Biophysical interactions control the size and abundance of large phytoplankton chains at the Ushant tidal front

Abstract

Phytoplankton blooms are usually dominated by chain-forming diatom species that can alter food pathways from primary producers to predators by reducing the interactions between intermediate trophic levels. The food-web modifications are determined by the length of the chains; however, the estimation is biased because traditional sampling strategies damage the chains and, therefore, change the phytoplankton size structure. Sedimentological studies around oceanic fronts have shown high concentrations of giant diatom mats (>1 cm in length), suggesting that the size of diatom chains is underestimated in the pelagic realm. Here, we investigate the variability in size and abundance of phytoplankton chains at the Ushant tidal front (NW France) using the Video Fluorescence Analyzer (VFA), a novel and non-invasive system. CTD and Scanfish profiling characterized a strong temperature and chlorophyll front, separating mixed coastal waters from the oceanic-stratified domain. In order to elucidate spring-neap variations in the front, vertical microstructure profiler was used to estimate the turbulence and vertical nitrate flux. Key findings were: (1) the VFA system recorded large diatom chains up to 10.7 mm in length; (2) chains were mainly distributed in the frontal region, with maximum values above the pycnocline in coincidence with the maximum chlorophyll; (3) the diapycnal fluxes of nitrate enabled the maintenance of the bloom in the frontal area throughout the spring-neap tidal cycle; (4) from spring to neap tide the chains length was significantly reduced; (5) during neap tide, the less intense vertical diffusion of nutrients, as well as the lower turbulence around the chains, intensified nutrient-depleted conditions and, thus, very large chains became disadvantageous. To explain this pattern, we suggest that size plasticity is an important ecological trait driving phytoplankton species competition. Although this plasticity behavior is well known from experiments in the laboratory, it has never been reported from observations in the field.

Figure 1. Study location and delimitation of the oceanographic regions.

Map of the study area off NE Atlantic region (A). Station position and hydrological typology at the Ushant tidal front: mixed, frontal, and stratified (B). nMDS plot based on hydrological data with superimposed cluster analysis at a Euclidean distance of 4.7 (–)(C). Each dotted circle represents a significant cluster (SIMPROF P<0.05).

Figure 2. Schematic representation of the Video Fluorescence Analyzer (VFA).

Imaging system components (A), and cartoon of the mechanical design of the VFA (B).

Figure 3. Snapshot of fluorescent particles imaged with the Video Fluorescence Analyzer (VFA).

Example of images taken in the tidal front during neap (A, B) and spring tides (C, D). The most representative phytoplankton chains are shown in the panels. The shape and cell connection type of the longest phytoplankton chains (red arrows) suggest that they belong to Pseudonitzschia spp. (C, D).

Figure 4. Cross-shelf sections of the water column using CTD-Scanfish.

Contour plots of temperature (A, B), chlorophyll a (C, D), and turbidity (E, F) during spring tide (A, C, E: 21/09/2009) and neap tide (B, D, F: 28/09/2009) along the transect. Numbers along the top of the panels A and B refer to sampling stations (Fig. 1). Italic bold numbers indicate stations located in the front.

Figure 5. Turbulence, diapycnal diffusivity, and nutrient fluxes in the study area.

Distribution of the turbulent kinetic energy dissipation rate (A, B), diapycnal eddy diffusivity (C, D), and vertical nitrate flux (E, F) along the transect during spring tide (A, C, E) and neap tide (B, D, F). The vertical dashed grey line shows the station 25 (unsumpled). The pycnocline is marked with a thick black line and corresponds to the σ_t = 26.6 and σ_t = 26.8 isopycnals during spring and neap tides, respectively. Numbers along the top of the panels A–D refer to sampling stations (Fig. 1). Italic bold numbers indicate stations located in the front. The turbulent kinetic energy dissipation rate and diapycnal eddy diffusivity are plotted on a base 10 log scale.

Figure 6. Nutrients in the study area.

Nitrate (A, B) and silicate concentrations along the transects (E, F). Mean size of the phytoplankton chains relative to depth and nitrate concentrations (C) and silicate (D). The panels located in the left and right refer to observations from spring and neap tide, respectively. Numbers along the top of the panels A, B, E, and F refer to sampling stations (Fig. 1). Italic bold numbers indicate stations located in the front. In panels C, D, G, and H, the mean size of chains at each sample are shown as proportional bubbles. Note that the stations with abundance lower than 5×10³ cells L⁻¹ were not added to the bubble plots. Vertical grey line = 13°C bottom thermocline.

Figure 7. Distribution of phytoplankton chains in the study using the Video Fluorescence Analyzer (VFA).

Abundance (A, B) and mean size (C, D) of chains. The panels located in the left and right refer to observation from spring and neap tides, respectively. The pycnocline is marked with a thick red line and corresponds to the σ_t = 26.6 and σ_t = 26.8 isopycnals during spring and neap tides, respectively. Numbers along the top of the panels A and B refer to sampling stations (Fig. 1). Italic bold numbers indicate stations located in the front.

Figure 8. Multivariate statistical analysis.

Principal components analysis (PCA) for all environmental variables labeling different factors: region (A), vertical position (B), and tidal phase (C). In panel D, the average abundance of phytoplankton chains of each sample is superimposed as proportional bubbles over the PCA. Fluor, chlorophyll; D, density; P, phosphate; Turb, turbidity; Sal, salinity; Si, silicate; TKE, dissipation rate; T, temperature.

Figure 9. Comparison of the phytoplankton chain length spectra with Kolmogorov length scale in the frontal area.

Spring tide (A). Neap tide (B). The bottom samples were removed from this analysis. In each box plot, the median (solid line) and mean (bold line) of the maximum axis length data are indicated in the center of the box and the edges of the box are the 25^th and 75^th percentiles; the whiskers extend to the most extreme data points that were not considered to be outliers.