Environmental Parameters
The highest rainfall was observed during the first half of June, the second half of July, and early September. The water temperature exhibited a bimodal pattern with peaks during the late PrM and early PoM, and the annual variation ranged between 23ºC and 34ºC. Salinity distribution followed a trend opposite to rainfall with the lowest values during MON (25 ± 6) and highest during the PrM (34.5 ± 1.3) and PoM (33 ± 2.5). The SD depth was relatively higher during the PoM (170 ± 43 m) and PrM (106 ± 56 m) than during the MON season (73 ± 30 m) (Fig. 2a-d). The NO3- + NO2-, PO43-, and SiO44- concentrations were higher during MON. The NH4+ concentrations were higher during MON and PrM (Table 2).
Seasonal variability in the chlorophyll-a concentrations
The total chl-a concentration during the MON season ranged between 1327 ng L-1 (July 29) and 5960 ng L-1 (August 6) (Fig. 3). During the PoM season, the chl-a concentrations ranged between 1195 ng L-1 (Nov 4) and 7246 ng L-1 (Nov 6). During PrM, the chl-a concentrations ranged between 828 ng L-1 (April 22) and 3308 ng L-1 (May 09). The seasonal trend was MON (3570 ± 1600 ng L-1) > PoM (3150 ± 1750 ng L-1) > PrM (1740 ± 830 ng L-1).
The concentration of chl-a in the < 3 µm fraction ranged from 216 ng L-1 (Sept 28) to 2775 (August 6th) ng L-1 during MON, 84 ng L-1 (Jan 30) to 1289 ng L-1 (October 15) during PoM, and 41 ng L-1 (May 25) to 577 ng L-1 (April 22nd ‘016) during PrM (Fig. 3). The seasonal trend was MON (860 ± 660 ng L-1) > PoM (400 ± 370 ng L-1) > PrM (290 ± 190 ng L-1).
The seasonal contribution of < 3 µm fraction to the total chl-a biomass was higher during MON (25.8 ± 13.8%) followed by PrM (22.1 ± 21.4%) and PoM (12.1 ± 6.7%) (Fig. 3). A significant positive correlation was observed between the total chl-a and < 3 µm a biomass (R2 = 0.7; P < 0.01).
Seasonal Variability in the Picophytoplankton Biomass
The CHEMTAX analysis of the < 3 µm fraction chl-a data revealed the contribution from 7 algal groups to the Pico chl-a biomass, comprising one group of Picocyanobacteria and six groups of PPEUK (Table 1). During the MON, a significant portion of the Pico chl-a biomass was contributed by the PPEUK (468 ng L-1; 58.3%) and the rest by Picocyanobacteria (388 ng L-1; 41.7%). During PoM, the PPEUK contributed 61.7% (234 ng L-1) and Picocyanobacteria 38% (166 ng L-1). During PrM, the PPEUK and Picocyanobacteria contributed 57.4% (173 ng L-1) and 42.6% (121 ng L-1), respectively (Fig. 4).
The PPEUK exhibited a variable community structure with respect to the seasons. The hydrography and environmental conditions during the MON season were driven by the precipitation intensity and freshwater runoff. At the initiation of the sampling period, under high saline (35) conditions, the Picoprasinophytes (33.8%) and Picochlorophytes (31.6%) were the dominant contributors to the PPEUK chl-a biomass. The dominance later shifted to the Picocryptophytes (42.8%) and Picodiatoms (50.5%), following a decline in salinity (20 to 25 and < 20, respectively). The subsequent rise in salinity, following a break-in precipitation (30), coincided with an increase in the Picoprasinophytes (36.7%) followed by the Picocryptophytes (43%; salinity 25) contribution. Towards the end of the MON season (salinity 30), the Picocryptophytes (52%) and Picodiatoms (43.4%) dominated the PPEUK community while the Picocryptophytes dominated during the PoM (37.2%) and PrM (54.7%) seasons (Fig. 4).
Seasonal Variability in the Picophytoplankton Abundance
Seasonally, during the MON, a major portion of the Pico abundance was contributed by SYN (35.6 x 103 cells mL-1; 74.3%) and rest by PPEUK (13.2 x 103 cells mL-1; 25.7%). During PoM, the SYN contributed 80.6% (44.6 x 103 cells mL-1) and PPEUK 19% (8.3 x 103 cells mL-1). During the PrM, SYN contributed 87.9% (95.6 x 103 cells mL-1) and PPEUK 12.1% (14.6 x 103 cells mL-1), respectively (Fig. 6a). Throughout the sampling, the cell abundance of the SYN (1.95 x 103 to 210 x 103 cells mL-1) and PPEUK (0.18 x 103 to 54.5 x 103 cells mL-1) exhibited a similar seasonal distribution pattern with a significant positive correlation (R2 = 0.74; P < 0.001). The seasonal trend for the SYN abundance was PrM (95 ± 78 x 103 cells mL-1) > PoM (45 ± 39 x 103 cells mL-1) > MON (36 ± 31 x 103 cells mL-1). The PPEUK abundance exhibited the following seasonal average trend; PrM (14.6 ± 17.5 x 103 cells mL-1) > MON (13 ± 13 x 103 cells mL-1) > PoM (8.3 ± 5.4 x 103 cells mL-1).
At the beginning of the sampling (MON season), before the decline in salinity due to precipitation and freshwater runoff into the Bay (salinity: 35), the Picochlorophytes-II were the dominant contributors (76%) to the PPEUK abundance. Subsequently, the decline of salinity (20 to 25) and temperature led to a change in the community dominance to Picoprasinophytes (50%). The further decline in salinity (< 20) resulted in the Picochlorophytes-I (52.3%) dominating the community with an increased contribution from the Picocryptophytes (18.4%). The dominance continued with the following rise in salinity (~30), but the contribution from Picochlorophytes-II (27.5%) and Picocryptophytes (24.8%) increased. A salinity of 25 led to an increased contribution from the Picoprasinophytes (22%) with dominance by Picochlorophytes-II (40.8%) that continued till the end of the MON season (69%) with an increase in salinity (30). The Picodiatoms increase (7%) during this period. During the PoM, the Picochlorophytes-I dominated the community (49.3%) with intermittent dominance of the Picochlorophytes-II (24.6%) and Picoprasinophytes (16%). During the PrM, the community was dominated by the Picochlorophytes-I (57%) (Fig. 6b-f).
Relationship of the Pico with the Environmental Parameters
During MON, the first 2 RDA axes explained 79.2% of the variability in the Pico groups' biomass and 94.2% of the relationship between biomass and the environmental variables (Fig. 7a). The temperature and NO3-+NO2- were the significant explanatory variables (Table 3). The Picocryptophytes positively correlated with nutrients (NH4+, NO3-+NO2-) and total chl-a and negatively with salinity, temperature, and SD depth. The Picodiatoms correlated positively with nutrients (NO3-+NO2-, SiO44-), rainfall, total chl-a, and negatively with salinity, temperature, and SD depth. The Picochlorophytes positively correlated with nutrients (NH4+, NO3-+NO2-) and negatively with salinity, temperature, and SD depth. The Picoprasinophytes were positively correlated with temperature, salinity, NO3-+NO2-, and SD depth. The Picocyanobacteria positively correlated with salinity, temperature, and SD depth. In the abundance triplot (Fig. 7b), the first 2 RDA axes explained 46.1% of the variability in the Pico groups' abundance and 96.2% of the relationship between abundance and the environmental variables (Fig. 7b; Table 4). The Picocryptophytes were positively correlated with nutrients (PO43-, NH4+) and negatively with salinity, temperature, and SD depth. The Picodiatoms correlated positively with salinity, temperature, SD depth, and SiO44-. The Picochlorophytes were positively correlated with salinity, temperature, and SD depth. The Picoprasinophytes were positively correlated with SD depth, PO43- and NH4+. The Picocyanobacteria positively correlated with temperature, SD depth, NH4+, and PO43-.
During PoM, the first 2 RDA axes explained 88.8% of the variability in the Pico groups' biomass and 91.7% of the relationship between biomass and the environmental variables (Fig. 7c). The Picocryptophytes were positively correlated with nutrients (NH4+, NO3-+NO2-). The Picodiatoms correlated positively with temperature, SiO44- and total chl-a and negatively with NO3-+NO2-. The Picoprasinophytes were positively correlated with NH4+ and PO43-. The Picochlorophytes did not exhibit any significant relationship. The Picocyanobacteria positively correlated with temperature and total chl-a. In the abundance triplot (Fig. 7d), the first 2 RDA axes explained 74.6% of the variability in the abundance of the Pico groups and 92.2% of the relationship between abundance and the environmental variables (Fig. 7d). The Picodiatoms positively correlated with the SD depth. The PPEUK groups and SYN did not show any significant correlation with the nutrients.
During PrM, the first 2 RDA axes explained 95.5% of the variability in the biomass of the Pico groups and 95.5% of the relationship between biomass and the environmental variables (Fig. 7e). The Picodiatoms positively correlated with PO43-. The Picocryptophytes, Picochlorophytes, and Picoprasinophytes positively correlated with NH4+, NO3-+NO2- and the SD depth. The Picoprymnesiophytes and picodinoflagellates correlated positively with NO3-+NO2-. The Picocyanobacteria were positively correlated with NO3-+NO2-. In the abundance triplot, the first 2 RDA axes explained 98.5% of the variability in the Pico groups' abundance and 98.5% of the relationship between abundance and the environmental variables (Fig. 7f). All the PPEUK groups correlated positively with PO43- and NO3-+NO2- except the Picoprasinophytes, which correlated with only PO43-. Also, the Picodiatoms positively correlated with SiO44-. The Picocyanobacteria positively correlated with PO43-.