Influence of mesoscale eddies on the spring phytoplankton groups in the Southern Gulf of Mexico

2.1 Study area

The GoM is the largest marginal sea in North America, with an extent of 1.6 million km² (Figure 1), shared by three countries (Mexico, Cuba, and the United States of America). Its bathymetry is complex (up to a depth of 3,900 m), including submarine canyons, abyssal plains, escarpments, and a wide continental shelf, both in the north (Florida) and south (Yucatan) (Goff et al. 2016) (Figure 1).

Figure 1:

Gulf of Mexico: (a) map with the main features of the circulation pattern such as the presence of the loop current that enters the Gulf between Cuba and the Yucatan peninsula and the presence of eddies, both cyclonic (C) and anticyclonic (A); (b) map with the sampling stations considered in this study. Bathymetry is shown in m.

The GoM is an extremely dynamic region characterized by a loop current that enters the Gulf through the Yucatan channel, forming a meander that leaves the Gulf through the Florida Strait. The Loop current releases a series of eddies (mainly anticyclonic) that move west and collide with the coasts of Veracruz and Tamaulipas (Díaz-Flores et al. 2017). During this displacement, these eddies modify the physical and chemical characteristics of the water column, thereby affecting phytoplankton communities (Linacre et al. 2015). Additional hydrodynamic processes occur inside the GoM at different scales that strongly influence planktonic ecosystems, including internal waves (Santiago-Arce and Salas-de-León 2012) and mesoscale cyclonic eddies (Durán-Campos et al. 2017).

In climatic terms, three main seasons are recognized in the GoM, 1) the “Nortes” season, which occurs from November to February, characterized by the presence of strong winds (up to 80 km h⁻¹) that cross the GoM from north to south, exerting an outstanding influence in the surface and subsurface layers of the gulf, 2) the dry season, from March to May, when the evaporation levels are maximum, modifying the surface salinity levels, and 3) the rainy season, which occurs from June to October, when the contribution of freshwater to the gulf of the main river systems (e.g. Grijalva-Usumacinta) substantially modifies the hydrographic structure of the water column (Durán-Campos et al. 2017; Ojeda et al. 2017).

2.2 Sampling

A total of 63 stations were sampled both day and night during a multidisciplinary research cruise (SOGOM-3) onboard the R/V Justo Sierra operated by the National Autonomous University of Mexico from April 21 to May 15, 2017 (Figure 1).

At each station, a CTD/Rosette system (SeaBird 9 plus/General Oceanics) equipped with a fluorescence sensor (ECO WET Labs) and 12 Niskin bottles (General Oceanics, 10-l capacity) were used to record the conductivity, temperature, pressure, and chlorophyll-a fluorescence, and to collect seawater samples at six different standard depths (5, 10, 20, 50, 75, and 100 m) for chemical determinations (nutrients and chlorophyll-a) and identification of phytoplankton cells. To avoid contamination of the samples at each station, the Niskin bottles were rinsed with purified water before each casting. The CTD was cast at a speed of 0.5 m s⁻¹, configured to acquire data at 24 Hz. For safety reasons, each cast was performed 100 m from the bottom.

Immediately after each CTD cast, subsamples of 100 ml for nutrient determinations were collected from Niskin bottles and filtered through two cellulose acetate syringe filters placed in series (0.45 and 0.22 µm, Merck Millipore), stored in acid washed polypropylene containers and finally frozen at −20 °C until analyses.

For chlorophyll-a determinations, subsamples of 2 l were collected and filtered through GF/F filters (Merck Millipore) 47 mm in diameter. The vacuum was created using a stainless-steel vacuum manifold system at 10 pounds per square inch. After filtration, the membranes were stored in Eppendorf Safe-Lock Tubes covered with aluminum foil to avoid the negative effects of light and frozen in liquid nitrogen until analysis. Special precautions were taken during the filtering process under low-light conditions to avoid possible degradation.

For phytoplankton cell determination, 1-l subsamples of seawater were collected in dark glass bottles, fixed with a Lugol acetate solution (4 ml), and kept in the dark until analysis.

2.3 Laboratory analyses

Immediately after the research cruise, the concentrations of nitrate, soluble reactive P (SRP), and soluble reactive Si (SRSi) were analyzed using a flow injection analysis/automation (FIA) system 300 equipped with a PerkinElmer Lambda 25 spectrophotometer, following the standard methods adapted by Grasshoff et al. (1983). The precision of the analysis with this system was 0.1 µM for nitrate, 0.04 µM for SRP and 0.1 µM for SRSi.

Chlorophyll-a was determined using fluorometry immediately after the research cruise. An extraction of the pigment was carried out with 10 ml of 90 % acetone (HPLC grade) under refrigeration for 24 h. To determine chlorophyll-a concentration, an aliquot was measured using a Trilogy Laboratory Fluorometer (Turner Designs) with a non-acidified chlorophyll-a module. During these processes, special precautions were taken to operate under extremely low-light conditions.

Seawater subsamples (1l) collected at each station/depth fixed with Lugol acetate solution were subjected to a decantation process until a volume of 100 ml. This volume was then transferred to a 50-ml Uthermöl sedimentation column for 24 h and kept in the dark to avoid light effects (Edler and Elbrächter 2010). Phytoplankton cells were counted and identified at the group level (Cyanophyceae, Bacillariophyceae, Dinophyceae, Chlorophyceae, Haptophyta, Raphidophyceae, Silicoflagellata, and Phytoflagellates) using an inverted light microscope with phase contrast (Carl Zeiss Axio Vert), following standard keys (e.g., Cupp 1943; Komárek and Anagnostidis 1986; Throndsen 1997; Throndsen et al. 2003; Tomas 1997). Finally, the identified cells were standardized to abundance units (cells l⁻¹), according to Edler and Elbrächter (2010).

According to the objectives of our study and owing to the large volume of samples (63 sampling sites and six depths), the identification of the organisms at the species level was not considered; however, we identified the organisms at the group level. Identification of organisms at the group level is appropriate for multivariate methods. In fact, aggregation into major groups has been reported to reflect well-defined gradients of impact more accurately than taxonomic lists at the species level (Ajmal Khan 2006). The information obtained in our study was extremely valuable for identifying vertical and horizontal distribution patterns and determining how the presence of physical structures affects phytoplankton groups. Studies in numerous environments (including the GoM) have applied similar approaches to identify organisms at the group level. Indeed, historical works have pointed out that some meteorological conditions (e.g. “Nortes” events) determine the presence of diatoms in the coasts off Veracruz (Santoyo and Signoret 1978). Two chlorophyll-a peaks have been identified in Campeche Canyon (southern GoM), one of them deep (below 60 m depth) and associated with the limit of the euphotic layer, which is dominated by coccolithophores, whereas the other peak associated with the bottom was described in the Campeche Bank and dominated by silicoflagellates (Durán-Campos et al. 2017). More recently, in waters off the coast of Campeche (southern GoM), it was identified that nanoflagellates are more abundant in regions that present a high concentration of ammonium and phosphates, diatoms dominate in waters rich in silicates, and cyanobacteria tend to be more abundant in waters with temperatures greater than 30 °C (Gómez-Figueroa et al. 2023).

2.4 Data analyses

The CTD data stored at each station were processed at different levels. They were initially converted using the nominal calibration files with the software provided by the manufacturer (SBE Data Processing 7.26.7) following different standard processing routines, applying filters to eliminate spurious or poor-quality data, and then averaging every 1 dB. The conservative temperature and absolute salinity were obtained following the routines of the thermodynamic equation of Sea Water – 2010 (IOC et al. 2010). Maps of the horizontal distributions of the hydrographic parameters were constructed at the standard depths considered in this study (5, 10, 20, 50, 75, and 100 m).

Finally, we performed canonical correspondence analysis (CCA) following standard routines using the CANOCO software (v. 4.5) (Clarke 1993; Ter Braak 1986) to analyze the influence of each environmental variable measured in this study on phytoplankton groups.

3.1 Hydrography

The horizontal distribution of conservative temperatures at different depths showed interesting variations. At 5 m depth, values ranged from 25.5 to 28.2 °C showing their maxima in a core that extends from the southwestern portion to the central part of the study area (Figure 2a). At 10 m depth, values ranged from 25.3 to 27.7 °C with a horizontal distribution pattern like that observed at 5 m depth, with a core located in the southwestern/central portion of the study area (Figure 2b). At 20 m depth, the conservative temperature ranged from 25.3 to 27.4 °C with their maximum values located in a core observed in the central portion of the study area coinciding with the core observed at 5 and 10 m depth, but an additional core was also observed near the southeast coast (Figure 2c). At 50 m, the horizontal distribution was quite different from that observed in the upper layer, with a range of 20.6–26.3 °C, and the presence of a cold core in the central region of the study area (Figure 2d). At a depth of 75 m, the presence of the cold core was evident in the central portion, which registered 19.3 °C (Figure 2e), while at a depth of 100 m two fragmented cold cores were observed, one located in the central portion of the study area and another towards the northwest region, that recorded 18.3 °C (Figure 2f).

Figure 2:

Horizontal distribution of conservative temperature (ºC) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

The absolute salinity also varied depending on the depth. At 5 m depth, the range was 36.3–37.1 g kg⁻¹ with a horizontal distribution pattern where the southern and southeastern portion were less saline than the northern portion. In addition, a core with higher salinity was observed in the northeastern region of the study area near the Yucatan Platform (Figure 3a). At 10 m depth, the absolute salinity varied between 34.4 and 37.0 g kg⁻¹ with a distribution pattern like that observed at 5 m, with an area of lower salinity in the southern region, and a core of higher salinity in the Yucatan platform region (Figure 3b). The same distribution pattern, with fewer saline waters in the southern portion and higher salinity cores in the eastern region, was repeated at depths of 20 m (Figure 3c) and 50 m (Figure 3d). At a depth of 75 m, the core with the highest salinity was observed in the central portion of the study area (Figure 3e), whereas at a depth of 100 m, a large core with less saline water was observed in the central portion, with values <36.53 g kg⁻¹ (Figure 3f).

Figure 3:

Horizontal distribution of absolute salinity (g kg⁻¹) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

The density values showed contrasting variability depending on the depth, which is in agreement with the conservative temperature distribution. At a depth of 5 m, the density values ranged from 23.4 24.4 kg m⁻³, with a large core of lower density in the central portion of the study area (Figure 4a). The horizontal distribution of density at 10 m depth was very similar to that observed at 5 m, with a large core of lower density reaching 23.6 kg m⁻³ in its center (Figure 4b). At 20 m depth, the density values ranged between 23.7 and 24.4 kg m⁻³ with a core of higher density in the northern region of the study area (Figure 4c). The density distributions in the deeper layers were very different. At a depth of 50 m, a high-density core was observed in the central portion, reaching values >25.5 kg m⁻³, which is in agreement with the core of the lower temperature observed at this depth (Figure 4d). At depths of 75 m (Figure 4e) and 100 m (Figure 4f), the same pattern was observed, with higher-density cores in the southern portion reaching values >26.0 kg m⁻³.

Figure 4:

Horizontal distribution of density values (kg m⁻³) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

3.2 Nutrients

The distribution of nitrates and nitrites showed high horizontal and vertical variability. As expected, the concentration increases with increasing depth. At 5 m depth, the lowest concentration was observed (1.13–2.78 µM) with a core of higher concentration in the central portion (Figure 5a). At a depth of 10 m, the concentration increases slightly, reaching a maximum value of 3.2 µM (Figure 5b). At a depth of 20 m, the range varied between 1.5 and 4.1 µM with the presence of fragmented cores of different concentrations in the central portion of the study region (Figure 5c). At 50 m depth, the maximum concentration reached 6.6 µM with the presence of cores of different concentrations throughout the sampling region (Figure 5d). The concentration of nitrates increased to 10.7 μM at 75 m depth (Figure 5e) and 15.7 μM at 100 m depth (Figure 5f); in both cases, cores with different concentrations were also observed throughout the study region.

Figure 5:

Horizontal distribution of nitrates + nitrites (µM) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

The concentration of phosphates was low, reaching 0.86 µM in the entire sampling region in all depths. At 5 m depth, the concentration presented a range of 0.08–0.47 µM with the highest values observed in a core in the northern region. A secondary core with a high concentration was observed in the central portion, reaching 0.40 µM (Figure 6a). At 10 m depth, two cores of high concentration were observed, one in the southeastern part of the study area with 0.30 µM, while the second one was observed in the central region with values of 0.28 µM (Figure 6b). At 20 m depth the concentration ranged from 0.07 to 0.25 µM with fragmented cores of different values that amounted 0.25 µM (Figure 6c). At 50 m depth, two well-defined high-concentration cores were observed, the first was in the central portion with values reaching 0.47 µM, while the second one was detected in the southeastern portion, with values >0.40 µM (Figure 6d). At a depth of 75 m, the maximum concentration was 0.68 µM located in the core located in the central portion of the study region (Figure 6e). The maximum concentration for all depths (0.83 µM) was recorded at 100 m depth, with different fragmented cores distributed along the study area (Figure 6f).

Figure 6:

Horizontal distribution of phosphates (µM) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

Similar to the above nutrients, the concentration of silicates showed wide variations, both in their concentrations and in their horizontal and vertical distributions. At 5 m depth, the concentration ranged between 1.1 and 6.5 µM, with the maximum values observed in a core in the central portion of the study area (Figure 7a). At 10 m depth, the concentration ranged between 1.1 and 2.9 µM with the highest values observed in the southern portion (Figure 7b). At 20 m depth, two cores of high concentration were observed, one in the northwestern portion that registered 4.4 µM and another one located in the central portion of the study area, with values >4 µM (Figure 7c). The Silicates values at 50 m depth ranged from 1.1 to 6.7 µM, with the maximum concentration located in a core in the central portion (Figure 7d). At the depth of 75 m, two cores with high concentrations were observed in the northern and central portions of the study area reaching 7.2 and 6.5 µM, respectively (Figure 7e). At 100 depth, the maximum concentration reached 10 µM in two cores located in the southwestern portion of the study area, one of them close to the coast (Figure 7f).

Figure 7:

Horizontal distribution of silicates (µM) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

3.3 Chlorophyll-a

Chlorophyll-a is still recognized as one of the best proxies for phytoplankton biomass. In our study, the concentration was heterogeneous, both vertically and horizontally.

The lowest concentrations were observed at the surface of the water column. At 5 m depth (Figure 8a), it registered a range of 0.03–0.37 mg m⁻³, while at 10 m depth (Figure 8b) it varied from 0.07 to 0.37 mg m⁻³. In both cases, cores with different concentrations were observed in the southeastern and central regions. At 20 m depth, the values ranged from 0.05 to 0.53 mg m⁻³, with two cores of high concentration located in the southeastern and central regions (Figure 8c). The chlorophyll-a concentration increased at 50 m depth, in a range of 0.07–0.89 mg m⁻³ with maximum values found in a wide portion of the southern region (Figure 8d). The maximum concentration was observed at 75 m, reaching 1.15 mg m⁻³, with a high-concentration core located in the southeastern portion (Figure 8e). Finally, at 100 m depth, the chlorophyll-a values varied from 0.02 to 0.58 mg m⁻³, with two cores of high concentration located in the northern portion of the study area (Figure 8f).

Figure 8:

Horizontal distribution of chlorophyll-a (mg m⁻³) at different depths: (a) 5 m, (b) 10 m, (c) 20 m, (d) 50 m, (e) 75 m and (f) 100 m.

3.4 Phytoplankton groups

The water samples from different depths allowed us to identify eight phytoplankton groups: Cyanophyceae, Bacillariophyceae, Dinophyceae, Chlorophyceae, Haptophyta, Raphidophyceae, Silicoflagellata and Phytoflagellates, which were found in high variable abundances.

Dinophyceae was the most abundant group with 454,160 cells l⁻¹. Phytoflagellates reached a total abundance of 171,939 cells l⁻¹, while Cyanophyceae presented 50,720 cells l⁻¹. Bacillariophyceae registered 25,457 cells l⁻¹, Haptophyta recorded 15,851 cells l⁻¹, and Silicoflagellata recorded 7,940 cells l⁻¹, whereas the two groups with the lowest abundances were Raphidophyceae and Chlorophyceae, with 1,557 and 882 cells l⁻¹, respectively (Figure 9).

Figure 9:

Total abundance (cells l⁻¹) of eight phytoplankton groups in the Southern Gulf of Mexico during the spring of 2017. Cyan, Cyanophyceae; Dino, Dinophyceae; Baci, Bacillariophyceae; Chlo, Chlorophyceae; Hap, Haptophyta; Rap, Raphidophyceae; Sili, Silicoflagellata; Phyto, Phytoflagellates.

The abundance of each phytoplankton group varied significantly depending on sampling depth. Cyanophyceae showed the highest abundance at the depths of 5 and 10 m. Dinophyceae were particularly abundant at depths of 10 m and 50 m, whereas the abundance of Bacillariophyceae was similar at all sampling depths. Chlorophyceae were abundant at 5 and 10 m depths, and Haptophyta was most abundant at 75 m depth. The abundance of Raphidophyceae was uniform across all sampling depths. Silicoflagellata showed maximum abundance values at 50 m depth, whereas phytoflagellates showed their highest abundance at 100 m depth (Figure 10).

Figure 10:

Abundance (cells l⁻¹) of eight phytoplankton groups in the Southern Gulf of Mexico during the spring of 2017 at each sampling depth. Cyan, Cyanophyceae; Dino, Dinophyceae; Baci, Bacillariophyceae; Chlo, Chlorophyceae; Hap, Haptophyta; Rap, Raphidophyceae; Sili, Silicoflagellata; Phyto, Phytoflagellates.

The horizontal distribution of the total abundances of the eight phytoplankton groups at all sampled depths at each station also showed high variability (Figure 11). The highest abundances (>10,000 cells l⁻¹) were found in three main large regions, one was the region of the Campeche Canyon, the second one was at the central portion of the study area, while a third region with high values was registered in the northern portion. Relatively high abundance values were also observed in the southern zone of the study area at stations near the coast. Notably, the high values coincided with regions where nutrient-rich cold cores were observed, particularly in the southern part of the study area.

Figure 11:

Horizontal distribution of the total abundance (cells l⁻¹) of the eight phytoplankton groups in all sampled depths at each station.

The CCA ordination analysis (Figure 12) showed that the two first axes explained the 74.5 % of the accumulated variance (Supplementary Table S1), with a correlation gradient among the abundances of the eight phytoplankton groups and the hydrographic and chemical parameters. The ordination biplot graph showed that the phytoplankton groups were associated with specific parameters: Dinophyceae and Bacillariophyceae were correlated with the conserved temperature, longitude, and concentration of silicates, whereas silicoflagellates were influenced by absolute salinity and phosphate concentration. Cyanophyceae and Chlorophyceae seemed to respond to the concentration of dissolved oxygen and the total depth of each station, whereas the remaining groups (Haptophyta, Raphidophyceae, and Phytoflagellates) were related to the concentration of phosphates and nitrate + nitrite. The correlation values are summarized in Supplementary Table S2.

Figure 12:

Canonical correspondence analysis of the total abundance of the eight phytoplankton groups in all sampled depths (black triangles) related to the hydrographic and chemical parameters recorder (red arrows). Cyan, Cyanophyceae; Dyn, Dinophyceae; Baci, Bacillariophyceae; Chlo, Chlorophyceae; Hap, Haptophyta; Rap, Raphidophyceae; Sili, Silicoflagellata; Phy, Phytoflagellates; DO, dissolved oxygen; T, conservative temperature (ºC); Z, total depth of each station; S, absolute salinity (g kg⁻¹); NO₃+NO₂, nitrates + nitrites; PO₄, phosphates; SiO₂, silicates; Long, longitude.