the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Mar 2026
Precision of Phytoplankton Pigment Analysis by High Performance Liquid Chromatography: An Assessment of the Global Ocean Color Validation Dataset Analyzed by NASA
Abstract. Space-borne ocean color sensors capable of measuring phytoplankton pigments, such as chlorophyll a, have greatly expanded our understanding of oceanic biological processes. The ability to generate such measurements in a way that satisfies the requirements of climate-quality data records is contingent in part on the quality of the in situ ground or sea truth observations that serve as datasets for vicarious calibration and algorithm validation activities. The National Aeronautics and Space Administration (NASA) has a mandate to collect and distribute in situ data of the highest quality to support data product validation for ocean color missions; hence the agency uses a centralized, quality-assured laboratory to perform high performance liquid chromatography (HPLC) analysis of pigment samples collected by NASA-affiliated investigators. Since its establishment in 2011, the facility at NASA’s Goddard Space Flight Center has processed over 30,000 samples collected in all the major ocean basins. We evaluated the replicate sample precision, measured as the percent coefficient of variation among replicates, for total chlorophyll a and all primary, secondary, and tertiary pigments to investigate the sources of variability in analytical measurements. Mean analytical precision (CV %) ranged from 3.2 % for divinyl chlorophyll a to 17.1 % for chlorophyllide a. The analytical precision performance benchmarks for total chlorophyll a (5 %) and primary pigments (8 %), established for legacy ocean color missions, were met for total chlorophyll a and for 10 of 12 primary pigments. Two primary pigments exceeded the 8 % benchmark: diatoxanthin (8.6 %) and peridinin (9.2 %). No performance benchmarks have been established for secondary or tertiary pigments. Precision was evaluated against average sample concentration, pigment mass injected into the HPLC instrument, filtered volume, estimated phytoplankton size-fractions (micro-, nano-, and picoplankton), and sample origin (coastal versus oceanic) using multivariate regression and non-parametric approaches. Neither concentration nor pigment mass appeared as significant drivers of precision variability across their ranges. Precision showed minimal variation across concentration ranges for total chlorophyll a and primary pigments but deteriorated toward detection limits for secondary and tertiary pigments when samples with invariant replicates (CV % = 0) were excluded from analysis. Filtration volumes >1000 mL generally improved precision, though it degraded at higher volumes for some pigments in censored datasets, suggesting physical stresses during extended filtration. For most pigments, sample precision was statistically poorer in coastal versus oceanic samples, though previous interlaboratory comparisons suggest this reflects methodological rather than biogeochemical factors. Multivariate regression models explained ≤ 3.0 % of precision variability in full datasets but up to 44 % when invariant replicates were excluded, indicating that analytical precision is primarily governed by methodological factors rather than systematic dependencies on sample characteristics. The distinction between analytical precision and sample heterogeneity emerged as a possible factor, with increased variability at low concentrations for rare taxa likely reflecting stochastic cell capture during filtration rather than analytical limitations. These findings demonstrate that rigorous quality assurance protocols achieve precision performance suitable for climate-quality ocean color validation, with pre-analytical sample processing and field replication strategies identified as priorities for further improvements.
- Preprint
(4182 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1540', Anonymous Referee #1, 24 Apr 2026
-
AC1: 'Reply on RC1', Joaquin Chaves, 22 Jun 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1540/egusphere-2026-1540-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Joaquin Chaves, 22 Jun 2026
-
RC2: 'Comment on egusphere-2026-1540', Anonymous Referee #2, 26 Apr 2026
General Comment
The aim of this paper is to assess the precision of high-performance liquid chromatography (HPLC) analysis of phytoplankton pigments, which are widely used as reference data for satellite validation and bio-optical algorithm development. This is a topic of clear relevance, as relatively few studies have specifically evaluated the uncertainties associated with this analytical method, despite the need for high-quality in situ pigment measurements in satellite ocean-color validation exercises. The work is therefore important and timely; however, several aspects of the manuscript require clarification and further development.First, important methodological information is missing regarding the limit of detection (LOD) and limit of quantitation (LOQ) of the analytical procedure. The manuscript does not clearly report the LOD and LOQ values for the different pigments, nor does it indicate how many samples fall below either threshold. It is essential to specify which observations are below the LOD and which are below the LOQ, since these two situations have different implications for data interpretation and uncertainty analysis.
Closely related to this point, it is not clear how values below these analytical thresholds were handled in the statistical treatment. Specifically, the manuscript should clarify whether concentrations below the LOD or LOQ were retained using substituted values, treated as censored data, or excluded entirely from the dataset. This information is nunclear, but foundamental because comparisons of analytical precision can be strongly influenced by the inclusion of values near or below detection capability.
A further issue concerns the classification of samples into oceanic, coastal, and estuarine waters. The concentration ranges associated with these three categories are not sufficiently defined, and the rationale behind the geographical definition of “coastal waters” is unclear. In particular, the manuscript refers to coastal waters as extending to 200 km from shore, but no justification or reference is provided for this criterion. This definition does not correspond to commonly adopted legal or oceanographic delimitations, and therefore requires explanation and appropriate citation (i.e., continental shelf and Exclusive Economic Area are commonly considered 200 nautical miles, c.a. 385km, territorial-coastal woter commonly 12 nautical miles). More importantly, it should be clarified whether these environmental classes are being used purely as geographical descriptors or whether they are implicitly associated with trophic regimes such as oligotrophic, mesotrophic, and eutrophic waters, as seems to emerge in later sections of the manuscript.
Another aspect that appears to be missing is a comparison with the precision obtained from the quantification of pigment standard mixtures, which has already been addressed in previous intercomparison exercises such as the SeaHARRE intercomparisons and also in the cited Canuti et al., 2025. Since those studies assessed analytical reproducibility under controlled standard conditions, it would be highly informative to evaluate whether the precision observed here on natural samples represents an improvement, a deterioration, or a consistency relative to repeated measurements of mixed standards or aliquots from the same analytical batch.
Finally, one fundamental point that must be clarified concerns the replication structure of the dataset. The manuscript indicate observations are based on duplicate analyses, and how many on triplicate or more analyses. However is not clear if it was evluated to which extend the statistical robustness of duplicate-based precision estimates differs from that of triplicate or higher-order replicate analyses. Although some differences between these groups appear to be suggested in the manuscript, the current presentation does not provide enough detail to properly assess their significance.
Overall, the study addresses an important analytical issue, but the manuscript requires clarification of these methodological points before the conclusions on HPLC precision can be considered fully supported.
Detailed Comments:
Abstract.
At line 14, the manuscript refers to primary, secondary, and tertiary pigments; however, at this stage of the paper this classification is not introduced, its basis is not explained, and the reader is left without a clear understanding of which compounds belong to each category. Since this terminology is central to the interpretation of the results, the classification scheme should either be briefly defined in the Abstract or postponed until it is properly introduced in the main text.
A few lines later, the expression “mean analytical precision” should be add "expressed as coefficient of variation (CV%), and should indicate explicitly whether this estimate is based on duplicate analyses, triplicate analyses, or a mixture of both. Similarly, the repeated mention of primary pigments shortly thereafter remains unclear because the compounds included in this category have not yet been identified.
Introduction.
At line 91, replicates are introduced, but once again it is not specified whether these correspond to duplicate measurements, triplicate measurements, or a combination of different replication levels. This information should be made explicit from the outset because it directly affects the interpretation of the analytical precision estimates.
Regarding Table 1, there is a nomenclature issue that should be addressed. For example, monovinyl chlorophyll a is listed among the secondary pigments, but in the associated calculations it appears as “monovinyl chlorophyll a + allomers + epimers.” A similar ambiguity is present for divinyl chlorophyll b and other compounds reported as sums of chemically related forms. This is confusing because the same pigment name seems to refer simultaneously to the pure compound in one column and to an aggregated analytical signal in another. The manuscript should therefore specify chemically and analytically what distinguishes these columns, clearly indicating whether the reported values refer to pure standards, co-eluting derivatives, degradation products, stereoisomers, or summed chromatographic peaks. Also should be specified on which basis the division in primary, secondary and tertiary pigments was established.
Methods.
At line 121, the injector configuration is described as “900 μL syringe head.” This wording appears inaccurate, since 900 μL likely refers to the injection loop volume rather than to the syringe head itself. The terminology should be checked and corrected according to the actual injector configuration.
The manuscript also refers to the use of deuterium and tungsten compartment lamps. Since the deuterium lamp is employed for UV detection whereas the tungsten lamp covers the visible range, it should be clarified that the detector was used not only for visible absorbance measurements of pigments but also for UV absorbance measurements of the internal standard (vitamin E or related compounds), as specified later.
For the solvents and chemicals, the suppliers and manufacturers are not reported. This information should be included for all solvents, standards, and reagents to ensure reproducibility.
The extraction procedure is insufficiently described. In particular, the manuscript does not indicate:
- the extraction solvent volume used,
- whether the same extraction volume was applied to all filter sizes,
- which filter diameters were processed,
- whether mechanical disruption or assisted extraction procedures (e.g., sonication, bead beating, grinding) were employed,
- and whether extraction consisted of a single or multiple extraction steps.Even a concise methodological description is necessary, because these parameters can strongly affect pigment recovery and therefore analytical precision.
At line 133, the manuscript seems to imply that the internal standard is used to calculate the extraction volume. This is conceptually inaccurate. An internal standard can be used to assess extraction recovery, injection reproducibility, or analytical efficiency, but not to directly determine the physical extraction volume unless additional gravimetric or volumetric measurements are performed. This statement should therefore be reconsidered and reformulated.
At line 131, the sentence indicating that “pigments samples” would be more appropriately phrased by stating that " water samples were extracted for subsequent HPLC analysis" or similar, since the current wording is analytically imprecise.
For vitamin E, the supplier is again missing
At line 181, where the dataset is described as encompassing coastal, estuarine, and oceanic waters, it would be useful to report at least the approximate chlorophyll a concentration ranges associated with these water types, so that the reader can better appreciate the trophic and analytical range covered by the study.
At line 198, diagnostic pigments are introduced, but the exact set of pigments considered diagnostic should be explicitly listed. This is important because different approaches exist in the literature—for example, the Uitz et al. framework approach uses weighted coefficients, whereas other formulations based on diagnostic pigment sums adopt different definitions (i.e., Vidussi et al.). The manuscript should therefore clearly state which convention is followed.
At line 209, the manuscript refers to the “extracted volume obtained.” This wording appears questionable, because in standard analytical practice one usually refers to the volume of solvent added for extraction unless the final recovered extract volume was actually measured after processing. If the latter was not done, the terminology should be corrected.
At line 225, adding the pigment mass as calculated would help.
The criterion for excluding observations with CV% = 0 (invariant replicates) also requires a more detailed explanation. In particular, it is not evident whether these zero-variance cases correspond to truly identical replicate quantifications, rounding effects, or values approaching/below detection limits. Since the treatment of such cases can influence the resulting precision statistics, the exclusion criterion should be justified more in details.
Concerning Figure 2d, which compares MODIS-derived and in situ chlorophyll a values, it would be important to specify how many total chlorophyll observations were used in the MODIS calbration/validation and whether these same observations were subsequently retained in the dataset used for the present figure.
In Figure 4, regression coefficients are reported for the logarithmically transformed chlorophyll a data. While this transformation is statistically reasonable, it would also be informative to report the corresponding coefficient of determination for the non-transformed chlorophyll a values, as this would provide a more direct appreciation of the absolute analytical variability.
At line 483, fucoxanthin and diadinoxanthin are discussed as carotenoids for which state-of-the-art analytical accuracy appears to be maintained, in SeaHARRE exercises dedicated to coastal samples. However, this interpretation should be nuanced. These pigments are typically present at relatively high concentrations in many coastal phytoplankton assemblages, which inherently facilitates more robust quantification. Their apparently superior analytical precision may therefore not depend solely on pigment identity, but also on the fact that they are more frequently measured well above the analytical detection and quantitation limits. By contrast, pigments such as 19'-butanoyloxyfucoxanthin or other minor accessory carotenoids are analytically more challenging, and several laboratories report greater difficulty in consistently quantifying them, partly because their concentrations often approach the limit of detection, as also highlighted in previous intercomparison SeaHARRE exercises. Consequently, differences in analytical precision among pigments should not be interpreted only as intrinsic pigment-specific behavior, but also in relation to concentration range and signal-to-noise conditions. In addition, chromatographic position may contribute to these differences. Fucoxanthin and diadinoxanthin generally elute in regions of the chromatogram that are analytically stable and well resolved in many standard HPLC methods, being neither among the earliest nor the latest eluting compounds where baseline instability, solvent front effects, or peak broadening may compromise quantification. Their relatively favorable elution characteristics may therefore also explain why they show consistently good reproducibility across methods.
At line 495, the manuscript states that in Canuti, 2025, the methods were evaluated on 957 samples from the Mediterranean Sea. This description appears inaccurate, as the dataset seems to include samples not only from the Mediterranean but also from other regions, including the Black Sea, the Iberian area, and possibly additional locations mentioned elsewhere in the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-1540-RC2 -
AC2: 'Reply on RC2', Joaquin Chaves, 22 Jun 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1540/egusphere-2026-1540-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Joaquin Chaves, 22 Jun 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 523 | 263 | 39 | 825 | 30 | 45 |
- HTML: 523
- PDF: 263
- XML: 39
- Total: 825
- BibTeX: 30
- EndNote: 45
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Review: “Precision of Phytoplankton Pigment Analysis by High Performance Liquid Chromatography: An Assessment of the Global Ocean Color Validation Dataset Analyzed by NASA” by Chaves et al.
This manuscript presents a comprehensive assessment of replicate-sample precision for phytoplankton pigment measurements obtained by HPLC. The authors make use of a very large and unique global dataset collected from 2011–2022 and provide a rigorous evaluation of analytical precision of pigments relevant to ocean color calibration and validation activities.
The article is clearly written, methodologically sound, and relevant in the context of prior SeaHARRE intercomparison exercises. The results show that the precision benchmarks for total chlorophyll a and most primary pigments are met, and that variability is dominated by pre-analytical and sampling-related factors rather than instrumental/analytical limitations. Overall, the manuscript is scientifically strong and represents a valuable contribution to the ocean color validation and phytoplankton pigment analysis and I recommend publication after a minor revision.
General comments:
Please provide information on the reporting practices, particularly for concentrations near LOD/LOQ, and on the number of significant digits reported. I am also missing an evaluation of how many datasets actually had CV=0%, and how many of these cases were associated with measurements at or below LOD/LOQ. Furthermore, please clarify how such numbers were reported: were concentrations replaced by LOD/LOQ values, e.g., 0.0001, or were zeros reported when pigments were not detected? This information is important for understanding the procedure used to exclude samples with invariant replicates (CV=0%).
Chlide a generally results from artificial degradation of Chl a during extraction and should therefore be added to Chl a. Since most Chlide a quantified by HPLC is not natural occurring it should not be interpreted independently and I suggest removing Chlide a from the figures and most of the text (see a later comment on it).
Specific comments:
L.80: “In the absence of standardized reference materials… intercalibrations are a necessary substitute”?: The use of standardized reference materials and participation in intercalibrations are complementary activities and should both be applied, where possible, to ensure robust quality assurance.
L. 110: one ancillary pigment? Which pigment? It is evident from Table 1. Also, the text states “two sets of pigment sums and ratios routinely reported”; however, Table 1 shows more than two pigment sums and no ratios. Please correct the text/table.
L. 114: “Tertiary pigments are a set of less frequently analyzed pigments” -less frequently reported pigments?
L. 114: Why specifically mention gyroxanthin diester (which is often present in certain regions) while other omitted pigments are not mentioned (dinoxanthin, myxoxanthophyll, C2-MGDG, etc.)?
L. 131: Please specify the extraction solvent (e.g. 90% acetone).
L. 144: 30,000 samples mentioned in the introduction?
L. 190: Correct “pigmen” to pigment.
Table 1: Correct Pheophtyin a to Pheophytin a.
L. 128: “a reference wavelength of 700 nm was added” Do you mean: “…a reference wavelength at 700 nm (±10 nm) was included and used for baseline correction of the 665 nm signal”?
L. 250, Fig. 2d: What is the rationale for comparing satellite data from 2002-2023 to TChl a from 2011-2022? Could the offset observed in Fig. 2d partly reflect the difference in time periods?
L. 257: Can you elaborate on why Diato and Perid in particular exceed the 8 % benchmark? Possible factors could include broad peaks, co‑elution, or generally small peak areas. Please expand on this in the discussion.
L. 293: (CV % >0 (excluding invariant replicates). One ”)” is missing.
Fig. 5, 6, and 7: It is difficult to see the details, especially the white lines. Can the figures be enlarged?
L. 383: As mentioned earlier, Chlide a is (mostly) an artifactual product from chlorophyll a formed during extraction due to water content in acetone. I suggest briefly describing this with references and then remove chlorophyllide a from the article (it would still be included in TChl a).
L. 400: “Volumes below 200 mL showed precision degradation in our dataset”: refer to Fig. 7? Where can it be seen that there will be minimal additional precision benefit above 1 L in high biomass waters?
L .430: “.The divergent behavior at high filtration volumes, where precision degrades despite greater sample volume, supports the interpretation that physical stresses during extended filtration (e.g., cell lysis, filter overloading) compound the natural heterogeneity already present”. When filtering large volumes in low Chl a concentration (oligotrophic) regions, e.g. 4-5000 mL, many pigments (e.g., prasinoxanthin, neoxanthin, lutein, alloxanthin) from algal species not abundant in these areas will still be detected near their LOD/LOQ, which means that there will be increased uncertainty on the determination of the peak areas. While natural pigment concentrations span ~4 orders of magnitude, the filtered volumes do not.
L. 447: At least part of the explanation may again relate to low cell numbers and the higher uncertainty associated with very small pigment peaks from sparsely represented taxa.
L. 452-455: Please provide further details on reporting practices near the LOD/LOQ (see general comments).
L. 463: The statement that larger filtration volumes reduce relative measurement error in oligotrophic environments is not clearly supported for all pigments in Fig. 7, although it may hold if even larger volumes were filtered (acknowledging practical limitations).
L. 479: iSeaHARRE-5 – should ”in” be added?
L. 487-which methodological factors are referred to here (e.g., calibration, peak-area integration)?
L. 503. Chlorophyllide a should not be mentioned here, as it is an extraction artefact (see comments above).