the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Apr 2024
Ecosystem characteristics of land covers with various anthropogenic impacts in a tropical forest region of Southeast Asia
Abstract. Given the severe anthropogenic pressure on tropical forests and the high demand for field observations of ecosystem characteristics, it is crucial to collect such data both in pristine tropical forests and in the converted deforested land cover classes. To gain insight into the ecosystem characteristics of pristine tropical forests, regrowth forests, and cashew plantations, we established an ecosystem monitoring site in Phnom Kulen National Park, Cambodia. Here, we present observed datasets of forest inventories, leaf area index, leaf traits of woody species, a fraction of intercepted photosynthetically active radiation, and edaphic and meteorological conditions. We examined how land-use and land-cover change affect species and functional diversity, stand structure, and edaphic conditions among the three land-cover classes. We further investigated relationships between diameters at breast height and tree height, estimated aboveground biomass (AGB), and explored relationships between ecosystem characteristics and AGB. We uncovered some key differences in ecosystem characteristics among the land-cover classes. We also demonstrated the feasibility of locally updating AGB estimates using power law functions. These datasets and findings can contribute to filling data gaps in tropical forest research, addressing global environmental challenges, and supporting sustainable forest management.
- Preprint
(2199 KB) - Metadata XML
-
Supplement
(3474 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2024-98', Anonymous Referee #1, 03 Jul 2024
This study provides observed datasets of meteorological conditions, forest inventory, leaf traits of woody species, leaf area index (LAI) and a fraction of photosynthetically active radiation (fPAR) of pristine tropical forests, regrowth forests, and cashew plantations within Phnom Kulen National Park. Nine forest plots were selected and these forest inventory results cover the period from Apr 2022 to Apr 2023. The authors compared the differences in these ecosystem characteristics among the three types of land-cover classes and provided a well-organized dataset of a tropical forest ecosystem. However, despite the hard monitoring work, I am not quite sure whether this dataset is valuable enough to reach the high standard of ESSD.
Major concern
This study compared the species diversity, leaf functional traits, biomass, and other characteristics of three types of forests. I doubt whether this dataset is unique and useful enough. Does the comparison of these indexes among three types of forests can provide us with any new insights that haven’t been investigated before? The authors claimed that this dataset fills data gaps in tropical forest research, so does it mean there were few studies investigating the ecosystem characteristics in tropical forests? I cannot see any new findings according to the data analysis section.
Three methods were used to calculate aboveground biomass, and the selection of the method can cause a large difference in the calculated results. How to prove which method is better than the other two as the authors claimed in the discussion 4.5.4?
Minor
The journal required a data link at the end of the abstract.
Please add the background of land cover map in Figure 1.
The titles of these tables are too long. Please move the note to the end of the table.
Authors in Table 1 is not a good representation of reference. Just use “reference”.
What is the difference in precipitation and temperature data between Figs. 2 and 3? Are they repeated results?
In Fig. 4, does the small plot represent the local amplifier of the plot outside? The green line seems different among the two plots.
Hard to distinguish the line form of AGBh and AGBf in Figure 5a.
Line 383-384. I doubt the dataset in this study can be used to investigate the effect of land use change from natural land to settlement areas. The land use change is a process, and it is not reflected in this dataset with a short period.
Line 400. Please add the supporting figures or table numbers.
Several figures in the supplementary were not cited in the main text, please check.
Citation: https://doi.org/10.5194/essd-2024-98-RC1 -
AC1: 'Reply on RC1', chansopheaktra sovann, 19 Sep 2024
Manuscript DOI: https://doi.org/10.5194/essd-2024-98
Comment’s citation: https://doi.org/10.5194/essd-2024-98-RC1
Response to Referee’s comments (RC1)
RC1. 'Comment on essd-2024-98', Anonymous Referee #1, 03 Jul 2024 reply
Comment 1: This study provides observed datasets of meteorological conditions, forest inventory, leaf traits of woody species, leaf area index (LAI) and a fraction of photosynthetically active radiation (fPAR) of pristine tropical forests, regrowth forests, and cashew plantations within Phnom Kulen National Park. Nine forest plots were selected and these forest inventory results cover the period from Apr 2022 to Apr 2023. The authors compared the differences in these ecosystem characteristics among the three types of land-cover classes and provided a well-organized dataset of a tropical forest ecosystem. However, despite the hard monitoring work, I am not quite sure whether this dataset is valuable enough to reach the high standard of ESSD.
Response 1. We thank the reviewer for the constructive evaluation of our manuscript and helpful comments. The reviewer raises many good points that have helped to improve the manuscript substantially.
Land use and land cover change is one of the most severe environmental challenges within the Earth system. In the context of tackling current global environmental challenges, field observations are necessary to assess the dynamic responses of ecosystems to changing environmental conditions on fine spatial and temporal scales. Hence, there is a high demand for field observations of ecosystem characteristics. We agree that there are other locations doing field observations of these highly valuable ecosystem characteristics but, 1) there is a great lack of field sites doing observations of ecosystem characteristics in the tropics in general in comparison to temperate and boreal biomes. Here we present the first data of a newly established field site in the tropics where measurements of ecosystem characteristics will be taken continuously from now on. 2) Even though there are other locations doing observations of ecosystem characteristics (see response below), it is very unusual with sites doing such comprehensive observations covering this many different aspects of the ecosystem characteristics. This is not only the case in the tropics but in general. This integrated dataset allows for studies on how different aspects affect the ecosystem functioning. 3) Finally, but most importantly, we are not only doing observations at one typical ecosystem type; but we are studying ecosystem characteristics of ecosystems with variable amounts of human influence. Given the severe anthropogenic pressure on tropical forests, this is an extremely important aspect, as we can then use the nearby pristine evergreen forest locations as a baseline, as how would the other ecosystems would function if not being affected by humans. We can thereby directly quantify the impact of human land use on the measured ecosystem characteristics. All these points have been emphasised in the revised manuscript (Pages 1–3, Lines 10–68; Pages 12–13, Lines 265–277; Page 20, Lines 378–387; Page 27, Lines 582–600).
Major concern
Comment 2. This study compared the species diversity, leaf functional traits, biomass, and other characteristics of three types of forests. I doubt whether this dataset is unique and useful enough. Does the comparison of these indexes among three types of forests can provide us with any new insights that haven’t been investigated before? The authors claimed that this dataset fills data gaps in tropical forest research, so does it mean there were few studies investigating the ecosystem characteristics in tropical forests? I cannot see any new findings according to the data analysis section.
Response 2. Thank you for pointing out that we were not clear on the insights provided by the analysis of the data. Our data show a substantial decrease in several of the ecosystem characteristics in the land-cover classes affected by anthropogenic land cover change. In the revised manuscript, we have made sure to emphasise these most important results (Pages 1–3, Lines 10–68; Pages 12–13, Lines 265–277; Page 27, Lines 582–600).
There are other studies on the characteristics of these types of ecosystems. But especially Southeast Asia, renowned for its biodiversity richness, suffers from a scarcity of integrated datasets that encompass such a broad spectrum of ecosystem characteristics across different land-cover classes. Previous studies by Gael Sola et al. (2014) and Theilade et al. (2022) primarily focused on species composition and aboveground biomass in old-growth forests in Cambodia. Van Do et al. (2020) focused on aboveground biomass and stand structure in undisturbed old-growth forests across various ecoregions in Vietnam, while Zin and Mitlöhner (2020) examined stand structure and species composition in primary and secondary evergreen forests in Myanmar. Furthermore, our study fills critical data gaps in tropical forest research by delivering detailed functional trait data alongside comprehensive abiotic environmental factors such as weather and soil conditions. Functional trait analysis represents an advanced concept crucial for understanding ecosystem functions and services (Díaz et al., 2013). Most existing tropical functional trait data are concentrated in specific regions like French Guiana, neglecting Southeast Asian tropical forests (Kattge et al., 2020). Hence, no previous studies covered such diverse ecosystem variables in the Southeast Asia, and none included the various stages of human intervention, as we do when also including cashew plantations, the dominant agricultural land use in the region of our study site.
Additionally, this paper is not only focusing on the novel insights provided by the analysis, it is also focusing on presenting a novel field site collecting unique data on ecosystem characteristics (see response above). We are certain that these data are going to be used in many future studies. For instance, the functional traits will be used for parameterisation of the dynamic vegetation model, LPJ-GUESS, to improve its predictive accuracy. Our fPAR and LAI data are crucial for remote sensing applications, particularly for modelling gross primary production (GPP) at regional and global scales. There is a great lack of field observations for these variables, especially covering the Sentinel era (after 2016). Hence, we will use these data, in combination with other data collected from the Southeast Asian region, for remote sensing studies on spatiotemporal dynamics in LAI, fPAR and GPP. We are currently making a remote sensing-based land use and land cover change time series since the 2000’s of the region; and will aim at linking these data with the various land use changes to see how ecosystem functioning and services have changed over time across this region.
In conclusion, our study’s dataset is a significant advancement in tropical forest research by offering comprehensive, region-specific data on species diversity, functional traits, biomass, and other ecosystem variables across diverse land covers in Southeast Asia. We are confident that our datasets will provide novel insights into tropical forest dynamics, bridging existing knowledge gaps and supporting both scientific understanding and conservation practices in the region.
Comment 3. Three methods were used to calculate aboveground biomass, and the selection of the method can cause a large difference in the calculated results. How to prove which method is better than the other two as the authors claimed in the discussion 4.5.4?
Response 3. Thank you for making us aware that we were not clear on this point. Our study introduces the AGBh method as an alternative to the widely used AGBf function for estimating aboveground biomass. Unlike AGBf, which relies on generalized assumptions of uniform species composition and stand structure, AGBh incorporates site-specific data, including species-specific wood density and localized diameter at breast height (DBH) and tree height (H) relationships, expecting to enhance accuracy. Our locally adopted AGBh method produced estimates ~ 30 % higher than the generic AGBf for both EF and RF (Table 2, Fig. 4b (Pages 12–13, Lines 265–275; Page 16, Lines 321)). This is likely due to the combined effects of higher mean wood density and a stronger DBH relationship, resulting in a more pronounced exponential growth response in AGB (Fig. 4a, Page 16). Still, these ~ 30 % higher values align with the range reported in previous studies (Tables A2–A3, Page 28–29). In contrast, in the CP case, our AGBh method produced estimates less than a quarter of the generic AGBf method. The reason is that the AGBh method is less reliable when a weak DBH-H relationship is detected as it fails to accurately capture the overall tree size and volume. This is also reflected in the substantially larger uncertainty in the CP AGBh method as indicated by the standardized errors of the parameters within the DBH-H relationship (Table A4, Table S7.1 (Page 29, Line 634; Page 26, line 240)). However, to fully validate the AGB allometric equations either destructive field-observed data or lidar scans of a terrestrial laser scanner would be necessary. We are currently planning a campaign where we are going to use a terrestrial laser scanner for scanning these ecosystems. Hopefully, future research will be able to include these direct field measurements of ecosystem structures to more accurately validate the AGB methods for these land-cover classes. We have clarified these points in the revised discussion (Pages 24–25, Lines 514–528).
Minor
Comment 4. The journal required a data link at the end of the abstract.
Response 4. The data links have been added at the end of the abstract section (Page 1, Lines 22–24).
Comment 5. Please add the background of the land-cover map in Figure 1.
Response 5. The land-cover background map has been added in revised Figure 1 (Page 4, Line 84).
Comment 6. The titles of these tables are too long. Please move the note to the end of the table.
Response 6. We agree. The titles of Table 1 and Table 2 in main text of the paper have been revised to make a short title. Their supplementary information has been relocated to the “Note” section at the end of each table (Page 9, Lines 206–213; Pages 12–13, Lines 265–277).
Comment 7. The authors in Table 1 are not a good representation of reference. Just use “reference”.
Response 7. We agree and it has been corrected. Additionally, the reference format in the column has also been changed to in-text citation format (Page 9, Line 206).
Comment 8. What is the difference in precipitation and temperature data between Figs. 2 and 3? Are they repeated results?
Response 8. Thank you for pointing this out. It is correct that they were the same. We have combined Fig. 2 and Fig. 3 to eliminate redundancy in the revised manuscript (Fig. 2 in revised manuscript, Page 11, Line 238).
Comment 9. In Fig. 4, does the small plot represent the local amplifier of the plot outside? The green line seems different among the two plots.
Response 9. Thank you for pointing this out, this was an error in the script. We have now corrected the zoomed plot so that they are the same (Fig. 3, Page 15, Line 310).
Comment 10. Hard to distinguish the line form of AGBh and AGBf in Figure 5a.
Response 10. Thank you for pointing this out; we have increased the size of the lines to increase visibility (Fig. 4a, Page 16, Line 323).
Comment 11. Line 383-384. I doubt the dataset in this study can be used to investigate the effect of land use change from natural land to settlement areas. The land use change is a process, and it is not reflected in this dataset with a short period.
Response 11. Thank you for pointing this out, we agree that it was an unclear formulation. But, we still think that these data can be used to study impact by land cover change, as these ecosystems are located near each other. We can thereby use the pristine ecosystems as a baseline, to see how the other land-cover classes differ in relation to what would have been there in case humans would not have affected the land. We have rephrased the sentence and clarified our point (Page 20, Lines 381–385)
Comment 12. Line 400. Please add the supporting figures or table numbers.
Response 12. We added a reference to Fig. 6 to support this text in the revised manuscript (Page 21, Line 402).
Comment 13. Several figures in the supplementary were not cited in the main text, please check.
Response 13. Thanks for pointing this out. We now cite all supporting info in the main text.
References
Díaz, S., Purvis, A., Cornelissen, J. H. C., Mace, G. M., Donoghue, M. J., Ewers, R. M., . . . Pearse, W. D. (2013). Functional traits, the phylogeny of function, and ecosystem service vulnerability. Ecology and Evolution, 3(9), 2958-2975. https://doi.org/10.1002/ece3.601
Gael Sola, Samreth Vanna, Lauri Vesa, Mathieu Van Rijnand, & Matieu Henry. (2014). Forest biomass in Cambodia: from field plot to national estimates. http://www.cambodia-redd.org/wp-content/uploads/2016/01/Forest-biomass-in-Cambodia-from-field-plots-to-national-estimates.pdf
Kattge, J., Bönisch, G., Díaz, S., Lavorel, S., Prentice, I. C., Leadley, P., . . . Wirth, C. (2020). TRY plant trait database – enhanced coverage and open access. 26(1), 119-188. https://doi.org/10.1111/gcb.14904
Theilade, I., Phourin, C., Schmidt, L., Meilby, H., van de Bult, M., & Friborg, K. G. (2022). Evergreen forest types of the central plains in Cambodia: floristic composition and ecological characteristics. 2022(8), e03494. https://doi.org/10.1111/njb.03494
Van Do, T., Yamamoto, M., Kozan, O., Hai, V. D., Trung, P. D., Thang, N. T., . . . Hung, B. K. (2020). Ecoregional variations of aboveground biomass and stand structure in evergreen broadleaved forests. Journal of Forestry Research, 31(5), 1713-1722. https://doi.org/10.1007/s11676-019-00969-y
Zin, I. I. S., & Mitlöhner, R. (2020). Species composition and stand structure of primary and secondary moist evergreen forests in the Tanintharyi Nature Reserve (TNR) Buffer Zone, Myanmar. Open Journal of Forestry, 10(4), 445-459. https://doi.org/10.4236/ojf.2020.104028
Citation: https://doi.org/10.5194/essd-2024-98-AC1
-
AC1: 'Reply on RC1', chansopheaktra sovann, 19 Sep 2024
-
RC2: 'Comment on essd-2024-98', Anonymous Referee #2, 27 Jul 2024
Tropical forests play an important role in global terrestrial biodiversity and biogeochemical cycles. It is crucial to collect the ecosystem characteristics of tropical forests after anthropogenic impacts. This manuscript presented observed forest inventory, leaf area index, leaf traits of woody species, a fraction of intercepted photosynthetically active radiation, and edaphic and meteorological conditions in a newly established ecosystem monitoring site of Kulen National Park. The dataset will be helpful in analyzing the forest response to human disturbance in Cambodia. However, the spatial representation of these field observation data is limited for the pantropical forest, and the observation time period is only one year. Thus, I don’t think this manuscript can be published on Earth System Science Data.
Main comments:
- “Land cover” in the title is confusing and unsuitable. The three land cover classes in this study are all forest land, but the tree species management intensities are different.
- The location of weather station is different with that of the nine forest inventory plots, and the observation sites of soil condition and radiation. Thus, I don’t think the meteorological station can represent the microclimate of three land cover classes due to the diverse vegetation cover, tree composition and stand structure.
- How was temperature and precipitation between April 2022 and April 2023 compared with the multi-year average? Was it a dry, wet, or normal year?
- In the discussion section, the authors compared many observed ecosystem characteristics with previous studies. However, most variables (e.g., edaphic conditions, species diversity, SLA, LDMC) area qualitative comparison. Considering the manuscript is a data description paper, the quantitative comparison is important to increase the confidence of the dataset. I suggest adding several tables about ecosystem characteristics comparison in the Appendix or supplementary materials.
Specific comments:
- Line 13: The time period of the dataset should be specified. For example, “…we established an ecosystem monitoring site and conducted about 1-year of field observation in Phnom Kulen…”
- Line 18: “…explore the relationships…” between ecosystem characteristics and AGB. What the ecosystem characteristics refers to and these variables need to be specified clearly.
- Line 60-62: Does the meteorological conditions include both microclimate and soil conditions? I noted that the observed dataset includes edaphic conditions in the abstract.
- Line 77: “…Approximately 60% and 13% ….”. It confused me.
- Line 80: It will be better to show the locations of nine forestry inventory plots or Phnom Kulen National Park in the whole Southeast Asia in Figure 1.
- Line 236-240: The decimal places of surface temperature should be consistent with the values in Table 2.
- Line 241-246: The subplots (P, Tair, and Rg) of Figure 6 are same as those in Figure 5.
- Line 274: The units of AGBf and AGBh in Table 2 are not consistent with those in the caption of Table 1.
- Line 275-287: The decimal places of SLA, LDMC and Chl should be consistent with the values in Table S5.1-5.4.
- Line 290: The mean and SD of DBH and H should be labeled in Figure S6.1.
- Line 306-309: The p-value of the fitted power-law relationship should be labeled in Figure 4.
- Line 314-316: Why the estimated AGB of this study was substantially lower than the other two methods?
- Line 322-326: The unit of y-axis in Figure 5a should be kg tree-1.
- Line 453: Why the Chl of CP in Kulen is higher by 2-2.5 times than that in India? It should be explained.
Citation: https://doi.org/10.5194/essd-2024-98-RC2 -
AC2: 'Reply on RC2', chansopheaktra sovann, 19 Sep 2024
Manuscript DOI: https://doi.org/10.5194/essd-2024-98
Comment’s citation: https://doi.org/10.5194/essd-2024-98-RC2
Response to Referee’s comments (RC2)
RC1: 'Comment on essd-2024-98', Anonymous Referee #2, 27 Jul 2024 reply
Comment 1: Tropical forests play an important role in global terrestrial biodiversity and biogeochemical cycles. It is crucial to collect the ecosystem characteristics of tropical forests after anthropogenic impacts. This manuscript presented observed forest inventory, leaf area index, leaf traits of woody species, a fraction of intercepted photosynthetically active radiation, and edaphic and meteorological conditions in a newly established ecosystem monitoring site of Kulen National Park. The dataset will be helpful in analyzing the forest response to human disturbance in Cambodia. However, the spatial representation of these field observation data is limited for the pantropical forest, and the observation time period is only one year. Thus, I don’t think this manuscript can be published on Earth System Science Data.
Response 1: We thank the reviewer for the constructive evaluation of our manuscript and helpful comments. The reviewer raises many good points that have helped to improve the manuscript substantially. It is absolutely correct that the spatial representation of this field site is very limited in a pantropical perspective. It is almost impossible for one research group to perform such a detailed field inventory of ecosystem characteristics across the tropical forest region. For pantropical studies, it is necessary to combine these data with other data sets, such as from the FLUXNET, ICOS, SpecNet, TRY, GBOV, etc.
In the context of tackling the current global environmental challenges, field observations are essential for understanding ecosystem responses to environmental changes at fine spatial and temporal scales. Given the intense anthropogenic pressures on tropical forests and the urgent need for comprehensive field data, our dataset provides valuable insights into the impacts of deforestation on biodiversity, carbon sequestration, and hydrological cycles. These fine-scale observations reveal key ecosystem processes, that can be used to enhance remote sensing products and dynamic vegetation models, and when applied on the pantropical scale support comprehensive global environmental monitoring. This importance emphasises the suitability of our dataset for publication in ESSD, as it supports more effective monitoring of responses to global environmental issues. We have made sure to emphasise these points in the revised manuscript (Pages 2–3, Lines 45–68).
It is correct that the length of the time series is rather short at the moment. This is a newly established field site, and as such this is the first data that is presented. However, we have also made sure to emphasise in the revised manuscript that this is a newly established site and that more data will come and will be uploaded regularly to https://zenodo.org/communities/cambodia_ecosystem_data/. Here, we have already included an additional year of weather data (Page 1, Lines 13–15, 24; Page 2, Lines 61–63, Pages 27, Lines 586–589).
Main comments
Comment 2: “Land cover” in the title is confusing and unsuitable. The three land cover classes in this study are all forest land, but the tree species management intensities are different.
Response 2: Thank you for pointing this out, we agree and have revised title to: “Characteristics of ecosystems under various anthropogenic impacts in a tropical forest region of Southeast Asia” (Page 1, Lines 1–2).
In our main text, we chose to retain the use of "land cover" to describe pristine tropical forests, natural regrowth forests, and cashew plantations, reflecting different levels of human impact. This terminology aligns with the Food and Agriculture Organization’s (FAO) definition of "land cover" as the physical and biological cover of the Earth's surface, including both vegetation and man-made features (FAO, 1997; FAO/UNEP, 1999). According to the REDD+ Program (Ministry of Environment (Cambodia), 2018), pristine tropical forests and natural regrowth forests are classified as forest land, whereas cashew plantations, which were originally natural forests, are categorized as cropland.
Comment 3: The location of the weather station is different from that of the nine forest inventory plots, and the observation sites of soil condition and radiation. Thus, I don’t think the meteorological station can represent the microclimate of three land cover classes due to the diverse vegetation cover, tree composition and stand structure.
Response 3: Thank you for pointing out that we were not clear on this point. We used the weather station data to represent the general meteorological conditions of the wider area (the Kulen National Park). The station was centrally placed, with distances to the nearest and farthest forest inventory plots ranging from 3 km to 6 km (RF1 to RF3, Fig. 1, Page 4). When measuring rainfall, incoming radiation, wind speed, and air temperature at 2 m height, it is necessary to place the station in an area without trees and other disturbing elements, and we believe that these measurements are representative for the general meteorological conditions at the Kulen plateau, across which the observed ecosystems are spread. We have made sure to clarify in the revised manuscript that the meteorological data is not ecosystem specific, but for the wider area (Page 3, Lines65–66; Pages5–6, Lines 122–129; Pages 10–12 in the Supplementary material).
Comment 4: How were temperature and precipitation between April 2022 and April 2023 compared with the multi-year average? Was it a dry, wet, or normal year?
Response 4: Thank you for pointing this out. This is the first established meteorological station at the Kulen plateau. Hence, we lack long-term historical weather data for a direct comparison. The recorded data indicate that the Kulen plateau experienced significantly higher rainfall than the nearby lowlands of Banteay Srei (22 km west) and Siem Reap City (40 km southwest). This information has been added to the revised manuscript (Page 1, Lines 13–15; Page 10, Lines 223–225; Page 20, Lines 385–387).
Comment 5: In the discussion section, the authors compared many observed ecosystem characteristics with previous studies. However, most variables (e.g., edaphic conditions, species diversity, SLA, LDMC) are qualitative comparisons. Considering the manuscript is a data description paper, the quantitative comparison is important to increase the confidence of the dataset. I suggest adding several tables about ecosystem characteristics comparison in the Appendix or supplementary materials.
Response 5: Thank you for pointing this out. We agree and have added a comparison table of species richness, Shannon index and LAI in the supplementary material (Table S10.1–S10.3; Pages 36–38).
Specific comments:
Comment 6. Line 13: The time period of the dataset should be specified. For example, “…we established an ecosystem monitoring site and conducted about 1-year of field observation in Phnom Kulen…”
Response 6. Thank you for pointing out that we were not clear enough on this point. Since this is not just a one year study site, we have instead clarified that this is the first data coming out of the site (Page 1, Lines 13–15, Lines 24).
Comment 7. Line 18: “…explore the relationships…” between ecosystem characteristics and AGB. What the ecosystem characteristics refer to and these variables need to be specified clearly.
Response 7. Thank you. We deleted the sentence and revised the abstract (Page 1, Lines 10–24).
Comment 8. Line 60-62: Does the meteorological conditions include both microclimate and soil conditions? I noted that the observed dataset includes edaphic conditions in the abstract.
Response 8. Thank you for pointing out that we were not clear on this point. The meteorological conditions refer to general conditions in the wider study area (see response above), whereas the soil conditions were collected at six of the ecosystems. This has been clarified in the revised manuscript (Page 2–3, Lines 63–66).
Comment 9. Line 77: “…Approximately 60% and 13% ….”. It confused me.
Response 9. We agree that this was confusing. We changed the sentence to “Approximately 60% of Kulen is covered by cashew plantations, another 13% consists of forestland, while the remainder comprises other land types (Singh et al., 2019)” (Page 3, Lines 81–82).
Comment 10. Line 80: It will be better to show the locations of nine forestry inventory plots or Phnom Kulen National Park in the whole Southeast Asia in Figure 1.
Response 10. We updated the overview map in Fig. 1 to show the location of the study area in Southeast Asia accordingly (Page 4).
Comment 11. Line 236-240: The decimal places of surface temperature should be consistent with the values in Table 2.
Response 11. Thank you for noticing this. We have adjusted the text to ensure consistency with the decimal places of surface temperature values in Table 2 (Pages 10, Lines 230–232).
Comment 12. Line 241-246: The subplots (P, Tair, and Rg) of Figure 6 are same as those in Figure 5.
Response 12. Thank you for pointing this out. We assume it was Fig. 2 and 3 that was referred to. We have combined these figures to eliminate redundancy in the revised manuscript (Fig 2 of the revised manuscript, Page 11).
Comment 13. Line 274: The units of AGBf and AGBh in Table 2 are not consistent with those in the caption of Table 1.
Response 13. Thank you for pointing out that this was not clear. We use Mg ha⁻¹ for AGBf and AGBh in Table 2 for biomass per hectare, and kg tree⁻¹ in Table 1 for stand-level biomass estimation, each unit is standard in its respective context. We have revised Section “2.3.3 Stand structural attributes” to clarify this point (Pages 9, Lines 204–205).
Comment 14. Line 275-287: The decimal places of SLA, LDMC and Chl should be consistent with the values in Table S5.1-5.4.
Response 14. Thank you for noticing this. Values of SLA, LDMC, and Chl in Section 3.3 and Table 2 of the revised manuscript (Page 14, Lines 279–290; Page 12–13) have been rounded to two decimal places for consistency with Table S6.1–6.4 (Page 20–24 in the supplementary material).
Comment 15. Line 290: The mean and SD of DBH and H should be labeled in Figure S6.1.
Response 15. They have been labelled in the revised Figure S7.1 (Page 25 in the supplementary material).
Comment 16. Line 306-309: The p-value of the fitted power-law relationship should be labeled in Figure 4.
Response 16. We agree and added the p-value for the fitted power-law relationship in the revised Figure 3 (Page 15, Line 309).
Comment 17. Line 314-316: Why the estimated AGB of this study was substantially lower than the other two methods?
Response 17. Thank you for pointing this out. Our study introduces the AGBh method as an alternative to the widely used AGBf function for estimating aboveground biomass. Unlike AGBf, which relies on generalized assumptions of uniform species composition and stand structure, AGBh incorporates site-specific data, including species-specific wood density and localized diameter at breast height (DBH) and tree height (H) relationships, expecting to enhance accuracy. Our locally adopted AGBh method produced estimates ~ 30 % higher than the generic AGBf for both EF and RF (Table 2, Fig. 4b (Pages 12–13, Lines 265–277; Page 16, Lines 323)). This is likely due to the combined effects of higher mean wood density and a stronger DBH relationship, resulting in a more pronounced exponential growth response in AGB (Fig. 4a, Page 16). Still, these ~ 30 % higher values align with the range reported in previous studies (Tables A2–A3, Page 28–29). In contrast, in the CP case, our AGBh method produced estimates less than a quarter of the generic AGBf method. The reason is that the AGBh method is less reliable when a weak DBH-H relationship is detected because it fails to accurately capture the overall tree size and volume. This is also reflected in the substantially larger uncertainty in the CP AGBh method as indicated by the standardized errors of the parameters within the DBH-H relationship (Table A4; Table S7.1 (Page 29–30, Line 635; Page 26 in the supplementary material)). However, to fully validate the AGB allometric equations destructive field-observed data or lidar scans of the ecosystem would be necessary. We are currently planning a campaign where we are going to use a terrestrial laser scanner for scanning these ecosystems. Hopefully, future research will be able to include these direct field measurements of ecosystem structures to more accurately validate these AGB methods for these land cover types. We have clarified these points in the revised discussion (Pages 24–25, Lines 514–528).
Comment 18. Line 322–326: The unit of y-axis in Figure 5a should be kg tree-1.
Response 18. Thank you for this suggestion; we agree and have changed it in the revised manuscript (Fig. 4a, Page 16, Line 323).
Comment 19. Line 453: Why the Chl of CP in Kulen is higher by 2-2.5 times than that in India? It should be explained.
Response 19. The difference between our result and those reported in India is attributed to methodological variations. We used a Chlorophyll Meter (SPAD 502 Plus; Konica Minolta Sensing Inc., Japan) for direct field measurements, whereas the Indian study utilized a laboratory extraction method with dimethyl sulfoxide (DMSO) and acetone, followed by absorbance analysis at multiple wavelengths (645 nm, 663 nm, 652 nm, and 470 nm) (Mog & Nayak, 2018). We deleted this sentence to avoid confusion.
References
Ministry of Environment (Cambodia). (2018). Cambodia forest cover 2016. Phnom Penh, Cambodia: Mininistry of Environment
Mog, B., & Nayak, M. G. (2018). Leaf Morphological and Physiological Traits and Their Significance in Yield Improvement of Fifteen Cashew Varieties in West Coast Region of Karnataka. International Journal of Current Microbiology and Applied Sciences, 7(07), 1455-1469. https://doi.org/10.20546/ijcmas.2018.707.173
Singh, M., Evans, D., Chevance, J.-B., Tan, B. S., Wiggins, N., Kong, L., & Sakhoeun, S. (2019). Evaluating remote sensing datasets and machine learning algorithms for mapping plantations and successional forests in Phnom Kulen National Park of Cambodia. PeerJ, 7, e7841. https://doi.org/10.7717/peerj.7841
FAO, 1997: State of the World's Forests. Food and Agriculture Organization, Rome, Italy, 200 pp.
FAO/UNEP, 1999: Terminology for Integrated Resources Planning and Management. Food and Agriculture Organization/United Nations Environmental Programme, Rome, Italy and Nairobi, Kenya.
Citation: https://doi.org/10.5194/essd-2024-98-AC2
Status: closed
-
RC1: 'Comment on essd-2024-98', Anonymous Referee #1, 03 Jul 2024
This study provides observed datasets of meteorological conditions, forest inventory, leaf traits of woody species, leaf area index (LAI) and a fraction of photosynthetically active radiation (fPAR) of pristine tropical forests, regrowth forests, and cashew plantations within Phnom Kulen National Park. Nine forest plots were selected and these forest inventory results cover the period from Apr 2022 to Apr 2023. The authors compared the differences in these ecosystem characteristics among the three types of land-cover classes and provided a well-organized dataset of a tropical forest ecosystem. However, despite the hard monitoring work, I am not quite sure whether this dataset is valuable enough to reach the high standard of ESSD.
Major concern
This study compared the species diversity, leaf functional traits, biomass, and other characteristics of three types of forests. I doubt whether this dataset is unique and useful enough. Does the comparison of these indexes among three types of forests can provide us with any new insights that haven’t been investigated before? The authors claimed that this dataset fills data gaps in tropical forest research, so does it mean there were few studies investigating the ecosystem characteristics in tropical forests? I cannot see any new findings according to the data analysis section.
Three methods were used to calculate aboveground biomass, and the selection of the method can cause a large difference in the calculated results. How to prove which method is better than the other two as the authors claimed in the discussion 4.5.4?
Minor
The journal required a data link at the end of the abstract.
Please add the background of land cover map in Figure 1.
The titles of these tables are too long. Please move the note to the end of the table.
Authors in Table 1 is not a good representation of reference. Just use “reference”.
What is the difference in precipitation and temperature data between Figs. 2 and 3? Are they repeated results?
In Fig. 4, does the small plot represent the local amplifier of the plot outside? The green line seems different among the two plots.
Hard to distinguish the line form of AGBh and AGBf in Figure 5a.
Line 383-384. I doubt the dataset in this study can be used to investigate the effect of land use change from natural land to settlement areas. The land use change is a process, and it is not reflected in this dataset with a short period.
Line 400. Please add the supporting figures or table numbers.
Several figures in the supplementary were not cited in the main text, please check.
Citation: https://doi.org/10.5194/essd-2024-98-RC1 -
AC1: 'Reply on RC1', chansopheaktra sovann, 19 Sep 2024
Manuscript DOI: https://doi.org/10.5194/essd-2024-98
Comment’s citation: https://doi.org/10.5194/essd-2024-98-RC1
Response to Referee’s comments (RC1)
RC1. 'Comment on essd-2024-98', Anonymous Referee #1, 03 Jul 2024 reply
Comment 1: This study provides observed datasets of meteorological conditions, forest inventory, leaf traits of woody species, leaf area index (LAI) and a fraction of photosynthetically active radiation (fPAR) of pristine tropical forests, regrowth forests, and cashew plantations within Phnom Kulen National Park. Nine forest plots were selected and these forest inventory results cover the period from Apr 2022 to Apr 2023. The authors compared the differences in these ecosystem characteristics among the three types of land-cover classes and provided a well-organized dataset of a tropical forest ecosystem. However, despite the hard monitoring work, I am not quite sure whether this dataset is valuable enough to reach the high standard of ESSD.
Response 1. We thank the reviewer for the constructive evaluation of our manuscript and helpful comments. The reviewer raises many good points that have helped to improve the manuscript substantially.
Land use and land cover change is one of the most severe environmental challenges within the Earth system. In the context of tackling current global environmental challenges, field observations are necessary to assess the dynamic responses of ecosystems to changing environmental conditions on fine spatial and temporal scales. Hence, there is a high demand for field observations of ecosystem characteristics. We agree that there are other locations doing field observations of these highly valuable ecosystem characteristics but, 1) there is a great lack of field sites doing observations of ecosystem characteristics in the tropics in general in comparison to temperate and boreal biomes. Here we present the first data of a newly established field site in the tropics where measurements of ecosystem characteristics will be taken continuously from now on. 2) Even though there are other locations doing observations of ecosystem characteristics (see response below), it is very unusual with sites doing such comprehensive observations covering this many different aspects of the ecosystem characteristics. This is not only the case in the tropics but in general. This integrated dataset allows for studies on how different aspects affect the ecosystem functioning. 3) Finally, but most importantly, we are not only doing observations at one typical ecosystem type; but we are studying ecosystem characteristics of ecosystems with variable amounts of human influence. Given the severe anthropogenic pressure on tropical forests, this is an extremely important aspect, as we can then use the nearby pristine evergreen forest locations as a baseline, as how would the other ecosystems would function if not being affected by humans. We can thereby directly quantify the impact of human land use on the measured ecosystem characteristics. All these points have been emphasised in the revised manuscript (Pages 1–3, Lines 10–68; Pages 12–13, Lines 265–277; Page 20, Lines 378–387; Page 27, Lines 582–600).
Major concern
Comment 2. This study compared the species diversity, leaf functional traits, biomass, and other characteristics of three types of forests. I doubt whether this dataset is unique and useful enough. Does the comparison of these indexes among three types of forests can provide us with any new insights that haven’t been investigated before? The authors claimed that this dataset fills data gaps in tropical forest research, so does it mean there were few studies investigating the ecosystem characteristics in tropical forests? I cannot see any new findings according to the data analysis section.
Response 2. Thank you for pointing out that we were not clear on the insights provided by the analysis of the data. Our data show a substantial decrease in several of the ecosystem characteristics in the land-cover classes affected by anthropogenic land cover change. In the revised manuscript, we have made sure to emphasise these most important results (Pages 1–3, Lines 10–68; Pages 12–13, Lines 265–277; Page 27, Lines 582–600).
There are other studies on the characteristics of these types of ecosystems. But especially Southeast Asia, renowned for its biodiversity richness, suffers from a scarcity of integrated datasets that encompass such a broad spectrum of ecosystem characteristics across different land-cover classes. Previous studies by Gael Sola et al. (2014) and Theilade et al. (2022) primarily focused on species composition and aboveground biomass in old-growth forests in Cambodia. Van Do et al. (2020) focused on aboveground biomass and stand structure in undisturbed old-growth forests across various ecoregions in Vietnam, while Zin and Mitlöhner (2020) examined stand structure and species composition in primary and secondary evergreen forests in Myanmar. Furthermore, our study fills critical data gaps in tropical forest research by delivering detailed functional trait data alongside comprehensive abiotic environmental factors such as weather and soil conditions. Functional trait analysis represents an advanced concept crucial for understanding ecosystem functions and services (Díaz et al., 2013). Most existing tropical functional trait data are concentrated in specific regions like French Guiana, neglecting Southeast Asian tropical forests (Kattge et al., 2020). Hence, no previous studies covered such diverse ecosystem variables in the Southeast Asia, and none included the various stages of human intervention, as we do when also including cashew plantations, the dominant agricultural land use in the region of our study site.
Additionally, this paper is not only focusing on the novel insights provided by the analysis, it is also focusing on presenting a novel field site collecting unique data on ecosystem characteristics (see response above). We are certain that these data are going to be used in many future studies. For instance, the functional traits will be used for parameterisation of the dynamic vegetation model, LPJ-GUESS, to improve its predictive accuracy. Our fPAR and LAI data are crucial for remote sensing applications, particularly for modelling gross primary production (GPP) at regional and global scales. There is a great lack of field observations for these variables, especially covering the Sentinel era (after 2016). Hence, we will use these data, in combination with other data collected from the Southeast Asian region, for remote sensing studies on spatiotemporal dynamics in LAI, fPAR and GPP. We are currently making a remote sensing-based land use and land cover change time series since the 2000’s of the region; and will aim at linking these data with the various land use changes to see how ecosystem functioning and services have changed over time across this region.
In conclusion, our study’s dataset is a significant advancement in tropical forest research by offering comprehensive, region-specific data on species diversity, functional traits, biomass, and other ecosystem variables across diverse land covers in Southeast Asia. We are confident that our datasets will provide novel insights into tropical forest dynamics, bridging existing knowledge gaps and supporting both scientific understanding and conservation practices in the region.
Comment 3. Three methods were used to calculate aboveground biomass, and the selection of the method can cause a large difference in the calculated results. How to prove which method is better than the other two as the authors claimed in the discussion 4.5.4?
Response 3. Thank you for making us aware that we were not clear on this point. Our study introduces the AGBh method as an alternative to the widely used AGBf function for estimating aboveground biomass. Unlike AGBf, which relies on generalized assumptions of uniform species composition and stand structure, AGBh incorporates site-specific data, including species-specific wood density and localized diameter at breast height (DBH) and tree height (H) relationships, expecting to enhance accuracy. Our locally adopted AGBh method produced estimates ~ 30 % higher than the generic AGBf for both EF and RF (Table 2, Fig. 4b (Pages 12–13, Lines 265–275; Page 16, Lines 321)). This is likely due to the combined effects of higher mean wood density and a stronger DBH relationship, resulting in a more pronounced exponential growth response in AGB (Fig. 4a, Page 16). Still, these ~ 30 % higher values align with the range reported in previous studies (Tables A2–A3, Page 28–29). In contrast, in the CP case, our AGBh method produced estimates less than a quarter of the generic AGBf method. The reason is that the AGBh method is less reliable when a weak DBH-H relationship is detected as it fails to accurately capture the overall tree size and volume. This is also reflected in the substantially larger uncertainty in the CP AGBh method as indicated by the standardized errors of the parameters within the DBH-H relationship (Table A4, Table S7.1 (Page 29, Line 634; Page 26, line 240)). However, to fully validate the AGB allometric equations either destructive field-observed data or lidar scans of a terrestrial laser scanner would be necessary. We are currently planning a campaign where we are going to use a terrestrial laser scanner for scanning these ecosystems. Hopefully, future research will be able to include these direct field measurements of ecosystem structures to more accurately validate the AGB methods for these land-cover classes. We have clarified these points in the revised discussion (Pages 24–25, Lines 514–528).
Minor
Comment 4. The journal required a data link at the end of the abstract.
Response 4. The data links have been added at the end of the abstract section (Page 1, Lines 22–24).
Comment 5. Please add the background of the land-cover map in Figure 1.
Response 5. The land-cover background map has been added in revised Figure 1 (Page 4, Line 84).
Comment 6. The titles of these tables are too long. Please move the note to the end of the table.
Response 6. We agree. The titles of Table 1 and Table 2 in main text of the paper have been revised to make a short title. Their supplementary information has been relocated to the “Note” section at the end of each table (Page 9, Lines 206–213; Pages 12–13, Lines 265–277).
Comment 7. The authors in Table 1 are not a good representation of reference. Just use “reference”.
Response 7. We agree and it has been corrected. Additionally, the reference format in the column has also been changed to in-text citation format (Page 9, Line 206).
Comment 8. What is the difference in precipitation and temperature data between Figs. 2 and 3? Are they repeated results?
Response 8. Thank you for pointing this out. It is correct that they were the same. We have combined Fig. 2 and Fig. 3 to eliminate redundancy in the revised manuscript (Fig. 2 in revised manuscript, Page 11, Line 238).
Comment 9. In Fig. 4, does the small plot represent the local amplifier of the plot outside? The green line seems different among the two plots.
Response 9. Thank you for pointing this out, this was an error in the script. We have now corrected the zoomed plot so that they are the same (Fig. 3, Page 15, Line 310).
Comment 10. Hard to distinguish the line form of AGBh and AGBf in Figure 5a.
Response 10. Thank you for pointing this out; we have increased the size of the lines to increase visibility (Fig. 4a, Page 16, Line 323).
Comment 11. Line 383-384. I doubt the dataset in this study can be used to investigate the effect of land use change from natural land to settlement areas. The land use change is a process, and it is not reflected in this dataset with a short period.
Response 11. Thank you for pointing this out, we agree that it was an unclear formulation. But, we still think that these data can be used to study impact by land cover change, as these ecosystems are located near each other. We can thereby use the pristine ecosystems as a baseline, to see how the other land-cover classes differ in relation to what would have been there in case humans would not have affected the land. We have rephrased the sentence and clarified our point (Page 20, Lines 381–385)
Comment 12. Line 400. Please add the supporting figures or table numbers.
Response 12. We added a reference to Fig. 6 to support this text in the revised manuscript (Page 21, Line 402).
Comment 13. Several figures in the supplementary were not cited in the main text, please check.
Response 13. Thanks for pointing this out. We now cite all supporting info in the main text.
References
Díaz, S., Purvis, A., Cornelissen, J. H. C., Mace, G. M., Donoghue, M. J., Ewers, R. M., . . . Pearse, W. D. (2013). Functional traits, the phylogeny of function, and ecosystem service vulnerability. Ecology and Evolution, 3(9), 2958-2975. https://doi.org/10.1002/ece3.601
Gael Sola, Samreth Vanna, Lauri Vesa, Mathieu Van Rijnand, & Matieu Henry. (2014). Forest biomass in Cambodia: from field plot to national estimates. http://www.cambodia-redd.org/wp-content/uploads/2016/01/Forest-biomass-in-Cambodia-from-field-plots-to-national-estimates.pdf
Kattge, J., Bönisch, G., Díaz, S., Lavorel, S., Prentice, I. C., Leadley, P., . . . Wirth, C. (2020). TRY plant trait database – enhanced coverage and open access. 26(1), 119-188. https://doi.org/10.1111/gcb.14904
Theilade, I., Phourin, C., Schmidt, L., Meilby, H., van de Bult, M., & Friborg, K. G. (2022). Evergreen forest types of the central plains in Cambodia: floristic composition and ecological characteristics. 2022(8), e03494. https://doi.org/10.1111/njb.03494
Van Do, T., Yamamoto, M., Kozan, O., Hai, V. D., Trung, P. D., Thang, N. T., . . . Hung, B. K. (2020). Ecoregional variations of aboveground biomass and stand structure in evergreen broadleaved forests. Journal of Forestry Research, 31(5), 1713-1722. https://doi.org/10.1007/s11676-019-00969-y
Zin, I. I. S., & Mitlöhner, R. (2020). Species composition and stand structure of primary and secondary moist evergreen forests in the Tanintharyi Nature Reserve (TNR) Buffer Zone, Myanmar. Open Journal of Forestry, 10(4), 445-459. https://doi.org/10.4236/ojf.2020.104028
Citation: https://doi.org/10.5194/essd-2024-98-AC1
-
AC1: 'Reply on RC1', chansopheaktra sovann, 19 Sep 2024
-
RC2: 'Comment on essd-2024-98', Anonymous Referee #2, 27 Jul 2024
Tropical forests play an important role in global terrestrial biodiversity and biogeochemical cycles. It is crucial to collect the ecosystem characteristics of tropical forests after anthropogenic impacts. This manuscript presented observed forest inventory, leaf area index, leaf traits of woody species, a fraction of intercepted photosynthetically active radiation, and edaphic and meteorological conditions in a newly established ecosystem monitoring site of Kulen National Park. The dataset will be helpful in analyzing the forest response to human disturbance in Cambodia. However, the spatial representation of these field observation data is limited for the pantropical forest, and the observation time period is only one year. Thus, I don’t think this manuscript can be published on Earth System Science Data.
Main comments:
- “Land cover” in the title is confusing and unsuitable. The three land cover classes in this study are all forest land, but the tree species management intensities are different.
- The location of weather station is different with that of the nine forest inventory plots, and the observation sites of soil condition and radiation. Thus, I don’t think the meteorological station can represent the microclimate of three land cover classes due to the diverse vegetation cover, tree composition and stand structure.
- How was temperature and precipitation between April 2022 and April 2023 compared with the multi-year average? Was it a dry, wet, or normal year?
- In the discussion section, the authors compared many observed ecosystem characteristics with previous studies. However, most variables (e.g., edaphic conditions, species diversity, SLA, LDMC) area qualitative comparison. Considering the manuscript is a data description paper, the quantitative comparison is important to increase the confidence of the dataset. I suggest adding several tables about ecosystem characteristics comparison in the Appendix or supplementary materials.
Specific comments:
- Line 13: The time period of the dataset should be specified. For example, “…we established an ecosystem monitoring site and conducted about 1-year of field observation in Phnom Kulen…”
- Line 18: “…explore the relationships…” between ecosystem characteristics and AGB. What the ecosystem characteristics refers to and these variables need to be specified clearly.
- Line 60-62: Does the meteorological conditions include both microclimate and soil conditions? I noted that the observed dataset includes edaphic conditions in the abstract.
- Line 77: “…Approximately 60% and 13% ….”. It confused me.
- Line 80: It will be better to show the locations of nine forestry inventory plots or Phnom Kulen National Park in the whole Southeast Asia in Figure 1.
- Line 236-240: The decimal places of surface temperature should be consistent with the values in Table 2.
- Line 241-246: The subplots (P, Tair, and Rg) of Figure 6 are same as those in Figure 5.
- Line 274: The units of AGBf and AGBh in Table 2 are not consistent with those in the caption of Table 1.
- Line 275-287: The decimal places of SLA, LDMC and Chl should be consistent with the values in Table S5.1-5.4.
- Line 290: The mean and SD of DBH and H should be labeled in Figure S6.1.
- Line 306-309: The p-value of the fitted power-law relationship should be labeled in Figure 4.
- Line 314-316: Why the estimated AGB of this study was substantially lower than the other two methods?
- Line 322-326: The unit of y-axis in Figure 5a should be kg tree-1.
- Line 453: Why the Chl of CP in Kulen is higher by 2-2.5 times than that in India? It should be explained.
Citation: https://doi.org/10.5194/essd-2024-98-RC2 -
AC2: 'Reply on RC2', chansopheaktra sovann, 19 Sep 2024
Manuscript DOI: https://doi.org/10.5194/essd-2024-98
Comment’s citation: https://doi.org/10.5194/essd-2024-98-RC2
Response to Referee’s comments (RC2)
RC1: 'Comment on essd-2024-98', Anonymous Referee #2, 27 Jul 2024 reply
Comment 1: Tropical forests play an important role in global terrestrial biodiversity and biogeochemical cycles. It is crucial to collect the ecosystem characteristics of tropical forests after anthropogenic impacts. This manuscript presented observed forest inventory, leaf area index, leaf traits of woody species, a fraction of intercepted photosynthetically active radiation, and edaphic and meteorological conditions in a newly established ecosystem monitoring site of Kulen National Park. The dataset will be helpful in analyzing the forest response to human disturbance in Cambodia. However, the spatial representation of these field observation data is limited for the pantropical forest, and the observation time period is only one year. Thus, I don’t think this manuscript can be published on Earth System Science Data.
Response 1: We thank the reviewer for the constructive evaluation of our manuscript and helpful comments. The reviewer raises many good points that have helped to improve the manuscript substantially. It is absolutely correct that the spatial representation of this field site is very limited in a pantropical perspective. It is almost impossible for one research group to perform such a detailed field inventory of ecosystem characteristics across the tropical forest region. For pantropical studies, it is necessary to combine these data with other data sets, such as from the FLUXNET, ICOS, SpecNet, TRY, GBOV, etc.
In the context of tackling the current global environmental challenges, field observations are essential for understanding ecosystem responses to environmental changes at fine spatial and temporal scales. Given the intense anthropogenic pressures on tropical forests and the urgent need for comprehensive field data, our dataset provides valuable insights into the impacts of deforestation on biodiversity, carbon sequestration, and hydrological cycles. These fine-scale observations reveal key ecosystem processes, that can be used to enhance remote sensing products and dynamic vegetation models, and when applied on the pantropical scale support comprehensive global environmental monitoring. This importance emphasises the suitability of our dataset for publication in ESSD, as it supports more effective monitoring of responses to global environmental issues. We have made sure to emphasise these points in the revised manuscript (Pages 2–3, Lines 45–68).
It is correct that the length of the time series is rather short at the moment. This is a newly established field site, and as such this is the first data that is presented. However, we have also made sure to emphasise in the revised manuscript that this is a newly established site and that more data will come and will be uploaded regularly to https://zenodo.org/communities/cambodia_ecosystem_data/. Here, we have already included an additional year of weather data (Page 1, Lines 13–15, 24; Page 2, Lines 61–63, Pages 27, Lines 586–589).
Main comments
Comment 2: “Land cover” in the title is confusing and unsuitable. The three land cover classes in this study are all forest land, but the tree species management intensities are different.
Response 2: Thank you for pointing this out, we agree and have revised title to: “Characteristics of ecosystems under various anthropogenic impacts in a tropical forest region of Southeast Asia” (Page 1, Lines 1–2).
In our main text, we chose to retain the use of "land cover" to describe pristine tropical forests, natural regrowth forests, and cashew plantations, reflecting different levels of human impact. This terminology aligns with the Food and Agriculture Organization’s (FAO) definition of "land cover" as the physical and biological cover of the Earth's surface, including both vegetation and man-made features (FAO, 1997; FAO/UNEP, 1999). According to the REDD+ Program (Ministry of Environment (Cambodia), 2018), pristine tropical forests and natural regrowth forests are classified as forest land, whereas cashew plantations, which were originally natural forests, are categorized as cropland.
Comment 3: The location of the weather station is different from that of the nine forest inventory plots, and the observation sites of soil condition and radiation. Thus, I don’t think the meteorological station can represent the microclimate of three land cover classes due to the diverse vegetation cover, tree composition and stand structure.
Response 3: Thank you for pointing out that we were not clear on this point. We used the weather station data to represent the general meteorological conditions of the wider area (the Kulen National Park). The station was centrally placed, with distances to the nearest and farthest forest inventory plots ranging from 3 km to 6 km (RF1 to RF3, Fig. 1, Page 4). When measuring rainfall, incoming radiation, wind speed, and air temperature at 2 m height, it is necessary to place the station in an area without trees and other disturbing elements, and we believe that these measurements are representative for the general meteorological conditions at the Kulen plateau, across which the observed ecosystems are spread. We have made sure to clarify in the revised manuscript that the meteorological data is not ecosystem specific, but for the wider area (Page 3, Lines65–66; Pages5–6, Lines 122–129; Pages 10–12 in the Supplementary material).
Comment 4: How were temperature and precipitation between April 2022 and April 2023 compared with the multi-year average? Was it a dry, wet, or normal year?
Response 4: Thank you for pointing this out. This is the first established meteorological station at the Kulen plateau. Hence, we lack long-term historical weather data for a direct comparison. The recorded data indicate that the Kulen plateau experienced significantly higher rainfall than the nearby lowlands of Banteay Srei (22 km west) and Siem Reap City (40 km southwest). This information has been added to the revised manuscript (Page 1, Lines 13–15; Page 10, Lines 223–225; Page 20, Lines 385–387).
Comment 5: In the discussion section, the authors compared many observed ecosystem characteristics with previous studies. However, most variables (e.g., edaphic conditions, species diversity, SLA, LDMC) are qualitative comparisons. Considering the manuscript is a data description paper, the quantitative comparison is important to increase the confidence of the dataset. I suggest adding several tables about ecosystem characteristics comparison in the Appendix or supplementary materials.
Response 5: Thank you for pointing this out. We agree and have added a comparison table of species richness, Shannon index and LAI in the supplementary material (Table S10.1–S10.3; Pages 36–38).
Specific comments:
Comment 6. Line 13: The time period of the dataset should be specified. For example, “…we established an ecosystem monitoring site and conducted about 1-year of field observation in Phnom Kulen…”
Response 6. Thank you for pointing out that we were not clear enough on this point. Since this is not just a one year study site, we have instead clarified that this is the first data coming out of the site (Page 1, Lines 13–15, Lines 24).
Comment 7. Line 18: “…explore the relationships…” between ecosystem characteristics and AGB. What the ecosystem characteristics refer to and these variables need to be specified clearly.
Response 7. Thank you. We deleted the sentence and revised the abstract (Page 1, Lines 10–24).
Comment 8. Line 60-62: Does the meteorological conditions include both microclimate and soil conditions? I noted that the observed dataset includes edaphic conditions in the abstract.
Response 8. Thank you for pointing out that we were not clear on this point. The meteorological conditions refer to general conditions in the wider study area (see response above), whereas the soil conditions were collected at six of the ecosystems. This has been clarified in the revised manuscript (Page 2–3, Lines 63–66).
Comment 9. Line 77: “…Approximately 60% and 13% ….”. It confused me.
Response 9. We agree that this was confusing. We changed the sentence to “Approximately 60% of Kulen is covered by cashew plantations, another 13% consists of forestland, while the remainder comprises other land types (Singh et al., 2019)” (Page 3, Lines 81–82).
Comment 10. Line 80: It will be better to show the locations of nine forestry inventory plots or Phnom Kulen National Park in the whole Southeast Asia in Figure 1.
Response 10. We updated the overview map in Fig. 1 to show the location of the study area in Southeast Asia accordingly (Page 4).
Comment 11. Line 236-240: The decimal places of surface temperature should be consistent with the values in Table 2.
Response 11. Thank you for noticing this. We have adjusted the text to ensure consistency with the decimal places of surface temperature values in Table 2 (Pages 10, Lines 230–232).
Comment 12. Line 241-246: The subplots (P, Tair, and Rg) of Figure 6 are same as those in Figure 5.
Response 12. Thank you for pointing this out. We assume it was Fig. 2 and 3 that was referred to. We have combined these figures to eliminate redundancy in the revised manuscript (Fig 2 of the revised manuscript, Page 11).
Comment 13. Line 274: The units of AGBf and AGBh in Table 2 are not consistent with those in the caption of Table 1.
Response 13. Thank you for pointing out that this was not clear. We use Mg ha⁻¹ for AGBf and AGBh in Table 2 for biomass per hectare, and kg tree⁻¹ in Table 1 for stand-level biomass estimation, each unit is standard in its respective context. We have revised Section “2.3.3 Stand structural attributes” to clarify this point (Pages 9, Lines 204–205).
Comment 14. Line 275-287: The decimal places of SLA, LDMC and Chl should be consistent with the values in Table S5.1-5.4.
Response 14. Thank you for noticing this. Values of SLA, LDMC, and Chl in Section 3.3 and Table 2 of the revised manuscript (Page 14, Lines 279–290; Page 12–13) have been rounded to two decimal places for consistency with Table S6.1–6.4 (Page 20–24 in the supplementary material).
Comment 15. Line 290: The mean and SD of DBH and H should be labeled in Figure S6.1.
Response 15. They have been labelled in the revised Figure S7.1 (Page 25 in the supplementary material).
Comment 16. Line 306-309: The p-value of the fitted power-law relationship should be labeled in Figure 4.
Response 16. We agree and added the p-value for the fitted power-law relationship in the revised Figure 3 (Page 15, Line 309).
Comment 17. Line 314-316: Why the estimated AGB of this study was substantially lower than the other two methods?
Response 17. Thank you for pointing this out. Our study introduces the AGBh method as an alternative to the widely used AGBf function for estimating aboveground biomass. Unlike AGBf, which relies on generalized assumptions of uniform species composition and stand structure, AGBh incorporates site-specific data, including species-specific wood density and localized diameter at breast height (DBH) and tree height (H) relationships, expecting to enhance accuracy. Our locally adopted AGBh method produced estimates ~ 30 % higher than the generic AGBf for both EF and RF (Table 2, Fig. 4b (Pages 12–13, Lines 265–277; Page 16, Lines 323)). This is likely due to the combined effects of higher mean wood density and a stronger DBH relationship, resulting in a more pronounced exponential growth response in AGB (Fig. 4a, Page 16). Still, these ~ 30 % higher values align with the range reported in previous studies (Tables A2–A3, Page 28–29). In contrast, in the CP case, our AGBh method produced estimates less than a quarter of the generic AGBf method. The reason is that the AGBh method is less reliable when a weak DBH-H relationship is detected because it fails to accurately capture the overall tree size and volume. This is also reflected in the substantially larger uncertainty in the CP AGBh method as indicated by the standardized errors of the parameters within the DBH-H relationship (Table A4; Table S7.1 (Page 29–30, Line 635; Page 26 in the supplementary material)). However, to fully validate the AGB allometric equations destructive field-observed data or lidar scans of the ecosystem would be necessary. We are currently planning a campaign where we are going to use a terrestrial laser scanner for scanning these ecosystems. Hopefully, future research will be able to include these direct field measurements of ecosystem structures to more accurately validate these AGB methods for these land cover types. We have clarified these points in the revised discussion (Pages 24–25, Lines 514–528).
Comment 18. Line 322–326: The unit of y-axis in Figure 5a should be kg tree-1.
Response 18. Thank you for this suggestion; we agree and have changed it in the revised manuscript (Fig. 4a, Page 16, Line 323).
Comment 19. Line 453: Why the Chl of CP in Kulen is higher by 2-2.5 times than that in India? It should be explained.
Response 19. The difference between our result and those reported in India is attributed to methodological variations. We used a Chlorophyll Meter (SPAD 502 Plus; Konica Minolta Sensing Inc., Japan) for direct field measurements, whereas the Indian study utilized a laboratory extraction method with dimethyl sulfoxide (DMSO) and acetone, followed by absorbance analysis at multiple wavelengths (645 nm, 663 nm, 652 nm, and 470 nm) (Mog & Nayak, 2018). We deleted this sentence to avoid confusion.
References
Ministry of Environment (Cambodia). (2018). Cambodia forest cover 2016. Phnom Penh, Cambodia: Mininistry of Environment
Mog, B., & Nayak, M. G. (2018). Leaf Morphological and Physiological Traits and Their Significance in Yield Improvement of Fifteen Cashew Varieties in West Coast Region of Karnataka. International Journal of Current Microbiology and Applied Sciences, 7(07), 1455-1469. https://doi.org/10.20546/ijcmas.2018.707.173
Singh, M., Evans, D., Chevance, J.-B., Tan, B. S., Wiggins, N., Kong, L., & Sakhoeun, S. (2019). Evaluating remote sensing datasets and machine learning algorithms for mapping plantations and successional forests in Phnom Kulen National Park of Cambodia. PeerJ, 7, e7841. https://doi.org/10.7717/peerj.7841
FAO, 1997: State of the World's Forests. Food and Agriculture Organization, Rome, Italy, 200 pp.
FAO/UNEP, 1999: Terminology for Integrated Resources Planning and Management. Food and Agriculture Organization/United Nations Environmental Programme, Rome, Italy and Nairobi, Kenya.
Citation: https://doi.org/10.5194/essd-2024-98-AC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 630 | 149 | 313 | 1,092 | 51 | 34 | 37 |
- HTML: 630
- PDF: 149
- XML: 313
- Total: 1,092
- Supplement: 51
- BibTeX: 34
- EndNote: 37
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1