Diatom Deformities and Tolerance to Cadmium Contamination in Four Species

Round 1

Reviewer 1 Report

This MS examines the effects of cadmium on the development and function of 4 photosynthetic diatom species that are used as biological sensors of environmental quality. The authors examine the effects of chronic and acute doses of Cd on morphology, photosynthetic function (measured as the quantum yield of photosystem II) and growth rate (measured as chlorophyll a concentrations). The goal is to determine which of the 4 diatom species surveyed are Cd-sensitive and -tolerant and select a species that is suitable for biomonitoring water quality over a wide range of Cd concentrations.   

The goals of the MS are laudable, as a better understanding of photosynthetic species’ responses to environmental contaminants will allow them to be further developed as sensitive biomonitors. This is especially important as it becomes more and more difficult to conduct toxicology research using animals, and vertebrates in particular.

My main issue is that the paper contains a number of statements and generalizations that cast doubt on the authors’ ability to generate data to support their research goal. In particular, many of the methods don’t adequately link experimental measurements to functional changes, and often use proxy measurements for fundamental biochemical processes that they would better measure directly (e.g., measuring quantum yield of PSII rather than [or in addition to] net photosynthesis; implying that Chl a concentrations are directly related to growth rate, rather than simply counting diatoms and calculating their concentrations). These problems are exacerbated by an apparent lack of background in some key aspects of experimental design, plant biology, biochemistry and (eco)toxicology. For example:

·      It’s stated that n=3 for all measurements, but it’s unclear whether these are actual experimental repetitions at different times with different cells. I suspect that each of the n repetitions is actually a subsample of a single experiment, given the very small error bars on most of the data. True biological replicates only very rarely give results with error bars smaller than the graph symbols like in fig. 2a.

·      There are a few statistical issues. For example, what do the error bars on the graphs represent: SEMs, SDs? Many of the numbers provided have too many significant digits given the small sample sizes (e.g., lines 201-2).

·      Many synonyms are used for identical phenomena, like teratologies, deformities and morphological abnormalities/aberrations for the effect of Cd on the species’ silica shells. It’s best to use a single term throughout.

·      The discussion of the effect of Cd on photosynthesis is rather perfunctory and suggests a lack of background in plant physiology. Why did you measure quantum yield? What’s the difference between quantum yield and “effective” quantum yield (line 354)? What is “photosynthetic capacity” (line 390)?

·      The authors need to use standard toxicology terms like chronic and acute exposures more consistently rather than alternate terms like dose-repsonse (e.g., line 194) for the latter. Why didn’t you calculate EC50 or LD50 values for Cd?

·      The source of tissue for the various measurements is unclear from the way the methods are written. This confusion is exacerbated by the use of non-standard terms when discussing part of the cells after their extraction (e.g., “Cd concentration in water” in line 221: do you mean growth medium?). A flowchart would help clarify tissue sources.

Another issue is that the MS at 20 pages is at least twice as long as it should be given the limitations in the data. This mostly stems from too much speculation about the significance of the results in an absence of actual experiments to support their point of view. Many of the figures could also easily be combined without sacrificing legibility, e.g., fig. 2 should be one graph with each species represented by different shaped data points. There are very lengthy passages in the results/discussion talking about the significance of trends in fluctuations in the values of data points during chronic exposures (e.g., 3.2.4 on quantum yield). Table 2 should be a supplemental table as the results are summarized in fig. 7.

Overall, the limited data obtained in the MS cast doubt on the ability of the authors to categorize these 4 species as Cd-tolerant or -sensitive.

Overall, I recommend that the MS undergo a major revision to address the points above prior to being considered for publication.

Other minor points are below.

·      No control picture for M. atomus between figs. 1s and t. Put a line with species name above each group of diatoms to separate them more obviously by species.

·      Line 171: How did you transform this phenotype data into numerical data that you could use for the statistical analyses?

·      Many concentrations (e.g., of EDTA) are in volumes. Should be in expressed as mM or mg/mL final concentrations.

·      Are frustule deformities the best morphological measure of environmental contamination?

·      Need to include some results in the abstract (e.g., numbers).

Author Response

Please find attached the responses to Review report number 1.

Author Response File: Author Response.docx

Reviewer 2 Report

General comments:

The authors present a very interesting work about the use of diatoms teratology to evaluate freshwater Cd contamination. They address an essential question of the (natural) variability of responses depending on the diatom species studied, which is of prime importance to correctly interpret teratology data in terms of water quality biomonitoring and environmental risk assessment. The results of this study will certainly help improving our capacity to monitor the risk of toxicity for aquatic organisms exposed to metals in their environment.

The paper is well written and easy to read. The objectives and scientific approach are well defined. The methodology is well described, and experiments seems to have been conducted rigorously. In my view, this article is acceptable in its present form, and I only have few comments to address to the authors, that are listed below.

Introduction

L67 – 68: Maybe the authors should mention here that 3 species where collected in the field and one came from a collection.

L73: In my opinion, this is a very good introduction. I just wonder why the authors referred to European river Cd contamination as, apparently, diatom species mostly originated from north America.

Materials and Methods

L110: Did the Cd concentrations in diatoms measured for the acute exposure experiment?

L125: Were the exposure solution renew during the experiment to allow a constant exposure?

L142: “For the evaluation of Cd effects at lower concentration” does it refer to the chronic experiment? If yes, please write it clearly.

L144: Maybe add “dissolved” to “then used for the determination of dissolved Cd concentrations”.

L187: The authors have tested the normality and homoscedasticity of the data before performing statistical analyses. But given the low number of replicates (n = 3), and to be statistically rigorous, only non-parametric tests should be used.

Results and discussion

Figure 2: As in the section 3.1 the authors mentioned C1, C2 etc. it should be great to let C1, C2 appear somewhere on the plots to make it easier for the reader to navigate between the text and the plots.

In addition, as the authors only selected C1, 2 and 3 to rank the four diatoms in terms of sensitivity (L210), a vertical line on the plots to separate C1, 2 and 3, from C4, 5 and 6.

L201: It is unclear to me on what criteria the ranking is based.

L2015-2016: How could the authors say that N. palea is more tolerant than G. truncatum since for the latter, no significant differences were observed (based on C1, 2 and 3)?

L219: A reference is missing.

L228: Were the container walls equilibrated with the contamination solution before the exposure of the diatoms? (e.g. fill the containers 48h before the exposure to allow the walls to equilibrate and renew the solution before the exposure)

L235 and Figure 3: For (c ) I would only say that Cd concentrations tented to continuously decrease because of the wide standard deviation for the last measure (day 21).

In my opinion, the authors should more discuss the fact that G. truncatum comes from a collection and responses observed for this species may not reflect what happens in environmental conditions. Moreover, the high frequency of deformities observed in control organisms from this species question about the relevance of using this species for water quality assessment.

Author Response

Please find attached the responses to Review Report number 2.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The revised version of the MS addresses many of my earlier concerns. Thanks to the authors for their revisions. The science is sound and the paper makes a very good contribution to the selection of diatom species for biomarkers of metal toxicity in aquatic environments.

I still would like to see a short justification for using quantum yield of PSII as a proxy for photosynthetic capacity, and of Chl a levels for diatom abundance and to estimate growth rate, with a reference for each. This can be done at L156. This helps to justify previously undefined statements like "photosynthesis inhibition" at L218 and "cell productivity" at L232 without your needing to measure net photosynthetic rate directly.

Re the growth phases of the cultures discussed in L274, how can you distinguish between linear/exponential/plateau phases without doing regression analysis and curve fitting on the growth curves in fig 4? Similarly in L280, you don't actually measure the growth rate (e.g., ug Chl a accumulated / d); either include these data or use a vaguer term like growth curve or kinetics, as in L371 and 375. Similarly, for "senescence phase", L390 and Fig 5a, final histograms: how do you know these cultures are senescent? Does the Chl-a level support this? At least provide a reference (e.g., for Chl-a levels in related senescent diatoms), and remove the "Ctrl and Cd100 in senescence" from the figure.  

I still find the discussion to be too long, due to the presence of perhaps too much speculation about reasons for the data you obtained without determining whether that might be true for your results, e.g., reasons for decreases in bioaccumulation in L336-360; L404: "suggests that N. palea maintained consistent photosynthetic mechanisms"... it would be good to measure CO2 uptake and/or O2 evolution in a Cd-exposed culture in the next set of experiments to justify this statement. L412: "detoxifying mechanisms" like "metal-binding thiolated peptides (L428)... like metallothionein synthesis perhaps... a western blot would demonstrate that MT protein is synthesized in response to chronic Cd exposure. Among organelles involved in metal isolation and thus detoxification, the vacuole may play a part (L436-7). I leave it up to the editor and authors whether they want to edit down the discussion.

Minor issues:

How was Cd provided (e.g., CdCl2? Supplier?), L117

Instead of "diatoms species/samples/cultures", "diatom" e.g, L14.

L128: ... prior to exposure, _as detailed below_.

L136: not covered, but capped.

L147: give x g, not rpm, also for spin at L165

L148, delete description of filtering and say "filtered as above".

L150... to determine bioaccumulation of Cd (e.g., intracellular...")

L236: Mean EC25 (...) effect on quantum yield of PSII for N. palea....

L237: if n=3, why present individual results and not means +/- SEMs?

Fig 3 legend: n=3 (?)

L314: not "superior to 50 ug Cd/L", but "of 100 ug Cd/L"

Table 2: units in column heading (ng Cd/mg Chl a) are different from those used elsewhere (ug Cd/ug Chl a).

L455: (b) internal valve view, _typical_ valve...

L467: approx 4x more deformities... appears to be significant in fig 7b, but did you do stats to support statements like "had the greatest degree of deformities" (L474),  "P. mesolepta was more impacted" (L479) and "was the most structurally impacted" (L514)? Use terms like significant, non-significant in discussion of these results.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf