Introduction

Anthropogenic pressures on aquatic systems have been more effective due to increased human population and industrialization, especially in the last century. Human-induced eutrophication and the effects of climate change are two major problems for coastal lagoons (Cloern 2001; Lloret et al. 2008) as in other freshwater lakes (Dugan et al. 2017; Woolway et al. 2020). Increased temperature and decreased precipitation due to climate change cause alterations such as salinity, lake water level, lake water temperature, and biota, especially in lagoons of arid and semi-arid regions (Angus 2017; Shalby et al. 2021). Particularly salinization is becoming a comprehensive environmental problem due to its ecological and economic effects (Rochette et al. 2010; Odountan et al. 2019). Besides climate change, deicing salts usage for safe driving during winter (Corsi et al. 2010), mineral fertilizers used for agricultural productivity (Barzegar et al. 2017), industrialization, and urbanization (Dugan et al. 2017) have increased the salinity level of aquatic systems.

Phytoplankton are effective indicators used in the evaluation of water quality, trophic status, and ecological conditions of coastal ecosystems as well as other ecosystems (e.g. lakes, open sea) due to their rapid response to environmental changes (Lugoli et al. 2012; Garmendia et al. 2013; Allende et al. 2019). Previous studies have shown that the response of phytoplankton to changing salinity is very diverse (Lancelot and Muylaert 2011; Wang et al. 2018). Depending on the interspecific differences in salinity tolerances, the change in salinity affects the phytoplankton diversity (Kirst 1989; Zhong et al. 2016). It has been reported that some species can survive with a wide range of salinity levels (Shikata et al. 2008; Bergesch et al. 2009). On the other hand, most species cannot tolerate increased concentrations or rapid changes in salinity, and consequently both species richness and diversity index decline with increasing salinity (Hammer 1986; Flöder and Burns 2004; Wang et al. 2018). Species number and diversity were detected as low in brackish waters such as lagoons compared with freshwater and marine water systems due to their highly variable salinity concentrations (Flöder and Burns 2004; Flöder et al. 2010). In addition, previous studies have suggested that taxonomic groups of phytoplankton shifted from Cyanobacteria and Chlorophyta to Bacillariophyta (diatoms) and Miozoa (Dinoflagellate) depending on salinity increase (Kies 1997; Muylaert and Sabbe 1999; Li et al. 2021). Although many studies have shown that salinity reduces phytoplankton biomass due to decreased growth rate (Kirst 1989; Evagelopoulos et al. 2007; Redden and Rukminasari 2008; Hernando et al. 2015), some other studies have shown that the high nutrient levels attenuated the inhibitory effect of salinity on phytoplankton biomass (Zhong et al. 2016; Yue et al. 2019).

Since the effect of the salinity gradient is more pronounced in estuarine systems and lagoons, they have been ideal systems for elucidating the effect of salinity on phytoplankton dynamics (Telesh and Khlebovich 2010; Zainol and Akhir 2019). Increased, decreased or absent freshwater inflow, increased evaporation, and seawater entrance by tidal flow during different hydrological periods cause significant spatial and temporal variability in salinity values (Zainol and Akhir 2019; Srichandan et al. 2019; Draredja et al. 2019). Therefore, the temporal and spatial distribution in phytoplankton community structure was observed as a result of changes in environmental parameters, and more dominantly changes in salinity (Haraguchi et al. 2015; Srichandan et al. 2019; Tarafdar et al. 2021). In most temperate and tropical lagoons, it was observed that higher freshwater inflow reduces the salinity concentration and increases the nutrients, and consequently, triggers the biomass increase (Froneman 2004; Gobler et al. 2005; Paerl et al. 2018). On the other hand, other studies have also stated biomass increases due to low or absent flow rates and disconnection with the sea which constitute higher residence time (Ortega-Cisneros et al. 2014).

Lagoons are located mostly in delta regions (Kızılırmak, Yeşilırmak, Büyük Menderes, Küçük Menderes, Gediz, Göksu deltas) in Türkiye. Although there are many lagoons in the Marmara Basin, they are not located on the same delta (Atalay et al. 2015). In recent years, the seasonal distribution of species composition, biodiversity, abundance, and dominant species was investigated for detecting water quality in Küçükçekmece (Yilmaz 2015; Yilmaz et al. 2021a)yükçekmece (Temel 2002; Aktan et al. 2009; Gülecal and Temel 2014; Yilmaz 2019) and Terkos (Yilmaz and Gülecal 2012; Yilmaz et al. 2021b) lagoons, their inlet streams or outlet channels in separate times. Many studies have been conducted on the effects of seasonality, hydrodynamic regime, climate change, and anthropogenic disturbance on the salinity and the phytoplankton community structure of a single coastal lagoon (Jacquet et al. 2006; Nche-Fambo et al. 2015; Caroppo et al. 2018; Derolez et al. 2020). However, there is a lack of studies that collectively evaluate the change in the salinity levels and its effect on phytoplankton of lagoons that are geographically close to each other and have the same trophic condition. The present study is the first study considering the 6 eutrophic lagoons in the Marmara Basin to determine the effect of salinity gradients among lagoons regarding the phytoplankton species richness, diversity, biovolume, the total number of species, taxonomic groups, and dominant species. In the present study: (1) We hypothesize that phytoplankton species richness, diversity, biovolume, the total number of species, and species composition would change in accordance with the salinity gradient among the 6 lagoons. (2) We further hypothesized that besides salinity, other environmental factors would also be effective in the spatial distribution of phytoplankton biovolume and dominant species. (3) Moreover, we hypothesize that as a result of seasonal changes in environmental parameters such as salinity, temperature, and nutrients, seasonal changes would occur in phytoplankton biovolume, species richness, diversity, and dominant species.

Materials and methods

Marmara Basin and studied lakes

The Marmara Basin with a catchment area of 24,100 km2 covers the precipitation areas of rivers which mostly pour into the Marmara Sea. Its catchment area comprises 3% of the Türkiye. It has boundaries with the Meriç-Ergene, Sakarya, Susurluk, and North Aegean basins. The basin includes the cities such as İstanbul, Kocaeli, and Bursa with the highest population density and industrialization in Türkiye (Özhan 2004). Therefore, domestic, and industrial wastewater have polluted the aquatic environments in the basin (Ergul and Karademir 2020). Due to the large population in İstanbul, the demand for usable water is increasing every year, and consequently, forces the existing water resources (ISKI 2023).

The Marmara Basin has a mild oceanic climate with rainy/snowy days in winter, but a long dry period in summer. The average precipitation is between 586 and 768 mm, and most rainfall occurs in the winter (Serengil et al. 2007). In the basin, during the winter period, the average temperatures were 5–6 °C while in the summer period, temperatures of 23–25 °C were reported (Directorate General of Water Management 2016).

Although there are natural lakes in the basin, many reservoirs were constructed due to the increasing water demand. There are also lagoons in the basin. The 6 eutrophic lagoons of Marmara Basin [Terkos Lagoon (TER), Büyük Çekmece Lagoon (BCK), Dalyan Lagoon (DAL), Hersek Lagoon (HER), Küçükçekmece Lagoon (KCK), and Mert Lagoon (MER)] were studied in the fall and spring of 2020 and 2021 (Fig. 1) (Table 1).

Fig. 1
figure 1

Location of sampling lagoons in the Marmara Basin (The full names of the lagoons were given in Table 1)

Full size image
Table 1 The general features of the 6 lagoons in the Marmara Basin (zmax denotes the measured minimum and maximum depths at the deepest station of each lake during the studied period)
Full size table

Terkos Lagoon (TER) which is situated 50 km northwest of Istanbul (Yilmaz and Gülecal 2012), was a natural lagoon until 1881. However, with the construction of a barrier between the lagoon and the Black Sea, it gained a freshwater status, and it has been used for the drinking water supply of Istanbul. It is fed by a lot of streams including Istranca, Sivasköy, and Çiftlikköy (Oğuz 1995). Büyük Çekmece Lagoon (BCK) has an important place among the surface waters within the province of Istanbul in terms of its size. It is located on the European side of Istanbul, near Çatalca (Yilmaz 2019). It was a natural lagoon due to its formation and discharges of its waters into the Sea of Marmara. However, due to the construction of a barrier at the connection point with the Sea of Marmara in 1985 to meet the water needs of Istanbul city, it turned into a freshwater lake over time (Özuluğ 1999). It is mainly fed by the Karasu Stream (Özuluğ 1999). Mert Lagoon (MER) is located in İğneada Longoz Forests on the Black Sea coast. This area has been announced as a Natural Site and National Park. There are three lagoons in the area which are fed by groundwaters and streams, and Mert Lagoon is the biggest one. This lagoon is separated from the Black Sea in the east by sandy dunes. However, with the rising of the waters, it sometimes merges with the sea (Altınsaçlı 2001; Güher 2003; Camur-Elipek et al. 2015). Küçükçekmece Lagoon (KCK) is located 15 km west of Istanbul, and the connection of the lagoon with the Sea of Marmara is provided by the Menekşe Stream, which is 1.5 m deep and 1 km long (Okumuş 2007). Due to the connection channel, lagoon water is saline character. The lagoon is fed by Nakkaşdere, Sazlıdere, and Ispartakule streams (Demirci et al. 2006). Due to the intense urbanization and industrialization around the lagoon, it is affected by different pollutants such as domestic and industrial wastes (Albay et al. 2005). Dalyan Lagoon (DAL) is situated on the Meriç River Delta (Enez district, Edirne) which is one of the nationally important wetlands for Türkiye. It is a natural habitat for many bird species. This lagoon is temporarily connected to the Aegean Sea with two narrow channels, therefore, the salinity of the water varies according to the seasons (Sezgin 2015; Senel et al. 2020). Hersek Lagoon (HER) which is separated from the sea with a sandy ridge, is located near to Gulf of Izmit in the eastern part of the Sea of Marmara (Dalkıran and Baki 2011). This sandy ridge has been reinforced with a concrete dike, but the grooves on the dike allow continuous water exchange. The Sea of Marmara, rainfall, and runoff are the main water resources of the lagoon (Uzun 2014). However, domestic, and industrial pollution due to Izmit Bay affected the water quality of the lagoon (Tolun et al. 2008).

Analysis of environmental variables

Three different depths were determined as surface (10 cm below), middle, and near the bottom for sampling and measuring physical and chemical variables, and the mean values of these three depths were used in the analyses. Hach-Lange HQ 40D water quality instrument was used for measurement in situ of electrical conductivity (EC), salinity (SAL), pH, dissolved oxygen (DO), and water temperature (T). In the laboratory of the Scientific and Technological Research Council of Turkey, Marmara Research Center (TUBİTAK-MAM), it was done measurements of the orthophosphate (PO4-P), total phosphorus (TP), nitrate-nitrogen (NO3-N), total nitrogen (TN), chemical oxygen demand (COD) concentrations and alkalinity (ALK) using standard methods. Water transparency was measured using a Secchi disk. Since sampling was conducted in September 2020 and April 2021 when the lagoons were in the mixing period, the mixing layer depth (zmix) was determined as the average depth of the studied lagoons (Naselli-Flores and Barone 2003). The light availability was estimated with the mixed layer to the euphotic zone (zmix/zeu) ratio (Jensen et al. 1994). Chlorophyll-a concentrations were measured with methanol extraction (Youngman 1978) to determine the trophic index of lagoons (Carlson and Simpson 1996). At the same time, Secchi disk depth and TP values were used to calculate the trophic index.

Phytoplankton analysis

A total of 28 samples were collected for phytoplankton analysis in September (fall) 2020 and April (spring) 2021. Depending on the surface area of each lagoon, two or three stations were chosen for sampling (Directorate General of Water Management 2015). Attention was taken to ensure that one of the stations was at the deepest point of the lagoon. The euphotic depth (zeu) was calculated as 2.5 times the Secchi depth (Cole 1994). Composite sampling was conducted according to the zeu using an integrated sampling tube. Lugol’s and formaldehyde solution was used for the fixation of phytoplankton. Diatoms were identified on permanent slides while other phytoplankton were identified from temporary preparations using an Olympus BX51 microscope (×400, ×600, ×1000) (European Committee for Standardization 2004). Cell counting was performed with an Olympus IX81 inverted microscope using standard methods (Utermöhl 1958). The average of three counting replicates was used to determine the final abundance of each species. Standard identification books and keys such as Trégouboff and Rose (1957), Hendey (1964), Sournia (1986), Round et al. (1990), Delgado and Fortuno (1991), Sims (1996), Throndsen (1997), John et al. (2003), Kramer and Lange-Bertalot (1986, 1991a, b, 1999), Lange-Bertalot (2001), Krammer (2000, 2002, 2003), Lange-Bertalot et al. (2017), Komarek and Anagnostidis (2008), Huber-Pestalozzi (1941, 1950, 1961, 1962, 1969, 1972, 1975, 1982, 1983) were used for the identification of algal species. The validity of species names and their habitat types were checked on the Algaebase website according to Guiry and Guiry (2023). Phytoplankton abundance was calculated from biovolume estimations with the number of cells and cell size measurements using geometric formulas (Sun and Liu 2003). At least 20 individuals were measured for cell size determination of each species in all samples (Brierley et al. 2007).

Data analysis

For the statistical analyses of phytoplankton and environmental variables, a total of 28 samples in 6 coastal lagoons were considered. The diversity index (H’) for phytoplankton was calculated according to Shannon and Weaver (1963). Environmental variables were logarithmically transformed except for pH. An analysis of variance (one-way ANOVA) test was applied to data for determining the statistical differences in environmental variables, species richness, diversity, and biovolume among the lagoons and seasons using SPSS 20.0 software. Spearman correlations between the environmental variables and the species richness, diversity, and biovolume were also determined using the SPSS 20.0 software. A linear regression model was also performed in SPSS 20.0 software between the log (salinity) and the log (species richness), log (diversity), and log (biovolume). Species that have a relative abundance of biovolume greater than 20% in one station of the specific lagoon were accepted as dominant species. Optimum levels of phytoplankton dominant species for salinity were calculated using the weighted averaging regression model. Cluster analysis was performed to detect the similarity of phytoplankton species among 6 coastal lagoons by using PAST 4.03 software. Canonical correspondence analysis (CCA) was carried out on the log-normal transformed abundance data using CANOCO software (ter Braak and Smilauer 2002) to reveal the relations between the biovolume (%) of the dominant species, sampling lagoons, and environmental variables since the response data have a gradient 4.56 units long. The statistical significance of the environmental predictor variables was assessed by 999 restricted Monte Carlo permutations. To analyze the relationship between the biovolume (%) of the dominant species and 12 environmental variables (T, pH, DO, SAL, EC, TN, NO3-N, TP, PO4-P, zmix/zeu, COD, ALK), we performed a CCA using biovolume (%) values of the 27 dominant species in both of the lagoons. CCA was performed, initially on the whole environmental and dominant species datasets. Forward selection indicated that 6 of the 12 environmental variables made a significant contribution to the variance in the biovolume (%) of the dominant species data.

Results

Environmental parameters

The minimum zmax values of the lagoons were measured in the fall of 2020, while maximum zmax values were measured in the spring of 2021 (Table 1). In Table 2, it was given the results of physical-chemical variables in 6 coastal lagoons of the Marmara Basin. The mean values of T were higher in fall 2020 (f = 319.86, p < 0.01). The mean values of DO (f = 42.10, p < 0.01) and NO3-N were higher in spring 2021 (f = 5.65, p < 0.01). However, T, DO, and NO3-N values were not significantly different among the lagoons. Also, some of the parameters such as pH, COD, TN, and zmix/zeu were not significantly different among the lagoons and the seasons. In DAL and HER lagoons, the mean EC and SAL values were higher (f = 31.89, f = 43.59, respectively, p < 0.01). The mean TP and PO4-P values were higher in the KCK lagoon (f = 32.76, f = 179.24, respectively, p < 0.01) as well as the mean EC and SAL values. The mean ALK values were different between MER and BCK (f = 3.51, p < 0.05). On the other hand, EC, SAL, TP, PO4-P, and ALK were not significantly different among the seasons.

Table 2 Mean values and standard deviations of environmental variables measured in 6 lagoons of the Marmara basin
Full size table

Phytoplankton

A total of 247 phytoplankton taxa were identified and the numbers of taxa in the TER, BCK, MER, KCK, DAL, and HER lagoons were 82, 63, 95, 31, 39, and 34 respectively. Among them, 199 freshwater, 31 brackish water, and 17 marine taxa were detected. The total percentages (%) of brackish and marine species of KCK, DAL, and HER lagoons were higher than in the other lagoons (Table 3). It was noticed clustering of phytoplankton assemblages of DAL and HER lagoons which are mesohaline, and TER and BCK lagoons which are freshwater characters (Fig. 2). Only species from Cryptophyta were detected in freshwaters, other taxonomic groups were observed in hyposaline and mesohaline lagoons, and their abundance (%) did not show a remarkable change compared to freshwater lagoons (Fig. 3a). However, Charophyta, Chlorophyta and Ochrophyta members did not found in the water greater than 36 ppt salinity concentrations. Relatively high SAL optima were observed for Chaetoceros sp., Chlamydomonas spp., Cylindrotheca closterium, Dunaliella obliqua, Gyrosigma sp., Gyrosigma wansbeckii, Mougeotia sp., Nitzschia acicularis, Prorocentrum micans, Protoperidinium sp., Pinnularia sp., and Stephanocyclus meneghinianus (Fig. 3b). Relatively low SAL optima were detected for Aulacoseira granulata, Aulacoseira subarctica, Gymnodinium catenatum, Gymnodinium inversum, Gyrosigma kuetzingii, Hymenomonas sp., Pediastrum duplex, Peridinium raciborskii var. palustre, Staurosira construens, and Trachelomonas manginii (Fig. 3b). The dominant taxa (> 20%) of the 6 coastal lagoons in fall 2020 and spring 2021 were summarized in Table 4.

Fig. 2
figure 2

A dendrogram of phytoplankton assemblages by cluster analysis in 6 coastal lagoons of Maramara Basin (The full names of the lagoons were given in Table 1)

Full size image
Fig. 3
figure 3

a Relative abundance (%) of phytoplankton taxonomic groups along the salinity gradient (b) Result of the weighted average model for salinity in the studied 6 coastal lagoons of Marmara Basin (Abbreviations of the species were given in Table 4)

Full size image
Table 3 The percentage (%) of habitat status of the identified species to the total number of species in each lagoon in the 6 coastal lagoons of Marmara Basin (The full names of the lagoons were given in Table 1)
Full size table
Table 4 The seasonal distribution, mean abundance (of stations) and their abbreviations of the dominant taxa in 6 coastal lagoons of Marmara Basin (The full names of the lagoons were given in Table 1)
Full size table

The species richness ranged between 4 and 33 during the studied periods in the 6 lagoons. Species richness was the highest in the MER lagoon (f = 12.77, p < 0.01) as shown in Fig. 4a. It was also higher in the TER lagoon. Species richness values were not significantly different in the fall of 2020 and spring of 2021 (Fig. 5). Species richness was negatively correlated with EC (r = -0.49, p < 0.01), and SAL (r = -0.47, p < 0.05). Moreover, it was positively correlated with diversity index (r = 0.71, p < 0.01) and biovolume (r = 0.47, p < 0.05). The Shannon diversity index ranged between 0.18 and 3.08 over the two periods. The diversity index was the highest in MER lagoon, while the lowest in KCK lagoon (f = 8.85, p < 0.01) (Fig. 4a). It was not significantly different among the seasons (Fig. 5). The minimum phytoplankton biovolume was recorded as 0.11 mm3 L−1, while the maximum was found as 27.89 mm3 L−1. Moreover, the mean phytoplankton biovolume was the highest in the KCK lagoon (7.40 mm3 L−1), while was the lowest in the HER lagoon (0.41 mm3 L−1). However, biovolume values were not significantly different among the lagoons (Fig. 4a). The biovolume was positively correlated with T (r = 0.52, p < 0.01), while negatively correlated with NO3-N (r = -0.41, p < 0.05). It was not significantly different among the seasons. However lower biovolume values were detected in spring 2021 of the lagoons except for MER lagoon (Fig. 5). The relationships between log (SAL) and log (Species Richness), log (Shannon Index), and log (Biovolume) were given in Fig. 4b. Regression results showed a higher negative relationship between log (SAL) and log (Species Richness) (r2= -0.13, p < 0.05).

Fig. 4
figure 4

a Distribution of Species Richness, Shannon Index (bits), and Biovolume (mm3/L ) along the 6 coastal lagoons of Marmara Basin (b) Relationships between log (Salinity) and log (Species Richness), log (Shannon Index), and log (Biovolume), respectively in the 6 coastal lagoons of Marmara Basin (The full names of the lagoons were given in Table 1)

Full size image
Fig. 5
figure 5

Seasonal distribution of Salinity, Species Richness, Shannon Index (bits), and Biovolume (mm3/L ) in the 6 coastal lagoons of the Marmara Basin (The full names of the lagoons were given in Table 1) (average values of the stations for each lagoon were used for determining Salinity, Species Richness, Shannon Index (bits), and Biovolume values)

Full size image

Phytoplankton and environmental parameters

The results of CCA using 6 environmental variables were given in Fig. 6. The dominant species–environmental correlations of CCA axes 1 and 2 are high, and the first two axes account for 94.8% of the variance in the dominant species–environmental relationships. In the positive part of the first axis, generally, DO was correlated with species such as Gomphonema sp., Gymnodinium catenatum, Gymnodinium inversum, Gyrosigma kuetzingii, Hymenomonas sp., Peridinium raciborskii var. palustre and spring period of TER and MER lagoons. NO3-N was correlated with species such as Chlamydomonas spp., Glenodinium alpestre, Pandorina morum, Protoperidinium sp., Stephanocyclus meneghinianus, and spring periods of the BCK, KCK, and HER lagoons. On the other hand, in the negative part of the first axis, ALK and SAL were correlated with species such as Cylindrotheca closterium, Gyrosigma wansbeckii, Nitzschia acicularis, Pinnularia sp., Prorocentrum micans, Trachelomonas manginii and fall periods of DAL, HER, and MER lagoons. Moreover, species such as Aulacoseira granulata, Aulacoseira subarctica, Cymatopleura elliptica, Pediastrum duplex, Staurosira construens were correlated with T and fall periods of TER and BCK lagoons. The last group which was found in both periods or correlated with both SAL and PO4-P consisted of species such as Dunaliella obliqua, Chaetoceros sp., Gyrosigma sp., and Mougeotia sp.

Fig. 6
figure 6

Ordination of the samples corresponding to the different lagoons, scores of phytoplankton biovolume by dominant species, and environmental variables, along the canonical correspondence analysis axes. Environmental variables: T: water temperature, DO: dissolved oxygen, SAL: salinity, PO4: orthophosphate, NO3: nitrate-nitrogen, ALK: alkalinity (red: Dalyan Lagoon, orange: Hersek Lagoon, purple: Mert Lagoon, blue: Küçükçekmece Lagoon, dark green: Terkos Lagoon, green: Büyükçekmece Lagoon)

Full size image

Discussion

Measured environmental parameters such as EC and main nutrients were similar to previous studies in DAL (Altinoluk-Mimiroglu and Camur-Elipek 2018), KCK (Yilmaz 2015), and MER (Altınsaçlı 2001) lagoons compared to our study. However, measured nutrients were slightly higher in previous studies of BCK (Gülecal and Temel 2014; Yilmaz 2019), and TER (Yilmaz and Gülecal 2012) lagoons. This may be related to the climatic factors which determine the water budget and therefore the chemical composition of water, or conservation strategies prepared to protect these drinking water areas in recent years.

In the lagoons, NO3-N values were generally higher in spring 2021. In this period, the increase in zmax values of the lagoons compared to fall 2020 indicated an increase in precipitation. Therefore, allochthonous inputs carried by streams from domestic and agricultural areas cause an increase in NO3-N values as observed in many coastal lagoons during the flooding period (Serrano et al. 2004; Li et al. 2014). Moreover, DO values were higher in spring 2021 when the precipitation started to increase. The higher DO values in spring 2021 might be related to river influx, and still low T values (Srichandan et al. 2015; Draredja et al. 2019; Akagha et al. 2020), while lower values in fall 2020 could be the reason for increased decomposition of organic matter due to higher phytoplankton biovolume (Domingos et al. 2012; Akagha et al. 2020). EC and SAL values in HER, DAL, and KCK lagoons were measured high due to their connection with the seawater by channels or surmountable dikes. In a lagoon that has a continuous or temporary connection with sea water, salinity values were reported higher than closed ones (Barnes et al. 2008; Navas-Parejo et al. 2020).

Phytoplankton biovolume values were not different significantly among lagoons, however, average values were lowest in HER lagoon where EC and SAL values were highest. Panigrahi et al. (2009) have observed the phytoplankton biomass decrease in seawater exchange areas of Chilika Lagoon (India) due to the increase in salinity values. On the other hand, the highest average biovolume in the KCK lagoon, where SAL values were also high, can be explained by the highest TP and PO4-P values in this lagoon. High nutrient values may have reduced the inhibitory effect of SAL on phytoplankton biovolume (Zhong et al. 2016; Yue et al. 2019). Phytoplankton biovolume was slightly higher in the fall of 2020 in most of the studied lagoons. As a result of higher temperature and higher residence time, biomass or biovolume increase was observed during the dry period of several coastal lagoons (Anandraj et al. 2008; Ortega-Cisneros et al. 2014; Srichandan et al. 2015; Tagliarolo and Scharler 2018). Only biovolume was lower in the fall 2020 of MER lagoon compared to the spring of 2021. The very high salinity values in this period compared to spring 2021 (~ 20 times high) might be effective in the reduction of the phytoplankton biovolume in this lagoon.

Species richness and diversity were detected as low in KCK and HER lagoons compared to other less saline lagoons. Moreover, KCK, DAL, and HER lagoons had the lowest recorded total numbers of species. Besides, a negative correlation was observed between SAL and species richness in the lagoons. These lagoons were affected by the wave movements and exposed to seawater at certain intervals, and consequently, the sudden increase in SAL values during these periods caused osmotic stress on the phytoplankton and triggered the decrease in species number (Kirst 1996). Flöder and Burns (2004) have also stated that intermittent salinity inputs into coastal ecosystems at weekly or monthly periods depending on the seawater inflow, will reduce the species richness compared to under constant freshwater or seawater conditions. Moreover, brackish water and marine species detected in these three lagoons were higher compared to other less saline lagoons. This finding shows that the distribution of both freshwaters, brackish, and marine species in the lagoons was affected by SAL gradients. Another piece of evidence showing the distribution of species affected by SAL was detected in the cluster analysis. According to this analysis, species composition was found as similar in DAL and HER lagoons which are mesohaline characters, while species composition was similar in TER and BCK lagoons which are freshwater characters. Many studies in coastal lagoons have already stated the important effect of SAL on the species composition of phytoplankton (Comin and Valiela 1993; López-Flores et al. 2006; Specchiulli et al. 2008). It has been stated in different studies that each species has a different SAL optima, and SAL affects the phytoplankton growth rate (Braarud 1951; Tyler and Seliger 1981). In our study, 12 of the 27 dominant species were determined to have high SAL optima and were distributed in more saline lagoons. In some previous studies, it has been stated that taxonomic groups of phytoplankton showed variation from Cyanobacteria and Chlorophyta to Bacillariophyta and Miozoa with salinity increase (Kies 1997; Muylaert and Sabbe 1999; Li et al. 2021). In our study, the individuals of 8 taxonomic groups contributed to the relative abundance of phytoplankton at different rates up to 36 ppt. However, only individuals belonging to the Cyanobacteria, Euglenozoa, Miozoa, and Bacillariophyta were detected over 36 ppt.

In the studied lagoons, the distribution of dominant species was affected by seasonality and was segregated into fall and spring periods. In the CCA diagram, the fall of 2020 period of more saline DAL, HER, and MER lagoons and high SAL optima species such as Cylindrotheca closterium, Gyrosigma wansbeckii, Nitzschia acicularis, Prorocentrum micans, and Pinnularia sp. were located near to SAL and ALK. It has been stated in different studies that these species have a high tolerance to SAL and were reported in brackish and marine environments (Marshall et al. 1981; Barillé et al. 2003; Van Bergeijk et al. 2003; Fatah et al. 2022; Guiry and Guiry 2023). Since the salinity values in the MER lagoon were much higher in fall 2020 than in spring 2021, the salinity-tolerant G. wansbeckii was detected as the dominant species in this environment. Species such as C. closterium, N. acicularis, and P. micans were reported in highly alkaline environments (Gosselain et al. 1994; El Gammal et al. 2017). The presence of low SAL optima species (Trachelomonas manginii) in this group can be explained by the high correlation of this species with ALK. This freshwater euglenophyte was also detected for the first time after the tropical cyclone in the outer channel area in the Chilika lagoon, where SAL values were measured high in previous studies (Srichandan and Rastogi 2020). This lagoon has also high alkalinity values (Srichandan et al. 2015). On the other hand, fall 2020 period of BCK and TER lagoons and low SAL optima species such as Aulacoseira granulata, A. subarctica, Cymatopleura elliptica, Pediastrum duplex, and Staurosira construens were correlated with T. These species were reported in many freshwater lakes during the early fall periods when the temperature is still high and mixing events increase (Padisák et al. 2009; Sevindik et al. 2017). The presence of benthic pennate diatoms such as C. elliptica and S. construens in the pelagic environment has also indicated wind-induced sediment resuspension (Padisák et al. 2009; Aubry et al. 2013; Pednekar et al. 2014).

In spring 2021, MER and TER lagoons, and low SAL optima species found in these lagoons (Gomphonema sp., Gymnodinium catenatum, Gymnodinium inversum, Gyrosigma kuetzingii, Hymenomonas sp., Peridinium raciborskii var. palustre) showed high correlation with DO. Due to the SAL decrease in MER lagoon, low SAL tolerant species have become dominant. On the other hand, DO values were much higher in these two lagoons (> 11 mg L−1) compared to other lagoons in the spring period. Ohtake et al. (1982) have attributed the high oxygen concentration detected in the spring to the vigorous photosynthetic activity of Gymnodinium spp. in the Nakanoumi lagoon, which was previously artificially freshened similar to TER lagoon to meet the irrigation need. In spring 2021, BCK, KCK, and HER lagoons and species such as Chlamydomonas spp., Glenodinium alpestre, Pandorina morum, Protoperidinium sp., and Stephanocyclus meneghinianus were correlated with NO3-N. These species have been detected in nitrate-rich eutrophic lakes (Happey-Wood 1976; Carli et al. 1994; Naselli-Flores and Barone 2000; Padisák et al. 2009; Zhang et al. 2013).

High SAL optima species such as Dunaliella obliqua, Chaetoceros sp., Gyrosigma sp., and Mougeotia sp. were separated from other groups in the CCA. In these species, Chaetoceros sp. and Mougeotia sp. were found in the fall of 2020 or spring of 2021 of KCK and DAL lagoons. They were correlated with PO4-P. Besides, PO4-P values in KCK were measured higher than in other lagoons. Chaetoceros species were reported in eutrophic coastal environments (Gotsis-Skretas and Friligos 1990; Annabi-Trabelsi et al. 2022), while Mougeotia species were detected in the eutrophic saline lake (Reati et al. 1996) with high PO4-P content. Dunaliella obliqua, which was detected in spring 2021 in HER lagoon, was indicated as a marine species (Guiry and Guiry 2023).

Conclusion

  1. 1)

    Although the phytoplankton biovolume did not change significantly due to SAL differences in 6 coastal lagoons, the average biovolume values were minimum in the most saline HER. Species richness and diversity were detected as low in more saline KCK and HER lagoons. Moreover, the total number of species was recorded as lowest in most saline KCK, DAL, and HER. In these lagoons, higher brackish and marine species were detected. Moreover, only individuals belonging to four taxonomic groups were detected over 36 ppt. In addition, species composition was found as similar in mesohaline lagoons, while species composition was similar in freshwater lagoons. SAL was also effective in the distribution of dominant species and species with higher SAL tolerance were found in most saline lagoons. These results show that the variation in SAL is the most significant factor in shaping the phytoplankton community structure among the lagoons.

  2. 2)

    Considering other environmental variables except for salinity, TP and PO4-P values affected the average biovolume in the KCK lagoon. Species distributed in a eutrophic environment with high PO4-P content were dominant in this lagoon. As a result of higher T, phytoplankton biovolume was slightly higher in most lagoons.

  3. 3)

    SAL values were not significantly different among the seasons. However, in the MER lagoon, high SAL values tended to decrease phytoplankton biovolume. The considerable decrease in SAL shifted the dominant species structure from high SAL optima species to low SAL optima species in this lagoon. Moreover, T, ALK, NO3-N, and DO were the main factors affecting the seasonal distribution of dominant species.

We can conclude that the changes in abiotic factors, particularly salinity, affect the spatial and temporal algal communities in the lagoons. Also, to better detect the effect of salinity on phytoplankton, these lagoons should be monitored with periodic studies. Therefore, the effects of climate change will be better understood.