Phytoplankton Community Composition Determined From Co-variability Among Phytoplankton Pigments From the NAAMES Field Campaign

Sasha J Kramer,David A Siegel,Jason R Graff

doi:10.3389/fmars.2020.00215

Abstract

Analysis of phytoplankton chemotaxonomic markers from high performance liquid chromatography (HPLC) pigment determination is a common approach for evaluating phytoplankton community structure from ocean samples. Here, HPLC phytoplankton pigment concentrations from samples collected underway and from CTD bottle sampling on the North Atlantic Aerosols and Marine Ecosystems Study (NAAMES) are used to assess phytoplankton community composition over a range of seasons and environmental conditions. Several data-driven statistical techniques, including hierarchical clustering, Empirical Orthogonal Function, and network-based community detection analyses, are applied to examine the associations between groups of pigments and infer phytoplankton communities found in the surface ocean during the four NAAMES campaigns. From these analyses, five distinguishable phytoplankton community types emerge based on the associations of phytoplankton pigments: diatom, dinoflagellate, haptophyte, green algae, and cyanobacteria. We use this dataset, along with phytoplankton community structure metrics from flow cytometric analyses, to characterize the distributions of phytoplankton biomarker pigments over the four cruises. The physical and chemical drivers influencing the distribution and co-variability of these five dominant groups of phytoplankton are considered. Finally, the composition of the phytoplankton community across the onset, accumulation, and decline of the annual phytoplankton bloom in a changing North Atlantic Ocean is compared to historical paradigms surrounding seasonal succession.

Highlights

The North Atlantic Ocean has long been a location of significant oceanographic interest due to its role in oceanic primary productivity, carbon sequestration, and climate mediation (Longhurst, 1998; Behrenfeld, 2014; Siegel et al, 2014)
Previous high performance liquid chromatography (HPLC) phytoplankton pigment-based analyses of phytoplankton successional processes for this region (i.e., Barlow et al, 1993; Sieracki et al, 1993; Taylor et al, 1993) have found that the onset and accumulation phases of the North Atlantic spring phytoplankton bloom are dominated by diatoms, which are hypothesized to thrive under turbulent physical conditions (Margalef, 1978)
The highest mean Fuco:Tchla and mean Perid:Tchla ratios were found in the surface ocean on North Atlantic Aerosols and Marine Ecosystems Study (NAAMES) 2, suggesting more diatoms and dinoflagellates compared with other cruises

Summary

Introduction

The North Atlantic Ocean has long been a location of significant oceanographic interest due to its role in oceanic primary productivity, carbon sequestration, and climate mediation (Longhurst, 1998; Behrenfeld, 2014; Siegel et al, 2014). HPLC analysis quantifies the composition and concentration of phytoplankton specific pigments, allowing for chemotaxonomic characterization of the phytoplankton community based on established relationships between pigments and various taxonomic groups. Applications of these different approaches have resulted in an understanding of seasonal trends in community structure that have been associated with both bottom-up (i.e., nutrients, light availability, turbulent mixing) and top-down factors (e.g., grazing by zooplankton). Previous HPLC phytoplankton pigment-based analyses of phytoplankton successional processes for this region (i.e., Barlow et al, 1993; Sieracki et al, 1993; Taylor et al, 1993) have found that the onset and accumulation phases of the North Atlantic spring phytoplankton bloom are dominated by diatoms, which are hypothesized to thrive under turbulent physical conditions (Margalef, 1978). Communities of haptophytes and dinoflagellates follow the peak of the diatom bloom, with background communities of green algae and cyanobacteria thriving in these lower-nutrient periods

Methods

Results

Conclusion