Phytoplankton pigment in situ measurements uncertainty evaluation: an HPLC interlaboratory comparison with a European-scale dataset

Abstract

Phytoplankton pigment data play a crucial role in ecological studies, enabling the identification of algal groups and estimation of primary production rates. Accurate measurements of chlorophyll a (TChl a) and other marine pigments are essential for the development of bio-optical algorithms and the validation of satellite data products. High-performance liquid chromatography (HPLC) is the gold standard method for quantifying multiple pigments in a single water sample. This study aims to investigate the uncertainties associated with phytoplankton pigment quantification by comparing duplicate sample analyses conducted by two laboratories, the Joint Research Centre of the European Commission (J) and the DHI Group, Denmark (D). The analyses were performed using the same HPLC method. The dataset comprised 957 natural samples collected between 2012 and 2017 from various European seas, representing different trophic conditions with TChl a concentrations ranging from 0.083 to 27.35 mg/m3. The study compared the results of the two independent analyses for TChl a and primary phytoplankton pigments, including chlorophyll b, chlorophyll c, carotens, fucoxanthin, 19′-butanoyloxyfucoxanthin, diadinoxanthin, diatoxanthin, 19′-hexanoyloxyfucoxanthin, peridin, and zeaxanthin. The percent difference between the two analyses was calculated to assess the uncertainties associated with pigment quantification. The mean percent difference observed between the two independent analyses of TChl a was 10.8%. For the primary phytoplankton pigments, the associated mean percent difference was 16.9%. These results meet the requirements of 15% and 25% uncertainties, respectively, which are applicable for the validation of satellite data products. The comparative analysis between the two laboratories demonstrates that the uncertainties associated with phytoplankton pigment quantification are within acceptable ranges for the validation of satellite data products. Moreover, the study investigates the propagation of uncertainties in diagnostic pigment values to phytoplankton indexes, which are derived using pigment-based algorithms to characterize phytoplankton populations according to functional types.