Introduction

Transitional waters such as bays are often exposed to a wide range of stresses associated with dense human population inhabited in coastal regions (Cloern and Jassby 2008; O’Boyle et al. 2015). Nutrient derived from increased fertilizer use in farms and discharge from municipal waste water treatment plants enriches water in bays (Boesch 2002; O’Boyle et al. 2015). Furthermore, changes in hydrologic regimes caused by dam construction and water withdrawal cause pronounced changes in nutrient input and fluctuations of salinity in bays (Muylaert et al. 2009). These changes consequently induce a complex pattern in phytoplankton composition and biomass in bays (Conley 2000; Vitousek et al. 1997).

Despite strong seasonal fluctuations in available sunlight and temperature expected at temperate latitudes, gradients of salinity and nutrients (nitrogen and phosphorous compounds) are paramount in determining phytoplankton growth in bay ecosystems, as they exhibit considerable change over time and space (Ralston and Stacey 2005). Depending on the temporal fluctuation of freshwater inflow, salinity as well as sediment and nutrient loading are affected in a bay system (Bruesewitz et al. 2013; Islam et al. 2014; Palmer et al. 2011; Ringuet and Mackenzie 2005). Parab et al. (2013) showed seasonal variation in freshwater discharge in an estuary resulted in variation in salinity accompanied by increases in nitrate concentration and water turbidity during the monsoon season. However, the salinity and nutrient fluctuations can also depend on whether bays are connected to or isolated from the sea (Pereira Coutinho et al. 2012). Likewise, the salinity and nutrient gradient can control the phytoplankton composition and succession because phytoplankton taxa have variable salinity growth optima and different requirements for resources. For example, nutrient enrichment influences phytoplankton dynamics and succession in Galveston Bay, facilitating diatom biomass accumulation in spring and summer, while it shifts the phytoplankton composition to cyanobacteria dominance in summer in Trinity Bay (Örnólfsdóttir et al. 2004). Similarly, diatoms were dominant phytoplankton in Jung Bay and they seemed more sensitive than dinoflagellates to the increase in nutrients (Hodgkiss and Lu 2004). Shifts in phytoplankton taxa in estuaries, also, correlate to salinity gradients, varying from a freshwater to marine-dominated communities (Quinlan and Phlips 2007). For example, diatoms dominated during periods of moderate to high freshwater inflows (low salinity) in winter/spring and were more abundant in the upper bay while summer/fall bloom in biomass of cyanobacteria occurred when inflow was low (more salinity and less turbulence) in Gulf of Mexico (Roelke et al. 2013; Dorado et al. 2015). To understand phytoplankton productivity, knowledge of phytoplankton composition and dynamics in relation to environmental conditions is prerequisite (Nazeer et al. 2018).

Phytoplankton assemblages are also known to be important for bays and estuaries ecosystem functioning (Cloern and Jassby 2010; Zingone et al. 2010). They are not only the most important primary producer supporting biochemical cycling, but they are also an indicator of aquatic pollution (Bellinger and Sigee 2010). Their short life spans enable quick responses to environmental changes (Ho et al. 2015; Liu 2008). Therefore, they provide an early warning of the environmental changes in transitional waters and indicate the health status of the ecosystems, which are often exposed to anthropogenic disturbances (O’Boyle et al. 2015).

Gorgan Bay is the largest bay (area of 400 km2) in the Caspian Sea and located in the south-eastern part of the sea, Iran. The bay provides nursery grounds for fish-supporting commercial fisheries and is a place for breeding and wintering of water birds (Bastami et al. 2012; Saghali et al. 2013; Taheri et al. 2012). During the last few decades, a significant increase in nutrient inputs due to the expansion of agriculture with the increased use of fertilizer and pesticide was observed in Gorgan Bay (Karbassi and Amirnezhad 2004; Roohi et al. 2010). At the same time, the wastewater from urban areas, Bandar-Torkaman and Bandar-Gaz, into the bay has exacerbated the nutrient over-enrichment (Shahryari et al. 2008). These anthropogenic stresses put Gorgan Bay at risk of eutrophication and may endanger fish and wildlife. However, the knowledge about phytoplankton assemblages in Gorgan Bay is almost absent. In the present study, spatial and temporal changes in phytoplankton composition in relation to environmental conditions and health status of Gorgan Bay are evaluated, based on monthly phytoplankton and physicochemical samples collected at multiple locations from March 2014 to September 2015. It was expected that the composition of the phytoplankton assemblage in Gorgan Bay strongly changes in relation to nutrients and salinity variations. Furthermore, it was hypothesized that the organic pollution Gorgan Bay experiences influences the phytoplankton, allowing genera which are tolerant to organic pollution to occur.

Materials and Methods

Site Description

Gorgan Bay is located along the Iranian coast of the Caspian Sea between 36° 48′ N, 53° 35′ E and 36° 55′ N, 54° 03′ E (Fig. 1). The bay is approximately 60 km long and 12 km wide, with a mean depth of 1.8 m (Taheri et al. 2012). It has restricted connections to the Caspian Sea through two channels: the Ashoradeh-Bandar Torkaman mouth in the northeastern corner with the approximate length of 3 km and width of 400 m and the Khozini waterway, which is an artificial channel located in 6 km to the west of the primary inlet with a width of approximately 200 m (Ghorbanzadeh Zaferani et al. 2017; Ranjbar and Hadjizadeh Zaker 2016).

Fig. 1
figure 1

Map of Gorgan Bay. Numbers indicate the sampling sites, and solid squares show the locations of mouths of rivers. The solid triangle shows the location of Bandar Torkaman Waste Water Treatment Plant

Full size image

The annual mean of 0.5 km3 water and 3.5 million ton sediment are discharged into the bay from ten rivers (Afshin 1994). The amount of water and its salinity in the bay are determined by water intrusion from the Caspian Sea, freshwater inflow (which is mainly from Gorgan-Rud with the annual mean outflow of 11.3 m3 s−1 in the north near the mouth of the bay and the Qaresoo River with the annual mean outflow of 0.8 m3 s−1 in the east), precipitation, and evaporation (Ghorbanzadeh Zaferani et al. 2017; Zonn et al. 2010; Jamshidi 2016; Kurdi et al. 2015). The rivers also carry nutrients from agricultural run-off and urban effluents into the bay (Bashari et al. 2014; Bastami et al. 2012; Nasrollahzadeh et al. 2008a, b; Ranjbar and Hadjizadeh Zaker 2016). Gorgan Bay is characterized as a low wave-energy system, for which physical processes are mostly controlled within the bay (Araghi et al. 2014; Bastami et al. 2012; Sharbaty 2012a; Taheri et al. 2012). Very little tide or wave is received from the Caspian Sea, and the water currents in the bay are produced by local winds (Ranjbar and Hadjizadeh Zaker 2016, 2018; Sharbaty 2012a, b). This current is weak (< 4 cm s−1) and mostly in the central part of the bay, traveling from the mouth of the bay toward inner parts (Sharbaty 2012a, b). There are currents along the coastal areas in the northern and southern coasts that flow from west to east (Sharbaty 2012a, b).

There are some seagrass assemblages (especially Roppia maritime) in shallow parts in the southeastern part of the bay where organic materials from the Qaresoo River potentially settle (Ghorbanzadeh Zaferani et al. 2016). Sediment type along the western coast is the coarsest consisting mostly of sand whereas the eastern and southern regions have finer sediments consisting more of silt. The deeper areas consist of clay (Lahijani et al. 2010).

Sampling

Sampling sites were selected to cover different regions of Gorgan Bay, which included the mouth of the bay and inner regions of the bay (Fig. 1). Sampling was carried out in two successive years (2014–2015). In the first period, a monthly sampling was conducted at 11 sites from March 2014 to February 2015. In the following period, sampling was conducted monthly at 6 of the original 11 sites (sites 5, 6, 7, 8, 10, and 11, which comprise the mouth of the bay, Fig. 1) from March to September 2015.

Temperature (°C), salinity (PSU), and pH were recorded using a portable water-quality multiprobe (HQ40 Hach, Hach Company, Colorado, USA), and nitrate (ppm) and orthophosphate (ppm) were measured with a spectrophotometer (Hach DR 2800, Hach Company, Colorado, USA). Water transparency (cm) was measured with a Secchi disc (Eaton et al. 1998). Chlorophyll-a (Chl-a, ppm) was measured using UV-visible spectrophotometry based on Environmental Sciences Section (ESS) standard method 150.1 (ESS 1991). Photosynthetically active radiation (PAR, Einstein m−2 day−1) was extracted from the satellite images, the Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua level 2 products (NASA Ocean Color 2014).

Phytoplankton was sampled with Van Dorn sampler (10 × 30 cm) and preserved in buffered formaldehyde with the final concentration of 2%. After a 10-day period of settling, samples were concentrated 33-fold with the sedimentation method (Utermöhl 1936). Phytoplankton abundance in a 1-ml subsample was then counted using a Sedgewick-Rafter cell using inverted microscopy (Nikon Eclipse TS100, Nikon Instruments Inc., New York, USA) (Sournia 1978). The volume of each cell was estimated using geometric shapes of cell morphologies and measures of appropriate linear dimensions (Vadrucci et al. 2007). Cellular volumes were then converted to cellular biomass (Hillebrand et al. 1999). Phytoplankton was identified following the methods in previous studies (Bagheri et al. 2014; Ganjian et al. 2010; Kideys et al. 2005; Nasrollahzadeh et al. 2014; Roohi et al. 2010). Species that were observed equal to or greater than 10% of the total sample biomass within any given sample was considered as dominant (Huszar and Caraco 1998).

Health Status

An algal species pollution index was calculated according to Palmer (1969). This index rates algal species on a scale from 1 (intolerant) to 6 (tolerant). The index was calculated by summing up the scores of the species present within the sample. The scores > 20 are consistent with high organic pollution, scores between 15 and 19 are considered probable evidence of organic pollution, and samples with pollution scores of < 10 are categorized as unpolluted water body (Jindal et al. 2014a). In this study, only dominant species (biomass ≥ 10% of total sample biomass) were included in the pollution score calculation.

Statistical Analysis

Environmental data were analyzed for temporal and spatial variations using ANOVA, where sampling location (west and east of the bay) and time (month or year) were used as factors (Nazeer et al. 2018). For these tests, environmental data were log10 transformed to normalize the skewness prior to the statistical test.

The phytoplankton assemblages were described in terms of mean monthly biomass, abundance and relative biomass of the species. The mean biomass and abundance were transformed taking log10 after adding 1 to all values to avoid obtaining negative infinity. Canonical correspondence analysis (CCA) was conducted on the subset of the transformed data that included only the phytoplankton taxa classified as dominant on one side and all environmental parameters on the other side (including Chl-a and PAR). Following the instruction in “vegan” R-package, species scores were scaled symmetrically with the square root of eigenvalues (Oksanen et al. 2015). This analysis maximizes the separation of taxa optima along canonical axes representing the primary environmental gradients (Dvoretsky and Dvoretsky 2017). All of the statistical analyses were carried out using statistics package R (R-Development Core Team 2017). The significant level of α = 0.05 was used throughout this study.

Results

Abiotic Parameters

Surface temperatures in the first survey ranged from 10.5 °C to 32.5°C, which was significantly different among sampling months (Table 1; Fig. 2a). It is consistent with the typical pattern observed in temperate seas. In the second sampling period, water temperature varied between 20.6 29.1 °C in spring (April–June) and 29.3–33.7 °C in summer (July–September). The maximum surface temperature was measured in August. According to the ANOVA, there was a significant difference in the mean surface temperature between the two sampling periods (Table 1).

Table 1 ANOVA on environmental data
Full size table
Fig. 2
figure 2

Spatiotemporal trend in mean monthly temperature (a), salinity (b), water transparency (c), pH (d), PAR (e), chlorophyll-a (f), nitrate (g), and orthophosphate (h) during the first (black and blue line) and second surveys (gray line) in Gorgan Bay. For the first year, sampling locations at mouth (black squares) and inner parts of the bay (blue circles) are separated. For the second year (6 months), only stations located at the mouth of the bay were sampled

Full size image

Surface salinity measured in the first sampling period changed temporally from 9.9 to 12.9 PSU (Fig. 2b). The salinity exhibited significantly higher values in the sites located in the mouth of Gorgan Bay (annual mean of 12 ± 0.7 PSU) in comparison with the inner (annual mean of 11 ± 0.6 PSU) (Table 1). The minimum and maximum salinities were observed in March and September, respectively. The surface salinity in the second period was significantly higher than the first sampling period (Table 1).

The minimum (25 ± 2.5 cm) and maximum (235 ± 65 cm) water transparencies were observed in July and February, respectively (Fig. 2c). Sampling month had a significant effect on mean water transparency (Table 1), but no significant difference was detected between sites in the mouth and inner parts of the bay (Table 1). Similarly, mean water transparency was not significantly difference between the two sampling periods (Table 1).

The values of pH were on the alkaline side in spring and summer (Fig. 2d). The pH showed significant differences among sampling months (Table 1). The location of sampling sites had significant effect on mean pH (Table 1). The sites in the mouth of the bay had higher mean pH.

PAR showed a seasonal pattern as the monthly mean PAR increased from 31.55 ± 0.69 Einstein m−2 day−1 in April to 58.2 ± 0.31 Einstein m−2 day−1 in July (Fig. 2e). The PAR showed significant differences among sampling month (Table 1), and the minimum PAR was observed in February. No significant difference in mean PAR values among locations was found (Table 1). The mean PAR value increased significantly in the second year in comparison with the first sampling period (Table 1).

Significant difference was detected in surface Chl-a concentrations among sampling months (Table 1). A biphasic pattern was observed in surface Chl-a concentration (Fig. 2f). The first strong peak was observed in September (sites at the mouth of the bay 0.037 ± 0.0051 ppm, inner parts 0.032 ± 0.0025 ppm) and the second peak was observed in December (sites at the mouth 0.019 ± 0.0009 ppm, inner parts 0.018 ± 0.001 ppm). The location had a significant effect on mean surface Chl-a concentration (Table 1) as the sampling sites at the mouth had higher Chl-a. Results showed sampling year had a significant effect (Table 1) as the mean Chl-a increased in the second sampling period.

A clear difference among sites was observed in nutrients, i.e., orthophosphate and nitrate (Fig. 2g, h). Sites in the mouth of the bay were enriched substantially with higher nitrate (sites at the mouth 1.38 ± 0.58 ppm, inner part 0.75 ± 0.43 ppm) and orthophosphate (sites in the mouth 0.30 ± 0.17 ppm, inner part 0.23 ± 0.16 ppm) (Table 1). Higher concentrations of nitrate and orthophosphate were observed in the second sampling period compared with the first (Table 1). The mean nitrate concentration was different among sampling month (Table 1).

Seasonal Changes in Phytoplankton Assemblages

Phytoplankton composition in Gorgan Bay included 51 genera in which 16 belonged to Bacillariophyta, which on average comprised ~ 39% of the phytoplankton abundance and ~ 59% of the biomass. Other phytoplankton were represented by the phyla Pyrrophyta that consisted of 5 genera (~ 16% abundance, ~ 24% biomass), Cyanophyta that consisted of 15 genera (~ 23% abundance, ~ 13% biomass), Chlorophyta that consisted of 7 genera (~ 8% abundance, ~ 3% biomass), Chrysophyta that consisted of 5 genera (~ 4% abundance, ~ 1% biomass), and Euglenophyta that consisted of 3 genera (~ 11% abundance, ~ 0.4% biomass).

The phytoplankton data revealed that Bacillariophyta showed no strong seasonal difference in biomass, ranging from ~ 33 to ~ 98% of the total phytoplankton biomass at the mouth and inner parts of the bay, respectively, in the spring (April to June) to ~58% and ~66% of the total phytoplankton biomass at the mouth and inner parts of bay, respectively, in autumn (October to December), but they slightly decreased in winter (January to March, ~21% of the total phytoplankton biomass). More than 90% of Bacillariophyta biomass during the spring were represented by two genera: Gyrosigma sp. (mean of 2.69 × 106 cells m−3, 129 mg m−3 at the mouth of the bay and 14.0 × 106 cells m−3, 918 mg m−3 in the inner parts), and Pleurosigma sp. (mean of 3.70 × 106 cells m−3, 175 mg m−3 at the mouth of bay, Fig. 3a). Skeletonema sp. was observed as one of the dominant genera at the mouth of the bay with an increase in biomass during the winter (mean of 1.80 × 106 cells m−3, 75.6 mg m−3). While at the inner parts of the bay, Chaetoceros sp. increased in winter (mean of 9.80 × 106 cells m−3, 13.8 mg m−3, Fig. 3a).

Fig. 3
figure 3figure 3

Mean abundance (line) and biomass (column) of phytoplankton genera that represented more than 90% of the biomass for that particular phylum in a Bacillariophyta, b Pyrrophyta, and c Cyanophyta. The values in percent (%) represent the corresponding relative biomass for the genus within the phylum. Mean abundance (line) and biomass (column) of phytoplankton genera that represented more than 90% of the biomass for that particular phylum in d Chlorophyta, e Chrysophyta, and f Euglenophyta. The values in percent (%) represent the corresponding relative biomass for the genus within the phylum

Full size image

At the mouth of the bay, phylum Pyrrophyta comprised 52.3% (spring) and 37.3% (autumn) of the total phytoplankton biomass. In the inner regions of the bay, their biomass as a percent of the total phytoplankton biomass showed an increase during winter (22.5%). At the mouth of the bay, Prorocentrum sp. was the dominant Pyrrophyta in terms of biomass from spring (81.5%) to autumn (90.1%). The maximum biomass of Prorocentrum sp. was observed in spring (mean of 21.2 × 106 cells m−3, 466 mg m−3) followed by winter (mean of 7.80 × 106 cells m−3, 171 mg m−3). Exuviella sp. was a dominant Pyrrophyta at other sites, increasing in the winter and comprising 83.8% of Pyrrophyta biomass (mean of 6.30 × 106 cells m−3, 177 mg m−3) (Fig. 3b).

The biomass as a percent of the total phytoplankton biomass of phylum Cyanophyta varied seasonally from 1.5% (winter) to 24.4% (summer) at the mouth of bay and from 0.2% (winter) to 35.2% (autumn) at the other sites. Microcystis sp. occurred only at the sites located at the mouth of bay and was dominant in summer with a mean of 18.0 × 106 cells m−3 and biomass of 397 mg m−3. Anabaena sp. comprised approximately 57.5% (at the mouth of bay) and 45.5% (other sites) of the biomass in autumn and spring, respectively. Anabaenopsis sp. supported 56.3% of the total biomass in sites located at inner parts of the bay during autumn with a mean of 11.8 × 106 cells m−3 and 11.8 mg m−3 biomass (Fig. 3c).

The maximum biomass as a percent of the total phytoplankton biomass of phylum Chlorophyta was found in spring (4.2%). Among Chlorophyta, Chlorella sp. was the only genera observed as dominant at sites in the mouth of the bay. It made up 90% of the biomass within its phylum in spring, summer, and fall but decreased in winter (27.9%) when Ankistrodesmus sp. increased in biomass and comprised 53.6% of the Chlorophyta (3.10 × 106 cell m−3 and 3.73 mg m−3). The seasonal mean biomass of Chlorella sp. decreased from 46.1 mg m−3 (11.2 × 106 cell m−3) in spring to 66.9 mg m−3 (10.2 × 106 cell m−3) in summer to 2.72 mg m−3 (0.38 × 106 cell m−3) in autumn (Fig. 3d).

Chrysophyta increased during the winter. Melosira sp. and Synedra sp. comprised 100% of the Chrysophyta biomass in winter (Fig. 3d). The mean biomass and abundance were 3.70 mg m−3 and 3.70 × 106 cells m−3 for Melosira sp., and 13.2 mg m−3 and 2.20 × 106 cells m−3 for Synedra sp. (Fig. 3e). At the inner part of the bay, Dinobryon sp. was observed during summer (93.3% of Chrysophyte biomass, 6.45 mg m−3 and 0.40 × 106 cells m−3).

Euglena sp. is the dominant genus of Euglenophyta during the whole study. Total biomass of Euglena sp. was higher near the mouth of bay with the maximum mean of 25.2 × 106 cell m−3 and 10.1 mg m−3 near the mouth of bay and 16.4 × 106 cell m−3 and 6.56 mg m−3 at the other sites during summer. The contribution of Euglenophyta to the total biomass of the phytoplankton assemblage varied from 0.01% in winter to 2.3% in autumn (Fig. 3e).

Canonical Correspondence Analysis of Phytoplankton and Environmental Variables

To assess the temporal and spatial distribution of phytoplankton assemblages in relation to environmental variables in Gorgan Bay, the data were analyzed using canonical correspondence analysis in two ways. The first subset of data explored temporal patterns by analyzing the data from six locations (sites 5, 6, 7, 8, 10, and 11 in Fig. 1) during 18 monthly sampling (first subset) because not all sampling sites were used in the second sampling period. For the analysis of spatial distribution, the second subset, the data from all 11 sampling locations sampled in the first sampling period (second subset) was analyzed (Fig. 1). These 11 sites are scattered across the whole bay.

According to CCA of the first subset (Fig. 4), the eigenvalues associated with first two canonical axes were much greater than those of the other axes, accounting for 64.5% of the total variability. Bacillariophyta showed diverse composition in response to environmental parameters. Some of the dominant Bacillariophyta such as Gyrosigma sp., Pleurosigma sp., and Skeletonema sp. showed no clear relationship with a specific environmental parameter. The Gyrosigma sp. was mostly influenced by observation in autumn. Whereas, Pleurosigma sp. and Skeletonema sp. were affected by summer and spring, respectively (Fig. 4b). The biomass of Rhizosolenia sp. and Stephanodiscus sp. were positively correlated with orthophosphate concentration and water transparency (Fig. 4a).

Fig. 4
figure 4

Canonical correspondence analysis for the Gorgan Bay with a species and environmental conditions weightings and b the observation weightings for the first two canonical components represented 75.47% of the total variability in the data. The keywords for parameter biplot include the following: Temp, temperature; Sal, salinity; PAR, photosynthetically active radiation; CHL, chlorophyll-a; NO3, nitrate; PO4, orthophosphate; Visib, water transparency; Anab, Anabaena sp.; Anabop, Anabaenopsis sp.; Chlor, Chlorella sp.; Coscin, Coscinodiscus sp.; Exuv, Exuviella sp.; Gyros, Gyrosigma sp.; Microc, Microcystis sp.; Navi, Navicula sp.; Nytzs, Nytzschia sp.; Perid, Peridinium sp.; Proroc, Prorocentrum sp.; Rhizos, Rhizosolenia sp.; Skelet, Skeletonema sp.; Stephan, Stephanodiscus sp.; Syned, Synedra sp.; Pleuro, Pleurosigma sp.; Th-nema, Thalassionema sp. The key to observation times and locations follows a number code, where the first digit represents the sampling sites and the second pair of digits represents the month of sampling. The shapes represent the following: empty square, spring (April–June); empty circle, summer (July–September), filled square, autumn (October–November), and filled circle, winter (January–March)

Full size image

The other relationship inferred from the CCA is that the biomass of the Pyrrophyta’s dominant genera such as Exuviella sp., Peridinium sp., and Prorocentrum sp. (Fig. 4a) was inversely related to temperature, PAR, chlorophyll-a, pH, and salinity. Pyrrophyta had a positive association with the orthophosphate concentration and water transparency, which showed a significant increase during autumn and winter (Fig. 2c, h). This relationship is shown on CC1 and was most influenced by observations from March and early April (Fig. 4b).

The biomass of Cyanophyta genus, Microsystis sp., was related to temperature and PAR (Fig. 4a). The high temperature and PAR were apparent during summer and peaked in August and July (Fig. 2a, e). Anabaenopsis sp. showed relationship with salinity, and Anabaena sp. was weakly correlated with nitrate. The salinity, PH, PAR, and temperature exhibited strong correlation with Chlorophyll-a. The relationship is depicted in the first CCA axis, which explained 41.3% the total variability and was mostly influenced by observations in spring and summer (Fig. 4b).

A similar relationship was also found for Chlorella sp. with temperature, pH, and PAR (Fig. 4a) and was observed mostly in spring and summer (Fig. 4b).

Among the Chrysophyta, Syndera sp. (phylum Chrysophyta) thrived when the orthophosphate increased and was mostly influenced by the autumn and winter observations (Fig. 4a, b).

The spatial distribution of phytoplankton assemblages in Gorgan Bay is shown in Fig. 5. CCA1 and CCA2 explained 33.99 and 23.37% of the total variation, respectively. The results revealed that the sampling site near the mouth of the bay (sites 5, 6, 7, 8, 10, and 11), scattered in the bottom side of CCA biplot, showed a positive correlation with temperature, salinity, chlorophyll-a, nitrate, and orthophosphate concentrations. Some phytoplankton genera such as Microcystis sp., Euglena sp., Cymbella sp., Suriella sp., and Chlorella sp. flourished in this part of the bay in association with high temperature, salinity, chlorophyll-a, and nitrate concentration (bottom left, Fig. 5a). Peridinium sp., Prorocentrum sp., and Stephanodiscus sp. were dominant in terms of biomass when the concentration of orthophosphate increased (bottom right, Fig. 5b).

Fig. 5
figure 5

Canonical correspondence analysis for the Gorgan Bay with a species and environmental conditions weightings and b the observation weightings for the first two canonical components represented 57.27% of the total variability in the data. The keywords for parameter biplot include the following: Temp, temperature; Sal, salinity; PAR, photosynthetically active radiation; CHL, chlorophyll-a; NO3, nitrate; PO4, phosphate; Visib, water transparency; Anab, Anabaena sp.; Anabop, Anabaenopsis sp.; Chlor, Chlorella sp.; Coscin, Coscinodiscus sp.; Cyclo, Cyclotella sp.; Cymb, Cymbella sp.; Cocco, Cocconeis sp.; Eug, Euglena sp.; Exuv, Exuviella sp.; Gymno, Gymnodinium sp.; Gyros, Gyrosigma sp.; Microc, Microcystis sp.; Navi, Navicula sp.; Nytzs, Nytzschia sp.; Perid, Peridinium sp.; Proroc, Prorocentrum sp.; Rhizos, Rhizosolenia sp.; Skelet, Skeletonema sp.; Stephan, Stephanodiscus sp.; Suri, Suriella sp.; Syned, Synedra sp. The key to observation times and locations follows a number code, where the first digit represents the sampling sites and the second pair of digits represents the month of sampling. The shapes represent sampling site: empty square, site 11; empty circle, site 10; empty triangle, site 6; empty diamond, site 7, inverted empty triangle, site 8, squared plus sign, site 5; error mark, site 4; asterisk, site 9; filled square, site 1; filled circle, site 2, and filled triangle, site 3

Full size image

The sampling sites in inner parts of the bay (sites 1, 2, 3, 4, and 9) were found in the top right of CCA biplot (Fig. 5b) and negatively correlated with orthophosphate, temperature, salinity, chlorophyll-a, and nitrate concentration. Exuviella sp., Gymnodinium sp., Rhizosolenia sp., and Skeletonema sp. were the species that had high biomass in this part of the bay (Fig. 5a).

The biomass of Exuviella sp., Skeletonema sp., Rhizosolenia sp., and Gymnodinium sp. (top right, Fig. 5a) were mostly in sites 1, 2, 3, 4, and 9. They inversely related to salinity, chlorophyll-a, nitrate concentration, pH, and temperature (Fig. 5a).

Health Status

The indicator phytoplankton taxa and corresponding pollution scores are shown in Table 2. Of the dominant species present in Gorgan Bay, Euglena sp. is the most tolerant taxa and several species such as Nitzschia sp., Navicula sp., Cymbella sp., and Chlorella sp. were tolerant to eutrophic conditions (Table 2). The Palmer pollution index (PI) ranged from 0 (western Gorgan Bay) to 24 (eastern Gorgan Bay) calculated for observation in January and July, respectively (Fig. 6). The PI values were lower in the inner Gorgan Bay throughout the year and are consistent with unpolluted water body. In contrast, the PI values at the mouth of the bay were mostly higher than the threshold value for a probable polluted area in spring and summer, experiencing polluted condition in July, September, and October. The values decreased during late autumn and winter months (Fig. 6).

Table 2 Palmer’s pollution index (ASPI) in Gorgan Bay
Full size table
Fig. 6
figure 6

Spatiotemporal trend in Palmer pollution index (PI) at the mouth (black line) and inner areas (gray line) of Gorgan Bay. Dashed green lines show the threshold of ASPI for unpolluted (ASPI = 10), probable pollution (ASPT = 15), and polluted water bodies (ASPI = 20)

Full size image

Discussion

Bacillariophyta was dominant phytoplankton phylum and had the highest number of genera (51 genera including 16 diatoms) in Gorgan Bay. Diatoms are considered euryhaline and eurythermal phytoplankton growing in moderate to high nutrient levels (Huang et al. 2004; Hubble and Harper 2002). In this study, the temperature ranged from 10.5 to 32.5 °C and salinity varied from 9.9 to 12.9 PSU (Fig. 2a, b); according to the OCED classification, these conditions are classified as mesotrophic to eutrophic (OECD 1982). Literature survey revealed similar results in other lagoons and wetlands in the Caspian Sea (Ganjian et al. 2010; Roohi et al. 2010; Nasrollahzadeh et al. 2008a, b). Generally, the number of taxa increases from the west to east of the southern Caspian Sea (Kideys et al. 2008) as it composed of 27 genera (12 diatoms) in the western coastal water (Bagheri et al. 2011) and increased to 68 genera (20 diatoms) in Miankaleh wetland in the east (Masoudi et al. 2012).

Results showed that there were seasonal differences in the composition of phytoplankton community, and this temporal variation was associated with environmental conditions. For example, Cyanophyta biomass is significantly associated with nitrate, salinity, PAR, and temperature (Fig. 4). These parameters during summer were within the ranges of conditions considered suitable for Cyanophyta (Jindal et al. 2014b). Similar relationship between environmental conditions and biomass of Cyanophyta was reported by previous studies. Ahn et al. (2011) showed that water temperature and nitrogen compounds played important roles in bloom formation of Cyanophyta, and Chomérat et al. (2007) showed Cyanophyta such as Anabaenopsis sp. favored increasing light and water column stability in late spring and early summer.

In this study, Microcystis sp. composed the highest proportion of the biomass of Cyanophyta during summer. One potential reason is that Microcystis sp. is competitively advantageous under different light regimes and nutrient conditions (Preece et al. 2017). Having gas vesicles, Microcystis sp. can either float up when the light is limited or move down when more light is available (Brookes and Ganf 2001). A similar pattern was also reported by Belokda et al. (2017) and Rakocevic (2012). The composition of Cyanophyta also differed depending on habitat characteristics (Paerl 1996). For example, colonial forms like Microcystis sp. were usually found in eutrophic brackish waters and lagoons (Gasinaite et al. 2005; Kanoshina et al. 2003).

Similarly to Cyanophyta, high biomass of Chlorophyta (Chlorella sp.) was observed in spring and summer. Temperature and light are considered essential factors for Chlorella sp. productivity (Singh and Singh 2015; Chinnasamy et al. 2009). Belokda et al. (2017) showed, in addition to temperature, the biomass of Chlorophyceae have a significant negative correlation with orthophosphate. This study is consistent with these results (Fig. 4).

Pyrrophyta occurred in winter and was correlated with orthophosphate concentration and water transparency (Fig. 4). Many previous studies recorded the occurrence of Pyrrophyta, Prorocentrum sp., during winter (Springer et al. 2005; Tango et al. 2005). Pyrrophyta such as Peridinium sp. and Prorocentrum sp. prefer oligotrophic condition, having difficulty surviving in a eutrophic condition often found in summer (Hinder et al. 2012). It has been demonstrated that high density of Prorocentrum sp. can occur under temperature of 12–22 °C, low to moderate salinities ranged from 5 to 10 PSU, and moderate wind (Tango et al. 2005). Autumn and winter are known to be a flood season in this study region, and the bay receives a large amount of freshwater leading to high nutrient availability, low temperature, and low salinity; moderate wind also occurs in late winter (Mehdipour et al. 2017). Additionally, Prorocentrum sp. can exhibit high growth rate, enabling it to benefit from pulses of nutrients (Smayda 1997), which may occur in late autumn (Fig. 2h). In addition, high phosphate concentration can benefit Peridinium sp. because it has ability to accumulate phosphorus in their cell and use it for rapid growth (Salmaso and Tolotti 2009). Similar to Pyrrophyta, Chrysophyta abundance also increased when salinity, temperature, and PAR decreased and increased with increasing orthophosphate concentration (Fig. 4). The dominance of Synedra sp., Chrysophyta, under low salinity was reported by Madhu et al. (2007) and Taukulis and John (2006). In this study, Gorgan Bay experienced lower salinity during the flood season (Fig. 2b). Bharathi et al. (2018) reported, Synedra sp. and Melosira sp. were dominant diatoms during a moderate discharge period when the estuary experienced reduced salinity, and these genera usually occurred in the upper and middle parts where salinity is lower. It was also suggested that the low N:P and N:Si ratios were associated with some diatom genera such as Synedra sp. and Melosira sp. (Bharathi et al. 2018). Slate and Stevenson (2007) observed that Synedra sp. grows in the phosphorus-enriched environment. These results are consistent with these studies (Fig. 4).

Spatial differences in the composition of phytoplankton community in association with the gradient of salinity and nutrients including nitrate and orthophosphate were found in Gorgan Bay. The sites in the mouth of the bay are characterized by higher salinity (Fig. 2b). The current from the sea, hydrodynamic process within the bay and freshwater inflow discharged into the bay control the salinity regime in Gorgan Bay (Ghorbanzadeh Zaferani et al. 2017). Although sites in the mouth of the bay and the eastern part of the bay receive freshwater from two major rivers (Gorgan-Rud and Qaresoo River, Fig. 1), the current received from the Caspian Sea affected the salinity more especially around the mouth of the bay. Jamshidi and Bin Abu Bakar (2011) reported that surface water temperature and salinity in the mouth of the bay, deeper sites, were similar to the adjacent seawater. The surface salinity in the southern Caspian Sea ranged from 10.23 to 13.19 ppt and from 12.7 to 14.8 ppt in the eastern littoral area of Gorgan Bay (Saghali et al. 2013; Nasrollahzadeh et al. 2011).

This study revealed that the sampling sites located in the mouth of the bay were also enriched with more nitrate and orthophosphate (Fig. 2g, h). According to Nasrollahzadeh et al. (2008a, b), nitrate was the dominant form of inorganic nitrogen in the southern Caspian Sea due to mineralization of organic compounds at the surface layer because of high concentrations of dissolved oxygen and higher temperature. Nitrate increases along the coast from west to east associated with the surface water temperature gradient (Nasrollahzadeh et al. 2008a, b). Moreover, sites in the mouth of the bay receive a large amount of nutrients and organic matters discharged into the bay from urban and agricultural areas especially in summer (Nasrollahzadeh et al. 2008a, b, Nasrollahzadeh et al. 2014; Shiganova et al. 2003). It has been reported that about 68% of the phosphorous originated from pollution sources such as wastewater treatment plants of Bandar Turkmen (with a population around 50,000) near the primary bay entrance (Ranjbar and Hadjizadeh Zaker 2016). Previous studies suggested that macrophyte communities in the littoral area of the southern and eastern parts of Gorgan Bay also reduced wave action and increased sedimentation and nutrient concentrations including phosphorous (Ghorbanzadeh Zaferani et al. 2016; Ozimek et al. 1990; Yin et al. 2001). Javani et al. (2014) indicated that the maximum concentration of nitrate was in summer (2.5 ppm) in the east and south parts of the bay. Conversely, the orthophosphate concentration increased during autumn and winter and the western part had the minimum orthophosphate level. These results are consistent with that of Javani et al.’s (2014).

Phytoplankton production directly impacts chlorophyll-a concentration. Since it is drived by a combination of nutrient enrichment and increase in temperature and light intensity (Nasrollahzadeh et al. 2008a, b), the mass of nutrient and organic matters discharged into the bay during spring and summer, especially near the mouth of bay, resulted in higher primary production (total biomass) and chlorophyll-a (Nasrollahzadeh et al. 2008a, b, Nasrollahzadeh et al. 2014; Shiganova et al. 2003). Chlorophyll-a also had a smaller peak in the late autumn, responding to orthophosphate enrichment and total biomass of the phytoplankton assemblage (Fig. 2f). However, other factors such as zooplankton grazing could effect which phytoplankton genus thrives (Vanni and Temte 1990). Microcystis sp. was mostly recorded in the mouth of the bay where salinity and nutrient concentrations were greater (Fig. 5). One potential reason is that Microcystis sp. has the highest salt tolerance of all Cyanophyta and can grow in brackishwater environment (Orr et al. 2004; Tonk et al. 2007; Verspagen et al. 2006). When temperature, salinity, and light intensity increased, and the system deals with less turbulence, Cyanophyta tends to grow faster (Reynold and Walsby 1975). This condition usually occurred during summer near the mouth of Gorgan Bay (Fig. 2a, b, and e). The number of reports on Microcystis sp. bloom in brackish water has been increasing (Paerl and Paul 2012). The southern Caspian Sea has experinced several Cyanophyta blooms during August 2005, October 2006, and summers of 2007, 2009, and 2010 in coastal waters (Nasrollahzadeh et al. 2008; Nasrollahzadeh et al. 2011).

Being enriched with more nutrients, the high biomass of Euglena sp. was recorded in the mouth of the bay (Fig. 5). Euglena sp. is considered a biological indicator of the high concentration of organic pollution (Mahapatra et al. 2011). Euglena sp. exhibits fast growth in an environment exposed to less treated domestic wastewater. As a mixotrophic algae, Euglena sp. can also utilize organic matter as another source of energy (Mahapatra et al. 2013). The sites near the mouth of the bay are exposed to domestic wastewater from sewage treatment plan of Bandar Torkaman (Fig. 1). Therefore, the high biomass of Euglena sp. in this region may be considered an indicator of the organic pollution. Ranjbar and Hadjizadeh Zaker (2018, 2016) reported that Bandar Torkaman wastewater plan had the highest contribution of pollution in Gorgan Bay.

The results of Palmer index (PI) confirmed that Gorgan Bay experienced some level of organic pollution near the mouth of bay, especially during the summer. The PI was mostly influenced by Euglena sp. as the most dominant tolerant taxa present in the mouth of Gorgan Bay. The index is recommended as a convenient but reliable method to detect water pollution (Salem et al. 2017). For example, Salem et al. (2017) observed spatiotemporal variation in pollution status of Nile Delta revealing higher pollution during summer. Their study also showed increased abundance of Euglenophyta and Chlamydomonas sp., which is consistent with this study.

Conclusion

This study presents the temporal and spatial variations of phytoplankton composition in Gorgan Bay and how the environmental conditions affect the variations. Results from this study show the phytoplankton composition in the bay is affected by the input of water from Caspian Sea and rivers along with the dynamics of physical conditions within the bay. These variations were controlled by nutrients, salinity, PAR, and temperature. Nutrient availability was controlled by freshwater input and decomposition of organic material within the bay, and salinity was controlled by relative importance of river discharge and currents from Caspian Sea. In addition, sediment type also affected the accumulation of organic material. Bacillariophyta was dominant phylum and occurred throughout the year. Higher biomasses of Chlorophyta (Chlorella sp.) occurred in the spring followed by Cyanophyta (Microcystis sp.) and Euglenophyta (Euglena sp.) maxima in the summer. In the winter, Chrysophyta (Synedra sp., and Melosira sp.) and Pyrrophyta (Prorocentrum sp., Peridinium sp., and Exuviella sp.) showed maxima. Microcystis sp. and Euglena sp. mostly occurred in the mouth of the bay with more saline and nutrient-enriched condition. Spatiotemporal variation of Palmer pollution index (PI) indicated Gorgan Bay experienced pollution in summer, and the sites near the mouth of the bay were in a polluted condition. It is concluded that these latter two species are probably associated with domestic wastewater and should be actively monitored in the future.