Ten-Year Estimation of Net Primary Productivity in a Mangrove Forest under a Tropical Monsoon Climate in Eastern Thailand: Significance of the Temperature Environment in the Dry Season

Round 1

Reviewer 1 Report

The submission by Poungparn and coauthors provide a generally well written paper with sound science. In my comments below, I mostly address the need to bring the main findings (and implications of those findings) to the forefront of the paper, particularly in the abstract, results, and discussion. See below for specific comments and suggestions.

Table 3, p.10 - Why are basal area and mortality rate correlated with dry season temperatures but not NPP?
Is it because mortality does not count against NPP for that year? But, presumably higher temperatures and higher mortality rates lead to longer term reduction in living biomass and thus the ability of the forest to maintain higher NPP over the long term.

On page 13, l.391-393, you state, "We investigated the climatic factors related to the basal area increment and found that the basal
area increment and biomass of the living trees was inversely correlated with the mean and maximum monthly temperature in the dry season (Table 3)." I suggest providing a bit more context and interpretation here. Again, from Table 3, I would ascertain that basal area is negatively correlated with dry season temperatures because more trees die when it is warmer in the dry season. The implication (at least to me) is that warmer dry seasons could lead to higher mortality, long term reduction in living biomass, and thus lowered health and function of mangrove forests in the region. Such interpretation is implied in the discussion but needs to be explicitly stated and highlighted.

Regarding the comment above, you do address this idea of increased mortality on p.14, lines 431 to 433, where you state, "We inferred that the maximum and mean monthly temperatures in the dry season may have influenced the magnitude of NPP by significantly decreasing the biomass increment in living trees and increasing the mortality rate." This is good, but should be stated early in the section as this is a major finding of your study.
Also, is biomass increment during warmer dry seasons decreasing in the trees that survive or is the decline in living biomass mostly the result of mortality? Said differently, is biomass increment lower because there are fewer trees per hectare that contribute to NPP (during years with warm dry seasons)?

Author Response

Response to Reviewer 1 Comments

 

Comments and Suggestions for Authors

The submission by Poungparn and coauthors provide a generally well written paper with sound science. In my comments below, I mostly address the need to bring the main findings (and implications of those findings) to the forefront of the paper, particularly in the abstract, results, and discussion.

• We revised the abstract that the main finding (as your suggested) was expressed in the abstract (Line 33-35) as “It implies that the hot-dry season may lead to high mortality, long-term reduction in the increment of living-trees biomass, and thus lowered the ability to maintain high NPP of mangrove forests in long-term period.”

 

See below for specific comments and suggestions.

Table 3, p.10 - Why are basal area and mortality rate correlated with dry season temperatures but not NPP? Is it because mortality does not count against NPP for that year? But, presumably higher temperatures and higher mortality rates lead to longer term reduction in living biomass and thus the ability of the forest to maintain higher NPP over the long term.

• We considered that basal area and mortality rate correlated with dry season temperatures but not NPP. NPP is a summation product of DY and L that we couldn’t find significant relationship between L and the dry season temperatures. Therefore, NPP didn’t show a significant relationship with the dry season temperatures. We accepted your suggestion that the high temperatures and high mortality rate may cause a long-term reduction in living biomass.

 

On page 13, l.391-393, you state, "We investigated the climatic factors related to the basal area increment and found that the basal area increment and biomass of the living trees was inversely correlated with the mean and maximum monthly temperature in the dry season (Table 3)." I suggest providing a bit more context and interpretation here.

• We reorganized the Discussion section and provided more context and interpretation in line 419-427 of the revised manuscript, “We investigated the climatic factors related to the basal area increment and found that the basal area increment and biomass of the living trees was inversely related with the mean and maximum monthly temperature in the dry season (Figure 6e and f). The higher the air temperature was in the dry season, the smaller the increment in the tree trunk diameter. The maximum temperature had a strong negative effect on stem growth over a 30-year period in a terrestrial tropical forest under the Thai monsoon climate; it increased respiration, reduced stomatal conductance and thus lessened high transpiration demand by directly decreasing photosynthesis [21]. To minimize the rate of water loss under high temperature conditions, plant maintained low stomatal conductance resulting in a decreased photosynthetic rate and consequently reduced plant growth.”.

 

Again, from Table 3, I would ascertain that basal area is negatively correlated with dry season temperatures because more trees die when it is warmer in the dry season. The implication (at least to me) is that warmer dry seasons could lead to higher mortality, long term reduction in living biomass, and thus lowered health and function of mangrove forests in the region. Such interpretation is implied in the discussion but needs to be explicitly stated and highlighted.

• We explicitly stated the implication in line 405-411 of the discussion in the revised manuscript. “We inferred that the maximum and mean monthly temperatures in the dry season may have influenced the magnitude of NPP by significantly decreasing the biomass increment in living trees and increasing the mortality rate. The warmer dry seasons could lead to higher mortality, long term reduction in living biomass, and thus lowered the ability to maintain high NPP of mangrove forests in this region in long term. Clark et al. [18] reported that the NPP decreased in the warmer years in a terrestrial tropical wet forest at La Selva in Costa Rica during two decades of study.”

 

Regarding the comment above, you do address this idea of increased mortality on p.14, lines 431 to 433, where you state, "We inferred that the maximum and mean monthly temperatures in the dry season may have influenced the magnitude of NPP by significantly decreasing the biomass increment in living trees and increasing the mortality rate." This is good, but should be stated early in the section as this is a major finding of your study.

• As your comment earlier, we moved that sentence to the earlier part of the Discussion section (3 Temperature as an initial factor regulating NPP magnitude).

 

Also, is biomass increment during warmer dry seasons decreasing in the trees that survive or is the decline in living biomass mostly the result of mortality? Said differently, is biomass increment lower because there are fewer trees per hectare that contribute to NPP (during years with warm dry seasons)?

• The biomass increment over one year was calculated based on total tree biomass at each time. So, the lower biomass increment during the dry season is a result of both low biomass increment of living trees and high mortality that could not be explained separately. Although the high mortality caused lower tree density, the basal area of living trees gradually increased (Figure 3 of the revised manuscript).

 

Reviewer 2 Report

This manuscript reports changes of tree density, composition, and annual NPP over ten years in mangrove forests in eastern Thailand. It finds that increments in basal area and living tree biomass are negatively correlated with the maximum and mean monthly temperatures in the dry season. And the annual mortality rate is positively correlated with annual rainfall and the maximum monthly temperature in the dry season. However, the current version missing a great piece of data analysis (only correlation test was conducted), and both Introduction and Method section need more work (see my comments below). 

Line 40, probably change to "Mangrove ecosystems are ecologically significant to coastal areas, i.e.,..."

Line 44, odd sentences: "..have reportedly been occurring..", change to "mangrove degradation and deforestation are occurring at an alarming rate."

Line 45, shorten "...as well as changing patterns of climatic events, such as low or high intensity of extreme events, have been reported" to "and enhanced extreme events"

Line 46, "Approximately 2% of the..." this sentence is not that alarming, please rephrase the sentence. And is this loss due to warming climate only or with extreme events? I assume more carbon will be released back to the atmosphere under extreme events, such as tornado and fires. 

Line 48-53, these sentences highlight that limited studies reporting NEP but not NPP. This help set the stage of studying NEP but NPP, however, authors reported NPP in this study. So please change this sentence structure.

Line 53-55, problem statement in the last sentence is not clear. do you mean more research is needed due to the large variations of NPP among different regions? 

Line 57, do you need to mention terrestrial forests in this study? since the focus is mangrove forest?

Line 69-71, were findings from these previous studies included in this study for Results or Discussion. If not, no need to mention it here. Because the in line 72, one thing has not been done in one plot is not a strong problem statement. 

Line 73-76, I suggested authors to reorganize this paragraph, starting by the objective and hypothesis. the sentence of "we performed correlation analysis..." belong to methods, is not necessary here. 

Line 87, Figure 1, these two panels (a, b) can be merged by creating a inset map.

Line 95, inconsistent digits for the mean and SE (and was it SE or SD?)

Line 111, Figure 2, please increase the font size.

Line 115-121, this measurement happen later (in 2017) than the below measurements of tree inventory. considering move down this section.

Line 166, more details needed here, for example, 1. did author log-transform the data to meet the normal residuals? 2. what are the climatic factors, were min, max, mean temp all being tested? and were the strongly correlated? 3. I assume the correlation test was based on the yearly data, so what is the meaning of showing seasonal temperature and rainfall in Figure 2b? if the monthly temp in dry season was included in the analysis, please describe it. 4. More importantly, authors should apply time series analysis/trend analysis to all the reported data, in addition to the correlation test.

Line 170, similar to the order in Method section, this should be moved down as the last paragraph of Result section. And why studying the relative elevation and inundation periods? I feel like this was not mentioned in the Introduction nor Methods to tell readers about its importance to help understand the temporal variations of NPP (in line 54, this sentence about spatial variation should be expanded and mentioning elevation, inundation periods, etc). And why only measured in 2017 but not every 4 or 5 years?

Line 189, authors did not mention tree density in the Method section, and more stats can be applied to tree density, for example, comparing the differences among years statistically. Same thing for dead tree, tree height, etc, reported in this paragraph. If no replicates was conducted within each year, trend analysis/time series analysis should be conducted for all the data reported in this study.

Line 261, Table 3, now I feel like if linear regression is better than the simple correlation test, it also helps authors to make graphs instead of tables which is more attractive to readers. And it was surprising that there were strong correlations for annual maxi temp and annual mean temp but not for annual mini temp, because looking at Figure 4b, they are very similar except Y7.

Line 283, this sentence sounds like authors conducted a correlation test between salinity and specific distribution spatially and seasonally. 

Line 321, why this T/R ratio was not included in the Methods/Results? but only in the Discussion? Discussion should be discussing the findings in the Result section.

Line 377, is this sentence adding new insights to the literature? of course the NPP will be influenced by the component of NPP and climatic factors...That's being said, the whole Discussion section should be reorganized to only discuss the important findings from this study that will help advance the literature. 

 

 

Author Response

Response to Reviewer 2 Comments

 

Comments and Suggestions for Authors

This manuscript reports changes of tree density, composition, and annual NPP over ten years in mangrove forests in eastern Thailand. It finds that increments in basal area and living tree biomass are negatively correlated with the maximum and mean monthly temperatures in the dry season. And the annual mortality rate is positively correlated with annual rainfall and the maximum monthly temperature in the dry season. However, the current version missing a great piece of data analysis (only correlation test was conducted), and both Introduction and Method section need more work (see my comments below).

 

Line 40, probably change to "Mangrove ecosystems are ecologically significant to coastal areas, i.e.,..."

• We changed to "Mangrove ecosystems are ecologically significant to coastal areas, i.e.,..." as your suggestion in line 39 of the revised manuscript.

 

Line 44, odd sentences: "..have reportedly been occurring..", change to "mangrove degradation and deforestation are occurring at an alarming rate."

• We made it short to "mangrove degradation and deforestation are occurring simultaneously [7]." In line 43 of the revised manuscript.

 

Line 45, shorten "...as well as changing patterns of climatic events, such as low or high intensity of extreme events, have been reported" to "and enhanced extreme events"

• We revised the sentence as “Mangrove mortality and loss of area due to climate change and enhanced extreme events, have been reported [8].” in line 43-44 of the revised manuscript.

 

Line 46, "Approximately 2% of the..." this sentence is not that alarming, please rephrase the sentence. And is this loss due to warming climate only or with extreme events? I assume more carbon will be released back to the atmosphere under extreme events, such as tornado and fires.

• We changed “at an alarming rate” to “simultaneously” in line 43. According to Hamilton and Friess (2018), they calculated the loss rate of carbon from deforestation of global mangrove ecosystem from 2000 to 2012. So, we revised this sentence as “Approximately 2% of the long-term stored carbon in mangrove ecosystem was released back into the atmosphere at the global scale due to mangrove deforestation from 2000 to 2012 [4]” (Line 44-46 of the revised manuscript).

 

Line 48-53, these sentences highlight that limited studies reporting NEP but not NPP. This help set the stage of studying NEP but NPP, however, authors reported NPP in this study. So please change this sentence structure.

• In order to highlight how important of NPP study, we revised this paragraph as “In relation to the carbon cycle in mangrove ecosystems, carbon budgets have been addressed in the form of net primary productivity (NPP), i.e., [9–13], net ecosystem production (NEP), i.e., [10,14], and soil respiration, i.e., [15,16]. Among these, the carbon budgets have been mainly reported in the form NPP; this metric has been estimated with the summation method using various components, such as the growth increment in the aboveground and belowground parts, coarse woody debris flux, and coastal phytoplankton NPP. The variation in the NPP estimation of mangrove forests depends on the components included in the NPP estimation and geographical regions with different climatic conditions [9,11].” as shown in line 47-54 of the revised manuscript.

 

Line 53-55, problem statement in the last sentence is not clear. do you mean more research is needed due to the large variations of NPP among different regions?

• We revised the sentence to “The variation in the NPP estimation of mangrove forests depends on the components included in the NPP estimation and geographical regions with different climatic conditions [9,11]”, as shown in line 52-54 of the revised manuscript. This will give a broad introduction to the reader that more research on NPP estimation is necessary to understand the response of NPP in different climatic conditions of different geographical regions.

 

Line 57, do you need to mention terrestrial forests in this study? since the focus is mangrove forest?

• Although our study focuses on the mangrove forest, we mentioned the study of terrestrial forests to show that fewer studies were conducted in the mangrove forest than the terrestrial one. This highlights an importance of the study on mangrove forest productivity in a relation to the climatic factors.

 

Line 69-71, were findings from these previous studies included in this study for Results or Discussion. If not, no need to mention it here. Because the in line 72, one thing has not been done in one plot is not a strong problem statement.

• All of the previous studies were investigated from the same study plot. According to your comment, we revised this part (line 67-70 of the revised manuscript) by selection only the previous studies whose findings were included in the Results or Discussion of the present study.

 

Line 73-76, I suggested authors to reorganize this paragraph, starting by the objective and hypothesis. the sentence of "we performed correlation analysis..." belong to methods, is not necessary here.

• According to your suggestion, we reorganized the paragraph to “In the present study, we aimed to investigate the variation in forest productivity from 2008–2018. It was hypothesized that the variation in forest productivity within the decade was related to climatic factors, including rainfall and temperature, under the monsoon climate.”, as shown in line 72-74 of the revised manuscript.

 

Line 87, Figure 1, these two panels (a, b) can be merged by creating a inset map.

• The two panels of Figure 1 were merged into an inset map.

 

Line 95, inconsistent digits for the mean and SE (and was it SE or SD?)

• We expressed the mean and SD of the annual rainfall with inconsistent digits (Line 94-95 of the revised manuscript).

 

Line 111, Figure 2, please increase the font size.

• The font size in Figure 2 was increased.

 

Line 115-121, this measurement happen later (in 2017) than the below measurements of tree inventory. considering move down this section.

• We moved down that section to 2.4 of the Method section as your suggestion.

 

Line 166, more details needed here, for example, 1. did author log-transform the data to meet the normal residuals?

• We revised the manuscript by adding more detail in the Method section (Data analysis, Line 167-172). The data were verified to be in the normal distribution by using log-transform and Shapiro-Wilk test before the further analysis.

 

2. what are the climatic factors, were min, max, mean temp all being tested? and were the strongly correlated?

- We used all parameters including min, max, and mean air temperatures in the

   analysis. The max and mean air temperatures showed strong correlations.

3. I assume the correlation test was based on the yearly data, so what is the meaning of showing seasonal temperature and rainfall in Figure 2b? if the monthly temp in dry season was included in the analysis, please describe it.

• We showed seasonal temperature and rainfall in Figure 2b because we want to express how seasonal variation occur in this region, and also because we used the maximum and mean monthly temperature in the dry and wet seasons as the climatic factor in the statistical analysis. We explained the duration of the dry and wet seasons as “The dry season began in November and lasted to April. The wet season usually started in May and extended to October; the rainfall was high during this period, at 87.3% on average (Figure 2b).” in line 97-99 of the revised manuscript.

 

4. More importantly, authors should apply time series analysis/trend analysis to all the reported data, in addition to the correlation test.

• We performed the time series analysis/trend analysis for the data shown in Table 1 of the previous version manuscript. The time series analysis/trend analysis was shown in Figure 3 of the revised manuscript. The data included density, DBH, and basal area of the living trees (Figure 3a and b), density and DBH of the dead trees (Figure 3c), and density and DBH of the newly recruited trees (Figure 3d). The results showed that the density of the living trees significantly decreased, while the DBH, and basal area of the living trees significantly decreased from 2008–2018.

 

Line 170, similar to the order in Method section, this should be moved down as the last paragraph of Result section. And why studying the relative elevation and inundation periods? I feel like this was not mentioned in the Introduction nor Methods to tell readers about its importance to help understand the temporal variations of NPP (in line 54, this sentence about spatial variation should be expanded and mentioning elevation, inundation periods, etc). And why only measured in 2017 but not every 4 or 5 years?

• The result of relative elevation and inundation periods was moved to the last section of the Results section. We studied the relative elevation and inundation periods to explain species distribution in the study plot. We also discussed the species distribution of mangrove in the study plot in a relation to salinity environment (reported by our previous study). The saline environment affected the tree growth and thus forest productivity. Unfortunately, we have available data only in 2017.

 

Line 189, authors did not mention tree density in the Method section, and more stats can be applied to tree density, for example, comparing the differences among years statistically. Same thing for dead tree, tree height, etc, reported in this paragraph. If no replicates was conducted within each year, trend analysis/time series analysis should be conducted for all the data reported in this study.

• In the revised manuscript, we mention the tree density in the Method section (Line 120-121). According to your suggestion, we conducted time series analysis for the parameters of density or number, average DBH, basal area of living trees, dead trees and newly recruited trees by using linear regression analysis. The results of this analysis were shown in Figure 3 of the revised manuscript instead of Table 1 of the previous manuscript. Each of graphs was also cited in the text of Results section.

 

Line 261, Table 3, now I feel like if linear regression is better than the simple correlation test, it also helps authors to make graphs instead of tables which is more attractive to readers. And it was surprising that there were strong correlations for annual maxi temp and annual mean temp but not for annual mini temp, because looking at Figure 4b, they are very similar except Y7.

• Instead of Table 3 of the previous version manuscript, we made graphs showing linear regression models as your suggested. The results were shown in Figure 6 (a-h) of the revised manuscript. We presented the linear regression lines in the graphs only for the significant relationship (P < 0.05).

 

Line 283, this sentence sounds like authors conducted a correlation test between salinity and specific distribution spatially and seasonally.

• We didn’t conduct a correlation test between salinity and specific distribution spatially and seasonally. This sentence is referred to a previous study, Komiyama et al. (2019), so that we cited the reference in this sentence, “Thus, the soil-water salinity showed both spatial and seasonal fluctuations [27] corresponding to the species distribution in our study plot.” Line 301-302 of the revised manuscript.

 

Line 321, why this T/R ratio was not included in the Methods/Results? but only in the Discussion? Discussion should be discussing the findings in the Result section.

• We respectively added the method of T/R ratio calculation (Line 133-134), and result of T/R ratio (Line 216) in the revised manuscript. The low T/R ratio indicates the biomass allocation into the root system.

 

Line 377, is this sentence adding new insights to the literature? of course the NPP will be influenced by the component of NPP and climatic factors...That's being said, the whole Discussion section should be reorganized to only discuss the important findings from this study that will help advance the literature.

• This sentence concluded the factors influencing NPP magnitude that will give information to the readers who are not in this field or new to NPP study.

 

Round 2

Reviewer 2 Report

I would like to thank authors' hardworking on revising this manuscript. I only have some suggestions on Figures. First, please increase the font size for all figures. Right now the X Y axis labeling are too small, so do the values. Second, Figure 6 is missing the labeling for Y axis, I know all 7 panels sharing the same Y axis, but authors need to label it. And if authors want to keep the blank spaces for panel b-h for the purposes of sharing the same scale of X axis, they should keep the scale the same or at least the same for each column. And increase the size for the legend.

 

Author Response

Reply to reviewer 2

 

Reviewer 2

 

I would like to thank authors' hardworking on revising this manuscript. I only have some suggestions on Figures.

• We would like to thank the reviewer for constructive comments and suggestions given to the manuscript.

 

First, please increase the font size for all figures. Right now the X Y axis labeling are too small, so do the values.

• For all of the Figures (2, 3, 5, and 6), we increased the font size from 14 to 16 points for all elements of the graphs in the original excel files.

 

Second, Figure 6 is missing the labeling for Y-axis, I know all 7 panels sharing the same Y axis, but authors need to label it. And if authors want to keep the blank spaces for panel b-h for the purposes of sharing the same scale of X-axis, they should keep the scale the same or at least the same for each column. And increase the size for the legend.

• We added the label for Y-axis of Figure 6. The panel a to h shared the same Y-axis with the label. We adjusted the scale of X-axis for each column (column d, f, h; column c, e, g), however we still keep the label for X-axis of each individual panel in order to easily recognize for readers. We increased the font size to 16 points for all elements of the graphs in the original excel files.