Temperature Dependence of Freshwater Phytoplankton Growth Rates and Zooplankton Grazing Rates
Round 1
Reviewer 1 Report
Review for the paper "Temperature dependence of freshwater phytoplankton growth rates and zooplankton grazing rates" by Jennifer Pulsifer and Edward Laws submitted to "Water".
General comment.
The authors conducted a series of dilution experiments to analyze the temperature dependence of freshwater phytoplankton growth rates and grazing rates of zooplankton in a hypereutrophic lake. They used standard methods to collect phytoplankton and measure environmental parameters in the lake. They confirmed the fact that phytoplankton growth rates exceed zooplankton grazing rates. In their case, this difference was 0.13 d−1 with the estimated Q10 values being 1.48 and 1.54, respectively. These results are important for modeling studies to predict possible effects of temperature on the growth rates of phytoplankton communities in a changing environment.
Specific remarks.
Line 26. Consider replacing "based an analysis " with " based on an analysis "
Consider replacing "in accord" with " in accordance" throughout the text.
The authors should indicate the method they used to compare the slopes of the regression lines (L 179-181).
Table 1 requires a footnote to define all abbreviations (u, g, Gmax, K, R2)
Line 235. Consider deleting " , Heidelberg,"
Line 287. Consider replacing "(2 Top. Stud. Oceanogr.)" with "II"
Author Response
Line 26. Consider replacing "based an analysis " with "based on an analysis"
Response: Changed to based on analysis
Consider replacing "in accord" with " in accordance" throughout the text.
Response: Done
The authors should indicate the method they used to compare the slopes of the regression lines (L 179-181).
Response: We have indicated that we used a t-test.
Table 1 requires a footnote to define all abbreviations (u, g, Gmax, K, R2)
Response: Good point. We have added footnotes in which we define all abbreviations
Line 235. Consider deleting " , Heidelberg,"
Response: Good point. Heidelberg has been deleted
Line 287. Consider replacing "(2 Top. Stud. Oceanogr.)" with "II"
Response: Good point. We have made this change.
Reviewer 2 Report
The authors report on about a year’s worth of roughly monthly estimates of phytoplankton growth and zooplankton grazing rates from a lake at LSU, with samples taken over an annual cycle of temperature (11-33°C) and standing-stock biomass concentrations (40-2000 mg chla/L). They use this data to study the temperature dependence of both rates. This is well placed in the literature between Eppley and Kremer et al. The authors conclude that Q10 for phytoplankton and their grazers is lower than often assumed (about 1.5 vs 1.88) and their results are consistent with Kremer’s recent analysis.
I thank the authors for a very clearly written and engaging paper. This is a well designed and described study. I would strongly encourage you to deposit the raw data and your analysis results in a data repository such as figshare.com.
Methods questions
A bit more detail on the light treatment would be helpful for me. Light varies from 38-182 µmol m-2 s-1, which presumably affects phytoplankton growth rate considerably. How was this variability used or averaged? The “replicates” seem to be experiencing different conditions. Can the different growth rates of phytoplankton bottles be estimated? Are the estimated grazing rates the same independent of irradiance?
I had trouble understanding the reasoning at lines 92-97. Why is the chl concentration at Station ALOHA relevant to the volume of culture required here? Are you appealing to similar grazing experiments from this location? I would be more concerned that large grazers might be excluded in a 60 mL bottle (particularly once diluted 32 fold)
Detailed questions
I assume that nitrate concentrations are very low; perhaps it would be worth mentioning? (line 52)
Is there a seasonal succession in phytoplankton functional groups or cell size in this lake that could confound the analysis (c.f., line 189)? If not, it might be worth mentioning. (c.f. Line 53 indicating 80% of chi is chlorophyceae or cyanophyceae; is there a seasonal / nutrient / temperature driven trend?)
Figure 2. I don’t see a symbol at f = 0. Line 146 suggests, if I understand correctly, that the value of the ratio is 1 at f=0. This would be a big deviation from the line shown on the graph. Please elaborate or clarify.
Your f axis extends beyond the range of observations. Please remove f = 63/64 and don’t extrapolate your fitted curve.
Since you have triplicates, consider showing error bars on this figure.
Were replicates used in the parameter estimation, or averages for each value of f?
Integration of differential equation at line 128. Since you are appealing to Euler integration, it might be helpful to explicitly include delta t = 1 h in Eqns (1) and (2).
How were mu, Gmax, and K estimated? The caption to figure 2 gives a partial explanation, but it would be nice to have this in the methods. (The curve, line 149, is not, I think, a numerical integration; rather it is a plot of Eq (1, 2) using the parameters estimated from an optimization and numerical integration.) How were replicates used? What was the sample size for the optimization? Do you in fact need the nonlinearity to make Gmax and K identifiable? If Eq (2) always applied, you would only find the ratio Gmax/K.
I can imagine other ways to analyze this data. It would be helpful to provide the raw data (replicates, f, chl at time 0 and 24) in a supplement / data repository.
Table 1. 4/26/2019. The chl at time 0 is much higher in this observation than any other. Please confirm.
Did you also measure ammonium and phosphate in the lake (line 52). Please consider reporting these data too.
Line 155-158. Please give a bit more information about these studies. How were they comparable or different from your work?
Figure 3 (line 166). Please explain a bit more about the rationale for binning and averaging the data. In favour of averaging I would argue that if samples are unevenly distributed along the temperature axis, that using the raw data might put undue emphasis on certain temperatures (e.g., it looks like most studies are around 20°C, with relatively few at the extremes.) On the other hand, the binning and averaging might obscure deviations from the pattern and certainly reduced the statistical power. In particular the claim at line 160 (r = 1, p < 0.0028) is suspicious since its derived from averaged data (I think). R2 will be much lower on raw data. But I’d rather read your views than rely on my speculations.
Line 175 (173-176). Why do you think that phytoplankton gains and losses are in balance? There are large fluctuations in biomass over the year. The bottle is not likely equivalent to what happens in the lake. Do you not think that viral lysis and apoptosis is happening in the bottles? Is sinking important in this lake?
Line 199-201. This does not make sense to me. What is the sufficient time frame? 24 h? Likely to be as true in bottles and the lake? Why is the lake in balance? This simple argument contradicted with a better argument in line 203-204.
Table 1, Fig. 3. I don’t think you explicitly connect grazing rate g to Gmax (presumably g = Gmax / P(t=0)). It would be helpful to be explicit.
Author Response
I thank the authors for a very clearly written and engaging paper. This is a well designed and described study. I would strongly encourage you to deposit the raw data and your analysis results in a data repository such as figshare.com
Response: I am not familiar with fig share.com, but I will be happy to do that. I think I should wait until the manuscript is formally accepted before going ahead, but I think this is a good suggestion.
A bit more detail on the light treatment would be helpful for me. Light varies from 38-182 µmol m-2 s-1, which presumably affects phytoplankton growth rate considerably. How was this variability used or averaged? The “replicates” seem to be experiencing different conditions. Can the different growth rates of phytoplankton bottles be estimated? Are the estimated grazing rates the same independent of irradiance?
Response: We averaged the results from the three bottles, i.e., we did not try to differentiate rates at the three different irradiances.
I had trouble understanding the reasoning at lines 92-97. Why is the chl concentration at Station ALOHA relevant to the volume of culture required here? Are you appealing to similar grazing experiments from this location? I would be more concerned that large grazers might be excluded in a 60 mL bottle (particularly once diluted 32 fold)
Response: I have added a sentence acknowledging that exclusion of relatively uncommon but perhaps important grazers is an issue with dilution experiments. Because the abundance of all kinds of organisms is much higher in a place like University Lake, I think that issue is not as big a concern in University Lake as it might be at a place like Station ALOHA, but it is a concern nonetheless.
I assume that nitrate concentrations are very low; perhaps it would be worth mentioning? (line 52)
Response: I have added the information about average nitrate concentrations.
Is there a seasonal succession in phytoplankton functional groups or cell size in this lake that could confound the analysis (c.f., line 189)? If not, it might be worth mentioning. (c.f. Line 53 indicating 80% of chi is chlorophyceae or cyanophyceae; is there a seasonal / nutrient / temperature driven trend?)
Response: Changes in the abundance and composition of the phytoplankton in University Lake are much more episodic than seasonal. Runoff from heavy rains raise the nutrient concentrations in the lake. Louisiana is second only to Hawaii in terms of annual rainfall.
Figure 2. I don’t see a symbol at f = 0. Line 146 suggests, if I understand correctly, that the value of the ratio is 1 at f=0. This would be a big deviation from the line shown on the graph. Please elaborate or clarify.
Your f axis extends beyond the range of observations. Please remove f = 63/64 and don’t extrapolate your fitted curve.
Response: Good points. I have replaced the old Fig. 2 with the results of the experiment from 11/6/18. That experiment includes data at f=0 as well as at f=63/64.
Since you have triplicates, consider showing error bars on this figure
Were replicates used in the parameter estimation, or averages for each value of f?
Response: We averaged the results from the three bottles. We do not have error estimates.
Integration of differential equation at line 128. Since you are appealing to Euler integration, it might be helpful to explicitly include delta t = 1 h in Eqns (1) and (2).
Response: I have included the Δt in these two equations.
How were mu, Gmax, and K estimated? The caption to figure 2 gives a partial explanation, but it would be nice to have this in the methods. (The curve, line 149, is not, I think, a numerical integration; rather it is a plot of Eq (1, 2) using the parameters estimated from an optimization and numerical integration.) How were replicates used? What was the sample size for the optimization? Do you in fact need the nonlinearity to make Gmax and K identifiable? If Eq (2) always applied, you would only find the ratio Gmax/K.
Response. Typically we had initial and final chlorophyll concentrations at f values of 0, 1/2, 3/4, 7/8, 15/16, 31/32, and 63/64. The model contains three parameters, the growth rate of the phytoplankton, the grazing rate of the zooplankton in the undiluted water, and the value of K, where the grazing rate is half the maximum value. We varied those three parameters to give the best fit to the data. The curve shown in Fig. 2 is the ratio of the final to initial chlorophyll concentration estimated by integrating the model versus the actual data (triangles). Note that the legend on the ordinate says Final/Initial biomass. We chose the three parameters that gave the best fit in a least squares sense to those data. I have added a sentence explaining this in the revised document (lines 140–141).
I can imagine other ways to analyze this data. It would be helpful to provide the raw data (replicates, f, chl at time 0 and 24) in a supplement / data repository.
Response: OK—I can include these data with the data requested by the first referee.
Table 1. 4/26/2019. The chl at time 0 is much higher in this observation than any other. Please confirm.
Response: Agreed—this was one of those episodic blooms
Did you also measure ammonium and phosphate in the lake (line 52). Please consider reporting these data too
Response: The ammonium and phosphate data are included along with the nitrate and chlorophyll a data.
Line 155-158. Please give a bit more information about these studies. How were they comparable or different from your work?
Response: The principal difference was that the chlorophyll concentrations in University Lake were much higher and required use of a nonlinear grazing rate model. The other studies relied on the traditional Landry model.
Figure 3 (line 166). Please explain a bit more about the rationale for binning and averaging the data. In favour of averaging I would argue that if samples are unevenly distributed along the temperature axis, that using the raw data might put undue emphasis on certain temperatures (e.g., it looks like most studies are around 20°C, with relatively few at the extremes.) On the other hand, the binning and averaging might obscure deviations from the pattern and certainly reduced the statistical power. In particular the claim at line 160 (r = 1, p < 0.0028) is suspicious since its derived from averaged data (I think). R2 will be much lower on raw data. But I’d rather read your views than rely on my speculations.
Response: I wanted to give equal weight to each temperature interval when looking for temperature dependence. The binning accomplished that. In some cases only a range (admittedly small, but a range) of temperatures was specified. Those results could be put in a bin but could not be associated with a particular temperature. The r value of 1 is the result of a nonparametric test in which data were assigned ranks. The ranks can align perfectly even if the data do not lie on a straight line.
Line 175 (173-176). Why do you think that phytoplankton gains and losses are in balance? There are large fluctuations in biomass over the year. The bottle is not likely equivalent to what happens in the lake. Do you not think that viral lysis and apoptosis is happening in the bottles? Is sinking important in this lake?
Response: Unless there is some kind of monotonic trend in phytoplankton biomass, phytoplankton gains and losses must balance when averaged over a sufficiently long timeframe. I would guess that the principal loss mechanism not taken into account in the dilution experiments is sinking. I have no reason to think that viral lysis and apoptosis do not occur in the incubation bottles.
Line 199-201. This does not make sense to me. What is the sufficient time frame? 24 h? Likely to be as true in bottles and the lake? Why is the lake in balance? This simple argument contradicted with a better argument in line 203-204.
Response: As noted above, unless there is some kind of monotonic trend in phytoplankton biomass, phytoplankton gains and losses must balance when averaged over a sufficiently long timeframe.
Table 1, Fig. 3. I don’t think you explicitly connect grazing rate g to Gmax (presumably g = Gmax / P(t=0)). It would be helpful to be explicit.
Response: Correct—I have noted this in the footnote to Table 1 and in the caption to Fig. 3.
Reviewer 3 Report
The paper represents significant contribution to better understanding of temperature effects on phytoplankton and zooplankton communities. Experimental and statistical analyses were well designed and performed. Only few spelling/style errors were noticed. I wish to offer my sincere accolades to authors.
Suggestions to authors:
26: "a now-" should be corrected to "a new"
Sentences that are written in active form should be transformed in the passive form which is more common in the scientific writing:
41: "we measured" should be replaced with: "...the growing rates of phytoplakton...were measured"
43: "We assumed" should be replace with "it was assumed"
44: Replace active form with passive form: "It was assumed..."
45: Replace active form with passive form:: "The used methodology..."
59: Replace active form with passive form: "Dilution technique...was used"
75, 76: Replace active form with passive form: "Because of the... it was hypothesized..."
96: Replace active form with passive form: "We felt..." should be replaced with "It was suspected (assumed)..."
123: Replace active form with passive form: "it was attempted to fit..."
125: Replace active form with passive form: "...differential equations, describing the rate of change of the... was integrated..."
154: Replace active form with passive form: "...the results of...were combined..."
158: Replace active form with passive form: "Data were bined..."
Possessive pronouns should be avoided:
198: "...but analyses presented here..."
205: Replace "our analyses" with "this analyses" or "analyses presented here"
Author Response
26: "a now-" should be corrected to "a new"
Response: I am afraid that I disagree with this suggestion. I think the 1972 paper by Eppley is now a classic paper (even though its conclusions about the temperature dependence of phytoplankton growth rates may be wrong).
Sentences that are written in active form should be transformed in the passive form which is more common in the scientific writing:
Response: I am afraid that I disagree with this comment as well. In general, the active voice is preferred because use of the active voice makes it clear who did the acting. I know that in the past there has been a tendency to prefer the passive voice in scientific writing, but my impression is that use of the active voice has become more common in recent years for the reason I have noted above.
Possessive pronouns should be avoided:
Response: I did not make these changes for similar reasons.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Using the dilution technique, this manuscript analyzes the dynamics of phytoplankton and zooplankton in an urban lake with high chlorophyll concentrations. The analysis of the results of these experiments together with the results obtained from similar experiments conducted in other freshwater systems, shows that in these systems phytoplankton growth rates consistently exceed zooplankton grazing rates.
In my opinion, the manuscript reads well and is worthy of publication.
I only find a few issues that I think need modification and / or clarification, in order to facilitate the understanding by the reader.
Material and methods
Considering the dilution technique, it can be thought that a pore size of 1.6 µm to obtain the filtered water is too large.
It can also be thought that the use of 25 ml experimental volumes with incubation times of 24 hours violates the postulates of the technique, which usually employs appreciably larger volumes. Could these two circumstances be justified-explained in the manuscript?
In chlorophyll determinations, the removal of filter debris via filtration through GF/F could cause loss of pigment, due to adherence to the filter. Can comment on this?
Line 122. Here it reads that in most cases to obtain changes in the net growth rate of phytoplankton (chlorophyll) it was necessary to dilute the water by a factor of 8. However, later, in the results, line 147 it reads that in most cases it was necessary to dilute by at least a factor of 4. Please revise this.
The purpose of the integration (equations 1 and 2) needs explanation. I think not all readers are familiar with the subject and need some guidance. How integration and the results in Table 1 are related?
Figure 2. The X-axis labels are not correct.
Discussion
The corollary “zooplankton grazing rates exceeds phytoplankton growth rates during the cooler months” does not seem evident from figure 3A. At least it is not immediate. Please explain.
The regression equations for figure 5 could be given. In this way, the reader could estimate the Q10 for growth and grazing rates.
In the last paragraph of the text: it could be recalled that the net phytoplankton growth rates were not significantly different from zero. This would help the reader to remember that gains and losses are in balance.
Reviewer 2 Report
Review of:
“Phytoplankton growth rates and zooplankton grazing rates in a eutrophic urban lake”
By Pulsifer and Laws
This manuscript reports some interesting results about phytoplankton growth rates, chlorophyll a and zooplankton grazing rates over a large period of observations from a highly eutrophic lake. Furthermore, the authors attempt to examine whether water temperature significantly affect these biotic parameters of the plankton community. However, the paper completely lacks of an ecological background. Moreover, the justification on the two hypotheses is very poorly developed. It seems like the study was designed well and that it could be of general interest but the writing and presentation of the major outcomes is very sloppy. I think the authors should rewrite the paper, beginning with specific hypotheses which will be related to an ecological background. Then develop an introduction that justifies the theoretical framework (first) and then the specific test case. Then the methods will make sense and the results and discussion can reflect on those hypotheses. Once the science is more clearly justified and described, the results could be of interest to the broader audience. Taking the above into consideration, although the work has some merit, I cannot recommend acceptance in the current form.
Figures and Tables
Likewise, the figures are not well design and major revision is needed. For example, the axis titles are not capitalized and the figure legends are not effective and do not help the figures stand alone.
