Distribution and community structure of microphytoplankton in relation to increasing anthropogenic impact along coastal waters of Jeddah, the central Red Sea

The present study was carried out in the coastal waters of Jeddah, which is a fast-growing metropolitan city located by the central Red Sea. This region provides many opportunities to study the effect of anthropogenic impact on the coastal marine ecosystems of the Red Sea. The presence of different human interferences makes this region vulnerable to environmental changes that can eventually affect biotic and abiotic interactions within the system. In order to study such changes, seven sites were selected along the Jeddah coastline (Fig. 1). They were selected in such a way that they ultimately represent various regions in the system, which are not similar in terms of anthropogenic interference. Site 1 was located in the northern part of Jeddah, where the magnitude of human interference is minimal and can be considered as a reference location for the study. Site 2, located in the central region, is exposed to human impact through wastewater, while site 3, also located in the central region, is mainly affected by the presence of a desalination plant. Sites 4 and 5 represent in the current study the most polluted areas, which are called Al-Shabab and Al-Arbaeen lagoons, respectively. Both sites are exposed to human intervention in many ways, mainly because of their location in the heart of the city. Site 6 was located in the active zone of the Jeddah international port, which can be considered as one of the largest ports functioning on the entire Red Sea coast. The last site (7) was located further south, near a sewage treatment plant.

Figure 1

The fieldwork was carried out in March, July and November 2017 using a mechanized boat. Physical parameters (such as surface temperature and salinity) were determined using a multiparameter water quality probe (Horiba U50). In order to determine inorganic nutrients and phytoplankton biomass (chlorophyll a), a water sample was collected from a depth of 5 m depth using a Niskin sampler (5 l). At least 10 l of seawater was collected and kept in a cool, dark place until reaching the laboratory. After reaching the shore laboratory, 8–10 l of seawater was filtered through a filter paper with a pore size of 0.7 μm and a diameter of 47 mm (Whatman GF/F) to determine the content of chlorophyll a. After the filtration, a few drops of magnesium carbonate (MgCO₃) were added to the filter paper to prevent the disintegration of chlorophyll molecules. All filter papers were then wrapped in aluminum foil and kept in a deep freezer (−80°C). To determine inorganic nutrients, about 1 l of seawater was filtered through a membrane filter paper with a pore size of 0.45 μm, and the filtered seawater was then placed in a deep freezer until further analysis. Both chlorophyll a and inorganic nutrients were determined according to the protocols specified by Parsons et al. (1984) using a UV spectrophotometer (Shimadzu UV 1700). Phytoplankton samples were collected using a plankton net with a mesh size of 20 μm (Hydrobios) equipped with a flowmeter to determine the volume of water filtered. Horizontal towing (0.5 m below surface water) was carried out for 5 min at a speed of about 1 knot. Immediately after the collection, the phytoplankton samples were preserved with Lugol’s iodine solution and a few drops of concentrated formaldehyde solution in an amber glass bottle (Kürten et al. 2015). All samples were then stored in a cool, dark place until further analysis. Triplicate phytoplankton counts were carried out in a Sedgewick Rafter counting chamber using an inverted microscope (Leica DMI 3000B) according to the standard procedures of LeGresley & McDermott (2010). Phytoplankton species were taxonomically identified with the help of identification catalogues (Taylor 1976; Tomas 1997; Gómez 2013). They were then validated with the help of WoRMS (World Register of Marine Species, www.marinespecies.org) and named according to the latest taxonomic nomenclature.

Interpretation of the obtained data was carried out using statistical software SPSS V23, and PRIMER 6. In order to determine the relationship between different physical, chemical and biological parameters, Pearson’s coefficient of correlation (r) was calculated. Moreover, one-way analysis of variance (ANOVA) was carried out to determine both spatial and temporal variation of different parameters. Biodiversity indices (Margalef’s species richness, Pielou’s evenness index and Shannon diversity index) were calculated to study the differences in the phytoplankton spatial and temporal distribution. Nonmetric multidimensional scaling (NMDS) was carried out to determine the similarity between different sites based on various environmental parameters. The dendrogram based on the Bray-Curtis similarity index of the phytoplankton abundance was constructed to determine similarities between different sampling locations.

Temperature ranged between 24.5°C at site 1 and 30.1°C at site 4 in March and July, respectively, with minimal spatial variation. High average temperature (29.5°C) was recorded in July at all sites. The surface salinity showed significant differences between the sites (p < 0.05), especially the lagoon sites (4 and 5), which showed lower values (on average 35.2 and 36.1, respectively) during the study period (Fig. 2). The maximum salinity of 39.9 was observed at site 3 in July, while the minimum (35.1) was observed at site 4 in July (Fig. 2). During the present study, the concentrations of all nutrient salts showed significant spatial variability (p < 0.05) with higher values at sites 4 and 5 compared to the other sites (Fig. 3). The nitrite concentration fluctuated between 0.04 and 20.27 μmol l⁻¹ at sites 1 and 5 in July, respectively, with an average value of 3.93 ± 6.82 μmol l⁻¹ (Fig. 3a). These values were relatively high in November (average 4.84 ± 8.2 μmol l⁻¹). Nitrate concentrations ranged between a minimum of 0.05 μmol l⁻¹ at site 2 and a maximum of 29.30 μmol l⁻¹ at site 4 in July with an average value of 4.33 ± 6.82 μmol l⁻¹ (Fig. 3b). The highest average value of 6.20 ± 11.37 μmol l⁻¹ was recorded in November, while the lowest in July. Ammonium was the largest source of inorganic nutrient derivatives for the studied months, mainly at the lagoon sites, but to a minimum extent also at the coastal sites. Its average concentration was 15.47 ± 26.63 μmol l⁻¹. The minimum concentration was observed at site 7 (0.22 μmol l⁻¹) and the maximum of 78.16 μmol l⁻¹ at site 5 in July (Fig. 3c). Its average seasonal values ranged between 13.15 and 18.24 μmol l⁻¹ in March and November, respectively. Higher average concentrations of phosphate were observed in November (5.35 μmol l⁻¹), with a maximum of 25.90 μmol l⁻¹ at site 4 (Fig. 3d). On the other hand, its minimum value (0.02 μmol l⁻¹) was observed in July at site 3 (average 2.62 μmol l⁻¹). Silicate concentrations did not show any seasonal variation and ranged between 0.43 μmol l⁻¹ at site 6 and 24.50 μmol l⁻¹ at site 4 in November with an overall average of 6.49 ± 8.38 μmol l⁻¹ (Fig. 3e).

The phytoplankton biomass in terms of chlorophyll a showed a significant variation among the sites (p < 0.05). At the coastal sites (1–3 and 6–7), the biomass followed mainly the oligotrophic nature of the Red Sea, while at the lagoon sites (4 and 5) it showed an opposite trend, indicating almost eutrophic conditions (Fig. 3f). The lowest phytoplankton biomass was determined at site 1 (0.13 mg m⁻³) in July and the highest at site 4 (269.14 mg m⁻³) in November with an average value of 41.03 ± 51.31 mg m⁻³ (Fig. 3f). Irrespective of the site, the highest average value of 56.59 mg m⁻³ was recorded in November, while the lowest concentration of 29.73 mg m⁻³ in March.

Figure 2

Figure 3

The total phytoplankton abundance varied significantly between the sites (p < 0.05), ranging from a minimum of 52.4×10³ cells m⁻³ at site 1 in July to 40 800 × 10³ cells m⁻³ at site 4 in November, with an overall average of 6249.9×10³ ± 10797×10³ cells m⁻³. The lagoon sites (4 and 5) showed the highest phytoplankton densities in all months of sampling with mean values of 23 638.2 × 10³ cells m⁻³ and 16469×10³ cells m⁻³ at sites 4 and 5, respectively (Fig. 4a). Other sites showed considerably lower numbers, ranging from 52.4 × 10³ cells m⁻³ at site 1 in July to 3507.7×10³ cells m⁻³ at site 2 in November (Fig. 4a). In the case of the lagoon sites, the phytoplankton density fluctuated between 6544.7×10³ and 40800×10³ cells m⁻³ at site 4 in July and November, respectively (Fig. 4a). On the other hand, less pronounced fluctuations in the total density of phytoplankton were found at other coastal sites (1–3 and 6–7). As for the seasonal variation in the phytoplankton density, November showed the highest value (on average 9937.2 ± 15144.5×10³ cells m⁻³), while the lowest average value was recorded in July.

Figure 4

A total of 174 species of phytoplankton were identified in this study (95 diatoms, 75 dinoflagellates, 3 cyanophytes and 1 silicoflagellates; Table 1). Diatoms were the most dominant taxa, accounting for 95.1% of the total phytoplankton at all sites and varied from 15.5 to 40 600 × 10³ cells m⁻³ at sites 7 and 5 in November, respectively (average 5943.2 × 10³± 10 717.2 × 10³ cells m⁻³; Fig. 4b). They were followed by dinoflagellates, which constituted 3.3% of the total phytoplankton density and varied between 6.6 × 10³ cells m⁻³ at site 1 in March and 2066.7×10³ cells m⁻³ at site 5 in July with an average value of 247.5 ± 276×10³ cells m⁻³ (Fig. 4c). On the other hand, the density of cyanophytes was minimal (average: 99.2 × 10³ ± 224.1 × 10³ cells m⁻³), with relatively high abundance at site 6 in November (Fig. 4d).

Five diatom species accounted for about 89.2% of the total phytoplankton population, namely: Skeletonema costatum (Greville) Cleve, 1873 (48.5%), Chaetoceros decipiens Cleve, 1873 (32.4%), Coscinodiscus radiates Ehrenberg, 1840 (3.7%), Chaetoceros curvisetus Cleve, 1889 (2.8%) and Chaetoceros affinis (1.8%). These species occurred with very high abundance at the lagoon sites (4 and 5). In March, the centric diatom species C. decipiens was the most dominant species (average: 121.7 × 10³ cells m⁻³), accounting for about 24.5% of the total phytoplankton. This high value resulted in a higher density (8500×10³ cells m⁻³) at site 5, which contributed 79.5% to the total phytoplankton density at this site.

Furthermore, C. curvisetus contributed about 19% to the total phytoplankton density at site 5 (2100×fi 10³ cells m⁻³). On the other hand, S. costatum occurred with the maximum density of 23 500 × 10³ cells m⁻³ (99.7% of the total phytoplankton) at site 4. In July, C. decipiens and C. radiates were the main diatom species responsible for the higher densities at sites 4 and 5, respectively. In addition, C. affinis and C. curvisetus occurred at site 5 in relatively large numbers – 1666.7 × 10³ and 1000 × 10³ cells m⁻³, accounting for 8.7% and 5.2% of the total density, respectively. It is worth mentioning that the harmful dinoflagellate Alexandrium tamiyavanichii Balech, 1994 dominated (1200 × 10³ cells m⁻³, accounting for 6.2% of the total phytoplankton) compared to the other dinoflagellate species. The massive bloom of S. costatum (40 000 × 10³ cells) and C. decipiens (18 899 × 10³ cells m⁻³) was observed at sites 4 and 5, respectively, in November. The blue green alga Trichodesmium erythraeum dominated at coastal sites 1–3 and 6–7 (average: 399×10³ cells m⁻³) in November and contributed 2.9% to the total population.

The coastal sites showed different patterns of the diatom species dominance: T. erythraeum, Chaetoceros sp., Thalassiothrix longissimi Cleve & Grunow, 1880, Pseudo-nitzschia delicatissima (Cleve) Heiden, 1928, C. decipiens and Guinardia flaccida (Castracane) H.Peragallo, 1892 accounted for 18.62%, 9.24%, 6.16%, 4.09%, 3.55% and 2.30% of the total phytoplankton, respectively. Tripos furca (Ehrenberg) F.Gómez, 2013 and Protoperidinium oceanicum (VanHöffen, 1897) Balech, 1974 dominated among the dinoflagellates and they represented 1.52 and 1.45% of the total phytoplankton at the coastal sites. The diatom flora comprised 67 Centrales and 28 Pennales. The centric diatom genus Chaetoceros dominated the diatom community with a representation of 20 different species. It was followed by another centric genus, Rhizosolenia, represented by 15 species, and then eight species of the pennate diatom Pseudo-nitzschia. Whereas, Thallassiothrix and Guinardia were represented by four species each. Dinoflagellates were represented by 75 species, with the genus Tripos (26 species) being the most important contributor to the community composition, followed by the genera Protoperidinium (19 species) and Dinophysis (six species).

Chlorophyll a shows a high positive correlation with all measured inorganic nutrients, with r values ranging from 0.91 (p < 0.05) to 0.97 (p < 0.05) for silicates and nitrates, respectively. Moreover, the total phytoplankton density and the diatom C. decipiens were positively correlated with all measured nutrients and negatively correlated with salinity (Table 2). Skeletonema costatum was positively correlated with nitrate, phosphate and silicate, and negatively correlated with salinity. The density of C. affinis was correlated with nitrite, ammonia and silicate. Alexandrium tamarense was positively correlated with both nitrate and ammonia (Table 2). On the other hand, C. radiates and T. erythraeum did not show any correlation with the measured parameters.

The average number of species at sites 2, 3 and 6 was 96, 90 and 85, respectively, while the lowest numbers of 19 and 30 were observed at sites 5 and 4, respectively (Fig. 5a). The values of Margalef’s species richness (d) fluctuated between 1.08 and 6.75 at sites 5 and 2, respectively. Based on Pielou’s evenness (J’) values, most sites showed an even distribution with the minimum value (0.15) at site 4 (Fig. 5b, c). The Shannon diversity index (H’) also showed a similar trend as the richness values, with a higher index observed at the sites with a larger number of species (Fig. 5d). Multidimensional scaling (MDS) based on the

Figure 5

Bray–Curtis similarity index of phytoplankton density showed less than 20% similarity between different sampling sites. Further, they also showed two clusters with 60% similarity (Fig. 6a, b). The results showed the presence of two main clusters, the first cluster comprises only the lagoon sites (4 and 5), which is characterized by high densities and low salinity and diversity. The other cluster comprises the neritic sites (1–3 and 6–7).

Figure 6

Marine ecosystems worldwide are subjected to both natural and anthropogenic changes, which in turn affect the structure and functioning of existing communities. Rapid industrialization and increasing urbanization add more pressure to the systems by destroying the ecological balance and the food web dynamics (Maso et al. 2006). The Red Sea, known for its extreme oligotrophic characteristics, is currently changing this status at least in the lagoons and coastal waters of the growing metropolitan cities, which receive sewage effluents (Al-Aidaroos et al. 2019). In the present study, salinity showed considerable spatial variation compared to temporal variation, with low values at the lagoon sites. This can be related to the geographical location of these two sites, which receive a continuous sewage discharge of about 103 000 m³ per day (Peña-García et al. 2014). Salinity showed a significantly high negative correlation with phytoplankton biomass, which was mainly due to the occurrence of waters with low saline at the lagoon sites. Other sites (2 and 7) receive a relatively large number of wastewater discharges and site 3 (which is located near the desalination plant), did not show any reduction in the salinity values throughout the study period. This can be associated with the mixing process as these sites and its directly connection with open waters. The present study shows abnormally high values of all measured inorganic nutrients at the lagoon sites. At these locations, the Red Sea is not oligotrophic, instead it is highly eutrophic in terms of nutrient availability. The high nutrient concentrations at the lagoon sites are in line with the previous studies (e.g. El-Rayis 1998; El-Sayed 2002a; Khomayis 2002; Al-Farawati et al. 2008; Al-Harbi & Khomayis 2010; Peña-García et al. 2014) and can be explained by the continuous inflow of sewage water into these lagoons. On the other hand, the low nutrient concentrations at other coastal locations are also comparable to studies that were carried out in different parts of the Red Sea (Dorgham et al. 2012; Kürten et al. 2015; Devassy et al. 2017, El-Sherbiny et al. 2019). As previously stated, sites 2 and 7 also received significant amounts of inorganic pollutants, but the effects were smaller. This is mainly due to the improved design of the outfall, consisting of a submerged multiport diffuser pipe dispersing the wastewater over a distance of 150 m from a water depth of 25–50 m (Peña-García et al. 2014).

In the present study, spatial differences in phytoplankton biomass values were observed, rather than the temporal ones. Chlorophyll a and nutrients showed a significant positive correlation (p < 0.05), indicating the effect of inorganic nitrogen and silicate on phytoplankton biomass. The phytoplankton biomass at the lagoon sites was very high, which was completely uncharacteristic of the typical oligotrophic nature of the Red Sea, described by Shaikh et al. (1986), Qurban et al. (2014; 2017), Kürten et al. (2015), Devassy et al. (2017) and Kheireddine et al. (2017). The chlorophyll a values obtained at the lagoon sites in the present study were 100–200 times higher than at the other coastal sites indicating unfavorable habitat conditions at these two sites. Stagnant conditions and continuous inflow of both treated and untreated sewage water directly toward these two sites increased the inorganic nutrient loading, which in turn allowed the available phytoplankton to proliferate and form eutrophic conditions within these lagoon locations. Sites 2, 3 and 7 are also exposed to anthropogenic impact in terms of sewage disposal, and the proximity of the desalination plant but the phytoplankton biomass was still slightly higher compared to other coastal sites.

The impact of high nutrient concentrations significantly contributed to the proliferation of marine phytoplankton at some of the study sites. Traditionally known for its lower phytoplankton abundance, the Red Sea has experienced this type of phytoplankton population explosion only in a few cases. A similar type of higher phytoplankton densities was reported by Touliabah et al. (2010) and Devassy et al. (2017) from lagoon locations in the central and northern Red Sea, respectively. The densities observed at the lagoon sites during this study were much higher compared to all the studies mentioned above and can easily be compared to a tropical phytoplankton bloom. They may be related to the continuous sewage discharge into the coastal lagoons as well as their isolation from the open water. Other sites witnessed lower numbers of phytoplankton except site 2, which showed higher densities during a certain period of the study. This may be related to the anthropogenic impact mainly in the form of sewage disposal (Peña-García et al. 2014). The phytoplankton densities observed at other coastal sites are comparable to the normal pattern of phytoplankton distribution reported in the previous studies, both in the Red Sea proper and the Gulf of Aqaba (Sommer 2000; Post et al. 2002; Al-Najjar et al. 2007; Kürten et al. 2015; Devassy et al. 2017, El-Sherbiny et al. 2019).

Diatoms dominated in the phytoplankton community throughout the study period, similarly to what was described in previous studies (Shaikh et al. 1986; Devassy et al. 2017). The count of dinoflagellates was significantly low in March and showed higher values in July and November, which is consistent with the observations reported by Shaikh et al. (1986), Touliabah et al. (2010), Ismael (2015) and Kürten et al. (2015). Cyanophytes marked their presence mainly in the autumn, which is similar to the findings of Shaikh et al. (1986). The summer and autumn dominance of cyanophytes (mainly Trichodesmium) was reported by Post et al. (2002) from the Gulf of Aqaba and by Devassy et al. (2017) from the central Red Sea. Diatoms are known for their ability to use available inorganic nutrients and can form blooms within a short period of time. In the present study, the overall conditions were favorable for the growth of diatoms, which was reflected in their proliferation. It is known that phytoplankton biomass (rather than species composition) in the coastal waters is enriched by increasing nutrients through coastal upwelling as well as anthropogenic and riverine inputs (Polat 2007). Moreover, Officer & Ryther (1980) and Mishra et al. (2006) stated that diatom species forming a bloom require silicate as the primary nutrient and the ratio of nitrate and silicate affects the composition of phytoplankton.

Diatoms were more diverse than both dinoflagellates and cyanophytes in March and November. The number of phytoplankton species identified in the current study (174 species) was higher than that reported by Touliabah et al. (2010) – 73 species, while relatively lower than those recorded by Shaikh et al. (1986), Kürten et al. (2015) and Devassy et al. (2017). The main reason for this low diversity is the selection of sampling locations, which was primarily based on the anthropogenic impact. Typically, in such environments, phytoplankton has higher abundance but low diversity, which was the case in this study. The main diatom genera observed in the present study (Chaetoceros and Rhizosolenia) are known to be common inhabitants of the Red Sea. Similarly, dinoflagellates were dominated mainly by Tripos and Protoperidinium (Dowidar et al. 1978; Shaikh et al. 1986; Ismael 2015; Kürten et al. 2015; Devassy et al. 2017). Cyanophytes were represented mainly by T. erythraeum, which dominated in November. Based on the present study, the degree of human activity in Jeddah coastal waters has a significant impact on the density and diversity of phytoplankton. Furthermore, this study can be useful as a baseline for the development of monitoring programs for Jeddah coastal waters and as an aid in the implementation of a primary alert system for harmful algal blooms.