Response to major comments:

I can add text to the conclusion about best practices for monitoring procedures to determine mixing classifications.

Response to comments on lines 39-50 and lines 120-:

Regarding the two above points, it appears that when I transferred the manuscript to the template provided by the journal, I copied the Introduction twice – once into section 1 and the second time into section 2. The Materials and Methods are missing. I have pasted them below. I apologize for this oversight.

“Formation of the four study lakes occurred during the late-Wisconsin glaciation ~12,000 years ago (marine isotope stage 2; Jennings and Johnson 2011). They occupy a tunnel valley that was formed beneath the Wadena lobe of the Laurentide Ice Sheet. Following glacial retreat, the melting of stagnant ice blocks within the tunnel valley left depressions in the landscape now occupied by lakes and wetlands (Wright Jr. 1993).

Today, the four lakes investigated in this study sit in the HUC-12 watershed that sources the headwaters of the Mississippi River (U.S. Geological Survey 2017). Budd (478.6 meters above mean sea level; MAMSL) is the highest elevation, while Arco (465.8 MAMSL) and Josephine (465.4 MAMSL) lie at similar elevations (Figure 1). Deming is the lowest elevation of the lakes (464.8 MAMSL).

Lake depth measurements were collected using a Garmin Striker 4 dual-beam transducer (sonar) attached to a rowboat or canoe. Depth and GPS measurements were taken every six seconds while the boat was in motion. A Garmin GLO 2 GPS receiver and ArcGIS Collector app was used to navigate, track the boat’s course, and ensure even coverage. The shoreline of the lakes was obtained by walking along accessible areas of the shore with the Garmin GLO 2 GPS receiver, or from Lidar-derived digital elevation models. Bathymetry rasters (1 m resolution) were generated from the depth measurements in ArcGIS Pro 3.0 using a 3rd-degree Local Polynomial Interpolation. These rasters were used to calculate lake volumes and contour maps. Rasters and volume data have been deposited with the Environmental Data Initiative (Swanner et al. 2022).

Chemical, physical, and biological parameters measured on the four lakes included depth, temperature, specific conductance, salinity, turbidity, pH, oxidation-reduction potential, dissolved oxygen, photosynthetically active radiation, chlorophyll-a, and phycocyanin. Major cations, anions, and isotopes of water (δ2H-H2O and δ18O-H2O) were determined on lake water retrieved from different depths within the four lakes. Taxon-specific chlorophyll-a fluorescence was collected with a Fluoroprobe (BBE Moldaenke). The data and description of methods are available in the Environmental Data Initiative (Swanner et al. 2022). Measurements were made and samples were collected from a boat anchored within the deepest basin of each lake.

A string of temperature loggers (HOBO Water Temp Pro v2) placed at different depths were deployed into the deep areas of Deming, Arco, and Budd Lakes for one year. A conductance logger (HOBO Conductivity Logger) was added near the bottom of the strings in Arco and Budd after six months. These sensors measured temperature every thirty minutes and specific conductivity every 2 hours. The sensor string was not retrieved from Budd, as it could not be located in May 2022. Conductance measurements with a Yellow Spring Instruments ProDSS temperature/conductivity sensor on deployment and removal were used to check for drift in the HOBO conductance logger. Hourly wind speed data for the duration of sensor deployment utilized the ITCM5 (47.2400, -95.1900, 1480 feet elevation) weather station in Itasca State Park. Data was downloaded from MesoWest (https://mesowest.utah.edu/). Plots and analyses were produced in Python or R Studio (2022.07.2) using the RLakeAnalyzer package v.1.11.4.1 (Winslow et al. 2019).

Major anions (CO32-, HCO3-, Cl-, and SO42-) and cations (Na+, K+, Ca2+, Mg2+) were used to produce a Piper diagram in Geochemist’s Workbench 15.0. The concentration of cations and anions was calculated as the percentage of total cations and anions in meq L-1.

The isotopes of water (δ2HH2O and δ18OH2O) were measured on spring or seep water that had been filtered with 0.45 micron nylon syringe filters and stored at 4 °C with minimal headspace until analysis. Samples were analyzed with a Picarro L1102-i Isotopic Liquid Water Analyzer at the Stable Isotope Laboratory at Iowa State University. The analytical uncertainty and average correction factor for δ18OH2O are ± 0.05 ‰ and ± 0.30 ‰ for δ2HH2O relative to V-SMOW.

Samples for microscopy and water color were collected into amber bottles with a Van Dorn sampler from three different depths in each lake, including the SCML, if present, as determined with the YSI ProDSS. Water color was determined on water filtered through a GF-75 (Advantec) glass fiber filter (Cuthbert and del Giorgio 1992). Absorbance was measured at 440 nm and 750 nm. The absorption coefficient (g; m-1) was calculated by subtracting the absorbance at 750 nm from the absorbance at 440 nm and dividing by the path length (m):

g440 = (2.303 * A440 – A750)/(path length)                                                 (2)

A conversion was necessary to determine the color (mg Pt L-1) of the lake water:

Color = 18.216*(g440) - 0.209                                                                       (3)

Water sampled from the SCML was preserved with 1% Lugol’s solution upon returning to the laboratory. Fixed samples were settled in the dark for three to seven days.          

Student reports from courses taking place over several decades at the Itasca Biological Station and Laboratories (IBSL) (Knoll and Cotner 2018), formerly the Itasca Biological Station, were acquired from the library at the University of Minnesota, Twin Cities.”

Response to comments on line 150: 

We did not sample from a fixed mooring. The methods now include this statement, “Measurements were made and samples were collected from a boat anchored within the deepest basin of each lake.”

Response to comments on line 212:

Adjusting the scale to each dataset allows the trends to be visibly resolvable, and makes for easy visual comparison of one time point to another. I could keep the scales the same but would need to change the number of plots in the paper width, so it would be harder to compare. It is just a trade-off.

Response to comments on line 251:

This acronym is defined in the second paragraph of the methods.

Response to comments on line 253:

Supplementary Figure 2 is the drought record. Supplementary Figure 1 contains a cross-sectional lake profile showing the lake level referred to in this figure, and corresponds to the cross-sections identified in Figure 1.

Response to comments on Figure 6:

Thanks for catching this. It should be Supplementary Table 2.

Response to comments on line 275:

The last sentence of the caption has been modified to clarify, “The cross closest to the lake data points is the Deming bog sample, and the cross closest to the intersection of the LEL and LMWL is the Nicollet Creek sample.”

Response to comments on line 317:

This should be Supplementary Figure 5.

Response to comments on line Supplementary Figure 19:

This is the 95% confidence interval. I have added this information to the caption.

RC2 General comments:

• More needs to be included to justify the work. What is the novelty of the study as well as the relevance to the wider field. At the moment, I don’t feel this is included in either the abstract, introduction, or discussion.

(AC2) Line 75 from introduction addresses this point: “The identification of meromictic lakes is important as they are critical analogues for understanding of the biogeochemistry of past oxygen-stratified oceans (Swanner et al., 2020) and alterations to global biogeochemical cycles that may result from climate change and anthropogenic impacts strengthening stratification in lakes.”

However, the introduction could be rearranged to lead with the (current) last paragraph and build a bit more about why these particular lakes are of interest. The introduction was initially drafted by students and so may read a bit more as a review of the characteristics of meromictic lakes, but this is easily fixed. A few sentences can also be added to emphasize the relevance of these lakes to e.g. lake trends due to climate change and the particular applicability of these lakes to understanding the biogeochemistry of stratified systems more broadly.

• The methods are incomplete and organization needs to be improved. See my specific comments on what is missing.

(AC2) Please see comment AC2 on the EGUsphere forum. When copying the manuscript to the Biogeosciences template, I mistakenly copied the introduction twice instead of the methods. I apologize for this mistake, but the editor recommended this instead of re-uploading a corrected version. I believe the methods will address many of your comments.

• Why do we care about the biology/chemistry and how does it contribute to the characterization of the mixing regimes and/or the impact of that mixing. Need to link together the physics and the biology/chemistry to make a more cohesive story and
• Conclusions on the mixing regimes of some of the lakes is drawn from a small sample (7 profiles) and incomplete data collection (not to the maximum depth of Arco). Some acknowledgment, at the least, needs to be included that discusses how the data availability impacts the conclusions that are being drawn.

(AC2) We acknowledge there is a data limitation with the seasonal monitoring approach and even with the deployment of sensors in these small lakes. This can be addressed with a few additional statements in the text.

RC2 Specific comments:          

Introduction:

• More needs to be made to highlight the novelty of the study and/or the relevant contribution that this will make to the field in the introduction. E.g. are meromictic lakes understudied, very numerous, relevant to global processes? This occurs slightly in the final paragraph but is not sufficient.

(AC2) Please see my earlier comments and suggested reorganization of the introduction. There is particular interest in the geochemical community to use meromictic lakes to understand the biogeochemistry of stratified systems through space and time, in addition to their relevance to understanding future states (e.g. enhanced stratification) of lakes globally due to climate change. More emphasis to these points can be added throughout.

• Line 39-45 references are needed for these statements and the equation.

(AC2) This equation was taken from Wetzel 2001 textbook and is assumed to be well-known, but this reference can be added.

• Lines 50-55: SCML are not distinct for meromictic lakes so why particularly are you interested in them for this study?

(AC2) As shown by the data, the SCML in these lakes are well-documented, and of particular interest for the abundance of cyanobacteria, which has relevance for their utility in understanding past stratified systems. This point can be more clearly made and referenced.

• Overall, the introduction reads more like a review or a summary of information about meromictic lakes. You should include more text about what is not known in the field, why Is it relevant that we understand mixing regimes for these individual lakes. This is done a bit in the final paragraph but should be done throughout to build toward the research questions and integrate the information from the preceding paragraphs.

(AC2) See above comments. This is fair, and to some extent reflects the contribution of student writers. The introduction can be revised to give more specific justification for why these lakes are of interest scientifically.

RC2 Methods:

• What is the accuracy/precision of the sensors?
• What depths are the sensors and water samples collected? (spatial resolution). This is reported in the results (line 189, 201) but should be in the methods. When did sampling occur (timings, frequency)? What was the duration of the sensor deployment?
• Include information on what/how/where pH, PAR, chlorophyll a and phycocyanin were measured (frequency/depths etc.)
• “Rasters and volume data have been deposited with the Environmental Data Initiative…” – and used to calculate the max/mean depths and relative depths as per equation 1? The methods to generate the information in Table 1 needs to be detailed
• “Student reports from courses taking place over several decades at the Itasca Biological Station and Laboratories (IBSL) (Knoll and Cotner 2018), formerly the Itasca Biological Station, were acquired from the library at the University of Minnesota, Twin Cities.” More information is needed on these reports - how many, the timing, frequency etc. Also, the Knoll & Colner (2018) citation is not in the list of references.

(AC2) This citation will be added.

• Precipitation data is also used but not detailed in methods
• Overall, the organization of the methods was difficult to follow. Use subheadings to guide the reader through the method types (e.g. manual sampling, lab methods, high-frequency sampling etc.)
• Physical metrics (schmidt stability, lake number, buoyancy frequency) calculations not in the methods.
• Spring/creek isotopic composition data collection not in the methods. Need information about these sites/locations and the rationale behind the collection.

(AC2) I believe all of the preceding comments are addressed in the methods, which were omitted from the uploaded template by mistake. I am pasting them here as well:

“Formation of the four study lakes occurred during the late-Wisconsin glaciation ~12,000 years ago (marine isotope stage 2; Jennings and Johnson 2011). They occupy a tunnel valley that was formed beneath the Wadena lobe of the Laurentide Ice Sheet. Following glacial retreat, the melting of stagnant ice blocks within the tunnel valley left depressions in the landscape now occupied by lakes and wetlands (Wright Jr. 1993).

Today, the four lakes investigated in this study sit in the HUC-12 watershed that sources the headwaters of the Mississippi River (U.S. Geological Survey 2017). Budd (478.6 meters above mean sea level; MAMSL) is the highest elevation, while Arco (465.8 MAMSL) and Josephine (465.4 MAMSL) lie at similar elevations (Figure 1). Deming is the lowest elevation of the lakes (464.8 MAMSL).

Lake depth measurements were collected using a Garmin Striker 4 dual-beam transducer (sonar) attached to a rowboat or canoe. Depth and GPS measurements were taken every six seconds while the boat was in motion. A Garmin GLO 2 GPS receiver and ArcGIS Collector app was used to navigate, track the boat’s course, and ensure even coverage. The shoreline of the lakes was obtained by walking along accessible areas of the shore with the Garmin GLO 2 GPS receiver, or from Lidar-derived digital elevation models. Bathymetry rasters (1 m resolution) were generated from the depth measurements in ArcGIS Pro 3.0 using a 3rd-degree Local Polynomial Interpolation. These rasters were used to calculate lake volumes and contour maps. Rasters and volume data have been deposited with the Environmental Data Initiative (Swanner et al. 2022).

Chemical, physical, and biological parameters measured on the four lakes included depth, temperature, specific conductance, salinity, turbidity, pH, oxidation-reduction potential, dissolved oxygen, photosynthetically active radiation, chlorophyll-a, and phycocyanin. Major cations, anions, and isotopes of water (δ²H-H₂O and δ¹⁸O-H₂O) were determined on lake water retrieved from different depths within the four lakes. Taxon-specific chlorophyll-a fluorescence was collected with a Fluoroprobe (BBE Moldaenke). The data and description of methods are available in the Environmental Data Initiative (Swanner et al. 2022). Measurements were made and samples were collected from a boat anchored within the deepest basin of each lake.

A string of temperature loggers (HOBO Water Temp Pro v2) placed at different depths were deployed into the deep areas of Deming, Arco, and Budd Lakes for one year. A conductance logger (HOBO Conductivity Logger) was added near the bottom of the strings in Arco and Budd after six months. These sensors measured temperature every thirty minutes and specific conductivity every 2 hours. The sensor string was not retrieved from Budd, as it could not be located in May 2022. Conductance measurements with a Yellow Spring Instruments ProDSS temperature/conductivity sensor on deployment and removal were used to check for drift in the HOBO conductance logger. Hourly wind speed data for the duration of sensor deployment utilized the ITCM5 (47.2400, -95.1900, 1480 feet elevation) weather station in Itasca State Park. Data was downloaded from MesoWest (https://mesowest.utah.edu/). Plots and analyses were produced in Python or R Studio (2022.07.2) using the RLakeAnalyzer package v.1.11.4.1 (Winslow et al. 2019).

Major anions (CO₃^2-, HCO₃^-, Cl^-, and SO₄^2-) and cations (Na⁺, K⁺, Ca²⁺, Mg²⁺) were used to produce a Piper diagram in Geochemist’s Workbench 15.0. The concentration of cations and anions was calculated as the percentage of total cations and anions in meq L^-1.

The isotopes of water (δ²H_H2O and δ¹⁸O_H2O) were measured on spring or seep water that had been filtered with 0.45 micron nylon syringe filters and stored at 4 °C with minimal headspace until analysis. Samples were analyzed with a Picarro L1102-i Isotopic Liquid Water Analyzer at the Stable Isotope Laboratory at Iowa State University. The analytical uncertainty and average correction factor for δ¹⁸O_H2O are ± 0.05 ‰ and ± 0.30 ‰ for δ²H_H2O relative to V-SMOW.

Samples for microscopy and water color were collected into amber bottles with a Van Dorn sampler from three different depths in each lake, including the SCML, if present, as determined with the YSI ProDSS. Water color was determined on water filtered through a GF-75 (Advantec) glass fiber filter (Cuthbert and del Giorgio 1992). Absorbance was measured at 440 nm and 750 nm. The absorption coefficient (g; m^-1) was calculated by subtracting the absorbance at 750 nm from the absorbance at 440 nm and dividing by the path length (m):

g₄₄₀ = (2.303 * A₄₄₀ – A₇₅₀)/(path length)                                                        (2)

A conversion was necessary to determine the color (mg Pt L^-1) of the lake water:

Color = 18.216*(g₄₄₀) - 0.209                                                              (3)

Water sampled from the SCML was preserved with 1% Lugol’s solution upon returning to the laboratory. Fixed samples were settled in the dark for three to seven days.

Student reports from courses taking place over several decades at the Itasca Biological Station and Laboratories (IBSL) (Knoll and Cotner 2018), formerly the Itasca Biological Station, were acquired from the library at the University of Minnesota, Twin Cities.”

RC2 Results/discussion:

• Could Figures 1 and 2 be combined to have the morphometry in the landscape context. Currently the scale bars on Figure 2 are difficult to read and to notice the differences in scales. It should be noted that the scale bars differ among the individual lake maps if it is to be kept in the current format (which would be my last option)

(AC2) This seems reasonable and we can work on this.

• Differences in units on the x-axis on Figure 3 should be noted in the caption or made consistent

(AC2) We can add a sentence telling the reader to note the variable x-axes.

• Line 200: dimictic is a subset of holomixis no separate.

(AC2) Holomixis can be removed from this statement.

• The data from high frequency sensors show similar patterns at Arco and Deming (isothermal in fall) but you make different conclusions about mixing regimes. Why?

(AC2) The specific conductance data show maintenance of a chemocline in Deming Lake (lines 150-151), while this is not the case for Arco (lines 195-196).

• Comparisons of Schmidt stability values among lakes is not appropriate due to differences in depth/volume/surface area (lines 213-214). These need to be normalized (see Winslow et al., 2017 as an example) or another unitless metric used.

(AC2) We implemented the Schmidt stability calculation in rLakeAnalyzer package, which “was formalized by Idso (1973) to reduce the effects of lake volume on the calculation (resulting in a mixing energy requirement per unit area).” The values of Schmidt stability are per surface area in the units reported (J/m^2). Furthermore, the volumes, depths, and surface areas of the lakes are quite comparable (Table 1). We do not have high resolution buoy data for all lakes, as in the Winslow et al., 2017 study referenced that would be necessary for the temporal normalizations described in that work. We do have the thermistor data for Arco and Deming, but don’t feel that is appropriate to compare with the seasonal profiles from Josephine and Arco, and thus present the Schmidt stability calculations only from the seasonal profiles in Supplementary Figure 6.

• Line 206 reference to lake ice out dates 2022, is this the right author? Use of “Likely” needs justification, why is it likely?

(AC2) This reference refers to a state database. It is how the reference manager interpreted the journal style but may be corrected during copy editing. Likely can be replaced with possibly, but the manuscript shows that thermal stratification is strong during summer and there is no physical mechanism for mixing under ice, so isothermal periods are the likely time for mixing.

RC2: Winslow, L.A., Read, J.S., Hansen, G.J., Rose, K.C. and Robertson, D.M., 2017. Seasonality of change: Summer warming rates do not fully represent effects of climate change on lake temperatures. Limnology and Oceanography, 62(5), pp.2168-2178.