the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jan 2026
Carbon flux responses to seasonal and annual hydroclimatic variability in a tropical dry forest in South Ecuador
Abstract. Tropical dry forests play an important role in the global carbon cycle, but their responses to climate variability are still not well understood. Using a three-year period (April 2022 to March 2025) of eddy covariance measurements, we studied seasonal and annual controls on carbon balances in a Tumbesian dry forest in Southern Ecuador. During the study period, the forest functioned as a net carbon sink, with a net ecosystem exchange (NEE) of -285 gCm−2 year−1. The strongest carbon uptake occurred during the wet period (Feb–May) with 173.86 ± 66 gCm−2 month−1, while it was reduced to 39.80 ± 8.12 gCm−2 month−1 in the dry season (August–November).
Light use efficiency (LUE) and water use efficiency (WUE) were used to characterize the functional controls on carbon fluxes at both seasonal and annual scales. WUE showed relatively stable water–carbon exchange, whereas LUE displayed clear seasonal variation, reflecting the strong influence of seasonal vegetation growth and greenness. Principle component analysis (PCA) was conducted to further analyze controlling mechanisms in carbon fluxes. Seasonal results showed that gross primary productivity (GPP) was mainly controlled by energy-related factors, while ecosystem respiration (Reco) was primarily driven by a moisture–temperature gradient. Annually, GPP was predominantly influenced by variations in vapor pressure deficit (VPD), soil temperature (Ts), and incoming radiation (Rg), reflecting a strong coupling between surface energy balance and atmospheric moisture demand. These drivers were further modulated by ENSO related climate variability, as reflected by shifts in their PCA loadings across years. Overall, the results reveal a decoupling between photosynthesis and respiration and show that tropical dry forests are highly vulnerable to increasing climate extremes, highlighting the need for improved representation of these processes in Earth system models.
- Preprint
(11182 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-222', Anonymous Referee #1, 15 Feb 2026
-
AC1: 'Reply on RC1', Charuta Murkute, 27 Mar 2026
This study examined fluxes from a seasonal dry forest at altitude in Southern Ecuador. The site was a sink with sensitivity to moisture and radiation levels. This is a well implement and reported flux study, ideal for the modelling community for calibration of ecosystem models dealing with ecosystem trajectories and climate change. It is a useful study in that it is data from under-represented ecosystem type (dry forest) and from an understudied region, and a forest at altitude.
The site appears to non-ideal and the energy balance closure reflects this at ~0.7 but this is openly reported and described. Unfortunately, it is only a 3 year study, hopefully the site can be supported at a decadal scale to develop a stronger data set capturing the significant inter- and intra-annual variability. PCA analysis was useful to determine the dominant environment drivers of C and water exchange, as well as separating the behaviours of WUE and LUE, key functional attributes to assess sensitivities to changes in rainfall distribution and increases in temperatures and atmospheric CO2.
So a somewhat stock-standard flux paper, but well acquitted and reported, a useful contribution. Worthy of publication.
Response: Thanks for your kind and thoughtful feedback. Your insights are encouraging in refining our work.
Specific comments
- L22 “… moderate annual carbon assimilation, but low mean respiration of this semi-arid ecosystems, dry forests are highly efficient carbon sequesters Poulter et al. (2014)”
Throughout the text the method to cite references is incorrect, looks like a ref software issue. Check and fix throughout.
Response: Thank you for your suggestion. We have applied the corrections.
- This text sentence should be “… moderate annual carbon assimilation, but low mean respiration of this semi-arid ecosystems, dry forests are highly efficient carbon sequesters (Poulter et al., 2014).”
Also re-write as “… dry forests have a high carbon sequestration efficiency (Poulter et al., 2014).”
Response: Thank you for your comment. Adapted as suggested.
- L54 “The authors highlighted that semi-arid ecosystems play an important role in driving trends and interannual variations in the global carbon cycle, primarily due to seasonal water availability, which was further supported by Biederman et al. (2016).”
A useful sentence is but is missing the fact that dry forest burns along with tropical savannas, especially high rainfall woody savannas, which behave similarly to these dry forests. A significant fraction of the interannual variation of global carbon cycle is driven by dry forest and savannas fire (see papers by van der Werf and colleagues) – especially large wet seasons that drive growth, followed by dry seasons that then burn this fuel with a major emission of CO2. So, the text needs to recognise the role of fire in dry forest carbon dynamics.
It would be also worth reporting the fire regime of the site and region. The word fire does not appear anywhere in the ms, odd for a paper on dry forests. The site is at altitude with perhaps low population density, with fire uncommon, but we need to know this. Is it possible to include a guesstimate of fire return interval? Think of fire as an herbivore, it consumes leaf area and primary productivity. No fire will turn a dry forest into a higher cover forest dry rainforest.
Response: Thank you for your comment. We agree that fire plays an important role in the carbon dynamics of many semi-arid ecosystems, particularly in dry forests and savannas where wet-season biomass production is often followed by dry-season burning, leading to significant CO2 emissions. Regarding the fire regime at our study site, we examined the NASA FIRMS (Fire Information for Resource Management System) satellite fire detection dataset for the study region during the observation period. The analysis indicates that no fire events were detected within the vicinity of the study site during the measurement period, suggesting that fire did not directly influence the observed ecosystem carbon fluxes during our study. The study site is located at relatively high elevation, which likely reduces the frequency of anthropogenic burning compared to lowland savanna systems.
- L90-91 Provide more detail on the typical extent of the seasonality - what is dry season rainfall/ less than 20-30 mm? The distribution of rainfall matters.
Response: Thank you for your comment. We identified the onset of seasons following the approach described by (Mendes et al., 2020) who defined seasonal transitions in the Caatinga biome based on the temporal behavior of key meteorological variables. Using this framework, we analyzed the observed patterns of precipitation in our dataset. In addition, we referred to (Stan and Sanchez-Azofeifa, 2019) who provide criteria for identifying the onset of the rainy season in tropical dry forest ecosystems. The delayed onset in 2022 - 2023 was determined based on a shift in these hydro-meteorological indicators compared to other years. We agree and have add this information to section 2 Material and Methods.
- L101 Provide more detail on the forest structure, type and extent. Link to Figure 1a, is this a widely distributed forest type. Fig 1a provides a green coloured relief map, or forest distribution? The caption does not state what the green shading is.
Response: Thank you for your comment. Figure 1a) shows the altitude of Ecuador and not the forested region. Figure 1b) shows a focus on the study site and constitutes the central part of the Tumbesian region.
Deciduous and semi-deciduous dry forests occur between 200 - 1100 m. Typical species include Ceiba trichistandra, Cavanillesia platanifolia, Eriotheca ruizi, Handroanthus chrysanthus, Terminalia valverdae, Bursera graveolens, and Piscidia carthagenensis (Ortiz et al., 2019).
In the 1960s–70s, heavy logging (especially of Handroanthus) supported the parquet industry. In 1978, southwestern Ecuador’s coastal forests were protected, reducing logging. However, from 1978–2008, forest cover declined sharply (33% in dry forests, 18% in shrublands), mainly due to conversion into pastures and expanding agriculture (Tapia-Armijos et al., 2019). This information has been added to the manuscript.
- Is there an herbaceous layer, what is the stem density, biomass and/or approximate seasonal range in LAI? I see you have used NDVI, but an LAI range would be useful as well.
Response: Thank you for your comment. We agree with the reviewer and we will add the LAI meteorological plot in Fig. 2 and regarding the stem density, biomass we will add that information in the manuscript.
- L112 “the Li-COR Inc. smart chamber (LI-870 CO2/H2O gas analyser and 8200-01S Smart chamber) was used in field campaigns.” How many campaigns, in the wet and dry seasons?
Response: Thank you for your comment. We have conducted one field campaign per season (wet and dry) and year. That means, we have conducted measurements during August 2022 (dry), March 2023 (wet), October 2023 (dry) and March 2024 (wet). The main goal was to obtain soil CO2 efflux during peak phases. We are aware that the conclusions are limited, but also give an idea of the contribution of the soil to the ecosystem respiration during these contrasting seasons.
- L117 “… standard corrections Foken et al. to …”. Add year for this citation.
Response: Thank you for your comment. Corrected.
- L167 “… account for the influence of slope and aspect on radiation measurements, corrections were applied to net radiation.” I have not seen a correction to Rnet for slope and aspect? The sensors are presumably level and are capturing the radiant load as a function of aspect. If you need to correct for slope, you should not be running an eddy covariance system on the site. And I don’t get why you need to correct for slope.
Response: Thank you for your comment. The net radiation sensor was installed horizontally, as is standard for eddy covariance measurements. However, because the site is located at a sloped surface, the surrounding surface receiving radiation is not horizontal. Therefore, the correction was not applied to the sensor measurements themselves, but rather to estimate the effective radiation received by the sloping terrain within the flux footprint. Several studies have shown that in complex terrain the incoming shortwave radiation received by the surface can differ from that measured by a horizontally mounted sensor because the angle of solar radiation depends on the local slope and aspect (Del Castillo et al., 2018; Hiller et al., 2008). Therefore, accounting for terrain geometry can improve the representation of the surface energy balance. For example, including slope and aspect in radiation calculations has been shown to improve energy balance closure.
- Figure 2 and caption add a space on the Y axis labels “mmmonth−1”, include a space “mm month-1). Easier to read, especially on a y axis label. Same for all units “Wm-2”, should be “W m-2” .
Response: Thank you for your comment. Corrected as suggested.
- The term ‘water balance’ in the cation as P-ET maybe water deficit as the term here, water balance implies you quantified recharge and runoff.
Response: Thank you for your comment. We will adjust as suggested.
- Figure 4 panel g) what is this plot, its not described in the caption
Response: Thank you for your comment. Panel g) describes PPFD and will be adapted.
- Unit label “gCm−2”, change to “g C m-2” throughout.
Response: Thank you for your comment. Adjusted as suggested.
- Table 3 Useful data but maybe move out of the body text and to the Supplementary Information, keep the cites to the Table.
Response: Thank you for your comment. We will adapt it in the manuscript.
- Figure 5 “Cumulative distribution of a) gross primary productivity …” This is a not a distribution, I would call this a cumulative time series plot. Plus the units are kind of odd, cumulative per hour? Why not per day given it’s a year i.e. 365 points.
Response: Thank you for your comment. We agree that the term cumulative distribution is not the most appropriate description for this figure. The plot represents the cumulative sum of GPP over time, so it is better described as a cumulative time series. We will revise the wording in the manuscript accordingly.
Regarding the units, the cumulative values were calculated from the original hourly flux data, which is why they are expressed per hour. However, we agree that presenting cumulative values on a daily basis would be clearer for annual time series.
- I though the Discussion read well, an interesting study
Response: Thank you for your kind comment.
References
Del Castillo, E. G., Paw U, K. T., and Sánchez-Azofeifa, A.: Turbulence scales for eddy covariance quality control over a tropical dry forest in complex terrain, Agricultural and Forest Meteorology, 249, 390–406, https://doi.org/10.1016/j.agrformet.2017.11.014, 2018.
Hiller, R., Zeeman, M. J., and Eugster, W.: Eddy-Covariance Flux Measurements in the Complex Terrain of an Alpine Valley in Switzerland, Boundary-Layer Meteorol, 127, 449–467, https://doi.org/10.1007/s10546-008-9267-0, 2008.
Mendes, K. R., Campos, S., Da Silva, L. L., Mutti, P. R., Ferreira, R. R., Medeiros, S. S., Perez-Marin, A. M., Marques, T. V., Ramos, T. M., De Lima Vieira, M. M., Oliveira, C. P., Gonçalves, W. A., Costa, G. B., Antonino, A. C. D., Menezes, R. S. C., Bezerra, B. G., and Santos E Silva, C. M.: Seasonal variation in net ecosystem CO2 exchange of a Brazilian seasonally dry tropical forest, Sci Rep, 10, 9454, https://doi.org/10.1038/s4159
Cueva Ortiz, J.; Espinosa, C.I.; Quiroz Dahik, C.; Aguirre Mendoza, Z.; Cueva Ortiz, E.; Gusmán, E.; Weber, M.; Hildebrandt, P. Influence of Anthropogenic Factors on the Diversity and Structure of a Dry Forest in the Central Part of the Tumbesian Region (Ecuador–Perú). Forests 2019, 10, 31. https://doi.org/10.3390/f10010031
Tapia-Armijos, M.F., Homeier, J., Espinosa, C.I., Leuschner, C. and De La Cruz, M., 2015. Deforestation and forest fragmentation in South Ecuador since the 1970s–losing a hotspot of biodiversity. PloS one, 10(9), p.e0133701.
Stan, K. and Sanchez-Azofeifa, A.: Tropical Dry Forest Diversity, Climatic Response, and Resilience in a Changing Climate, Forests, 10, 443, https://doi.org/10.3390/f10050443, 2019.
Citation: https://doi.org/10.5194/egusphere-2026-222-AC1
-
AC1: 'Reply on RC1', Charuta Murkute, 27 Mar 2026
-
RC2: 'Comment on egusphere-2026-222', Anonymous Referee #2, 02 Mar 2026
I hereby provide my review of the manuscript entitled ‘Carbon flux responses to seasonal and annual hydroclimatic variability in a tropical dry forest in South Ecuador’ by Murkute and others. In this work, the authors report three years of eddy covariance flux measurements at a tropical dry forest in the Southern Hemisphere. This dataset is unique, and the study is important. However, the manuscript in its present form has several lacunae. The language needs significant improvement. The figures are inappropriate. The referencing is done poorly. The results are interesting, but not explained. Neither are they compared against previous and relevant studies. The data processing lacks critical details, casting doubt on the validity of its findings. Keeping these serious limitations in mind, I recommend a ‘Rejection’ of this manuscript. My detailed comments are provided below.
Major comments:
- The language used in this manuscript needs severe improvement. In its present form, it is difficult to understand at several places what the authors want to convey. There are multiple instances of incorrect sentence formation, dangling sentences, typos, etc., which make the reading incomprehensible.
- All the references are incorrectly formatted. For example, Siyum (2020), instead of (Siyum, 2020), Miles et al. (2006) instead of (Miles et al. 2020), Foken et al., etc.
- Lines 37-38: How can the dry seasons be related to reduced microbial activity? Usually, water stress accelerates respiration.
- Figure 1: It is required to show the flux footprint at the site, preferably overlaid on the site picture.
- Lines 103-104: Any EC instrument must be installed at a sufficiently higher height to ensure it is above the surface roughness height. This is done to ensure that the EC measurement principles are met and that they are carried in the well-mixed boundary layer. As a rule of thumb, a measurement height of at least 1.5 times the average height of the surrounding canopy is recommended. Not meeting this condition can result in a gradient-driven exchange, which will not accurately represent the ecosystem-atmosphere exchanges of carbon and water.
- Line 103: Site latitude and longitude are not appropriately written.
- Section 2.3: The high-frequency EC measurements are susceptible to random noise, which is evident in the time series as spikes. How were these handled? Also, were the trends in the data removed?
- Lines 127-128: It is an acceptable practice to remove the outliers in any dataset that fall beyond (mu plus-minus 3 sigma) if the dataset is randomly distributed. It is not clear how the authors arrived at the strategy of removing flux values that are less than or equal to 2 times the dataset's sigma. This raises serious questions about the further processing of the dataset and the conclusions drawn.
- Equation 3: The components in the definition of TKE are wrong.
- Line 139: It is not clear to me what is meant by "... magnitude of turbulent ... and energy.".
- Lines 147-149: It is unclear what is intended to be stated in these lines. Two references are mentioned without explaining the context. Table 1 does not report these parameters as claimed in the text. What are b0, b1, and b2?
- Lines 151-154: The process described here is unclear. Please write in simple sentences and in an elaborate manner. What are the dimensionless filters? How are they inversely proportional to the effectiveness of turbulent transport?
- Lies 155-157: It is not appropriate to say that the boundary layer is weak or absent. What is the isothermal stratification? In fact, at nighttime, stable stratification in the atmospheric boundary layer is observed due to a cooler Earth surface in the absence of thermal heating.
- Line 163: Please elaborate on the moving point method and its significance for the present study.
- Sections 2.5 and 2.6: Please provide equations relating Rnѱ, Rn, and Rgѱ.
- At what height was PPFD measured? Mention the measurement heights of all the variables used in this study.
- Please report SM in VWC (volumetric water content).
- Lines 234-235: Could you elaborate more on the variability of VPD with rainfall? Usually, it is directly related to temperature and humidity.
- Line 239: Before using it further in this section and thereon, could you explain more about the values of the Niño 1+2 index and its interpretation?
- Figure 2: The units on the Y-axis are not written properly, for example, Wm-2 instead of W m-2.
- Table 1: This table should be improved. It should be unambiguously stated whether the values reported in this table are based on half-hourly data or otherwise. The slope is actually the degree of EBC. The intercept has a unit and should be mentioned.
- Line 253: The authors report higher closure of the surface energy budget during the wet season than the dry season. This is counterintuitive to several previous studies that reported a lower closure of the energy budget in wetter periods, for example, https://dx.doi.org/10.1127/0941-2948/2006/0167, https://dx.doi.org/10.1007/s12040-019-1158-x. Can the authors explain this? Perhaps, a comparison across sites will be more informative to the readers.
- Figure 3: GPP and Reco are barely discernible from each other. Please use a better colour scheme.
- Line 284: "Interestingly, the onset of the respective seasons starts later in 2022-2023". This is an interesting observation; however, it is not clear how the onsets of the seasons were identified. The authors should explain this explicitly with the formulation of any objective measures, if used.
- Figure 6 seems to be redundant.
- Figure 8: The unit of LUE is wrong. Markers, instead of MMarkers.
- Section 3.6: It is not clear at what scale the PCA was implemented. It should ideally be done at a half-hourly scale for a more realistic result.
- Lines 317-318: I do not understand this sentence.
- Lines 319-325: I do not quite follow the argument here. PC1 during the wet season shows positive loading from VPD and LE. Despite this, it is marked as energy-driven. On the contrary, PC2 in the dry season has positive loadings from Ts. but marked as moisture-driven. Strictly speaking, the results of the PCA are not decisive here and need further investigation. more specific analyses to be doubly sure.
- Line 326: ", but Reco contributed little to this axis (0.05)". Isn't Reco the target variable here? What am I missing?
- Figure 9: The captions are wrong and indicate to wrong figures in a misleading manner. The biplots are congested, and the variable labels can not be read.
- The Discussion section is written poorly and lacks depth. Mostly, the results are repeated and not explained sufficiently. The reasons behind observed variabilities are not objectively probed, but have somehow been attempted to justify by speculative means.
- Lines 375-380: These discussions are mostly speculative and lack a concrete evidence-based connection to the observations. In fact, they seem disjoint to a large extent.
- Lines 381-383: "Ta showed minimal ... changes in Rg". This is standard textbook information, which may be redundant to the audience of a scientific journal. However, it is more important to discuss the probable reasons governing the variabilities reported in this paper.
- Please explain first the Birch Effect to the audience with appropriate reference(s). Moreover, this explanation again seems to be heuristic in nature.
- A section in the Discussion section attempts to link El Niño with the observed features in this study. It is to be noted that ENSO is a planetary-scale event, and 3 years of record is too short to study its impact. Thus, it will probably result in wrong attribution and explanation.
- Lines 430-431: "As conditions ... canopy light absorption". How? Why do leaf shedding and reduced light absorption by the canopy affect LUE but not WUE? The exchange of light, carbon, and water all depend on the leaf area and should be impacted equally by any change in the leaf area.
Minor comments:
- Line 23: incomplete sentence.
- Lines 44-45: Garcia et al. (2017) did not study any Indian ecosystem.
- Line 105: ... at a frequency of …
- Line 109: What is TDR?
- Line 114: ... at a 15-minute interval …
- Line 200: Version 2 of which product? Please include a link at which the readers can access this dataset.
- Line 216: ".. mentioned above." Where?
- Line 219: Replace "climatic factors" with "climate variables".
- Line 273: "variables", instead of "factors".
- Lines 302-303: "In the remaining .... lower values". Rephrase this sentence. The same with the next one.
Citation: https://doi.org/10.5194/egusphere-2026-222-RC2 -
AC2: 'Reply on RC2', Charuta Murkute, 27 Mar 2026
I hereby provide my review of the manuscript entitled ‘Carbon flux responses to seasonal and annual hydroclimatic variability in a tropical dry forest in South Ecuador’ by Murkute and others. In this work, the authors report three years of eddy covariance flux measurements at a tropical dry forest in the Southern Hemisphere. This dataset is unique, and the study is important. However, the manuscript in its present form has several lacunae. The language needs significant improvement. The figures are inappropriate. The referencing is done poorly. The results are interesting, but not explained. Neither are they compared against previous and relevant studies. The data processing lacks critical details, casting doubt on the validity of its findings. Keeping these serious limitations in mind, I recommend a ‘Rejection’ of this manuscript. My detailed comments are provided below.
Major comments:
- The language used in this manuscript needs severe improvement. In its present form, it is difficult to understand at several places what the authors want to convey. There are multiple instances of incorrect sentence formation, dangling sentences, typos, etc., which make the reading incomprehensible.
Response: Thank you for your comment. We acknowledge the concern that the manuscript requires improvement and apologize for this inconvenience. We will correct typos, improve sentence construction and ensure that the scientific content is presented more clearly and coherently.
- All the references are incorrectly formatted. For example, Siyum (2020), instead of (Siyum, 2020), Miles et al. (2006) instead of (Miles et al. 2020), Foken et al., etc.
Response: Thank you for your comment. We apologize for this inconvenience and will of course correct the formatting.
- Lines 37-38: How can the dry seasons be related to reduced microbial activity? Usually, water stress accelerates respiration.
Response: Thank you for your comment. During the dry season in tropical dry forest ecosystems, soil moisture becomes very low and reaches down to 10 % .vol. Please also see Fig. 2a. This very low water availability limits microbial processes with a decline in microbial activity and decomposition rates as the diffusion of substrates and nutrients in soil water films becomes limited and many microorganisms enter dormant states (Castro et al., 2018 ; Sarma et al., 2022). We will add these points to the Introduction for clarification.
- Figure 1: It is required to show the flux footprint at the site, preferably overlaid on the site picture.
Response: Thank you for your comment. We agree with the reviewer and will include the flux footprint in Figure 1.
- Lines 103-104: Any EC instrument must be installed at a sufficiently higher height to ensure it is above the surface roughness height. This is done to ensure that the EC measurement principles are met and that they are carried in the well-mixed boundary layer. As a rule of thumb, a measurement height of at least 1.5 times the average height of the surrounding canopy is recommended. Not meeting this condition can result in a gradient-driven exchange, which will not accurately represent the ecosystem-atmosphere exchanges of carbon and water.
Response: Thank you for your comment. We agree with the reviewer that the height of the measurements is important for the quality and reliability of the flux data. Our measurement height is generally sufficient to capture turbulent fluxes above the canopy, although measurements may still be influenced by processes within the roughness sublayer. However, the EC system is installed at a sloped surface where the effective canopy height relative to the airflow is reduced due to topographic exposure. In such environments, the local flow structure and turbulence can differ from flat terrain, and studies have shown that EC measurements over slopes can still provide reliable flux data (Hammerle et al., 2007). Additionally, quality control procedures have been thoroughly applied to avoid non-reliable flux data. Please also see subsection 2.3 . We will add these points to the respective subsection.
- Line 103: Site latitude and longitude are not appropriately written.
Response: Thank you for your comment. We will write appropriately.
- Section 2.3: The high-frequency EC measurements are susceptible to random noise, which is evident in the time series as spikes. How were these handled? Also, were the trends in the data removed?
Response: Thank you for your comment. For the quality control and assurance we followed standard protocols using Li-COR EddyPro Software (7.0.9). For the high-frequency flux data that includes standard procedures for spike detection and removal to reduce the influence of random noise in the high-frequency time series. The software applies automated quality control routines such as despiking, coordinate rotation, time lag compensation, and frequency response corrections following established eddy covariance processing protocols. Please see Subsection 2.3. We will go through the respective subsection and be more transparent for the description of the data quality procedures.
- Lines 127-128: It is an acceptable practice to remove the outliers in any dataset that fall beyond (mu plus-minus 3 sigma) if the dataset is randomly distributed. It is not clear how the authors arrived at the strategy of removing flux values that are less than or equal to 2 times the dataset's sigma. This raises serious questions about the further processing of the dataset and the conclusions drawn.
Response: Thank you for this comment. We acknowledge that the mu plus-minus 3 sigma criterion is commonly used for removing outliers in normally distributed datasets. However, eddy covariance flux data often show strong seasonal variability and are not always normally distributed. In our case, the dataset was first separated into dry and wet seasons to account for the different environmental conditions and flux magnitudes during these periods. The (mu plus-minus 2 sigma) criterion was not chosen a priori but was arrived at through an iterative, data driven quality control procedure. An initial (mu plus-minus 3 sigma) filter was applied within each seasonal subset, following (Vickers and Mahrt, 1997). Visual inspection confirmed that this threshold was insufficient to remove all non-physical values a known limitation of single pass sigma methods caused by the masking effect (Vitale et al., 2021), where large spikes inflate the sigma estimate and allow moderately anomalous values to pass undetected. The threshold was subsequently tightened iteratively to mu plus-minus 2 sigma until non-physical spikes were no longer present upon visual inspection.
- Equation 3: The components in the definition of TKE are wrong.
Response: Thank you for your comment. Turbulent Kinetic Energy (TKE) represents the kinetic energy per unit mass associated with turbulent fluctuations in atmospheric flow and serves as a direct measure of the mixing efficiency of the atmosphere above the ecosystem. In this study, TKE was not applied directly but through a modified turbulence velocity scale u_TKE (del Castillo et al., 2018; Wilczak et al., 2001). In equation 3 where u′², v′², and w′² are the variances of the streamwise, cross-wind, and vertical wind velocity fluctuations, respectively, measured at 10 Hz by sonic anemometer. Expressing TKE as a velocity scale (m s-1) allows direct comparison with other turbulence indicators such as friction velocity (u*) and the standard deviation of vertical wind. We will clarify the meaning to avoid any confusion.
- Line 139: It is not clear to me what is meant by "... magnitude of turbulent ... and energy.".
Response: Thanks for the comment. The u_TKE-based dimensionless indices Iᵤ and Iw are inversely proportional to the magnitude of turbulent transport of mass and energy where mass transport refers to the vertical turbulent flux of CO2 that constitutes the measured ecosystem carbon exchange, and energy transport refers to the turbulent sensible heat and latent heat fluxes that together must balance the available radiative energy at the surface. Please also see (Del Castillo et al., 2018). We will clarify the meaning to avoid any confusion in the manuscript.
- Lines 147-149: It is unclear what is intended to be stated in these lines. Two references are mentioned without explaining the context. Table 1 does not report these parameters as claimed in the text. What are b0, b1, and b2?
Response: Thank you for your comment. To estimate the vertical wind component used in the filtering procedure, we applied a planar fit rotation to the wind field. In this approach, the mean vertical velocity is estimated using a multiple regression with the horizontal wind components. The coefficients b0, b1, and b2 are the regression parameters that define the fitted plane representing the mean flow surface for different wind directions (Del Castillo et al., 2018; Paw U et al., 2000; Wilczak et al., 2001). We will add this information to clarify the meaning and avoid any confusion.
- Lines 151-154: The process described here is unclear. Please write in simple sentences and in an elaborate manner. What are the dimensionless filters? How are they inversely proportional to the effectiveness of turbulent transport?
Response: Thanks for your comment. Based on the turbulence intensity scale, two dimensionless indices (Iu and Iw) were calculated. These indices compare the strength of the mean wind flow with the strength of turbulence. Because they are ratios, they are called dimensionless filters. Iu represents the ratio between the mean horizontal wind speed and the turbulence intensity scale. Iw represents the ratio between the vertical wind component and the turbulence intensity scale (Del Castillo et al., 2018).
These indices help to determine whether turbulent transport or mean flow dominates the exchange processes. When the values of these indices are small, turbulence is strong relative to the mean wind, and turbulent mixing is effective. In contrast, when the values are large, the mean wind is stronger than the turbulence. This indicates weak turbulent mixing and a higher possibility that fluxes are influenced by advection. For this reason, the indices are considered inversely related to the effectiveness of turbulent transport: higher index values correspond to weaker turbulent mixing.
To apply the filtering procedure, we examined the relationship between the measured CO2 flux and the turbulence indices. As turbulence increases, the magnitude of the measured flux typically increases until it reaches a stable level. Beyond this point, further increases in turbulence do not significantly change the flux values. This point was used as a threshold value to separate periods with adequate turbulence from those with insufficient turbulence.
All data points with turbulence index values beyond the threshold were removed from further analysis. This filtering step ensures that the final dataset includes only measurements obtained under conditions where turbulent transport dominates, which is a key assumption of the eddy covariance method.
We will add these explanations to the respective subsection for a more transparent and clear description.
- Lies 155-157: It is not appropriate to say that the boundary layer is weak or absent. What is the isothermal stratification? In fact, at nighttime, stable stratification in the atmospheric boundary layer is observed due to a cooler Earth surface in the absence of thermal heating.
Response: Thanks for the comment. We agree with the reviewer and will clarify the statement. During nighttime, a nocturnal boundary layer is developed, resulting in weak turbulent activities that limits vertical mixing and resulting in net ecosystem exchange (NEE) values unrepresentative of actual gas exchange in the ecosystem (Wutzler et al., 2018; Loescher et al., 2003). We also observed these processes at our study site, where using Hobo loggers we measured temperature and relative humidity at five different levels (1m - 22m) development of stable stratification layer during nighttime conditions.
- Line 163: Please elaborate on the moving point method and its significance for the present study.
Response: Thank you for your comment. In this study, the moving point method was applied to calculate separate u* thresholds for different seasonal conditions, specifically the dry season, wet season, wet-to-dry and dry-to-wet seasons. The flux data were grouped into bins according to increasing u* values, and the average CO2 flux was calculated for each bin. As turbulence increases, the measured flux typically increases until it reaches a stable value. The u* threshold was identified as the point beyond which further increases in u* do not lead to significant changes in the measured flux. Determining season specific u* thresholds were important because ecosystem processes and atmospheric conditions vary between different seasons. Using separate thresholds allows a more accurate identification of periods with insufficient turbulence for each seasonal condition, thereby improving the reliability of the filtered flux dataset also done by (Barr et al., 2013) and (Papale et al., 2006). We will adapt the description in the respective subsection.
- Sections 2.5 and 2.6: Please provide equations relating Rnѱ, Rn, and Rgѱ.
Response: Thank you for your comment. We will adapt it in the manuscript.
Rn is defined as follows:
Rn = SWin + SWout + LWin + LWout
The procedure requires calculations of the solar zenith angle (θz) and the angle between the horizontal to the inclined surface and the direction of the sun (ψ^2; all angles are expressed in radians). We calculated the corrections using an additional EC system in the study site that had been installed over a plane surface. Using measurements of SWin radiation at the reference site, the incoming solar radiation for an inclined plane (SWψ) can be calculated by applying the following equation:
SW_ψ= SW exp〖[-kt( ψ^2- θ_z^2 )][1+Asin^2 (ψ/2)]〗
Here, kt is the clearness index and A is the surface albedo. kt values are calculated using the ratio of SW and estimates of the extra-terrestrial radiation and A values are calculated using SWin and SWout measurements from the net radiometers, the solar zenith angle (θz), the angle between the horizontal and the inclined surfaces, and the direction of the sun using NOAA’s Solar Calculator based on astronomical algorithms . Similarly, LWin, LWout, and SWout were also corrected for calculation of net-radiation, which is defined as follows:
Rnψ = SWinψ + SWoutψ + LWinψ + LWoutψ
We agree with the reviewer and will adjust the manuscript by adding this description.
- At what height was PPFD measured? Mention the measurement heights of all the variables used in this study.
Response: Thank you for your comment. All sensors measuring environmental conditions, i.e. PPFD, net radiometer, precipitation, air temperature and relative humidity are installed in the same height level as the Irgason, i.e. 20 m above ground. Soil sensors for measuring volumetric soil water content and soil temperature as well as ground heat flux plate are installed in a depth of 20 cm. We will add this information in the description of the instrumentation.
- Please report SM in VWC (volumetric water content).
Response: Thanks for your comment. We will adjust it as suggested.
- Lines 234-235: Could you elaborate more on the variability of VPD with rainfall? Usually, it is directly related to temperature and humidity.
Response: Thank you for your comment and good point. We agree that VPD is directly controlled by temperature and relative humidity. In our study, rainfall was used mainly as an indicator of seasonal moisture conditions, as it shows a clear seasonal pattern at the study site. We also observed relatively low variability in temperature, while humidity changed more strongly between wet and dry periods. Therefore, the seasonal variation in VPD in our study is mainly associated with changes in atmospheric moisture related to rainfall patterns. Rainfall was used to illustrate this seasonal trend rather than implying a direct causal relationship with VPD. After rainfall, increased soil moisture and evaporation lead to higher atmospheric humidity, which reduces VPD. In contrast, during dry periods with little or no rainfall, lower humidity and higher temperatures typically result in higher VPD values. Please also see Fig. 2. We will include this point to the manuscript.
- Line 239: Before using it further in this section and thereon, could you explain more about the values of the Niño 1+2 index and its interpretation?
Response: Thank you for your comment. The Niño 1+2 index represents sea surface temperature (SST) anomalies in the eastern equatorial Pacific Ocean near the coasts of Peru and Ecuador. The values show how much the ocean temperature deviates from the long-term average. Positive values indicate warmer than normal SSTs, which correspond to El Niño conditions. Negative values indicate cooler than normal SSTs, which correspond to La Niña conditions. Values close to zero represent neutral conditions. We agree and will add the description to subsection 2.8.
- Figure 2: The units on the Y-axis are not written properly, for example, Wm-2 instead of W m-2.
Response: Thank you for your comment. We apologize and will correct it in the manuscript.
- Table 1: This table should be improved. It should be unambiguously stated whether the values reported in this table are based on half-hourly data or otherwise. The slope is actually the degree of EBC. The intercept has a unit and should be mentioned.
Response: Thank you for your comment. We follow standard procedures and the values reported in the table are based on half-hourly data. This information as well as the intercept of the regression equation is now reported in the manuscript with its corresponding unit (W m-2).
- Line 253: The authors report higher closure of the surface energy budget during the wet season than the dry season. This is counterintuitive to several previous studies that reported a lower closure of the energy budget in wetter periods, for example, https://dx.doi.org/10.1127/0941-2948/2006/0167, https://dx.doi.org/10.1007/s12040-019-1158-x. Can the authors explain this?
Response: Thanks for the comment. We agree that the results do not align with other studies. However, in our study site we observed an increase of the slope to 0.74 during the wet period indicating a relatively higher energy balance closure efficiency. This is likely due to enhanced latent heat flux (LE) driven by increased evapotranspiration. The higher RMSE (55.22 W m-2) suggests a higher variability in energy partitioning. Further, we also observed a higher energy balance closure during the wet season in the montane tropical rainforest (Murkute et al., 2024). This is consistent with findings from comparable tropical ecosystems (Campos et al., 2019), while studies from temperate, mid-latitude sites report lower EBC during wetter periods (e.g. Mauder and Foken, 2006). The discrepancy is related to fundamental differences in energy partitioning between tropical and temperate systems. In tropical rainforest sites the wet season is characterised by intense evapotranspiration and a strong dominance of LE in the energy budget (Campos et al., 2019). The same is true for our site see in Figure 1 in the supplementary. We will add a paragraph in the discussion to address this point.
- Perhaps, a comparison across sites will be more informative to the readers.
Response: Thank you for your comment. We agree with the reviewer and will add a comparison with other sites such as (Castro et al., 2018; García et al., 2017;Mendes et al.,2020).
- Figure 3: GPP and Reco are barely discernible from each other. Please use a better colour scheme.
Response: Thank you for your comment. We agree and will adapt to a better color scheme.
- Line 284: "Interestingly, the onset of the respective seasons starts later in 2022-2023". This is an interesting observation; however, it is not clear how the onsets of the seasons were identified. The authors should explain this explicitly with the formulation of any objective measures, if used.
Response: Thank you for this comment. We identified the onset of seasons following the approach described by (Mendes et al., 2020), who defined seasonal transitions in the Caatinga biome based on the temporal behavior of key meteorological variables. Using this framework, we analyzed the observed patterns of precipitation in our dataset. In addition, we referred to (Stan and Sanchez-Azofeifa, 2019), who provide criteria for identifying the onset of the rainy season in tropical dry forest ecosystems. The delayed onset in 2022 - 2023 was determined based on a shift in these hydro-meteorological indicators compared to other years. We agree and will add this information to section 2 Material and Methods.
- Figure 6 seems to be redundant.
Response: Thank you for your comment. The idea was to show the variability of the relationships between carbon fluxes and the environmental conditions. We hope the reviewer agrees to keep this figure in the manuscript.
- Figure 8: The unit of LUE is wrong. Markers, instead of MMarkers.
Response: Thank you for your comment. It will be corrected. We apologize for this inconvenience.
- Section 3.6: It is not clear at what scale the PCA was implemented. It should ideally be done at a half-hourly scale for a more realistic result.
Response: Thank you for your comment. Following studies by (Zhou et al., 2022) and (Zhu et al., 2022) we performed the PCA on daily data. We will add this information to subsection 2.9. However, we also have conducted the PCA on half-hourly data, but the results were biased by the strong diurnal cycle of the radiation components which result in a dominant energy driven gradient. This behavior is consistent with previous eddy-covariance studies where radiation strongly controls short-term variability in carbon and energy fluxes (Baldocchi, 2003). Please see Figure 2 in supplementary.
- Lines 317-318: I do not understand this sentence.
Response: Thank you for your comment. We have corrected the sentence. We apologize for this inconvenience.
Fig. 9 presents PCA bi-plots showing the relationships between environmental variables and ecosystem fluxes (GPP and Reco) across the four defined seasons (wet, dry, wet-dry, dry-wet ). The corresponding variable loadings and explained variances are provided in Table B1.
- Lines 319-325: I do not quite follow the argument here. PC1 during the wet season shows positive loading from VPD and LE. Despite this, it is marked as energy-driven. On the contrary, PC2 in the dry season has positive loadings from Ts. but marked as moisture-driven. Strictly speaking, the results of the PCA are not decisive here and need further investigation. more specific analyses to be doubly sure.
Response: Thanks for the comment. We will clarify the statement accordingly. The high loading of VPD on PC1 (0.81) does not contradict the energy-driven interpretation. During the wet season, when soil moisture is abundant and vegetation is not under water stress, VPD behaves primarily as a thermodynamic consequence of solar heating rather than as an independent moisture stress signal. As incoming solar radiation (Rg) increases through the day, air temperature (Ta) rises, and the saturation vapour pressure increases proportionally widening the gap between the moisture-holding capacity of the air and its actual water vapour content, thereby elevating VPD (Baldochi et al., 2001,Grace et al., 2010). This means that Rg, Ta, and VPD all rise and fall together following the radiation variation, which is precisely why they co-load on the same principal component. Similarly, the positive loading of LE on PC1 reflects energy-driven evapotranspiration, during the wet season, SM is not limiting and the primary control on evaporation is the availability of solar energy, so LE increases and decreases in direct response to Rg rather than in response to water availability. PC1 therefore captures the shared diurnal co-variation of the entire radiation atmosphere system.
- Line 326: ", but Reco contributed little to this axis (0.05)". Isn't Reco the target variable here? What am I missing?
Response: Thanks for the comment. For Reco, PC1 was dominated by the same radiation atmosphere variables (Ta = 0.89, Rg = 0.88, VPD = 0.87, H = 0.84), but Reco loaded negligibly on this axis (0.05) indicating that wet-season ecosystem respiration is decoupled from the diurnal radiation forcing that drives GPP, and is instead primarily captured by PC2 (Reco loading = 0.81) alongside soil moisture (0.82) and soil temperature (-0.68). This confirms substrate moisture rather than radiation control of Reco during the wet season. We will clarify the statement accordingly.
- Figure 9: The captions are wrong and indicate to wrong figures in a misleading manner. The biplots are congested, and the variable labels can not be read.
Response: Thank you for your comment. We apologize and have adapted the figure caption.
- The Discussion section is written poorly and lacks depth. Mostly, the results are repeated and not explained sufficiently. The reasons behind observed variabilities are not objectively probed, but have somehow been attempted to justify by speculative means.
Response: Thank you for your comment. We regret that the reviewer has this impression. However, to improve our manuscript, we will thoroughly revise the discussion, deepen the discussion on mechanistic understanding of the results and a better comparison to previous studies.
- Lines 375-380: These discussions are mostly speculative and lack a concrete evidence-based connection to the observations. In fact, they seem disjoint to a large extent.
Response: Thank you for this comment. The seasonal shift affects the timing and intensity of rainfall, which subsequently influences water related variables such as SM, RH, and VPD. Similar relationships between atmospheric circulation patterns, ITCZ dynamics, and ecosystem responses have also been reported in previous studies conducted in tropical montane regions (Spannl et al., 2016; Stan and Sanchez-Azofeifa, 2019). These references have now been included in the revised manuscript to better support the interpretation of the observed seasonal patterns. We hope the reviewer agrees.
- Lines 381-383: "Ta showed minimal ... changes in Rg". This is standard textbook information, which may be redundant to the audience of a scientific journal. However, it is more important to discuss the probable reasons governing the variabilities reported in this paper.
Response: Thank you for your comments. We agree and will add a paragraph in the discussion concerning the mentioned points.
Ta exhibited limited seasonal variability, reflecting relatively stable large-scale radiative forcing and advective controls. In contrast, Ts showed pronounced seasonal dynamics driven by variations in Rg, SM availability, and associated shifts in surface energy partitioning. The divergence between Ta and Ts thus emerges from fundamentally different coupling mechanisms: while Ta is buffered by atmospheric mixing and regional-scale processes, Ts responds directly to local land - atmosphere exchange processes. In particular, SM exerts a first-order control on the partitioning of available energy into LE and H thereby modulating surface temperature variability and amplifying thermal responses under water-limited conditions (Gevaert et al., 2018; Williams and Torn, 2015). This leads to enhanced H and reduced evaporative cooling during dry periods, strengthening the decoupling between surface and atmospheric temperatures.
- Please explain first the Birch Effect to the audience with appropriate reference(s). Moreover, this explanation again seems to be heuristic in nature.
Response: Thank you for your comments. We agree with the reviewer and will add a description of the Birch effect.
The Birch Effect describes the explosive pulse of CO2 released from soil when the first rainfall rewets the dry soil after a prolonged drought caused by the simultaneous reactivation of dormant soil microbes, breakdown of microbial osmolytes, and decomposition of months of accumulated dry-season litter producing a transient period where ecosystem respiration far exceeds GPP and the forest temporarily becomes a net carbon source (Birch, 1958; Jarvis et al., 2007). This process is reported in many dry forest sites such as (Castro et al., 2018; García et al., 2017). Please also see lines 395 -397.
- A section in the Discussion section attempts to link El Niño with the observed features in this study. It is to be noted that ENSO is a planetary-scale event, and 3 years of record is too short to study its impact. Thus, it will probably result in wrong attribution and explanation.
Response: Thank you for your comment. We agree that ENSO is a planetary scale phenomenon and that a three year dataset is too short to draw robust conclusions concerning ENSO. However, we can observe a strong interannual variability in the carbon fluxes as well as in the precipitation.
Even if the ENSO is a planetary phenomenon, we are in the ENSO core region, which has more clear relations to rainfall in our dry forest site (wet during EN and dry during LN) (Bendix et al., 2017; Bendix & Bendix, 2006) than more remote areas (as e.g. Europe or Africa) where teleconnections are modified by many confounding factors. Thus, we can learn general behaviour from already one El Niño. Of course, because the individual events are somewhat different in timing and intensity, the one year cannot be the full picture.
Looking at the typical ENSO indices, defining EL, LN and neutral conditions, a link to ENSO can be observed. The idea is thus, to explain interannual variations in the carbon exchange impacted by large scale conditions as also applied by various studies (Olchev et al., 2015; Castro et al., 2018).
- Lines 430-431: "As conditions ... canopy light absorption". How? Why do leaf shedding and reduced light absorption by the canopy affect LUE but not WUE? The exchange of light, carbon, and water all depend on the leaf area and should be impacted equally by any change in the leaf area.
Response: Thanks for the comment. We thank the reviewer for this important comment. Although leaf shedding reduces leaf area and therefore affects light, carbon, and water exchange simultaneously, LUE and WUE respond at different rates because they are governed by different physiological mechanisms. LUE declines earlier because photosynthetic efficiency is directly tied to chlorophyll content, both of which begin deteriorating during early leaf senescence (before leaves are physically shed) (Xu & Baldocchi 2003). WUE by contrast, remains relatively stable during this early senescence phase because stomatal closure triggered by rising VPD and declining soil moisture reduces both carbon assimilation and transpiration in near-equal proportions, preserving the carbon water ratio (Medlyn et al., 2011). LUE is therefore more sensitive to leaf biochemical quality and canopy phenological state, while WUE is more sensitive to stomatal regulation and atmospheric moisture demand. We have revised the manuscript accordingly to clarify the statement.
Minor comments:
- Line 23: incomplete sentence.
Response: Thanks. Adjusted.
- Lines 44-45: Garcia et al. (2017) did not study any Indian ecosystem.
Response: Thanks for your comment. We apologize for the inconvenience. We will add appropriate citations to Rodha et al., 2021.
- Line 105: ... at a frequency of …
Response: Thank you for your comment. We will adapt in the manuscript.
- Line 109: What is TDR?
Response: Thanks for your comment. Time-domain reflectometry (TDR). We will adapt in the manuscript.
- Line 114: ... at a 15-minute interval …
Response: Thank you for your comment. The measurements were taken at a 15 minutes interval for the soil respiration.
- Line 200: Version 2 of which product? Please include a link at which the readers can access this dataset.
Response: Thank you for your comment. The link is provided in the references.
- Line 216: ".. mentioned above." Where?
Response: Thank you for your comment. Adapted. It is mentioned in section 2.3 in L161.
- Line 219: Replace "climatic factors" with "climate variables".
Response: Thanks. Adjusted as suggested.
- Line 273: "variables", instead of "factors".
Response: Thanks. Adjusted as suggested.
- Lines 302-303: "In the remaining .... lower values". Rephrase this sentence. The same with the next one.
Response: Thanks. In the other seasons, values were lower throughout the year.Thus, variations in WUE are more pronounced between years than within a single year. Adjusted as suggested.
References:
Baldocchi, D. D.: Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of ecosystems: past, present and future, Global Change Biology, 9, 479–492, https://doi.org/10.1046/j.1365-2486.2003.00629.x, 2003.
Baldocchi, D. D. and Xu, L.: What limits evaporation from Mediterranean oak woodlands – The supply of moisture in the soil, physiological control by plants or the demand by the atmosphere?, Advances in Water Resources, 30, 2113–2122, https://doi.org/10.1016/j.advwatres.2006.06.013, 2007.
Barr, A. G., Richardson, A. D., Hollinger, D. Y., Papale, D., Arain, M. A., Black, T. A., Bohrer, G., Dragoni, D., Fischer, M. L., Gu, L., Law, B. E., Margolis, H. A., McCaughey, J. H., Munger, J. W., Oechel, W., and Schaeffer, K.: Use of change-point detection for friction–velocity threshold evaluation in eddy-covariance studies, Agricultural and Forest Meteorology, 171–172, 31–45, https://doi.org/10.1016/j.agrformet.2012.11.023, 2013.
Bendix J, Beck E. Environmental change and its impacts in a biodiversity hotspot of the south Ecuadorian Andes–monitoring and mitigation strategies. Erdkunde. 2016 Jan 1:1-4.
Bendix A, Bendix J. Heavy rainfall episodes in Ecuador during El Niño events and associated regional atmospheric circulation and SST patterns. Advances in Geosciences. 2006 Jan 9;6:43-9.
Birch, H.F. 1964. Mineralisation of plant nitrogen following alternate wet and dry conditions. Plant Soil 20:43–49.
Campos, S., Mendes, K. R., Da Silva, L. L., Mutti, P. R., Medeiros, S. S., Amorim, L. B., Dos Santos, C. A. C., Perez-Marin, A. M., Ramos, T. M., Marques, T. V., Lucio, P. S., Costa, G. B., Santos E Silva, C. M., and Bezerra, B. G.: Closure and partitioning of the energy balance in a preserved area of a Brazilian seasonally dry tropical forest, Agricultural and Forest Meteorology, 271, 398–412, https://doi.org/10.1016/j.agrformet.2019.03.018, 2019.
Castro, S. M., Sanchez-Azofeifa, G. A., and Sato, H.: Effect of drought on productivity in a Costa Rican tropical dry forest, Environ. Res. Lett., 13, 045001, https://doi.org/10.1088/1748-9326/aaacbc, 2018.
Del Castillo, E. G., Paw U, K. T., and Sánchez-Azofeifa, A.: Turbulence scales for eddy covariance quality control over a tropical dry forest in complex terrain, Agricultural and Forest Meteorology, 249, 390–406, https://doi.org/10.1016/j.agrformet.2017.11.014, 2018.
García, A. G., Di Bella, C. M., Houspanossian, J., Magliano, P. N., Jobbágy, E. G., Posse, G., Fernández, R. J., and Nosetto, M. D.: Patterns and controls of carbon dioxide and water vapor fluxes in a dry forest of central Argentina, Agricultural and Forest Meteorology, 247, 520–532, https://doi.org/10.1016/j.agrformet.2017.08.015, 2017.
Gevaert, A. I., Miralles, D. G., De Jeu, R. A. M., Schellekens, J., and Dolman, A. J.: Soil Moisture‐Temperature Coupling in a Set of Land Surface Models, JGR Atmospheres, 123, 1481–1498, https://doi.org/10.1002/2017JD027346, 2018.
Grace, J. B., Anderson, T. M., Olff, H., & Scheiner, S. M. (2010). On the specification of structural equation models for ecological systems. Ecological monographs, 80(1), 67-87.
Hammerle, A., Haslwanter, A., Schmitt, M., Bahn, M., Tappeiner, U., Cernusca, A., and Wohlfahrt, G.: Eddy covariance measurements of carbon dioxide, latent and sensible energy fluxes above a meadow on a mountain slope, Boundary-Layer Meteorol, 122, 397–416, https://doi.org/10.1007/s10546-006-9109-x, 2007.
Jarvis, Paul, Ana Rey, Charalampos Petsikos, Lisa Wingate, Mark Rayment, João Pereira, João Banza et al. "Drying and wetting of Mediterranean soils stimulates decomposition and carbon dioxide emission: the “Birch effect”." Tree physiology 27, no. 7 (2007): 929-940.
Loescher, H. W., Oberbauer, S. F., Gholz, H. L., and Clark, D. B.: Environmental controls on net ecosystem‐level carbon exchange and productivity in a Central American tropical wet forest, Global Change Biology, 9, 396–412, https://doi.org/10.1046/j.1365-2486.2003.00599.x, 2003.
Medlyn, B. E., Duursma, R. A., and Zeppel, M. J. B.: Forest productivity under climate change: a checklist for evaluating model studies, WIREs Climate Change, 2, 332–355, https://doi.org/10.1002/wcc.108, 2011.
Mendes, K. R., Campos, S., Da Silva, L. L., Mutti, P. R., Ferreira, R. R., Medeiros, S. S., Perez-Marin, A. M., Marques, T. V., Ramos, T. M., De Lima Vieira, M. M., Oliveira, C. P., Gonçalves, W. A., Costa, G. B., Antonino, A. C. D., Menezes, R. S. C., Bezerra, B. G., and Santos E Silva, C. M.: Seasonal variation in net ecosystem CO2 exchange of a Brazilian seasonally dry tropical forest, Sci Rep, 10, 9454, https://doi.org/10.1038/s41598-020-66415-w, 2020.
Murkute, C., Sayeed, M., Pucha-Cofrep, F., Carrillo-Rojas, G., Homeier, J., Limberger, O., Fries, A., Bendix, J., and Trachte, K.: Turbulent Energy and Carbon Fluxes in an Andean Montane Forest—Energy Balance and Heat Storage, Forests, 15, 1828, https://doi.org/10.3390/f15101828, 2024.
Olchev, A., Ibrom, A., Panferov, O., Gushchina, D., Kreilein, H., Popov, V., Propastin, P., June, T., Rauf, A., Gravenhorst, G., and Knohl, A.: Response of CO2 and H2 O fluxes in a mountainous tropical rainforest in equatorial Indonesia to El Niño events, Biogeosciences, 12, 6655–6667, https://doi.org/10.5194/bg-12-6655-2015, 2015.
Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W., Longdoz, B., Rambal, S., Valentini, R., Vesala, T., and Yakir, D.: Towards a standardized processing of Net Ecosystem Exchange measured with eddy covariance technique: algorithms and uncertainty estimation, Biogeosciences, 3, 571–583, https://doi.org/10.5194/bg-3-571-2006, 2006.
Paw U, K. T., Baldocchi, D. D., Meyers, T. P., and Wilson, K. B.: Correction Of Eddy-Covariance Measurements Incorporating Both Advective Effects And Density Fluxes, Boundary-Layer Meteorology, 97, 487–511, https://doi.org/10.1023/A:1002786702909, 2000.
Rodda, S. R., Thumaty, K. C., Praveen, M., Jha, C. S., and Dadhwal, V. K.: Multi-year eddy covariance measurements of net ecosystem exchange in tropical dry deciduous forest of India, Agricultural and Forest Meteorology, 301–302, 108351, https://doi.org/10.1016/j.agrformet.2021.108351, 2021.
Sarma, D., Burman, P. K. D., Chakraborty, S., Gogoi, N., Bora, A., Metya, A., Datye, A., Murkute, C., and Karipot, A.: Quantifying the net ecosystem exchange at a semi-deciduous forest in northeast India from intra-seasonal to the seasonal time scale, Agricultural and Forest Meteorology, 314, 108786, https://doi.org/10.1016/j.agrformet.2021.108786, 2022.
Spannl, S., Volland, F., Pucha, D., Peters, T., Cueva, E., and Bräuning, A.: Climate variability, tree increment patterns and ENSO-related carbon sequestration reduction of the tropical dry forest species Loxopterygium huasango of Southern Ecuador, Trees, 30, 1245–1258, https://doi.org/10.1007/s00468-016-1362-0, 2016.
Stan, K. and Sanchez-Azofeifa, A.: Tropical Dry Forest Diversity, Climatic Response, and Resilience in a Changing Climate, Forests, 10, 443, https://doi.org/10.3390/f10050443, 2019.
Vickers, D. and Mahrt, L.: Quality Control and Flux Sampling Problems for Tower and Aircraft Data, J. Atmos. Oceanic Technol., 14, 512–526, https://doi.org/10.1175/1520-0426(1997)014<0512:QCAFSP>2.0.CO;2, 1997.
Vitale, D., Fratini, G., Bilancia, M., Nicolini, G., Sabbatini, S., and Papale, D.: A robust data cleaning procedure for eddy covariance flux measurements, Biogeosciences, 17, 1367–1391, https://doi.org/10.5194/bg-17-1367-2020, 2020.
Wilczak, J. M., Oncley, S. P., and Stage, S. A.: Sonic Anemometer Tilt Correction Algorithms, Boundary-Layer Meteorology, 99, 127–150, https://doi.org/10.1023/A:1018966204465, 2001.
Williams, I. N. and Torn, M. S.: Vegetation controls on surface heat flux partitioning, and land‐atmosphere coupling, Geophysical Research Letters, 42, 9416–9424, https://doi.org/10.1002/2015GL066305, 2015.
Wutzler, T., Lucas-Moffat, A., Migliavacca, M., Knauer, J., Sickel, K., Šigut, L., Menzer, O., and Reichstein, M.: Basic and extensible post-processing of eddy covariance flux data with REddyProc, Biogeosciences, 15, 5015–5030, https://doi.org/10.5194/bg-15-5015-2018, 2018.
Zhou, L., Zhang, X., Zhang, Y., Song, Q., Myo, S. T. Z., Zhou, R., Lin, Y., Liu, Y., Bai, K., Gnanamoorthy, P., and Balasubramanian, D.: The cumulative drought exert disruptive effects on tropical rainforests in the northern edge of Asia - Based on decadal dendrometric measurements and eddy covariance method, Agricultural and Forest Meteorology, 316, 108858, https://doi.org/10.1016/j.agrformet.2022.108858, 2022.
Zhu, S., Clement, R., McCalmont, J., Davies, C. A., and Hill, T.: Stable gap-filling for longer eddy covariance data gaps: A globally validated machine-learning approach for carbon dioxide, water, and energy fluxes, Agricultural and Forest Meteorology, 314, 108777, https://doi.org/10.1016/j.agrformet.2021.108777, 2022.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 177 | 125 | 22 | 324 | 18 | 34 |
- HTML: 177
- PDF: 125
- XML: 22
- Total: 324
- BibTeX: 18
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
egusphere-2026-222
This study examined fluxes from a seasonal dry forest at altitude in Southern Ecuador. The site was a sink with sensitivity to moisture and radiation levels. This is a well implement and reported flux study, ideal for the modelling community for calibration of ecosystem models dealing with ecosystem trajectories and climate change. It is I a useful study in that it is data from under-represented ecosystem type (dry forest) and from an understudied region, and a forest at altitude.
The site appears to non-ideal and the energy balance closure reflects this at ~0.7 but this is openly reported and described. Unfortunately, it is only a 3 year study, hopefully the site can be supported at a decadal scale to develop a stronger data set capturing the significant inter- and intra-annual variability. PCA analysis was useful to determine the dominant environment drivers of C and water exchange, as well as separating the behaviours of WUE and LUE, key functional attributes to assess sensitivities to changes in rainfall distribution and increases in temperatures and atmospheric CO2.
So a somewhat stock-standard flux paper, but well acquitted and reported, a useful contribution. Worthy of publication.
Specific comments
L22 “… moderate annual carbon assimilation, but low mean respiration of this semi-arid ecosystems, dry forests are highly efficient carbon sequesters Poulter et al. (2014)”
Throughout the text the method to cite references is incorrect, looks like a ref software issue. Check and fix throughout.
This text sentence should be “… moderate annual carbon assimilation, but low mean respiration of this semi-arid ecosystems, dry forests are highly efficient carbon sequesters (Poulter et al., 2014).”
Also re-write as “… dry forests have a high carbon sequestration efficiency (Poulter et al., 2014).”
L54 “The authors highlighted that semi-arid ecosystems play an important role in driving trends and interannual variations in the global carbon cycle, primarily due to seasonal water availability, which was further supported by Biederman et al. (2016).”
A useful sentence is but is missing the fact that dry forest burns along with tropical savannas, especially high rainfall woody savannas, which behave similarly to these dry forests. A significant fraction of the interannual variation of global carbon cycle is driven by dry forest and savannas fire (see papers by van der Werf and colleagues) – especially large wet seasons that drive growth, followed by dry seasons that then burn this fuel with a major emission of CO2. So, the text needs to recognise the role of fire in dry forest carbon dynamics.
It would be also worth reporting the fire regime of the site and region. The word fire does not appear anywhere in the ms, odd for a paper on dry forests. The site is at altitude with perhaps low population density, with fire uncommon, but we need to know this. Is it possible to include a guesstimate of fire return interval? Think of fire as an herbivore, it consumes leaf area and primary productivity. No fire will turn a dry forest into a higher cover forest dry rainforest.
L90-91 Provide more detail on the typical extent of the seasonality - what is dry season rainfall/ less than 20-30 mm? The distribution of rainfall matters.
L101 Provide more detail on the forest structure, type and extent. Link to Figure 1a, is this a widely distributed forest type. Fig 1a provides a green coloured relief map, or forest distribution? The caption does not state what the green shading is.
Is there an herbaceous layer, what is the stem density, biomass and/or approximate seasonal range in LAI? I see you have used NDVI, but an LAI range would be useful as well.
L112 “the Li-COR Inc. smart chamber (LI-870 CO2/H2O gas analyser and 8200-01S Smart chamber) was used in field campaigns.” How many campaigns, in the wet and dry seasons?
L117 “… standard corrections Foken et al. to …”. Add year for this citation.
L167 “… account for the influence of slope and aspect on radiation measurements, corrections were applied to net radiation.” I have not seen a correction to Rnet for slope and aspect? The sensors are presumably level and are capturing the radiant load as a function of aspect. If you need to correct for slope, you should not be running an eddy covariance system on the site. And I don’t get why you need to correct for slope.
Figure 2 and caption add a space on the Y axis labels “mmmonth−1”, include a space “mm month-1). Easier to read, especially on a y axis label. Same for all units “Wm-2”, should be “W m-2” .
The term ‘water balance’ in the cation as P-ET maybe water deficit as the term here, water balance implies you quantified recharge and runoff.
Figure 4 panel g) what is this plot, its not described in the caption
Unit label “gCm−2”, change to “g C m-2” throughout.
Table 3 Useful data but maybe move out of the body text and to the Supplementary Information, keep the cites to the Table.
Figure 5 “Cumulative distribution of a) gross primary productivity …” This is a not a distribution, I would call this a cumulative time series plot. Plus the units are kind of odd, cumulative per hour? Why not per day given it’s a year i.e. 365 points.
I though the Discussion read well, an interesting study.