Integrating phytoplankton pigment andDNAmeta‐barcoding observations to determine phytoplankton composition in the coastal ocean

Abstract

Quantifying phytoplankton composition is critical to predicting marine ecosystem structure and function. DNA meta‐barcoding and high‐performance liquid chromatography (HPLC) pigment analysis are two widely used methods for assessing phytoplankton composition; however, comparing their performance has been done only rarely. Here, we integrate DNA meta‐barcoding and HPLC pigment observations to determine eukaryotic phytoplankton composition in the Santa Barbara Channel, California. We find that both methods identify the same four dominant eukaryotic phytoplankton taxa (diatoms, dinoflagellates, chlorophytes, and prymnesiophytes), but inter‐ and intra‐lineage variability in biomarker pigmentation (associated with both a lack of taxonomic specificity of biomarker pigments and intrinsic differences in accessory pigmentation) drives substantial disagreement between the methods. Covariation network analysis circumvents this disagreement and reveals that diverse assemblages of phytoplankton and other protists covary with distinct suites of biomarker pigments. Our results highlight the strengths and weaknesses of each method in characterizing phytoplankton composition and reveal novel insights into phytoplankton physiology that could only be gained by integrating the two methods. Finally, we suggest a path to monitor eukaryotic plankton communities on unprecedented spatiotemporal scales based on the covariation of unique phytoplankton and protistan assemblages with remotely sensible phytoplankton pigment concentrations.

Toward a synthesis of phytoplankton community composition methods for global-scale application

“…Third, some genera or species may co-occur in the environment, leading to the covariation of unexpected taxa with a pigment that they may not contain, but others in the group do. Finally, phytoplankton pigments may vary in composition and concentration due to phytoplankton physiological responses to the physical environment (Thompson et al, 2007), which includes light history (particularly as many pigments have photoprotective functions, including Allo and Zea) and nutrient availability (Schlüter et al, 2000; Henriksen et al, 2002; Catlett et al, 2022).…”

“…If accessory pigments can be accurately modeled from satellite measurements and the comparisons between pigments and amplicon sequencing or IFCB datasets continues on broader spatiotemporal scales, more comprehensive relationships can be developed between pigments and phytoplankton groups determined via other methods throughout the global ocean (Catlett et al, 2022; Chase et al, 2022). By comparing performance across methods, we also encourage consistency in sampling approaches, laboratory analysis, and method development.…”

“…For instance, dinoflagellates are an important phytoplankton group (particularly in coastal regions, where they may form toxic blooms), but changes in their relative abundance was not well exclained using Perid/Tchla in the current study. Datasets that have collected samples across a broader range in biomass and under varying physical and biogeochemical conditions, including both coastal (Lin et al, 2019; Catlett et al, 2022) and open ocean systems (Chase et al, 2022) will be ideal for further comparison.…”

“…Similar results were found in the West Antarctic Peninsula, where high relative contributions from cryptophytes, diatoms, and prymnesiophytes resulted in significant positive relationships between pigments and 18S (Lin et al, 2019), and in the Neuse River Estuary, where relatively high chlorophyte sequence abundances from 18S correlated well with pigments (Gong et al, 2020). Alternately, in regions with highly dynamic phytoplankton communities and year-round monitoring, the relationships between pigments and other PCC methods were not as clearly defined, due in part to the variability in accessory pigment composition and concentration across the seasonal cycle (Catlett et al, 2022). The capacity of pigments to separate different phytoplankton groups is similarly very dataset-dependent (Kramer and Siegel, 2019), with more and different groups occurring on local scales than global scales.…”

“…The assumptions inherent in some methods limit the interpretation of the results, such as the challenge of unequal copy numbers of the 16S and 18S genes across taxa (e.g., Godhe et al, 2008; de Vargas et al, 2015; Needham and Fuhrman, 2016). Comparisons between and among PCC methods are relatively rare, and reveal variability when different methods are compared (e.g., Not et al, 2008; Couple et al, 2015; Gong et al, 2020; Campbell et al, 2022; Chase et al, 2022; Catlett et al, 2022; Nardelli et al, 2023). In one example, amplicon sequencing and light microscopy each provide high resolution taxonomic information for larger phytoplankton, but abundance patterns did not agree in genus- to species-level comparisons (Abad et al, 2016).…”