during both July and August. The depth of the composite sample from the phytoplankton production layer was determined as twice the Secchi depth. The composite sample was collected into a container using a two-meter-long tube sampler, ensuring that subsamples were taken equally from all parts of the production layer. Phytoplankton biomasses and cell counts were determined at the species level from the phytoplankton samples in accordance with the water and marine management method guidelines [26]. The analysis is carried out using the Utermöhl method [27,28] with an inverted microscope using brightfield optics, phase contrast optics, or differential interference contrast (DIC) optics. The samples are analyzed for autotrophic and mixotrophic phytoplankton, cyanobacteria (blue-green algae), all (including heterotrophic) dinoflagellates and flagellates, and in marine samples, the mixotrophic ciliate Mesodinium rubrum.

For the measurement of phytoplankton counting units, an eyepiece micrometer attached to the microscope eyepiece is used. The phytoplankton sample is settled in a settling chamber, the bottom surface area of which is known. The quantitative microscopic counting method provides results on species composition, densities of counting units, and biomass. The method manual [26] includes, among other things, conversion factors, which are used to convert the marine sample counting results into units per liter. From the biovolume data, the biomass (wet weight) is simply derived by a rough assumption of a plasma density of 1 g cm⁻³ [28]. Biomass (wet weight) data are preferred for characterizing spatial and temporal phytoplankton patterns and for modeling [28].

2.3.2. Processing of Datasets

The water quality analysis presented in this study was primarily based on total phosphorus concentrations and on algae production-related chlorophyll a concentration in the surface water (0–10 m) during the ecological classification period (1 July–7 September). In addition, time series of ammonium nitrogen, nitrate nitrogen, and phosphate phosphorus concentrations in surface water (0–10 m) during winter (1 January–31 March) were examined at various observation stations in the coastal area off Turku. Wintertime concentrations of dissolved nutrients can be used as an index for summer conditions when they cannot be directly measured from surface water because during the summer season in the northern Baltic Sea inorganic nutrients are rapidly (within hours) assimilated by organisms [29]. DIN:DIP nutrient ratios were also analyzed using the ecological classification period data from station 220 for data below the production layer (deeper than 20 m). The Redfield ratio (N:P 16, based on molar concentrations) is often used as an indicator of which nutrient puts a limit to the growth: when the ratio is above 16, phosphorus is the limiting factor, and when it is below 16, nitrogen is the limiting factor [30].

A total of 34 monitoring stations provided the analyzed data (Figure 1). For comparison, some water quality data from the intensive stations Seili (Figure 1) and Brändö were also analyzed. Brändö is located at the northern boundary of the Archipelago Sea (N6732217-E176832, ETRS-TM35FIN). The first observations of total phosphorus concentrations date back to 1965. Measurements of chlorophyll a concentrations in surface water began in the 1980s. In this study, the time series of various variables were analyzed in more detail at stations in Raisio Bay (Turm 260) and near Rauvola Bay in Kaarina (Turm 175) (Figure 1;

Table 1. The main environmental characteristics of the selected monitoring stations. Position provided by coordinates (N, E) in the system ETRS-TM35FIN. Chl represents the average chlorophyll-a concentration during the ecological classification period (1 July–7 September) for the years 2021–2024. Ecol. status refers to the ecological status class determined based on the average chlorophyll-a concentration. In the inner archipelago, the class boundaries are good/moderate = 3.0 µg chl/L, moderate/poor = 7.0 µg chl/L, and poor/bad = 17 µg chl/L. In the middle archipelago, the corresponding boundaries are 2.5, 5.8, and 14.5 µg chl/L. The Seili station is in the middle archipelago, while the other stations are in the inner archipelago.

Station	Location ETRS-TM35FIN	Depth (m)	Chl (µg/L)	Ecol. Status
260	6712516 231712	3	11.9	poor
190	6709098 237337	8	16.5	poor
175	6704772 240056	7.5	16.7	poor
210	6796584 231541	21.6	9.5	poor
220	6703331 230194	51.8	6.98	moderate
Seili	6690561 220867	50.6	5.3	moderate

). Wastewater discharge at these locations ended in 2009, and the wastewater was redirected to the Turku Kakola treatment plant. Responses were also studied at station 190, located off the treatment plant (Table 1). The extent of the wastewater impact area was assessed by examining water quality and phytoplankton responses at observation stations 210 and 220, which are located in Airisto but still belong to the inner archipelago zone (Table 1).

The statistical processing of water quality data, such as the calculation of phosphorus–chlorophyll a regressions or statistical summary metrics of the datasets, was carried out using Microsoft^® Excel^® for Microsoft 365 MSO (version 2502 Build 16.0.18526.20264). The Mann–Kendall (MK) statistical test [31] was used in Excel XLSTAT to evaluate the presence of positive or negative trends in the time series of water quality or phytoplankton data. The null hypothesis (H0) is that there is no trend in the series; as an alternative hypothesis (HA), the MK test supposes that there is an upward or a downward trend in the tested sample. A significant level of 0.05 was selected in this study. The Mann–Kendall (MK) test can be used as an alternative to parametric linear regression analysis for trend testing. Unlike regression analysis, which assumes that the residuals from the fitted line are normally distributed, the MK test does not require this assumption. Instead, it is a non-parametric (distribution-free) test.

3. Results

3.1. Nutrient Loads

The total phosphorus load from wastewater to the sea in the coastal area off Turku was an average of 24 t/a in the 1980s. After that, between 1991 and 2008, it remained at around 16 t/a. After the Kakola central wastewater treatment plant began operations, the phosphorus load was halved and continued to decrease, now being less than 4 t/a (Figure 2).

The total nitrogen load has decreased more steadily than the phosphorus load since the 1980s, when it peaked at over 900 t/a (Figure 3). Today, the total nitrogen load from wastewater to the sea is just over 200 t/a. The ammonium nitrogen load has decreased even more significantly than the total nitrogen load (Figure 4). At the turn of the 1980s and 1990s, 60% of the total nitrogen load (around 800 t/a) was in the form of ammonium nitrogen. When the Kakola treatment plant started operating, the share of ammonium nitrogen had decreased to 30%, and in recent years, it has been only around 10%. The ammonium nitrogen load from Turku’s treatment plant had already dropped to 40 t/a between 2005 and 2008 but tripled when the new treatment plant was commissioned in 2009 and began receiving wastewater from Raisio and Kaarina as well (Figure 5). After 2013, the NH₄-N load started to decrease again and was only 23 t/a in 2021–2023.

3.2. Responses in Water Quality

The history of the entire wastewater load is most clearly visible in the water quality of Raisio Bay at the Turm 260 station (Figure 6). The phosphorus concentrations in the early 1970s indicate a level (300–400 µg TP/L) that was present before wastewater treatment began. In the 1980s, when the treatment plant’s operations began to stabilize, the concentrations decreased to 110 µg TP/L, and in the 1990s, they decreased to 70 µg TP/L. In the early 2000s, the treatment efficiency declined, and the phosphorus concentrations in Raisio Bay averaged 92 µg TP/L before the load completely ended in 2009. The average phosphorus concentration of 62 µg TP/L from 2010 to 2024 still indicates the lowest, or bad, quality level in ecological classification.

The decrease in total nitrogen content is also reflected in changes in its ecological classification (Figure 7). In the early 1970s, nitrogen concentrations were at 1200 µg TN/L, from which they decreased by one-third in the 1980s. At this level (800 µg TN/L), the concentrations remained on average until 2009, when wastewater loading ceased. The threshold for poor status (575 µg TN/L) was crossed in 2010. The lowest nitrogen concentrations were recorded in 2017–2018, averaging 329 µg TN/L, which was already close to the upper limit for good status (325 µg TN/L). From 2019 to 2024, the average was 447 µg TN/L, which corresponds to a poor status.

The wintertime decrease in soluble nitrogen concentrations (NH₄ and NO₂₃) in Raisio Bay corresponds to the development of wastewater loading (Figure 8). Between 1990 and 1998, NH₄-N concentrations had already decreased to a level of 200 µg NH₄-N/L from the previous peak values (for example, 1984: 1785 µg NH₄-N/L). Then, they increased again by approximately four times during the period from 1999 to 2005. During the period 2005–2024, a statistically significant decreasing trend in NH-N concentrations was observed (

Table 2. The result of the Mann–Kendall trend test for different water quality parameters monitored in the sea area off Turku (Figure 1). winter = the data were collected from 1 January to 31 March; summer = 1 July to 7 September. Kendall’s Tau is a non-parametric measure of the relationships between the columns of ranked data. p-value < 0.05 indicates rejection of null hypothesis of no trend, thus revealing the existence of the trend. S-value is transformed p-value.

Station	Years	Season	Parameter	Kendall’s Tau	p-Value	S-Value	Result
Raisio, 260	2005−2024	winter	NH4-N	−0.776	<0.0001	−147	decreasing trend
	2004−2024	summer	Chl	−0.558	0.001	−116	decreasing trend
Kaarina, 175	1969−1985	summer	TP	−0.578	0.001	−78	decreasing trend
	1986−2024	summer	TP	0.140	0.338	42	no trend
Turku, 190	1983−2008	summer	Chl	−0.365	0.012	−108	decreasing trend
	2009−2024	summer	Chl	−0.127	0.527	−15	no trend
Airisto, 210	1981−2007	summer	TP	0.466	0.001	151	increasing trend
Airisto, 220	1978−1998	summer	Chl	0.521	0.002	87	increasing trend
	1998−2010	summer	Chl	0.520	0.017	−40	decreasing trend

). After the cessation of wastewater loading, NH₄ concentrations decreased rapidly and were on average only 18 µg NH₄-N/L during the years 2021–2024. The cessation of wastewater loading did not seem to affect NO₂₃-N nitrate concentrations anymore, and they remained at a level of 280 µg NO₂₃-N/L.

The chlorophyll a concentration, which represents algal production, appears to decrease in Raisio Bay in the same proportion as the summer total nitrogen concentrations and wintertime NH₄-N concentrations (Figure 9; Table 2). The decrease in phosphorus concentration has also influenced the development. Between 2000 and 2009, the chlorophyll a concentration was 38.5 µg chl/L, which is classified as bad. After the cessation of wastewater loading, the concentration decreased by 68% and has averaged 12.3 µg chl/L, which is classified as poor. The poor status lower threshold (moderate/poor) is 7.0 µg chl/L.

In Kaarina’s Rauvola Bay (nearest monitoring station Turm 175), where wastewater loading also ended in 2009, a statistically significant decrease in total phosphorus concentration was observed during the period 1969–1985 (Table 2); however, no statistically significant change was detected in the subsequent period from 1986 to 2024. The total phosphorus concentration has averaged 59.5 µg TP/L during the years 1986–2024, still classified as bad (Figure 10). The chlorophyll a concentration was 26.7 µg chl/L between 2001 and 2007, classified as bad (Figure 11). After the cessation of wastewater discharge, it halved and temporarily dropped to the poor category. After that, the concentrations became more variable, and between 2011 and 2024, the average was 17.1 µg chl/L, right on the borderline between poor and bad (17 µg chl/L).

At monitoring station Turm 190, near the wastewater discharge site of Turku’s central treatment plant and Kakola treatment plant, wintertime ammonium nitrogen concentrations have varied according to the load (Figure 12; see also Figure 5). The chlorophyll-a concentration at station 190 was on average 23.1 µg chl/L between 1983 and 1986, classified as bad (Figure 13). After that, concentrations began to decrease statistically significantly (Table 2), reaching its lowest level in 2007–2008, with an average of 8.2 µg chl/L, classified as poor but close to the upper limit of moderate (7 µg/L). When the Kakola treatment plant began operations, the chlorophyll a concentration increased again and was 17.3 µg chl/L between 2011 and 2020, slightly above the bad threshold (17 µg chl/L). In 2021–2023, the concentrations had slightly decreased to 16.5 µg chl/L, placing them in the poor category.

In the northern part of Airisto, at observation station Turm 210, the impact of wastewater on total phosphorus concentration was evident in the 1970s (Figure 14). At that time, the average phosphorus concentration in surface water was 33 µg TP/L, classified as poor. As wastewater treatment improved, the phosphorus concentration stabilized at 25 µg TP/L in the 1980s, falling into the moderate category. Phosphorus concentrations then began to rise from the early 1980s until the early 2000s, so that during the years 1999–2007, the average (32.5 µg TP/L) indicated a return to the poor category. The increase in phosphorus concentration during the period 1981–2007 was statistically significant (Table 2). This increase was not related to wastewater discharge but was instead driven by the overall development of the Archipelago Sea. Since 2010, the situation has improved and based on the average phosphorus concentration of 29 µg TP/L, the classification is moderate (with the upper limit for moderate being 32 µg TP/L).

NH₄-N concentrations in winter (Figure 15) followed a similar trend and variation as observed at station Turm 190, near Turku’s wastewater discharge site (Figure 12). The difference lies in the concentration levels, which indicate dilution. While the NH₄ concentrations at station Turm 190 averaged 340 µg NH₄-N/L around the turn of the 2000s, the corresponding value at station 210 was 97 µg NH₄-N/L. And while the NH₄ concentration at station 190 was 45 µg NH₄-N/L during the years 2021–2024, the corresponding value at station 210 was 15 µg NH₄-N/L.

Chlorophyll a concentration was at its lowest in the early 1990s and again during the years 2008–2011, averaging 6.2 µg chl/L, which falls into the moderate category (Figure 16). The highest concentration was observed during the years 1998–2007, with an average of 10.2 µg chl/L (classified as poor). From 2012 to 2024, the chlorophyll a concentration remained in the poor category, with an average of 9.3 µg chl/L.

At observation station 220, located about 3 km south of station 210 in Airisto, chlorophyll a concentrations increased significantly during the period 1978–1998 (Table 2), but remained lower than those in the inner archipelago, even though the variation was similar to that at station 210 (Figure 17). Following the transitional period in the late 1990s, chlorophyll a concentrations exhibited a downward trend, with the decline being statistically significant over the period 1998–2010 (Table 2). The 2000s minimum occurred during the years 2008–2011. At station 220, the average was 3.7 µg chl/L, which was 40% lower than at station 210. From 2012 to 2024, chlorophyll a concentrations averaged 6.4 µg chl/L, falling into the moderate category. The difference compared to station 210 was about 30%.

3.3. Nutrient Ratios and Cyanobacteria

At Raisio Bay station 260, the wintertime nutrient ratio DIN:DIP indicates clear phosphorus limitation throughout the entire measurement period from 1997 to 2024, even though the ratio dropped to less than half by 2007, a couple of years before the complete cessation of wastewater discharge. The DIN:DIP ratio averaged 34 during the years 2007–2021 and 21 during 2022–2024, exceedingly still the phosphorus limitation threshold, above 16 (Figure 18).

In northern Airisto, at stations 210 and 220, the wintertime (1 January–31 March) DIN:DIP ratio decreased from the 1990s until the 2010s, dropping from a level of 45 to near the Redfield ratio of 16 (Figure 19). The decrease in the nutrient ratio was statistically significant during the years 1991–2009 but the change has not been statistically significant since then (

Table 3. The result of the Mann–Kendall trend test for different water quality parameters and the biomass of cyanobacteria monitored in the sea area off Turku (Figure 1). winter = the data were collected from 1 January to 31 March; summer = 1 July to 7 September. Kendall’s Tau is a non-parametric measure of the relationships between the columns of ranked data. p-value < 0.05 indicates rejection of null hypothesis of no trend, thus revealing the existence of the trend. S-value is transformed p-value.

Station	Years	Season	Parameter	Kendall’s tau	p-Value	S-Value	Result
Airisto, 210	1991–2009	winter	DIN:DIP	−0.626	<0.0001	−107	decreasing trend
	2010–2024	winter	DIN:DIP	0.219	0.276	23	no trend
	1991–2009	winter	DIN	−0.287	0.093	−49	no trend
	2010–2024	winter	DIN	0.067	0.767	7	no trend
	1991–2009	winter	DIP	0.582	0.001	99	increasing trend
	2010–2024	winter	DIP	−0.078	0.727	−8	no trend
	2011–2024	summer	cyanob. Nostocales	0.615	0.004	48	increasing trend
Airisto, 220	1999–2014	summer	DIN:DIP	−0.567	0.003	−68	decreasing trend
	2015–2024	summer	DIN:DIP	0.200	0.474	9	no trend
	1999–2024	summer	cyanob.	0.654	<0.0001	151	increasing trend
	2011–2024	summer	cyanob. Nostocales	0.615	0.004	48	increasing trend

). For a few years, the ratio fell below 16, but not significantly. After 2010, the ratio appears to have increased slightly, averaging 26 at station 210 and 17 at station 220 during the years 2021–2024. The ratios mainly indicate phosphorus limitation, but occasionally also combined N + P limitation, for example, in the years 2009–2011. The changes observed in the DIN:DIP ratios can be explained by the different trends in the concentrations of dissolved nitrogen and dissolved phosphorus (Figure 20). From 1991 to 2009, DIN concentrations decreased by about 30%, while DIP concentrations increased by the same amount. The decline in DIN concentrations was statistically significant at the 90% confidence level but did not reach the conventional 95% threshold (p = 0.093; Table 3). In contrast, the increase in DIP concentration was statistically significant at the 95% confidence level (Table 3). The decrease in dissolved nitrogen is partly related to the reduction in nitrogen load from wastewater, while the increase in dissolved phosphorus is linked to other factors. Since 2010, both concentrations have shown considerable variability between years, which is also reflected in their ratios, and the datasets do not show any statistically significant directional changes (Table 3).

In the summer, during the ecological classification period (1 July–7 September), a clearer change in the DIN:DIP ratio below the production layer (deeper than 20 m) has occurred since 2008 compared to the wintertime ratios (Figure 21). In the 1990s, the DIN:DIP ratio averaged 15.5, and during the years 2000–2007, it was only slightly lower at 14.8. However, since 2008, it has halved to an average of 7.4, which is clearly below the Redfield ratio of 16. The decrease in the DIN:DIP ratio observed between 1999 and 2014 was statistically significant, but no further change has occurred since then (Table 3).

At the comparison stations in the outer archipelago, Seili and the northern boundary of the Archipelago Sea in Brändö, the DIN:DIP ratios in wintertime have indicated mainly either combined N + P limitation or purely nitrogen limitation (Figure 22). At the Seili station in the 1990s, the DIN:DIP ratio was still 19, indicating phosphorus limitation. From 2001 to 2024, the DIN:DIP ratio at the Seili station has averaged 14.5, while at Brändö, it has averaged 9.5.

In northern Airisto, at stations 210 and 220, the biomass of cyanobacteria increased statistically significantly throughout the 2000s (Figure 23; Table 3). Most of them belong to the order Nostocales, with the dominant species being Aphanizomenon sp., and they can utilize atmospheric nitrogen for growth.

In northern Airisto, at stations 210 and 220, the first abundant observations of nitrogen-fixing cyanobacteria from the order Nostocales were made in 2009 and 2010, coinciding with the wintertime DIN:DIP ratio being close to the Redfield value of 16 (Figure 24). During that period, the average biomass of Nostocales was approximately 600 µg/L (wet weight), representing a tenfold increase compared to the mean levels recorded in the early 2000s. A statistically significant increase in the biomass of Nostocales cyanobacteria was observed at both northern Airisto stations between 2011 and 2024, with mean levels reaching 636 µg/(ww) during the July–August period of 2021–2024.

3.4. Summary: Data Analysis of 34 Monitoring Stations

During the years 2021–2024, the average total phosphorus concentration measured in the surface water within the impact area of Turku’s wastewater during the ecological classification period (1 July–7 September) was 42.1 µg TP/L. This concentration falls within the poor classification, which ranges from 32 µg TP/L to 52 µg TP/L. In the 1970s, the corresponding total phosphorus concentration averaged 79.3 µg TP/L, which fell into the bad classification. In the 1980s, it decreased to 50.9 µg TP/L, placing it in the poor classification. The chlorophyll a concentration in the 1980s was approximately the same as in recent years, at 11.7 µg chl/L. The chlorophyll a concentration peaked at an average of 15 µg chl/L during the years 1991–2009, when the phosphorus concentration was 48.3 µg chl/L.

The total phosphorus concentration (µg/L) explains 66.4% of the observed variation in chlorophyll a concentration in a statistically significant regression model (p < 0.001) based on the entire dataset (number of observations = 1179; Figure 25). The 95% confidence interval for the regression coefficient 0.314 is 0.301–0.327.

4. Discussion

The total phosphorus load in the marine area off Turku was at its highest before the introduction of wastewater treatment in the late 1960s. At that time, the phosphorus load from wastewater into the sea was estimated at 120 t/a, while the phosphorus load from the Aura River was slightly lower than today, at 40 t/a [10], totaling 160 t/a. The peak nitrogen load, on the other hand, occurred in the early 1980s, when Turku’s population had risen to 160,000 and nitrogen removal at treatment plants was inefficient. At that time, the nitrogen load from wastewater into the sea was approximately 1000 t/a, while the nitrogen load from the Aura River was at a level like today, around 600 t/a, totaling 1600 t/a. In recent years, the total phosphorus load to the sea off Turku has been 60 t/a, and the total nitrogen load has been 820 t/a. The phosphorus load has decreased by 62.5%, and the nitrogen load has decreased by 49%. Nevertheless, the nutrient load is still too high, as the ecological status of the marine area off Turku in the inner archipelago has, on average, only improved from bad to poor. Only at the outer boundary of the inner archipelago (station 220) has the total phosphorus concentration (22.7 µg TP/L) during the years 2021–2024 been just barely within the good classification (threshold 23 µg TP/L). However, even there, the chlorophyll a concentration (7 µg chl/L) has remained at a moderate level.

Based on previous studies, it is known that the total phosphorus concentration in the surface water of the Archipelago Sea has in recent years been at a level of 20 µg TP/L in the intermediate and outer archipelago zones [7]. This can be considered a background concentration, below which the total phosphorus concentration in the inner archipelago off Turku would not decrease, even if all local loads from the catchment area were to cease. By combining this information with the previously described reduction in phosphorus load and the measured average total phosphorus concentrations, a simple load model is obtained, y = 0.369 x + 19.95, where y is the average phosphorus concentration (µg/L) and x is the phosphorus load into the sea (t/a) (n = 3, p = 0.0015, R² = 1; data 20,0; 42,60; 79,160).

Using the equation, it can be calculated that achieving a phosphorus concentration of 32 µg TP/L (the upper limit of the moderate status) would require reducing the phosphorus load to 32.7 t/a from the current 60 t/a, a 45.5% reduction. Similarly, reaching the threshold for good status, a concentration of 23 µg/L, would require lowering the load to 8.3 t/a, representing an 86.2% reduction. Achieving good status in phosphorus concentration would therefore require reducing the phosphorus load from the Aura River from the current 56 t/a to 5 t/a, a 91% reduction, since further phosphorus removal from wastewater is practically no longer feasible. Such a requirement is undoubtedly impossible.

Moreover, even achieving good status in total phosphorus concentration would not result in a good status for phytoplankton. Responses in chlorophyll-a concentration to reduced total phosphorus levels can be examined using the previously presented regression equation y = 0.314 x − 1.19, where y is the average chlorophyll-a concentration (µg chl/L) and x is the average phosphorus concentration (µg TP/L) (Figure 25). Using the equation, it can be calculated that a phosphorus concentration of 32 µg TP/L (the upper limit of moderate status) would correspond to a chlorophyll-a concentration of 8.9 µg chl/L, while a phosphorus concentration of 23 µg TP/L (the upper limit of good status) would correspond to 6.0 µg chl/L. However, the ecological class threshold is much stricter, especially for the good status, which requires a chlorophyll-a concentration of 3 µg chl/L, while the threshold for moderate status is 7 µg chl/L.

And conversely, achieving the good status threshold for chlorophyll-a concentration would require a total phosphorus concentration of 13.3 µg TP/L, while the satisfactory status threshold would require 26 µg TP/L. The requirement for good status is so strict that, in practice, it would demand extensive measures across the entire Baltic Sea region, including a 50% reduction in the Baltic Sea’s total phosphorus concentration from current levels, in addition to significant local emission reductions. In fact, very similar conclusions were reached when using the FICOS model to calculate nutrient load ceilings for different Finnish marine areas [7,32]. The FICOS modeling approach presented by Lignell et al. [33] yielded a somewhat more optimistic outcome. Their results from Bayesian nutrient reduction scenarios showed in turn that reaching the EU WFD objectives for chlorophyll-a requires reductions of 40–70 % of anthropogenic riverine and point loads to the Aurajoki basin by which they mean the sheltered basin spanning from estuary to inner and middle archipelago [33]. Nonetheless, the modeling indicated that these objectives could only be met if atmospheric and internal loads were reduced by 50%. In practice, these boundary conditions are considered unrealistic [7].

When developing an ecological classification system, reference concentrations of chlorophyll-a were calculated by substituting the historical values of Secchi depth into the type-specific equations on the relationships between chlorophyll-a and Secchi depth [34]. However, statistical modeling could not be used for inner coastal types because historical Secchi depth values were missing. For inner coastal types, reference values have been estimated by relating the 5th percentile of current monitoring results to the reference values of outer coastal types [34]. While it seems obvious that, in the inner archipelago zone, the requirements for good status chlorophyll-a would be too strict and impossible to achieve, one option would be to refine the classification thresholds by reanalyzing the observation data. In the inner archipelago, it might be sufficient to adjust the current upper limit for moderate status (7 µg chl/L) to be the threshold for good status, redefine the lower limit for poor status (17 µg chl/L) as moderate, and set the poor status threshold at around 30 µg chl/L.

In any case, the wastewater treatment implemented has led to positive developments in water quality in the marine area off Turku, even though the goal of good status is still a ways off. In Raisio Bay, the chlorophyll a concentration, which indicates algal production, decreased by 68% when wastewater discharge was transferred to Kakola. Similarly, in Rauvola Bay near Kaarina, the reduction in chlorophyll a concentration was estimated to be 36%. At the outer boundary of the inner archipelago, in Airisto at stations 210 and 220, the observed decrease in chlorophyll a concentration in the early 2000s—approximately 40% between 2000 and 2009—appears to be linked to a significant reduction in ammonium nitrogen loading from wastewaters.

What is interesting at Airisto stations 210 and 220, however, is the development since 2009. Chlorophyll a concentrations started to rise again, and the variation between years also increased, even though ammonium nitrogen and phosphorus loading from wastewater continued to decrease. Between 2009 and 2011, NH₄-N loading to the sea was about 120 tons per year, while between 2021 and 2023, it was 23 tons per year. It is also worth noting that before the Kakola treatment plant started operating, in 2007–2008, NH₄ loading had decreased to 35 tons per year. Phosphorus loading, on the other hand, has steadily decreased throughout the entire observation period. Between 2009 and 2011, it was 6.3 tons per year, while between 2021 and 2023, it was 3.7 tons per year.

Since 2009, cyanobacterial biomass at the Airisto stations 210 and 220 has clearly increased compared to previous levels. Nutrient ratio analyses show that, at the same time, the system seems to have shifted towards nitrogen limitation, which favors nitrogen-fixing cyanobacteria, such as species of the Nostocales order like Aphanizomenon sp. [35,36] Due to the ability of nitrogen fixation, cyanobacteria are a nitrogen source for the system [37]. So, we can think that they, in part, nullify the work of wastewater treatment plants in reducing nitrogen removal. If internal phosphorus loading increases simultaneously and more dissolved phosphorus becomes available for phytoplankton, the result is an increase in algal production. This kind of development has been observed in the chlorophyll a concentrations at Airisto stations 210 and 220 since 2009 (Figure 16 and Figure 17). The temporal development of internal P loading intensity has not been studied in detail in the Archipelago Sea. The observed increase in wintertime dissolved phosphorus concentration at station 201 (Figure 20) does indeed suggest a strengthening of internal loading. It is also known that if the relative loading is calculated based on the amount of phosphorus reaching the upper surface layers (0–20 m), it is highest in the inner archipelago [7].

In their study on the responses of phosphorus fluxes in Stockholm’s inner archipelago to reduce wastewater loading, Walve et al. [13] identified a similar long-term decline in phosphorus concentrations in the water phase, as observed in the current study. Their results also indicate that estuarine and coastal sediments can respond rapidly to load reductions and exhibit small long-term legacy effects. However, their study did not investigate the effects of reduced phosphorus concentrations on algal production or cyanobacterial levels in the area, although the text does mention that chlorophyll a concentrations have also generally decreased. Based on the total phosphorus concentration (28 µg TP/L from 2006 to 2015; Table 1 [13]), the good status threshold (23 µg TP/L) was still far off also in Stockholm’s inner archipelago. If similar water quality data to that used in this study were available for the inner archipelago of Stockholm, a more detailed analysis of phosphorus and chlorophyll a concentrations could allow for the conclusion of whether achieving good ecological status is even possible there.

In my study, the interpretation of the minimum nutrient was based on the dissolved nutrient DIN:DIP ratios derived from monitoring data. This approach has limitations compared to, for example, similar experimental studies [38]. During the summer, dissolved nutrients cannot be measured from the productive surface layer using traditional water analyses, as all the available nutrient fractions are utilized by phytoplankton species and bacteria [29]. Therefore, nutrient ratios measured at other times of the year must be used as indicators for assessing nutrient availability. In this study, winter nutrient data were primarily used, as they are suitable for evaluating concentrations caused by steady wastewater loading. The data from Airisto station 220 were also used to calculate DIN:DIP ratios which were based on molar concentrations of deep-water layers (over 20 m) during the ecological classification period (1 July–7 September). Based on this, a clearer change was observed from 2009 onwards compared to the DIN:DIP ratios based on winter data. Earlier analyses like [P] have in any case shown that rather simple ratios can reflect the phytoplankton requirement for nutrients. However, from the management point of view, it is important to be aware that prevailing N-limitation will be more difficult to detect from nutrient data than prevailing P-limitation [30].

Cyanobacteria blooms and N₂ fixation have been closely linked to eutrophication of the Baltic Sea. Enhanced internal loading of phosphorus and the removal of dissolved inorganic nitrogen leads to lower nitrogen to phosphorus ratios, which are one of the main factors promoting nitrogen-fixing cyanobacteria blooms [39]. Vahtera et al. described the coupled processes inducing internal loading, nitrogen removal, and the prevalence of nitrogen-fixing cyanobacteria as a potentially self-sustaining “vicious circle” [39]. They concluded that to effectively reduce cyanobacteria blooms and overall signs of eutrophication, reductions in both nitrogen and phosphorus external loads appear essential.

But on the other hand, at least the results of ecosystem modeling by Neumann et al. [40] predicted that, in a medium-term perspective, a proportional reduction of nitrogen and phosphorus at the same time does not have the desired reducing effect on phytoplankton development in the open sea. In summer, the shortage of nitrogen has inhibitory effects on all phytoplankton groups, with the exception of cyanobacteria. Neumann et al. concluded that a possible solution can be an early and increased reduction in phosphorus load [39]. All phytoplankton groups are limited by phosphorus availability. An increased reduction of phosphorus can be supported by the fact that the increase in phosphorus load was twice compared with nitrogen in the last century [41].

5. Conclusions

The wastewater treatment implemented has led to positive developments in water quality in the marine area off Turku, even though the goal of good status is still a ways off. It also seems evident that, in the inner archipelago zone of the Archipelago Sea, the requirements for good ecological status would be too strict and practically impossible to achieve through any load reduction measures. Based on the research, it appears that the biomass of nitrogen-fixing cyanobacteria has increased in the northern Airisto over the past 15 years. This is driven by a decrease in external nitrogen loading and an increase in internal phosphorus loading. All phytoplankton groups are limited by phosphorus availability. Therefore, further reduction in all external phosphorus loading can be recommended. In the long term, this may also lead to a decrease in internal loading.