Pigment and fluorescence proxies to estimate functional diversity of phytoplankton communities

Abstract

Given the global loss of biodiversity, trait-based studies are needed to assess the associated consequences for ecosystem functions and services. Many studies focus on the assessment of functional diversity of natural communities as a mechanistic link between biodiversity and ecosystem functioning. In freshwater ecosystems, diversity of primary producers is crucial for resource use efficiency and trophic transfer of energy. Furthermore, one indicator of the ecological status of surface waters is the composition of natural phytoplankton communities. The number of available techniques for the quantification and discrimination of different phytoplankton groups have increased in recent years. For example, phytoplankton community composition can indirectly be assessed via CHEMTAX, a matrix factorization program, which calculates the contribution of different phytoplankton taxa to the total chlorophyll-a using concentrations of pigments analysed via liquid chromatography. A more direct, in vivo assessment can be achieved with instruments based on spectral fluorometry, such as the Algae Lab Analyser, which allows for a differentiation of four phytoplankton groups depending on spectral fluorescence signatures. In this study, we compared both methods by analyses of phytoplankton biomass and functional diversity from phytoplankton communities of three lakes of different trophic states, while a subset of biomass and diversity estimates derived from microscopic counts served as a reference. We found marked differences in biomass estimates of all assessed phytoplankton groups, with cyanobacteria being significantly underestimated by the Algae Lab Analyser. Furthermore, we show that the level of agreement between the methods somewhat depends on the trophic state of the lake. We conclude that both methods are suitable to estimate phytoplankton functional diversity with specific advantages and disadvantages. Here we provide users with a flow chart to help them find the most suitable method for their respective purposes.

A tribute to Winfried Lampert: how he injected an integrative vision of evolutionary ecology into the aquatic sciences

“…And how has Winfried Lampert contributed to advance our understanding of the importance of those factors? As indicated above, Winfried's early work was focused on carbon as the determining factor, but as early as 1987, in a paper on feeding and nutrition in Daphnia (Lampert 1987), he acknowledged that "salad is not always just salad" (see also , and that different algal species represent different qualities of food for Daphnia (Ilić et al 2023).…”

Differences in pigment composition and concentration between phototrophic, mixotrophic, and heterotrophic Chrysophyceae

“…The pigment extraction was performed following Ilić et al (2023). Briefly, the frozen filters were transferred into separate 10 ml glass test tubes and 4 ml acetone (100%, HiPerSolv Chromanorm, HPLC Grade) were added.…”

“…We used a modified high-performance liquid chroma tography (HPLC) protocol based on the protocol by Garrido & Zapata (1993), see Ilić et al (2023) for details on the methodology. We applied reversed-phase HPLC using the Shimadzu Prominence series system (DGU-20A3 degasser, LC-20AB binary pump, CTO-10AC column oven set to 40 °C, SPDM20A diode array detector).…”

“…For the calibration, previously available and new data were used. Calibration curves were established based on dilution series (at least five dilution steps in triplicates) of chlorophyll-c 2 and c 3 and diadinoxanthin, neoxanthin, and violaxanthin, while for all other pigments, calibration curves from previous analyses were used (Ilić et al 2023). All pigment peak areas were in tegrated at 436 nm and corrected for the internal standard, and the pigment concentration was calculated based on established calibration curves.…”

“…Data analysis. The peaks of each chromatogram were integrated using LabSolution, peaks with a height lower than 75 mAU (milliabsorption units) were not processed further (limit of detection as established from previous measurements (Ilić et al 2023). We analysed the data using R (version 3.1.1, R Core Team 2014; ggplot2 version 2.2.1, Wickham 2009;reshape2 version 1.4.2, Wickham 2007) and RStudio (version 1.0.153, RStudio Team 2020).…”