Spatio-temporal dynamics of phytoplankton community in a well-mixed temperate estuary (Sado Estuary, Portugal)

Physico-chemical characterization of the Sado Estuary

The seasonal cycle of water temperature in the Sado Estuary in 2018 and 2019, showed the expected pattern, with maxima temperature observed in summer and minima in winter (Fig. 2A). During summer, warmer temperatures were found in the inner regions of the estuary (AC and MC) and lower temperatures near the mouth of the estuary (EM). During winter, there was an inversion of the pattern, with the coolest waters recorded inside the estuary (Fig. 2A). Near the estuary mouth (EM), salinities recorded were always between 35 and 36 (Fig. 2B). In the inner stations, higher salinities (> 30) were found during summer/early-autumn of 2018 and late-spring/summer of 2019. Maxima salinities (> 36) were recorded in the summer of 2019 (Fig. 2B). The lowest salinities were always found in the upper region (AC), reaching a minimum of 12 in March 2018 (Fig. 2B).

The water turbidity was substantially higher in AC, reaching values above 10 NTU in the summer of 2018 and since spring of 2019, with a maximum of 40 NTU recorded in March 2018 (Fig. 2C). The turbidity values and seasonal pattern for stations MC and MR were similar, with a maximum of 10 NTU recorded in MR during spring of 2018 (Fig. 2C). Lower turbidity was observed during winter in stations AC, MC, and MR (Fig. 2C). Water turbidity was always lower than 1.5 NTU at EM (Fig. 2C). The CDOM was higher in the upper region and lower in the downstream area (Fig. 2D). At AC, a CDOM value < 0.45 m⁻¹ was recorded between winter 2018 and spring 2019. In the remaining period, values above 0.60 m⁻¹ were observed only in May, July, and August of 2018. At MC and MR, CDOM varied between 0.15 m⁻¹ and 0.65 m⁻¹ (Fig. 2D). In the downmost station (EM), CDOM was always below 0.15 m⁻¹, except for February 2019, which reached 0.25 m⁻¹ (Fig. 2D).

The pH observed in both studied years varied mainly from 7.8 to 8.2, except for July of 2019 at AC, where the minimum value was recorded (7.6) (Fig. 2E). Only the station located near the estuary mouth (EM) registered pH values always above 8.0 (Fig. 2E). The dissolved oxygen (DO) concentration ranged between 6 and 10 mg L⁻¹ in all stations (Fig. 2F). From April to October of 2018, higher DO concentrations were observed near the estuary mouth, when compared with the other stations. The minimum DO concentration was observed in May of 2018 (3.3 mg L⁻¹), simultaneously at AC and MC (Fig. 2F).

An increase in nutrient concentrations (DIN, phosphate, and silicate) was found from downstream towards upstream stations, with higher concentrations in 2018 than in 2019 (Fig. 2G,H,I). DIN reached maxima above 15 µmol L⁻¹ in spring (AC, MC, and MR) and autumn (AC) of 2018. In 2019, lower DIN concentrations were observed in the entire estuary, reaching maxima values during autumn in AC (> 10 µmol L⁻¹) (Fig. 2G). Phosphate concentrations were below 1.5 µmol L⁻¹ in the entire estuary, with higher values in the inner stations, and lower (< 0.6 µmol L⁻¹) near the estuary mouth (Fig. 2H). In the inner stations, higher concentrations were recorded during spring and autumn of 2018, and since mid-summer of 2019 (Fig. 2H). Silicate concentrations had two peaks each year in AC (in spring and autumn), being higher in 2018 than in 2019 (Fig. 2I). These two peaks could also be found during 2018 in MC and MR. Near the estuary mouth, silicate concentrations were always low (< 10 µmol L⁻¹) (Fig. 2I).

The daily precipitation measured in the Sado Estuary showed higher values in 2019 than in 2018 (Fig. 3). Precipitations above 5 mm day⁻¹ were recorded in winter and autumn of both years, with maxima (above 15 mm day⁻¹) observed in late-October of 2018, January, and December of 2019 (Fig. 3).

Figure 4 shows the representation of molar quotients between the concentrations of potentially limiting nutrients for phytoplankton growth for each station, which are delimited by the Si:N = 1, N:P = 16 and Si:P = 16 lines^40,41. Inner stations of the estuary (AC, MC and MR) had a higher availability of silicate, as Si:N and Si:P were mostly recorded above the ratio (Fig. 4A,B,C), whereas, near the estuary mouth (EM), there was a high number of samples with Si:N lower than 1, and Si:P lower than 16 (Fig. 4D). The N:P ratios above 16 were recorded especially during spring and summer of 2018, namely in March, May, and June at AC (Fig. 4A), in March, April and June at MC (Fig. 4B), in March and April at MR (Fig. 4C) and in March, April and June at EM (Fig. 4D).

Intra- and inter-annual variation of phytoplankton

Phytoplankton biomass (Chl-a) was consistently higher at AC in both years, reaching concentrations above 8 mg m⁻³ in May and July 2018 and July 2019 (Fig. 5). In the other three studied sites, Chl-a was always below 6 mg m⁻³. In these stations, during 2018, Chl-a maxima were observed in May at EM and MR, and in July at MC. During 2019, both AC and MC peaked in July, whereas in EM it was no clear peak (Fig. 5). There were three Chl-a peaks at MR during 2019, which were recorded in March, May, and July (Fig. 5).

Phytoplankton cell abundance for each studied site, as well as relative abundances of the phytoplankton groups identified are presented in Fig. 6. During 2018, AC and MC reached phytoplankton maximal abundances in May (above 400 × 10³ cell L⁻¹) (Fig. 6A,B), the station located downstream (EM) reached the maximum abundance in June (ca. 250 × 10³ cell L⁻¹) (Fig. 6D), while MR had low abundances all year round, with a maximum recorded in March (80 × 10³ cell L⁻¹) (Fig. 6C). In 2019, a peak was recorded in February at AC (ca. 250 × 10³ cell L⁻¹), but in the remaining year phytoplankton abundance was always below 50 × 10³ cell L⁻¹ (Fig. 6A). In MC and MR, the maxima were recorded in March (respectively 100 × 10³ cell L⁻¹ and 150 × 10³ cell L⁻¹) (Fig. 6B,C). A high cell concentration was also recorded in March in the downmost station, with the peak being observed in September (200 × 10³ cell L⁻¹) (Fig. 6D).

Results of the phytoplankton community at the class level evidenced the differences between the estuary mouth and the stations located inside the estuary, with cryptophytes being a highly dominant group recorded inside the estuary (AC, MC, and MR) (Fig. 6A,B,C). Several inter-annual differences could be observed. At AC, cryptophytes (Cryptophyceae) were more abundant during 2018, while in 2019 diatoms (Bacillariophyceae) dominated (Fig. 6A). A change in the community during late-autumn/winter could also be observed, with a reduction of diatoms and an increase in cryptophytes (Fig. 6A). At this station, it was also possible to observe a higher contribution of dinoflagellates (Dinophyceae) during 2019 than in 2018 (Fig. 6A). At MC, cryptophytes dominated during the studied period, and diatoms constituted the second most abundant group in most months, presenting higher concentrations in 2019 than in 2018 (Fig. 6B). The dominant phytoplankton groups at MR were similar with the ones observed at MC. The main difference was a higher contribution of diatoms in 2018 at MR, than in MC (Fig. 6C). Near the estuary mouth (EM), diatoms were, in general, the dominant group (Fig. 6D). The main exceptions occurred in April and October of 2018 when dinoflagellates dominated, and in July of 2019 when cryptophytes dominated (Fig. 6D). In the above-mentioned situations of dinoflagellate dominance recorded at EM in 2018, the phytoplankton assemblage was mostly composed by Prorocentrum cordatum in April (ca. 150 × 10³ cell L⁻¹) and dominated by Gymnodinium spp. and Gymnodinium catenatum in October (65 × 10³ cell L⁻¹) (Table S2).

For the other identified groups, in general, Euglenophyceae were more abundant inside the estuary, and had higher abundances in 2019 (Fig. 6A,B,C). Cyanophyceae occurred only sporadically, reaching higher abundances during spring inside the estuary (Fig. 6A,B), and during autumn near the estuary mouth (Fig. 6C). Other groups also formed sporadic blooms, as Prasinophyceae in May 2018 at MC and Prymnesiophyceae (i.e., Phaeocystis spp.) in November 2018 at EM (Fig. 6B).

The Chl-a concentrations (maximum and average) obtained during this study, for the different regions of the Sado Estuary, were smaller than those obtained for other periods and similar locations (Table 1). Higher differences were observed in the inner regions of the estuary (e.g., maximum of 12.6 mg m⁻³ obtained in 2018–2020, compared with 25 mg m⁻³ obtained during the 1990s in the Alcácer channel), than near the estuary mouth (e.g., maximum of 3.9 mg m⁻³ obtained in 2018–2020 and 5 mg m⁻³ obtained for the other periods). Although using different metrics (Chl-a maxima in this study and Chl-a 90th percentile for the WFD), the Chl-a maxima here obtained in the different regions of the estuary agree with the ‘High’ classification (Table 1).

Phytoplankton abundances obtained in the lower region showed similar results with those observations made for the 1960s (around 250 × 10³ cell L⁻¹) but were lower than the obtained in the 1990s (450 × 10³ cell L⁻¹) (Table 1). In the mid and upper estuarine regions, the phytoplankton abundances obtained in this study were much lower than the obtained during the 1990s (e.g., 490 × 10³ cell L⁻¹ recorded in the Alcácer channel in 2018/19, while 5300 × 10³ cell L⁻¹ were recorded in the same site in 1992/93). Phytoplankton maximal abundance recorded in this study were always below the thresholds established previously for the determination of the occurrence of a phytoplankton bloom in the Sado Estuary, as well as for the Portuguese upwelling coast (Table 1).

Spatial and temporal patterns of phytoplankton assemblages

A total of 10 phytoplankton groups were identified in the Sado Estuary during this study, with an increasing number of taxonomic entities identified from upstream (87 at AC) to downstream stations (179 at EM, Table S2). The phytoplankton groups with a higher number of taxa identified were the Bacillariophyceae and the Dinophyceae, with nearly 50 more taxa of dinoflagellates observed near the estuary mouth when compared to the stations inside the estuary (Table S2).

The year 2018 was characterized by having a higher number of taxonomic entities identified in all stations, reaching the maximum at EM (Table 2). Except for MR, 2018 was also characterized by higher phytoplankton abundance, with the maximum recorded at AC (Table 2). Table 2 also shows the phytoplankton diversity (H’) and evenness (J) indices obtained for the studied years and stations in the Sado Estuary. The highest average diversity and evenness were obtained in 2019 at EM (2.11 and 0.55, respectively). However, the maximum diversity and evenness were recorded at MR during April 2018 (Shannon–Wiener diversity of 3.23 and Evenness of 0.83), when 49 taxonomic entities were identified, all present in low abundances (Table 2).

The PERMANOVA analysis of the phytoplankton assemblages indicated significant differences between years (p = 0.004, Table 3), sampling sites (p = 0.001, Table 3) and seasons (p = 0.001, Table 3). Since the results also indicated significant interactions between factors “Years (times) Season” (p = 0.023, Table 3) and “Site (times) Season” (p = 0.001, Table 3), pairwise comparisons are shown in Table 4. The “Years (times) Season” comparison only indicated differences between the years during summer (p = 0.003, Table 4). The “Site (times) Season” comparison did not indicated differences between AC/MC and MC/MR during spring, as well as for all the study site comparisons during winter (Table 4).

PCO ordinations were used to analyze the spatial and temporal patterns of the phytoplankton assemblages (Figs. 7,8). Spatial patterns in the phytoplankton community composition were better observed when considering the axes 1 and 2 of the PCO analysis, with both axes explaining 37.3% of total phytoplankton variance (Fig. 7). The separation between the stations EM, MR and MC/AC can be observed along axis 1, while the separation between MC and AC occurred along axis 2. The results showed that a higher contribution of several chain-forming diatoms and dinoflagellates species were more related with the downstream station (EM) (e.g., the diatoms Chaetoceros spp., Guinardia delicatula and Leptocylindrus cf. danicus, and the dinoflagellates Gyrodinium spp., Scrippsiella group and athecate unidentified species) (Fig. 7A). These assemblages appeared to be more correlated with higher pH levels (Fig. 7B). In upstream stations (AC and MC), phytoplankton assemblages were dominated by several pennate diatom species (e.g., Nitzschia cf. sigma, Gyrosigma spp./Pleurosigma spp.) and by some centric diatoms (e.g., Coscinodiscus spp. and Melosira nummuloides) (Fig. 7A). This phytoplankton assemblage was correlated with warmer waters, richer in nutrients and with higher turbidity levels (Fig. 7B). The station located in the middle of the estuary (MR) showed higher similarity with phytoplankton assemblages recorded upstream (Fig. 7).

This PCO also showed that the phytoplankton community composition patterns observed between the years (2018 and 2019) were less evident than the spatial patterns (Fig. 7). To better identify the taxonomic entities responsible for the phytoplankton community structure between the factor Year, the SIMPER analysis was performed (Table S3). The average dissimilarity observed on phytoplankton community between 2018 and 2019 (51.7%) resulted from the overall assemblage, with each species contributing for a small percentage of the dissimilarity between both years. Some of the taxonomic entities that contributed most for the dissimilarity between the years were small unidentified flagellates (< 15 µm), G. delicatula, Thalassiosira spp. (< 10 µm), Skeletonema marinoi, Thalassionema nitzschioides, Chaetoceros spp. and L. cf. danicus (Table S3).

PCO analyses conducted for each sampling site showed the seasonal patterns in the phytoplankton communities at each location (Fig. 8), with the first two axes explaining 41.2%, 37.2%, 37% and 35.6% of total phytoplankton variance, for stations AC, MC, MR and EM, respectively. At AC there was a higher correlation of winter/spring phytoplankton assemblages with drivers such as nutrient concentrations (DIN and Si) and precipitation (Fig. 8A). During winter and spring, phytoplankton assemblages had a high contribution of athecate unidentified dinoflagellates and the diatom S. marinoi. During winter, cryptophytes and dinoflagellates belonging to the group Heterocapsa spp./Azadinium spp. were also important, while in spring there was a higher correlation with the dinoflagellates of Scrippsiella group and with several chain-forming diatoms species (e.g., Thalassiosira spp. (< 10 µm)—Fig. 8A, G.. delicatula and Rhizosolenia cf. imbricate—not shown). Summer/autumn phytoplankton community composition had a higher correlation with higher temperature and turbidity levels (Fig. 8A). During summer, there was a higher correlation with several pennate diatoms (e.g., T. nitzschioides, Diploneis cf. bombus, cf. Scoliotropis spp. and Cylindrotheca closterium). In autumn dominated a mixed community of pennate diatoms (e.g., C. closterium and unidentified species > 80 µm) and small dinoflagellates (e.g., Heterocapsa spp./Azadinium spp. and unidentified species < 15 µm) (Fig. 8A).

At MC the explanatory physico-chemical variables were almost the same that were found in AC, although with precipitation having a higher correlation with winter samples (Fig. 8B). The winter phytoplankton community was better correlated with Gymnodinium spp. and other athecate unidentified dinoflagellates, as well as with the diatom Grammatophora oceanica (Fig. 8B). Spring had a higher contribution of the dinoflagellates belonging to the Scrippsiella group and several diatoms, such as Trieres mobiliensis, Ditylum brightwellii and the chain-forming G. delicatula (Fig. 8B). The summer/autumn phytoplankton assemblage had a higher contribution of several pennate diatoms, most of them benthic species (e.g., Gyrosigma spp./Pleurosigma spp., cf. Scoliotropis spp. and N. cf. sigma) (Fig. 8B).

At MR, turbidity and temperature were again highly correlated with summer/autumn phytoplankton assemblages, but in this location there was also a correlation of 0.30 with precipitation (especially recorded during autumn, cf. Figure 3) and with phosphate concentration (Fig. 8C). In terms of phytoplankton assemblages, it was possible to distinguish two different groups, the winter/spring and the summer/autumn communities. The dinoflagellate Prorocentrum cordatum and several diatom species (e.g., D. brightwellii, G. delicatula, G. striata and Asterionellopsis glacialis) had higher correlation with spring communities, and the diatoms of the centric group (20–40 µm) and S. marinoi with the winter assemblage (Fig. 8C). The summer and autumn phytoplankton assemblages had a higher contribution of pennate diatoms (e.g., N. cf. sigma, Bacillaria paxillifera), dinoflagellates (e.g., Karenia spp. (< 20 µm), Dinophysis caudata and Prorocentrum micans) and the chlorophyte Monoraphidium cf. griffithii.

At EM, higher nutrient concentrations were most correlated with winter and some spring samples, temperature with summer and autumn samples, and precipitation and pH mainly with autumn samples (Fig. 8D). At this station, winter phytoplankton community composition had a high correlation with pennate diatoms, such as Amphora spp. and cf. Delphineis surirella. In spring, a community similar to that recorded in MR was found (dinoflagellate P. cordatum and several chain-forming diatoms species (e.g., D. brightwellii, G. delicatula, G. striata and A. glacialis). Summer and autumn were mainly correlated with high abundance of several dinoflagellate species (e.g., Dinophysis acuta, Protoperidinium cf. curtipes—Fig. 8D, Torodinium robustum, D. caudata, P. micans—not shown). During summer, a mixed community of dinoflagellates was found together with diatoms (e.g., R. cf. imbricata, Proboscia alata, Nitzschia spp., L. cf. danicus), and during autumn there was also a high correlation with the cyanobacteria Merismopedia cf. elegans (Fig. 8D).

Source: Ecology - nature.com