Search the data

AMT19 (JC039) pigment composition by HPLC analysis from CTD bottle samples

Data Acquisition and Analysis

Samples (1-2L) were vacuum filtered onto GF/F filters, flash frozen in liquid nitrogen and stored at -80°C on board ship. Samples were shipped back to UK on dry ice. On receipt at PML, a small amount of dry ice remained around samples. At PML, samples were stored at -80°C until analysis. Pigments were extracted from the sample filters into 2 mL 100% acetone using an ultrasonic probe (35 s; 50W). An internal standard was used (Trans-B-Apo-8'-carotenal) to correct for water content of filters. Extracts were centrifuged to remove filter and cell debris (5 min at 4000 rpm) and analysed by reversed phase HPLC, using a method based on Barlow et al. 1997. Analyses were performed using a Thermo Accela Series HPLC system with chilled autosampler (4°C) and photodiode array detector. The HPLC was calibrated using a suite of standards purchased from DHI (Denmark). Pigments were identified based on retention time and spectral match using photo-diode array.

Instrumentation Description

Thermo Accela Series HPLC system

BODC Data Processing Procedures

These data supersede data previously supplied to BODC without quality assurance (see the Data Quality Report section for more details). Data were submitted to BODC in Microsoft Excel spreadsheet format and saved to the archive with reference PML130050. The file was provided to BODC with date, latitude, longitude, station number and depth as metadata. Sample metadata were checked against information held in the database - there were no discrepancies.

The following parameters were provided in the file but only the parameters that were not derived from the summation of other parameters in the dataset were loaded into the database. The absent data value was recorded as the effective limit of detection (LOD). The effective limits of detection, calculated for a filter volume of 3L are given in the table below.

There were duplicate data provided for 12 samples (cast 5s, depth 1 m and 5 m; cast 8s, depths 1 m, 10 m, 17 m and 38 m; cast 15t, depths 1 m, 10 m and 20 m; cast 71s, depths 10 m, 20 m and 35 m) and the mean and standard deviation for these values were entered into the database for each sample.

Parameter codes defined in BODC parameter dictionary were assigned to the variables as shown in the table below. The data were provided in µg l ^-1 units that were converted to ng l ^-1 for all parameters but chlorophyll-a.

Data loaded into BODC's database without any further changes.

Data Quality Report

Dataset originally obtained using inadequate quality assurance (QA) procedures (for an example of adequate QA procedures during pigment analysis, see Van Heukelem and Hooker (2011)). A series of duplicate samples spanning the chl-a concentration range observed in the dataset were analysed using a quality assured, validated HPLC method by an experienced pigment scientist. A comparison between the original and duplicate sample pigment concentrations was performed, and systematic errors for individual pigments identified. The AMT19 pigment dataset was corrected for the systematic errors identified, and validated against surface pigment samples collected during AMT19 , and analysed independently by a validated pigment method in a NASA laboratory (Hornpoint). The root mean square deviation (RMSD) of the two sample sets (PML and Hornpoint) was determined and normalized to the concentration range for each pigment. The average normalized root mean square deviation (NRMSD) for the primary pigments was 7.9% (see table below). Notably, the highest NRMSD value for the primary pigments was observed for diatoxanthin, which was present at low concentration in all but one of the compared samples.

The precision for replicate filters (gives whole method precision, from filtering to analysis) was provided in a separate file and is available on request to BODC. The average coefficients of variation (CV%) were 7.97 and 15.04 for [TChl a] and [TPig], respectively.

Samples values provided with the effective LOD were flagged '<' for each parameter in the database.

Problem Report

This section is not relevant to this data set.

References Cited

Barlow R.G., Cummings D.G. and Gibb S.W., 1997. Improved resolution of mono- and divinyl chlorophylls a and b and zeaxanthin and lutein in phytoplankton extracts using reverse phase C8 HPLC. Marine Ecology Progress Series, 161 , 303-307.

Van Heukelem L. and Hooker S.B. 2011. The importance of a quality assurance plan for method validation and minimizing uncertainties in the HPLC analysis of phytoplankton pigments. In: Phytoplankton Pigments: Characterization, Chemotaxonomy and Applications in Oceanography, eds. Suzanne Roy, Carole A. Llewellyn, Einar Skarstad Egeland and Geir Johnsen. Cambridge University Press.