Review Reports - Decadal Shifts in Zooplankton Size Structure and Biomass in a Warming Tropical Lake

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Review of Busumtwi 4309989

Lake Busumtwi has been an important source of information about climate change in West Africa, and this report promises some insights into recent changes linked to anthropogenic forcing. I applaud the effort but I believe the report needs additional details about methods and results. For instance, although the authors report using a Secchi disk to record water transparency, those data do not appear nor is it evident whether corresponding data exist for Period 1. Differences in light attenuation might be contributing to observed mixed layer depths as well as surface radiation balance.

I was disappointed by the lack of taxonomic detail regarding Copepoda, particularly whether there have been any shifts in relative abundance of calanoids and cyclopoids. I recommend that vertical profile data for temperature, D.O. and chlorophyll be added to the supplementary materials.

Readers need to know if there were any differences in pigment filtration and extraction methods between periods, as well as whatever is known about fish populations over the same periods

Specific points:

Line 171: Unless the titration alkalinity is presumed to be entirely bicarbonate, which is unlikely at pH 9, it would be preferable to report it as Acid Neutralizing Capacity in mEq or mN.

Line 187: Temperature-compensation of measured electrical conductivity to specific conductance typically uses a presumption that ionic strength and ionic composition is representative of seawater, which is not true for Busumtwi. Did the authors compare the instrument estimates to actual measurements of conductivity using a water bath regulated to 25C for verification?

Line 205: Was chlorophyll determined from filtered samples? If so, how long after collection were the filters produced?

Line 228: What sampling gear and protocol were used in Period 1?

Table 1: Were the same allometric equations applied to Calanoida and Cyclopoida? Which were the dominant taxa?

Section 3.2: Were these volume-weighted averages? If not, they would represent biased estimates of true average values.

Author Response

Title of Manuscript: Decadal shifts in zooplankton size structure and biomass in a warming tropical lake

Manuscript ID: limnolrev-4309989

General Response to the Reviewer

We sincerely thank the Reviewer for the careful, constructive, and highly relevant comments. The observations have enabled us to correct all issues identified, which in our view has improved the manuscript considerably. Addressing these concerns has substantially improved the internal consistency, scientific robustness, and overall quality of the manuscript. We believe that the revised version now provides a considerably stronger and more defensible analysis of decadal changes in the zooplankton community of Lake Bosumtwi.

Specific Comments and Responses

Comment: Lake Bosumtwi has been an important source of information about climate change in West Africa, and this report promises some insights into recent changes linked to anthropogenic forcing. I applaud the effort, but I believe the report needs additional details about methods and results. For instance, although the authors report using a Secchi disk to record water transparency, those data do not appear, nor is it evident whether corresponding data exist for Period 1. Differences in light attenuation might be contributing to observed mixed-layer depths as well as surface radiation balance.

Response: Results of the Secchi depth measurements have been included in Section 3.2 with a corresponding statement in Section 2.2 to show how the data were acquired through standardised protocols across the two study periods.

Comment: I was disappointed by the lack of taxonomic detail regarding Copepoda, particularly whether there have been any shifts in the relative abundance of calanoids and cyclopoids. I recommend that vertical profile data for temperature, D.O. and chlorophyll be added to the supplementary materials.

Response: Previous surveys found no calanoid copepods in Lake Bosumtwi. Adult copepods were identified as the endemic cyclopoid Mesocyclops bosumtwii, while nauplii and copepodites were assigned to this species by inference because no other copepod species are known from the lake. This distinction is now stated in the Methods and table notes. Profile data on limnological parameters are available (48 profiles for the historical period and 28 for the recent study), but the dataset will be too large as supplementary materials. We therefore added comparable summaries of temperature, dissolved oxygen, habitat-layer depth, nutrients, and chlorophyll a to the Supplementary Materials.

Comment: Readers need to know if there were any differences in pigment filtration and extraction methods between periods, as well as whatever is known about fish populations over the same periods

Response: The same ethanol-extraction procedure was used for both periods. Samples were transported on ice, filtered within 24 h, frozen, and analysed within 48 h. These details were added to Sections 2.4 and 2.5. The results underwent rigorous QA/QC procedures in the analytical lab of the Department of Ecoscience, Aarhus University, Denmark. Analysis of fish populations was outside the scope of the present study, and therefore, fish predation is not analysed.

Comment: Line 171: Unless the titration alkalinity is presumed to be entirely bicarbonate, which is unlikely at pH 9, it would be preferable to report it as Acid Neutralizing Capacity in mEq or mN.

Response: We agree that at pH values near 9, alkalinity is not composed solely of bicarbonate, and contributions from carbonate and hydroxide are expected. In the original dataset, alkalinity was reported in mg L⁻¹ CaCO₃ following standard limnological convention, but this implicitly assumes bicarbonate dominance. We acknowledge that expressing alkalinity as Acid Neutralising Capacity (ANC, mEq L⁻¹) would more accurately reflect the acid‑neutralizing components present in Lake Bosumtwi. We have revised the terminology accordingly to avoid misinterpretation.

Comment: Line 187: Temperature-compensation of measured electrical conductivity to specific conductance typically uses a presumption that ionic strength and ionic composition are representative of seawater, which is not true for Bosumtwi. Did the authors compare the instrument estimates to actual measurements of conductivity using a water bath regulated to 25 °C for verification?

Response: Conductivity values were automatically corrected to 25 °C using the YSI Exo2 internal temperature‑compensation algorithm. Although this algorithm assumes a standard ionic-temperature coefficient, which may not fully reflect the unique ionic composition of Lake Bosumtwi, the sonde was calibrated with certified standards before field deployment. Verification in a temperature-controlled water bath was not performed. Because both historical and recent datasets applied standard 25 °C compensation, the measurements are comparable for temporal analysis. Because Lake Bosumtwi’s ionic composition may differ from the standard compensation assumption, we do not base the main ecological conclusions on small conductivity differences between periods.

Comment: Line 205: Was chlorophyll determined from filtered samples? If so, how long after collection were the filters produced?

Response: Water samples for chlorophyll a were immediately placed on ice after collection, transported to the laboratory, and filtered within 24 hours of sampling. After which, the filters were frozen and analysed using ethanol extraction and spectrophotometric analysis within 48 hours.

Comment: Line 228: What sampling gear and protocol were used in Period 1?

Response: The same sampling gear (70 μm mesh plankton net of mouth diameter 25 cm) and protocol (central and deepest lake sampling point and time of sampling) were used in Period 1, except that in Period 1, sampling frequency was biweekly.

Comment: Table 1: Were the same allometric equations applied to Calanoida and Cyclopoida? Which were the dominant taxa?

Response: There are no known Calanoid taxa inhabiting the lake. Hence, the allometric equations were applied to only Cyclopoida, which were the dominant taxa. Furthermore, for all dry weight estimates, taxon-specific length-weight equations developed and used in the historical study were applied verbatim in the current study without adjustment.

Comment: Section 3.2: Were these volume-weighted averages? If not, they would represent biased estimates of true average values.

Response: The values reported in Section 3.2 were not volume‑weighted averages. Instead, they were calculated as simple absolute values based on integrated sampling of the oxygenated stratum inhabited by zooplankton. This approach was chosen to maintain consistency with the historical 2005–2006 dataset, which used the same whole oxic water column method. Because our objective was to compare temporal changes between two periods using a harmonised analytical framework, applying volume‑weighted averaging to only the recent dataset would have introduced methodological inconsistency. We have clarified this in the Methods and in the caption of Table 3.

End of Responses

******************************************************************************

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The content and nature of the manuscript are considered to be a very good fit for this journal. But, I would like to inform you that publication is not possible in the current state of the manuscript. I suggest resubmitting the manuscript after undergoing a thorough error verification and correction process.

I would like to briefly summarize the major revisions or parts requiring top priority verification procedures as follows.

Re-evaluation of the number of survey points

It appears that the number of survey points consists of one sample per point. In such cases, a verification procedure is required to determine whether the data accurately represents the community variability of zooplankton in the study area. It is judged that considerations regarding heterogeneity by survey point and heterogeneity by water depth were not included in the results of this manuscript.

Assessment of suitability for changes in zooplankton community dynamics due to water temperature differences

I have doubts as to whether variations in water temperature by depth and changes in zooplankton community dynamics act as distinct factors of change in this manuscript. The variation in water temperature by depth during the survey periods of 2005-2006 and 2018-2020 is lower than the diurnal or monthly variation, and I question whether it is an appropriate evaluation method to include these temperature differences as direct factors or influencing changes in the dynamics and density of major zooplankton communities when writing the manuscript.

Zooplankton collection period

It appears that the sample survey method for the zooplankton collection period of 2005-2006 is not included in this manuscript. Verification and supplementation are required.

Checking of data measurements

In Table 4, regarding the difference between the 2005-2006 and 2018-2020 average values, average density, dry weight, and average length show higher values in the 2018-2020 period, whereas average biomass actually shows higher figures in 2005-2006. It is generally understood that longer body length and higher dry weight are typical, but the figures are assessed as containing an error. Careful verification and correction are required.

Author Response

Title of Manuscript: Decadal shifts in zooplankton size structure and biomass in a warming tropical lake

Manuscript ID: limnolrev-4309989

General Response to the Reviewer

We sincerely thank the Reviewer for the careful, constructive, and highly relevant comments. The observations prompted us to conduct a comprehensive re-examination of the source dataset, biomass-calculation workflow, statistical analyses, tables, figures, and interpretation of the results. This process revealed several inconsistencies in the original biomass calculations, particularly in the application of the length–weight and biovolume-derived dry-weight equations. In response, we carefully verified the source data, corrected the identified calculation errors, and revised Table 1 and Table 4 to provide the original equations used for each zooplankton taxon and developmental stage in the historical study. The original equations were then applied consistently to both the historical and recent datasets to ensure standardised cross-period comparisons. Following these corrections, we regenerated all individual dry-weight and biomass estimates, repeated the relevant statistical analyses, and updated all associated tables, figures, supplementary materials, Results, Discussion, and Conclusions in accordance with the revisions.

These revisions resulted in substantial changes to some numerical results and to the overall ecological interpretation of the study. In particular, the revised calculations eliminated biologically inconsistent length–weight patterns and provided a more internally coherent assessment of taxon- and stage-specific biomass estimates. Addressing these concerns has substantially improved the internal consistency, scientific robustness, and overall quality of the manuscript. We believe that the revised version now provides a considerably stronger and more defensible analysis of decadal changes in the zooplankton community of Lake Bosumtwi.

Specific comments and responses

I would like to briefly summarize the major revisions or parts requiring top-priority verification procedures as follows.

Comment: Re-evaluation of the number of survey points. It appears that the number of survey points consists of one sample per point. In such cases, a verification procedure is required to determine whether the data accurately represents the community variability of zooplankton in the study area. It is judged that considerations regarding heterogeneity by survey point and heterogeneity by water depth were not included in the results of this manuscript.

Response: We acknowledge the importance of considering spatial heterogeneity when interpreting zooplankton community data. In this study, zooplankton were sampled using vertical tows from a fixed central station over 24 months. This approach was chosen for three reasons. First, it ensures methodological consistency with the historical 2005–2006 dataset, which was collected using the same single‑station vertical‑tow protocol. Maintaining identical sampling procedures is essential for reliable inter‑period comparisons. Second, the sampling design captures substantial ecological variability. Each monthly sample represents a full‑water‑column integrated tow, and the 24‑month time series includes vertical, seasonal, interannual, and hydrodynamic variation. The resulting dataset reflects a wide range of environmental conditions and zooplankton responses. Third, previous studies of Lake Bosumtwi and other small, deep tropical crater lakes have shown that horizontal variation in zooplankton distribution is minimal, with communities largely structured by vertical gradients rather than by spatial heterogeneity across the lake surface. The lake’s small surface area, steep basin morphology, and strong stratification promote horizontal homogeneity in zooplankton assemblages, making a central‑station vertical tow an appropriate and widely used method for characterising whole‑lake zooplankton dynamics. Notwithstanding, we have revised the methodology section to include this clarification.

Comment 2: Assessment of suitability for changes in zooplankton community dynamics due to water temperature differences. I have doubts as to whether variations in water temperature by depth and changes in zooplankton community dynamics act as distinct factors of change in this manuscript. The variation in water temperature by depth during the survey periods of 2005-2006 and 2018-2020 is lower than the diurnal or monthly variation, and I question whether it is an appropriate evaluation method to include these temperature differences as direct factors or influencing changes in the dynamics and density of major zooplankton communities when writing the manuscript.

Response: We agree that the inter‑period temperature differences observed across depth strata are smaller than the diurnal and monthly variability typical of tropical lakes. This is expected in warm, low‑latitude systems where surface waters experience substantial short‑term fluctuations. As noted in the manuscript, the magnitude of the long‑term temperature change remains well within the physiological tolerance ranges of the zooplankton taxa present in Lake Bosumtwi, and we therefore do not interpret temperature as a direct driver of zooplankton biomass patterns. The ecological significance of warming lies instead in its indirect effects. Even modest increases in water temperature can strengthen thermal stratification, reduce mixed‑layer depth, and suppress vertical nutrient exchange. Our results show clear evidence of these processes, including a more stable and shallower epilimnion, reduced nutrient circulation to surface waters, and a shift toward cyanobacteria dominance, under stable and calmer waters indicative of warming. These changes collectively affect food quality and quantity for zooplankton, constrain vertical habitat space, and increase metabolic costs, mechanisms widely recognised to affect zooplankton growth, development, and biomass.

Comment: Zooplankton collection period. It appears that the sample survey method for the zooplankton collection period of 2005-2006 is not included in this manuscript. Verification and supplementation are required.

Response: Details of the sampling period and frequency for the 2005-2006 period have been included in the methodology section.

Comment 4: Checking of data measurements. In Table 4, regarding the difference between the 2005-2006 and 2018-2020 average values, average density, dry weight, and average length show higher values in the 2018-2020 period, whereas average biomass actually shows higher figures in 2005-2006. It is generally understood that longer body length and higher dry weight are typical, but the figures are assessed as containing an error. Careful verification and correction are required.

Response: We acknowledge the Reviewer’s concern regarding the unrealistic patterns produced by the operational length–weight equations previously applied. In response, we revisited the biomass calculations and re‑evaluated the underlying length–weight framework. Initially, we applied the original published, taxon‑specific length–weight regressions to the 2018–2020 dataset, but this produced what we thought were unrealistically high biomass estimates. This was the reason the earlier analysis used locally adjusted “operational” equations inherited from the historical workflow. However, as the Reviewer correctly notes, these adjusted equations generated biologically inconsistent outcomes, most notably the combination of increased copepod body length with a six‑fold decrease in estimated dry weight. To resolve this issue, we recalculated biomass using the original published length–weight equations for all taxa. The revised results now show positive and internally consistent length–weight relationships among the various taxa, eliminating the previously contradictory patterns. Importantly, these recalculated values reveal a substantial increase in total zooplankton biomass in the recent period (2.33-fold increase), rather than a decline. This revised pattern is strongly supported by the literature: in warming, cyanobacteria‑dominated systems, cyclopoid copepods can achieve exceptionally high biomass by shifting to alternative protozoan and microzooplankton prey that proliferate under bloom conditions. Given that cyclopoids constitute ~98% of the zooplankton community, it is plausible that this adaptive feeding strategy underlies the observed biomass increase. We have updated the manuscript accordingly, in all relevant sections (abstract, methods, results and discussion), clarified the methodological rationale, and ensured that all interpretations are now based on the recalculated, biologically consistent length–weight estimates.

End of Responses

******************************************************************************

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

In the manuscript entitled “Decadal shifts in zooplankton size structure and biomass in a warming tropical lake”, the authors analysed physical and chemical properties of the water column as well as zooplankton size structure and biomass in a tropical lake in Ghana over two periods separated by 15 years. While this analysis is not directly a long-term analysis of the trends over these 15 years, it provides very useful information on shifts in zooplankton over a decadal time scale and its relationship with environmental and biological changes. The analyses are conducted carefully. But some parts of the manuscript could be improved. Notably, the Introduction section is too repetitive, and some parts of it could be shortened and combined to make the section clearer and easier to read. Similarly, the discussion section is interesting but must be improved by making it clearer, more concise, and by avoiding the repetitions that are very numerous right now. In addition, some parts could benefit from a clearer link between obtained results from this study and interpretations. I also have some specific comments that are listed below:

L37. Technically, « trophic interactions » are not environmental, climatic, nor anthropogenic factors. I suggest to add « biological factors » in the list.

L41-43. You need to specify that this statement is about zooplankton.

L61. “Higher temperatures decrease survival rates”. Not always, not true for every species or ecosystems. I suggest to change the sentence to Higher temperatures could decrease survival rates”

L58-74. I think that this paragraph can be shortened and more concise, notably by avoiding too much repetitions.

L75-81. Same, too many repetitions here. These three sentences could be shortened and made more concise.

L92. Technically it’s not “future” conditions if the period started in 1980. Please reformulate the sentence.

L104-105. This sentence is not needed here and can be removed.

L115. “indicative of warming”. Same, this part can be removed.

L117. What are “warming parameters”? Please list the parameters that were tested.

L134. Could you indicate the average depth as well here?

L160. “Nutrients principally dissolved inorganic nitrogen”. A comma is missing after “Nutrients”.

L162. SRP needs to be spelled out completely the first time the term is used.

L161-163. A reference is needed for this statement.

L184. What was the depth interval between two measurements during the depth profiles?

L191. The Hydrolab measurements were not mentioned earlier.

L222-224. How was the spectrophotometer calibrated? How many times during the study period?

L283-308. With what software did you perform statistical analyses? Please mention it (and the most important packages if the analyses were done with R).

L311-316. I think that warming values should be presented here with only two significant numbers after the comma.

L321-330. I suggest to indicate in parentheses the name of statistical test for which the p-values are presented in the text.

L403. I guess that cladocerans are represented with a barplot instead of a boxplot to highlight the strong temporal variability in their biomass?

L468-491. I suggest that results from statistical analyses are directly integrated into the description of the results, and not separated into a specific sub-section.

L525-541. I agree with the statements in this paragraph, but I think that the actual findings of the study regarding zooplankton could be better incorporated in the framework of lower dissolved oxygen concentration. For now, while the paragraph is strictly speaking well-written, it looks more like a review on the literature regarding effects of anoxic zones on zooplankton in lakes than an actual discussion of the obtained results.

L626-628. This sentence is repetition from the part in L611-613. Overall, I think that this part of the discussion section is very interesting but could be made clearer and more concise, notably by avoiding the numerous repetitions.

Author Response

Title of Manuscript: Decadal shifts in zooplankton size structure and biomass in a warming tropical lake

Manuscript ID: limnolrev-4309989

General Response to the Reviewer

Specific comments and responses

In the manuscript entitled “Decadal shifts in zooplankton size structure and biomass in a warming tropical lake”, the authors analysed physical and chemical properties of the water column as well as zooplankton size structure and biomass in a tropical lake in Ghana over two periods separated by 15 years. While this analysis is not directly a long-term analysis of the trends over these 15 years, it provides very useful information on shifts in zooplankton over a decadal time scale and its relationship with environmental and biological changes. The analyses are conducted carefully. But some parts of the manuscript could be improved. Notably, the Introduction section is too repetitive, and some parts of it could be shortened and combined to make the section clearer and easier to read. Similarly, the discussion section is interesting but must be improved by making it clearer, more concise, and by avoiding the numerous repetitions that are present right now. In addition, some parts could benefit from a clearer link between the obtained results from this study and the interpretations. I also have some specific comments that are listed below:

Comment: L37. Technically, « trophic interactions » are not environmental, climatic, nor anthropogenic factors. I suggest to add « biological factors » in the list.

Response: “Biological factors” has been added to the list.

Comment: L41-43. You need to specify that this statement is about zooplankton.

Response: The statement has been modified to relate to zooplankton.

Comment: L61. “Higher temperatures decrease survival rates”. Not always, not true for every species or ecosystems. I suggest to change the sentence to Higher temperatures could decrease survival rates”

Response: The sentence has been changed to “Higher temperatures could decrease survival rates”. We revised the sentence to clarify that warming may lower zooplankton survival under some conditions rather than implying a universal response.

Comment L58-74. I think that this paragraph can be shortened and more concise, notably by avoiding too much repetition.

Response: The paragraph has been shortened, repetitions removed, and conciseness improved.

Comment: L75-81. Same, too many repetitions here. These three sentences could be shortened and made more concise.

Response: The paragraph has been shortened, repetitions removed, and conciseness improved.

Comment: L92. Technically, it’s not “future” conditions if the period started in 1980. Please reformulate the sentence.

Response: The sentence has been reformulated.

Comment: L104-105. This sentence is not needed here and can be removed.

Response: We thank the Reviewer for this suggestion. However, we believe the two sentences convey related but distinct points. The first highlights that African lakes are inherently sensitive to climate variability, while the second emphasises that this sensitivity, specifically warming trends, has been extensively documented.

Comment L115. “Indicative of warming”. Same, this part can be removed.

Response: The phrase “indicative of warming”, now on L110 due to the shortening of some preceding paragraphs, has been deleted.

Comment L117. What are “warming parameters”? Please list the parameters that were tested.

Response: The “warming parameters” have been specified

Comment L134. Could you indicate the average depth as well here?

Response: The average depth has been indicated with a supporting reference.

L160. “Nutrients principally dissolved inorganic nitrogen”. A comma is missing after “Nutrients”.

Response: A comma has been introduced

Comment L162. SRP needs to be spelt out completely the first time the term is used.

Response: SRP has been spelt out in full the first time it is used.

Comment L161-163. A reference is needed for this statement.

Response: The statement has been referenced.

Comment L184. What was the depth interval between two measurements during the depth profiles?

Response: The depth interval, which also corresponds to the downward rate of the YSI Exo II Sonde, has been indicated. The sonde was lowered at approximately 10 cm s⁻¹, and measurements were logged continuously rather than at a fixed depth interval. The wording in Section 2.3 was corrected accordingly.

Comment L191. The Hydrolab measurements were not mentioned earlier.

Response: The statement being referred to is the preceding statement, as follows: “on L185-187, which states 'The equivalent historical data (2005-2006) were collected using a Hydrolab H₂O Multiprobe 6SBP for water column profiling from the surface to a depth of 50 m at the same site as the present investigation”

Comment L222-224. How was the spectrophotometer calibrated? How many times during the study period?

Response: The Shimadzu UV‑1800 spectrophotometer was calibrated before each analytical session using certified reference standards and reagent blanks to verify wavelength accuracy and absorbance stability. Calibration checks were performed throughout the study period (2018–2020) at the start of every batch of chlorophyll a, TN, and TP analyses to ensure consistent instrument performance.”

Comment L283-308. With what software did you perform statistical analyses? Please mention it (and the most important packages if the analyses were done with R).

Response: The software package used for the statistical analyses has been indicated.

Comment L311-316. I think that warming values should be presented here with only two significant numbers after the comma.

Response: Warming values have been presented to two decimal places.

Comment L321-330. I suggest to indicate in parentheses the name of the statistical test for which the p-values are presented in the text.

Response: The name of the statistical tests for which the p-values are presented has been indicated in parentheses.

Comment L403. I guess that cladocerans are represented with a barplot instead of a boxplot to highlight the strong temporal variability in their biomass?

Response: The cladoceran distributions in Figure 3 were presented as boxplots, but zero-occurrence months caused them to appear bar-like on the logarithmic y-axis. We therefore revised Figure 3 to remove this visual ambiguity. Boxplots are now calculated from positive biomass values only, while zero-occurrence months are shown separately. This treatment has also been clarified in the figure caption.

Comment L468-491. I suggest that results from statistical analyses are directly integrated into the description of the results, and not separated into a specific sub-section.

Response: We agree that integrating statistical results directly into the narrative of each results subsection improves clarity and readability. For this reason, the key statistical findings (t‑tests, ANOVA results, effect sizes, and post‑hoc comparisons) have already been incorporated into the relevant thematic sections of the Results (e.g., temperature changes, oxygen trends, biomass differences). Section 3.12, however, serves a different purpose. It provides a concise synthesis of the overall statistical outcomes across all analyses, including assumptions testing (normality, homogeneity of variance), model selection decisions (e.g., Welch ANOVA), and effect‑size interpretations. These elements do not fit naturally within any single results subsection because they apply to the entire analytical framework rather than to one specific variable or pattern.

However, we have shortened Section 3.12 to a brief synthesis, thereby reducing repetition while retaining a consolidated statistical overview.

Comment L525-541. I agree with the statements in this paragraph, but I think that the actual findings of the study regarding zooplankton could be better incorporated in the framework of lower dissolved oxygen concentration. For now, while the paragraph is strictly speaking well-written, it looks more like a review of the literature regarding effects of anoxic zones on zooplankton in lakes than an actual discussion of the obtained results.

Response: We acknowledge that the paragraph (L525–541) reads partly as a general synthesis of zooplankton responses to hypoxia. However, the two oxygen‑related paragraphs serve distinct purposes: the first outlines how declining oxygen and hypolimnetic anoxic expansion restrict zooplankton habitat, while the second links these conditions to the physiological constraints on zooplankton observed in our data. We have shortened the general discussion of hypoxia and now begin with the observed 26.5% decline in dissolved oxygen. The revised text relates this directly to the sampled habitat and the simultaneous increase in copepod-dominated biomass. We also clarify that oxygen alone cannot explain the contrasting taxon-specific responses.

Comment L626-628. This sentence is a repetition of the part in L611-613. Overall, I think that this part of the discussion section is very interesting but could be made clearer and more concise, notably by avoiding the numerous repetitions.

Response: The section has been shortened, reducing repetition and redundancy.

End of Responses

*******************************************************************************

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript presents an interesting and timely study based on two decadal time series of zooplankton trophic structure and biomass in the tropical crater lake Bosumtwi, with a focus on the effects of warming and intensified stratification. Overall, the work gives the impression of a carefully conducted study with a solid field and laboratory basis, but it requires substantial revision in the interpretation of results and in the treatment of methodological limitations.

I consider the manuscript a valuable and potentially publishable contribution after major revision

A particular concern is that several key conclusions (strengthening of a trophic “bottleneck”, the role of cyanobacteria) are drawn without a direct analysis of phytoplankton community composition in the recent period and currently rely on logical inference and previously published studies rather than on data from this dataset. These points should be explicitly acknowledged as hypotheses supported indirectly by past work, and, where possible, backed up by additional evidence or clearly framed as speculative.

The Introduction is a strong part of the manuscript. It is well structured, provides a broad and up-to-date context on the effects of warming on thermal regime, stratification and lake biota, with a clear emphasis on tropical and African lakes, and it states the aims of the study clearly. The Materials and Methods section is also detailed and mostly clear. However, an important practical issue is not addressed: it is not stated whether any cross-calibration was performed between the Hydrolab and YSI instruments (like overlapping measurements or laboratory comparisons). Given that the Discussion puts substantial weight on relatively small inter-period differences in water temperature, the question of instrument comparability is critical and should be addressed explicitly

In my view, the key problematic part of the study is section 2.8 (Biomass determination, body length and dry weight measurements) and the chain of results and interpretations that depend on it. The authors openly state that they use an “operational” set of length-weight equations inherited from a historical biomass workflow, with local adaptations of published regressions, and that direct comparison with some of the original published equations is not always possible. Thus, the equations effectively represent a historically calibrated system applied consistently to both periods. On the one hand, this indeed ensures internal comparability between the two periods, as the authors correctly emphasize. On the other hand, the resulting values for some groups are clearly unrealistic! for example, a roughly 20% increase in adult copepod body length coupled with an approximately 6x decrease in estimated dry weight. The authors themselves acknowledge that these are “equation-derived” quantities and advise caution in interpreting them, yet they then make really strong use of these results when discussing changes in size structure

I strongly recommend substantially strengthening the discussion of methodological limitations associated with the length-weight approach. The authors should clearly state that absolute dry-weight values for particular taxa or developmental stages may be systematically biased, and that the main robust conclusions of the study rest on overall trends in total zooplankton biomass and in the ZB/PB, rather than on detailed interpretations such as “increasing body length with decreasing mass” in particular copepod stages. It would also be very helpful to provide at least one numerical example comparing the operational equations with the original published length-weight regressions, to give the reader a sense of the magnitude and direction of discrepancies. Without such clarification, the relisability of the detailed statements about “contrasting length-mass patterns in copepods” in the Results and Discussion sections remains questionable

Another methodological point is the treatment of temporal structure in the data. The time series have a pronounced seasonal component and likely temporal autocorrelation (monthly observations), yet the statistical tests are applied as if observations were independent. At minimum, this should be acknowledged and discussed in the limitations section, and the authors should be cautious not to overstate the strength of evidence where effective sample size may be smaller than the nominal number of observations.

Related issues appear in the Results and Discussion. For rotifers and cladocerans, the ANOVA does not reveal statistically significant differences between periods, yet parts of the Discussion refer to marked declines, especially for Moina micrura. Here, the wording should be carefully aligned with the statistics: the authors should clearly distinguish between descriptive trends and changes that are statistically supported. In its current form, Table 4 appears overloaded: many parameters are listed, but it is not clear which of them are central to the main narrative. The authors should either highlight key metrics or more clearly explain in the text which parameters are most important for interpretation. The subsection that interprets “increasing length with decreasing mass” in adult copepods is particularly problematic, as it reads as a biological explanation of what is very likely a methodological artefact arising from the length-weight equations. This part should be substantially shortened and reframed, explicitly recognising that stage-specific patterns may be dominated by methodological effects.

Despite these issues, the manuscript is generally well written, with good English and a coherent logical flow. The text, however, is rather long, and some parts of the Discussion repeat similar arguments. Repetitions must be reduced, giving space to a more explicit and structured discussion of the study’s limitations and uncertainties. I also recommend moderating the wording regarding the role of cyanobacteria and the strengthening of the bottleneck, repeatedly emphasizing that phytoplankton composition was not measured in the present dataset and that these interpretations rely on indirect evidence and previous studies rather than direct observations.

Author Response

Title of Manuscript: Decadal shifts in zooplankton size structure and biomass in a warming tropical lake

Manuscript ID: limnolrev-4309989

General Response to the Reviewer

Response to Specific Comments

Comment: A particular concern is that several key conclusions (strengthening of a trophic “bottleneck”, the role of cyanobacteria) are drawn without a direct analysis of phytoplankton community composition in the recent period and currently rely on logical inference and previously published studies rather than on data from this dataset. These points should be explicitly acknowledged as hypotheses supported indirectly by past work, and, where possible, backed up by additional evidence or clearly framed as speculative.

Response: We agree that phytoplankton composition was not measured concurrently during 2018–2020 and that recent cyanobacterial dominance cannot be demonstrated directly. We therefore revised the Introduction, Discussion, and Conclusions to remove the previous one-sided emphasis on cyanobacteria and strengthening of a trophic bottleneck. The revised manuscript distinguishes chlorophyll a as an indicator of phytoplankton standing biomass from direct information on taxonomic composition, food quality, or primary production. Additional Lake Bosumtwi studies show that cyanobacteria historically dominated biomass, while chlorophytes, dinoflagellates, diatoms, cryptophytes, and other groups were also present. A 2015 survey suggests that cyanobacterial dominance may have persisted near the recent study period, but this cannot be confirmed for 2018–2020. We now present a testable combined bottom-up and top-down hypothesis: higher nutrients and chlorophyll a may have increased direct and indirect food resources for copepods, while increased copepodite and adult biomass may have strengthened predation pressure on smaller zooplankton. These pathways are explicitly identified as hypotheses because phytoplankton composition, microbial prey, copepod diet, and predation rates were not measured directly.

Comment 2: The Introduction is a strong part of the manuscript. It is well-structured, provides a broad and up-to-date context on the effects of warming on thermal regime, stratification and lake biota, with a clear emphasis on tropical and African lakes, and it states the aims of the study clearly.

Comment 3: The Materials and Methods section is also detailed and mostly clear. However, an important practical issue is not addressed: it is not stated whether any cross-calibration was performed between the Hydrolab and YSI instruments (like overlapping measurements or laboratory comparisons). Given that the Discussion puts substantial weight on relatively small inter-period differences in water temperature, the question of instrument comparability is critical and should be addressed explicitly.

Response: Direct cross‑calibration between the Hydrolab (2005–2006) and YSI Exo2 (2018–2020) was not possible because the historical instrument (Hydrolab) was no longer available. The YSI Exo2 was factory‑calibrated and verified with certified standards before field deployment, and both instruments have high temperature accuracy and low drift. This information has been added to Section 2.3. Temperature is the most stable parameter across sondes, with a typical accuracy of ±0.01–0.1 °C. The observed warming signal in our study (+0.27–0.30 °C across strata) exceeds the combined uncertainty of both instruments. Because our analysis focuses on relative inter‑period differences and both datasets were processed using the same stratification‑based framework, the temperature comparisons remain robust. The observed warming signal is further supported by independent indicators such as thermocline shoaling, reduced oxygen content of surface waters and expansion of the anoxic hypolimnion. We also acknowledge in the Discussion that the relatively small temperature differences should be interpreted with this comparative instrumentation in mind.

Comment: In my view, the key problematic part of the study is section 2.8 (Biomass determination, body length and dry weight measurements) and the chain of results and interpretations that depend on it. The authors openly state that they use an “operational” set of length-weight equations inherited from a historical biomass workflow, with local adaptations of published regressions, and that direct comparison with some of the original published equations is not always possible. Thus, the equations effectively represent a historically calibrated system applied consistently to both periods. On the one hand, this indeed ensures internal comparability between the two periods, as the authors correctly emphasise. On the other hand, the resulting values for some groups are clearly unrealistic. For example, a roughly 20% increase in adult copepod body length, coupled with an approximately 6x decrease in estimated dry weight. The authors themselves acknowledge that these are “equation-derived” quantities and advise caution in interpreting them, yet they then make really strong use of these results when discussing changes in size structure.

Response: We appreciate the Reviewer’s concern regarding the unrealistic patterns produced by the previously applied operational length–weight equations. In response, we re‑examined the biomass calculations and reassessed the underlying length–weight framework. Our initial attempt to apply the original published, taxon‑specific regressions to the 2018–2020 dataset yielded biomass values that appeared unrealistically high, which is why the earlier analysis relied on locally adjusted “operational” equations inherited from the historical workflow. As the Reviewer correctly notes, however, these adjusted equations produced biologically inconsistent outcomes, most notably the combination of increased copepod body length with a six‑fold decrease in estimated dry weight.

To resolve this, we recalculated all taxon and stage‑specific biomass using the original published equations listed in Table 1. The revised results now show positive, internally consistent length–weight relationships, eliminating the earlier contradictions. Importantly, the recalculated values indicate a substantial increase, not a decline, in zooplankton biomass in the recent period. This pattern aligns with evidence that in warming, cyanobacteria‑dominated systems, cyclopoid copepods can attain very high biomass by shifting to protozoan and microzooplankton prey that proliferate under bloom conditions. Given that cyclopoids constitute ~98% of the zooplankton community, this adaptive feeding strategy plausibly explains the observed biomass increase. We have updated the manuscript accordingly across the abstract, methods, results, and discussion, clarified the methodological rationale, and ensured that all interpretations are based on the recalculated, biologically consistent estimates. All values in the main text and Supplementary Materials have been revised. The updated analysis shows a 2.33‑fold increase in mean total zooplankton biomass, driven predominantly by copepods. We also clarify that body length was measured directly, whereas dry weight and biomass were equation‑derived. Absolute values may therefore contain some uncertainty, but applying the same equations to both periods ensures a valid internal comparison. Stage‑specific differences are presented descriptively unless inferential statistics are reported.

Comment: I strongly recommend substantially strengthening the discussion of methodological limitations associated with the length-weight approach. The authors should clearly state that absolute dry-weight values for particular taxa or developmental stages may be systematically biased, and that the main robust conclusions of the study rest on overall trends in total zooplankton biomass and in the ZB/PB, rather than on detailed interpretations such as “increasing body length with decreasing mass” in particular copepod stages. It would also be very helpful to provide at least one numerical example comparing the operational equations with the original published length-weight regressions, to give the reader a sense of the magnitude and direction of discrepancies. Without such clarification, the reliability of the detailed statements about “contrasting length-mass patterns in copepods” in the Results and Discussion sections remains questionable

Response: This concern has been addressed in the methodological revisions carried out, thus eliminating the need to address it in the discussion.

Comment: Another methodological point is the treatment of temporal structure in the data. The time series has a pronounced seasonal component and likely temporal autocorrelation (monthly observations), yet the statistical tests are applied as if observations were independent. At minimum, this should be acknowledged and discussed in the limitations section, and the authors should be cautious not to overstate the strength of evidence where the effective sample size may be smaller than the nominal number of observations.

Response: We agree that monthly limnological time‑series data typically exhibit both strong seasonal characteristics and temporal autocorrelation, and that this can reduce the effective sample size relative to the nominal number of observations. In the present study, our primary objective was to compare overall long‑term conditions between two multi‑year periods rather than short‑term temporal dynamics. For this reason, we applied standard parametric tests to the aggregated datasets, acknowledging that these tests assume independence, particularly as the measurements in the recent period are not influenced by the historical period. We have revised the Discussion to explicitly acknowledge this limitation and to clarify that the statistical evidence should be interpreted in light of potential temporal autocorrelation. Importantly, the principal conclusions of the study, including differences in thermal structure, lower dissolved oxygen, and substantially higher copepod-dominated zooplankton biomass in the recent period, are supported by large effect sizes, consistent multivariate patterns, and convergent ecological indicators, which reduce the likelihood that the results are artefacts of temporal dependence alone.

Comment: Related issues appear in the Results and Discussion. For rotifers and cladocerans, the ANOVA does not reveal statistically significant differences between periods, yet parts of the Discussion refer to marked declines, especially for Moina micrura. Here, the wording should be carefully aligned with the statistics: the authors should clearly distinguish between descriptive trends and changes that are statistically supported. In its current form, Table 4 appears overloaded: many parameters are listed, but it is not clear which of them are central to the main narrative. The authors should either highlight key metrics or more clearly explain in the text which parameters are most important for interpretation. The subsection that interprets “increasing length with decreasing mass” in adult copepods is particularly problematic, as it reads as a biological explanation of what is very likely a methodological artefact arising from the length-weight equations. This part should be substantially shortened and reframed, explicitly recognising that stage-specific patterns may be dominated by methodological effects.

Response: The revisions presented so far address all these concerns. Furthermore, Table 4 has been revised with the new internally consistent and biologically plausible results. These revisions address all the Reviewers’ concerns. We would prefer to retain Table 4, as documenting shifts in both size structure and biomass between the historical and recent periods is a central objective of the study.

Response: We revised the Results and Discussion to distinguish statistically supported differences from descriptive trends. Changes in M. micrura are now consistently described as descriptive because cladoceran biomass did not differ significantly among years. Table 4 and its note now clarify the occurrence-based summaries and treatment of zero values. The contradictory adult copepod length–weight interpretation was removed, and the corresponding text was shortened. We also reduced repetition in the Discussion and moderated the cyanobacterial interpretation.

End of Responses

******************************************************************************

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Review of limnolrev-4309989-peer-review-v2

This ms. is much improved. The methods have been clarified sufficiently so that my methodological questions have been answered.

Minor points:

Obviously, Table 1 needs to be formatted better.

Table 3 would be a good place to summarize Secchi transparency across periods.

The small individual size differences between periods for rotifers is as expected for eutely, although Brachionus sometimes displays polymorphism.

The six significant digits reported in lines 463-464 are excessive and surely not significant.

Author Response

Obviously, Table 1 needs to be formatted better.

Response: Thank you for this useful suggestion. We reformatted Table 1 to improve readability. The table now separates the taxon/developmental stage, equation used, and source/notes into distinct columns. We also revised the explanatory note below the table to avoid unnecessary repetition of sources already provided in the table.

Table 3 would be a good place to summarize Secchi transparency across periods.

Response: We added Secchi depth to Table 3 and revised the table caption accordingly, so that water transparency is now summarised together with the other habitat-layer and phytoplankton-related variables.

The small individual size differences between periods for rotifers is as expected for eutely, although Brachionus sometimes displays polymorphism.

Response: We agree that the small differences in individual size between periods for rotifers should not be overinterpreted. We added a cautionary sentence noting that such differences are expected to be limited by eutely and that Brachionus calyciflorus may exhibit polymorphism, which can contribute to body-size variation.

The six significant digits reported in lines 463-464 are excessive and surely not significant.

Response:We revised the excessive number of significant digits in the relevant lines and checked the manuscript for similar cases. Numerical values in the narrative text have been rounded to an appropriate level of precision, while retaining sufficient detail for interpretation.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for having considered my comments on the previous version of the manuscript. Numerous changes were done in the manuscript, and I think that it is now clearer, and the Methods section is more understandable. Overall, the quality of the manuscript has greatly improved. I only have a few minor comments that I listed below.

Specific comments:

L42. "bottom-up (e.g., food availability) forces regulate biomass". I suggest to change to "bottom-up (e.g., food availability) forces regulate zooplankton biomass"

L47-50. "Cyclopoid copepods of the genus Mesocyclops are trophically flexible omnivores that can exploit suitable phytoplankton, ciliates, rotifers, and small cladocerans, allowing them to respond to changes in both basal resources and animal prey availability". I suugest to add "For example" at the begining of this sentence.

L593. "chlorophyll a indicates standing biomass rather than production rate or food quality". I suggest to change to "chlorophyll a is a proxy for standing biomass rather than production rate or food quality".

Author Response

Specific comments:

L42. "bottom-up (e.g., food availability) forces regulate biomass". I suggest to change to "bottom-up (e.g., food availability) forces regulate zooplankton biomass"

Response:
Thank you for the suggestions. We have revised the sentence as recommended and changed the wording to “bottom-up (e.g., food availability) forces regulate zooplankton biomass”.

Response:
We have added “For example” at the beginning of the sentence, as suggested.

Response:
We have revised the sentence accordingly to “chlorophyll a is a proxy for standing biomass rather than production rate or food quality”.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for your thorough and thoughtful revision of the manuscript and for the detailed responses to my comments. I appreciate the considerable work you invested in re‑examining the source dataset, correcting the biomass calculations, and clarifying the methodological framework. In its revised form, the manuscript is much stronger. From my side, I have no remaining substantive concerns. I believe the manuscript is now suitable for publication, subject only to minor editorial polishing by the journal.

Author Response

Dear Authors,

Thank you for your thorough and thoughtful revision of the manuscript and for the detailed responses to my comments. I appreciate the considerable work you invested in re‑examining the source dataset, correcting the biomass calculations, and clarifying the methodological framework. In its revised form, the manuscript is much stronger. From my side, I have no remaining substantive concerns. I believe the manuscript is now suitable for publication, subject only to minor editorial polishing by the journal.

Response:
We sincerely thank the reviewer for the positive evaluation of the revised manuscript and for acknowledging the revisions made to the dataset, biomass calculations, and methodological framework. We are pleased that the reviewer considers the manuscript suitable for publication. We will carefully address any remaining minor editorial polishing requested by the journal.

Author Response File: Author Response.docx