MAJOR COMMENTS

In this manuscript, the authors assess the main driving factors of variations in tree diversity, tree species composition and tree species wood density, all over the Amazon. To accomplish this objective, they used an astonishingly large dataset of 443 plots comprising 210,801 individuals of 3,527 species. As response variables, they considered Fisher’s Alpha of tree species, genus and family diversity as proxies of variations in tree diversity, the coordinates in axes 1 and 2 of a NMDS ordination of species abundances as proxies of variations in species composition, and the Community Weighted Mean (CWM) of wood densities at species level, as proxy of variations in wood density. As driving factors, they considered the Maximum Cummulative Water Deficit (MCWD), the Height Above the Nearest Drainage (HAND), a variable summarizing soil fertility, the % Clay, four regions representative of the geological age of the substrate. They performed Generalized Least Squares (GLS) models, Non-metric multidimensional scaling (NMDS), and permutational (i.e. non-parametric) analyses of variance (PERMANOVA), among other statistical techniques.

The objective of their research is very appropriate and ambitious. However, and honestly, I have a number of concerns regarding the study questions and hypotheses, the methods used and the way they are presented. Below, I give a number of major and minor comments in order to help the authors to improve their manuscript.

In the INTRODUCTION, is very necessary that the authors clarify which are the variables on which they are going to study the effect of topographically-driven local hydrological conditions; (i.e. species richness, species composition, forest structure, plant functional traits). For instance, they do not state that they don’t mention the Fisher’s Alpha (which is the index they use) until the Methods. Clarifying all these terms and commenting on the suitability of the Fisher’s Alpha would make the text much more understandable from the very beginning. More importantly, two of the hypotheses posed at the end of this section are not derived from the text of the Introduction and the authors should provide the appropriate justifications.

In the METHODS, there is a continuous confusion between the variables of interest (e.g. tree diversity, tree species composition and wood density), and the response variables used to represent them (i.e. Fisher’s Alpha, NMDS axes and CWM), which makes very difficult for the reader to understand the analyses performed. Therefore I would appreciate it if the authors make this differentiation clearer.

The RESULTS are very hard to follow and understand, for several reasons. First, because the tables and figures in which they appear are indistinctively mixed with the main text of the manuscript. Second, because the paragraphs and texts presenting the results of the different analyses are disorganized. For instance, there is no a clear reference to the most important driving factors of tree diversity (i.e. Fisher’s Alpha for species, genus and families) in order of significance. Third, the results are not presented in an organized way (i.e. Figure 4 is cited before Figure 3) and there is no a clear link between each of the analyses and the figures presenting their results. In fact, one of the analyses (PCoA; Table S4, Fig. S2) is not even mentioned in the main body of the manuscript. I think the authors should; (1) present the results of the different analyses in an organized way, clearly indicating the table/figure in which they are given, (2) comment just the significant results in the main text, placing the non-significant ones in the supplementary materials, (3) consider if all figures given are necessary (e.g. as far as I understand, Table 1 and Figure 4 present the same results, so please, remove Figure 4 or put it in the Supplementary materials) (4) place all tables and figures in a unique, specific section, after the references.

MINOR COMMENTS

Lines 28-29: The sentence “The water availability actually experience by plants”, sounds weird. I suggest writing better “Water availability for plants, within…”

Lines 31-32: I suggest rewriting: “which in turn affect the hydrological patterns and the large-scale distribution of plants”.

Lines 33-34: The term “Amazonia” refers more to a vaguely defined region than to a forested area. Please, write better “In the Amazon basin, where the world largest tropical forest is located”.

Line 37: The appropriate term is “global warming”.

Line 38: Please, write better “water availability for plants”.

Lines 37-42: This is a very long, difficult-to-follow and complex subordinate sentence. I suggest split it in two; (1) The first one, to state that global warming is leading to an increase in the frequency (and intensity) of droughts, which in turn is provoking a reduction in water availability for plants, as well as higher tree mortality and biomass loss, in tropical environments. (2) The second one, to state that, as a consequence of the former statement, results crucial to assess the influence of local hydrological conditions on forest diversity, structure and composition.

Line 50: You cannot write “On the other hand” here, because you have not previously written “on one hand” in line 47. I suggest start this sentence with “In contrast”.

Line 53: Which environmental conditions? I guess the later ones (i.e. those in the upper areas). If affirmative, please, write it.

Line 63: Do you refer to “species richness” and “species composition”? because the term “diversity” includes both. Are you going to study species richness of trees and species composition of all growth forms (i.e. including treelets, shrubs and herbs)? Please, clarify.

Lines 64-65: I don’t agree with the terms used here. Hydro-topographic conditions influence diversity in functional traits (no “functional composition”). I also strongly suggest mentioning the leaf traits studied by Kraft et al. (2008), namely: specific leaf area, leaf nitrogen concentration and leaf size.

Line 71: I guess you refer to “structure” and “functional traits diversity” of forests, don’t you?. Moreover, in lines 63-65 you don’t mention structure, just species richness, species composition and functional composition (i.e. “functional traits diversity”).

Lines 72-75: The first part of this sentence is awkward. Which is “non-seasonal”? I guess the rainfall, isn’t it?. Then, please, clarify that you are referring to different zones within the Amazon basin in which rainfall patterns vary.

Line 84: Which do you mean with “Geomorphological formations”? Please, clarify.

Line 91: The terms “tree diversity” and “species composition” are ambiguous (see comments in lines 63 and 71). Here you mention just wood density, not other plant functional traits, and you don’t mention structure. Please, clarify which are the focus variables on which you are going to study the effect of topographically-driven local hydrological conditions.

Lines 95-97: The terms used here are unclear (see comments in lines 63, 71 and 91). Do you refer to tree species richness, species composition and wood density? Then, please, clarify.

Lines 102-203: Which are these “opposing patterns”?. Please, clarify.

Lines 103-107: This hypothesis does not arise from the statements previously made in the Introduction. You should have explained what we know about the effects of areas with ancient edaphic structures and well-dissected reliefs vs those with younger edaphic structures and flatter topographies on tree species composition. Then, at the end of the Introduction, you should have posed the hypotheses derived from such statements.

Lines 108-109: Again, this hypothesis is not derived from the text of the Introduction, where wood density appears just at the end, in a very brief sentence (lines 92-94). Please, clarify (in the previous parts of the Introduction), why do you expect that forests on soils with greater drainage capacity and less water availability will have a higher wood density.

Line 120: Please, substitute “diversity” by “species richness”.

Lines 148-149: Which is “bilinear extraction”? Please, comment or at least give a reference.

Lines 149-151: Idem lines 148-149. Please, explain which the Strahler’s sixth order drainages are.

Lines 151-152: For which analysis was necessary that HAND data meet normality assumptions? Please, explain.

Line 157: I would split the sentence in lines 156-159 in two, making a point in “year” (i.e. where you finish explaining MCWD). Then, I would start a new sentence explaining what is WD and saying that it was calculated for every month.

Lines 158-159: I suggest placing this formula in its own, new paragraph, and just afterwards (in the next paragraph), explaining it terms with “where (term)…is… and (term)… is…”, as it is commonly used in scientific articles. To explain the different terms of the equation you can use the text in lines 159-162.

Line 168: Which do the initials “SCC” mean?

Lines 193-196: I guess you refer to the logarithmic series index, don’t you? In this case, you should briefly comment on this index in the Introduction, justifying why it is particularly appropriate for studies over large-scale regions such as yours. I also recommend including the formula of the Fisher’s Alpha at the end of this paragraph.

Lines 206-209: I strongly suggest placing this sentence in line 199, as it justifies the use of NMDS to “summarize” species composition in order to perform non-parametric multiple analyses of variance (PERMANOVAs).

Lines 236-237: The text in these lines contradicts somehow what is written in lines 240-241; i.e. that you assessed the influence of MCWD, soil fertility, soil texture and HAND (independent variables) on species diversity (response variable). Please, clarify.

Lines 244-245: Exponential variance terms are usually included in Multiple Regression models in order to quantify non-linear relationships. Please, could you explain or provide a reference on how it is possible to reduce heterocedasticity by adding an exponential variance structure?.

Lines 251-252: I recommend mentioning the independent variables in the same order as you used in previous lines; soil fertility, soil texture and HAND.

Line 257: When referring to the supplementary figures, such as in this sentence, instead of “Supporting information” you should clearly indicate the Figure. Otherwise, it is really difficult to find the results reported.

Lines 257-258: Which NMDS axes you used as response variables, representing species composition? Please, detail.

Lines 258-259: All over the manuscript and particularly in the Results section, it is a continuous confusion between the sariables of interest (e.g. tree diversity, tree species composition and wood density), and the response variables used to represent them (i.e. Ficher’s Alpha, NMDS axes and CWM). When referring to statistical models, I strongly suggest making clear that you are referring to the later ones. For instance, here I would write; “We modelled the NMDS axes and CWM as a function of…”, because these are really the variables you are modelling!!.

Lines 260-261: I suggest joining this sentence with the former, by a comma, and writing “including an exponential”. Moreover, you should justify why you included such a structure in the Methods.

Line 261-267: This sentence is very long and confusing, please, split in two, clearly indicating the models used in the GLS analyses for species composition (NMDS axes as response variable) and wood density (CWM as response variable).

Lines 268-269: See comment in lines 258-259.

Lines 268-269: Again, see comment in lines 258-259. Here you are mixing one variable of interest (i.e. tree diversity) which is quantified by the Fisher’s alpha for species, genus and families, with two response variables (i.e. NMDS axes and CWM) which are calculated to consider tree species composition and wood density variation. Please, be much more consistent in the use of terms.

Lines 272-273: The term “bivariate” is very used in Spatial Point Pattern Analyses, but it is weird here. I suggest writing; “We also examined the pair-wise correlations between the response variables considered, through Spearman’s correlation tests”. Moreover, you didn’t justified why you performed these correlations… and you should do it.

Line 278-279: There is a grammatical mistake here; you should use the same preposition in the two sub-sentences: “associated with the interaction” and “with the individual”. Moreover, I guess the results indicated in these lines 278-279 are also given in Table 1, but this table is firstly mentioned in line 281. Please, add a sentence at the beginning of this subsection clearly indicating that the results of the GLS models are given in Table 1.

Lines 278-281: The way of presenting these results is very confusing. Better, give sentences for Fisher’s Alpha diversity of species, genus and families, separately. In each of this sentences, indicate the significant driving factors in order of significance. For instance, Fisher’s Alpha diversity for species was significantly determined by MCWD and then, by the interaction between HAND and MCWD. Last but not least, if you give the P-values in the table, it is not necessary to repeat them in the main text!.

Lines 281-282: Write better “the partial effect of the interaction between”. Otherwise, it is very difficult to understand that you are going to comment other figure.

Line 292-293: Which is the analysis which results are you going to comment here? Please, say it properly, otherwise, it is really difficult to know what you are writing about. Moreover, why you first refer to Figure 4 than to Figure 3 (which is firstly mentioned in line 364)? If the results of Figure 4 are going to be presented first, this figure should be numbered as Figure 3. Moreover, I think that the size of the standardized effects is not important at all, if they are not given together with their significance, and both results are appropriately given in Table 1. Therefore, I recommend placing current Figure 4 in the Supplementary materials.

Line 295: There are seven figures in the Supplementary materials. To which of these figures you refer to? Please, clarify.

Lines 295-298: I don’t understand why you are giving these results here. According to Table 1, soil fertility had a positive effect on Fisher’s Alpha diversity of species, genus and family. However, just the effect for FA-family was significant (0.247**). Then, why you mention the results for Fisher’s Alpha diversities of species and genus?. Please, refer strictly to the significant results, otherwise, it is really difficult to differentiate what is important and what is not within your results.

Lines 302-303: Do you think that, with this low proportion of variance captured, it is worthy to model the relationships between the coordinates in NMDS axes 1 and 2, and the driving factors mentioned in Table 1? Why.

Lines 309-310: According to Table 1, tree species composition, represented by the coordinates of NMDS axis 1, was firstly determined by soil fertility, which shows the largest standardized coefficient and is the most significant factor (i.e. -0.56***; Table 1). I think this is the most important result to indicate here, but you don’t mention it up to lines 317-318. Please, correct this sentence (and re-structure the entire paragraph) to present your results in order of importance.

Lines 310-313: You have commented the results given by Table 1, then, those given by Figures 2 and 4, and now, you go back to Figure 2 again. Please, make clearer the links between each study question (Introduction), the analyses (Methods) to broach them, and the Tables and Figures (Results) in which they are presented. Otherwise, it is very difficult to follow the manuscript.

Line 316-317: To which analysis are these results referring to?

Lines 320-321, 323, 326, 337: See comment in line 295.

Lines 326-327: If the effect of this interaction is neither strong nor significant, don’t mention it at all.

Lines 327-328: This is very strange, although there is a clear difference on species composition (i.e. coordinates of NMDS axis 2) between wetter (blue) and drier (red) forests, there is no effect of HAND: MCWD on NMDS axis 2. What could explain this apparent lack of correlation among both results?

Lines 329-331: Then, simply delete this sentence.

Line 336: I guess that in Figure 2 you examine the partial effects of the interaction terms on those response variables that are strongly and significantly determined by interaction terms in Table 1, don’t you? In this case, I strongly suggest removing subpanels 2C (FA-family) and 2E (NMDS 2) from this figure as these two response variables are no significantly determined by interaction terms at all. The much simpler resulting figure would have just four panels; two for the effect of the interaction HAND: MCWD on FA-species (2A) and FA-genus (SB), and two for the effect of the interaction HAND: Region on NMDS-1 (2C) and CWM (2D).

Line 338: Wood density (CWM) was mainly (and negatively!) determined by Soil fertility and Region. Please, indicate that both effects were significant and negative.

Lines 342-344: See comment in lines 329-331.

Lines 347-354: See comment in line 336. Re-structure this figure leaving just four subpanels. In the Figure’s caption you should indicate that it explores the partial effects of interaction terms on FA-species, FA-genus, NMDS-1 and CWM.

Lines 360-368: The results referring to pair-wise correlations among response variables are given in Table S3, completely hidden in the manuscript, and not even mentioned at the beginning of this section. If you finally maintain these analyses (you should appropriately justified them), please, place Table S3 in the main body of the manuscript (as Table 2).

Lines 361-362: According to Table S3, there are 14 significant correlations among response variables. Why you have specifically focused in the two of them represented in Figure 3? Moreover, note that, from 25 possible correlations, 14 of them are significant. This result could be related to the fact that in multiple correlation tests, many correlations appear as significant by chance. If you keep this analysis, you should perform a Rice correction for multiple correlation tests.

Lines 390-394: In Table 1’s caption, please, indicate that you performed GLS models for tree diversity (i.e. Fisher’s alpha for species, genus and family; FA-species, FA-genus and FA-family), tree species composition (i.e. coordinates of NMDS axes 1 and 2; NMDS 1 and NMDS 2) and wood density (Community Weighted Mean; CWM). Do not refer to “Supporting information”, but just to Table S2, and indicate (in the caption, not below the table), that , and ** indicated significant coefficients with P < 0.05, 0.001 and 0.0001, respectively. In the first column of the table you should indicate that “Species”, “Genus” and “Family” refers to Fisher’s Alpha (e.g. FA-species, FA-genus, FA-families). Instead of “Wood density”, use CWM, which is the variable considered in the analysis.

Supporting information

Table S1: In my opinion, this table is not necessary at all, please, remove it.

Table S2: In the caption, indicate that Fisher’s Alpha represent tree species, genus and families diversity, the coordinates of NMDS axes 1 and 2 represent tree species composition, and the Community Weighted Mean of wood density per species, represent variations in wood density at species level. The second sentence of the caption more appropriate for the Results section, and the last one, is better placed in the Methods section. Please, remove these two sentences and just write that AIC and ∆AIC are indicated.

In Table S3, please, indicate significance levels as n.s. for non-significant ones, and , and * for those in which P ≤ 0.05, 0.001 and 0.0001.

Table S4: This analysis (PCoA) is firstly mentioned here in all the manuscript. If you maintain this table (and Figure S2), please, mention (and justify!) the analysis in the main body of the manuscript.

Dear authors,

I have now completed the revision of the manuscript ECOG-06-0125.R1, entitled “Local hydro-topographic influences tree diversity and composition across the Amazon basin”.

The manuscript has slightly improved from its previous version (ECOG-06-0125), and I appreciate the effort made by the authors to focus on their more significant results. Unfortunately, many sections are still poorly structured and difficult to follow, some aspects needs clarification, and more consistency in the use of terms is needed. I also recommend a general revision and correction of English writing.

Moreover, I am somewhat disappointed because the authors had not adequately responded to many of the questions I posed in my first revision (e.g. lines 327-328 and 361-362 of ECOG-06-0125), just answering with “done”. They have not accomplished some of the changes that I recommended, without providing any explanation.

I spent a lot of time revising the former (and current) version of the manuscript, and I honestly appreciate if the authors appropriately respond to all my queries and perform a careful correction considering all my concerns.

MAJOR COMMENTS

In the Introduction, the main hypotheses and variables of interest are now much clearer, however, the paragraphs stating the role of local hydrological conditions to determine changes in species richness and composition are still very messy. I suggest the authors to structure the Introduction as follows: (1) one general paragraph explaining what are local hydrological conditions and why they are, in general, important to determine changes in species richness, composition and plant traits, (2) one paragraph presenting the importance of local hydrological conditions to determine species richness (giving support to hypothesis 1), (3) one paragraph presenting the importance of local hydrological conditions to determine species composition (giving support to hypothesis 2), and (4) a paragraph presenting the importance of local hydrological conditions to determine variations in plant traits, and specifically wood density. The first and fourth paragraphs are now more or less appropriately framed in lines 28-44 and lines 81-97, but the second and third (which could be framed in lines 63-80 and 81-97) are still poorly structured. Below, in the minor comments, I give the authors some tips in order to improve them.

METHODS

The methods are clearer than in the previous version. Anyway, I would still suggest some re-structuring. First, I think that this section will be easier to follow if they explain first the dependent variables (i.e. now termed vegetation metrics), and then the independent ones (i.e. environmental variables). They should also revise the writing of some difficult-to-follow paragraphs such as the one on species composition (lines 221-249). Moreover, they mention proportions of sand and clay (not silt) in order to account for texture, but they don’t refer to these proportions anymore all over the main text.

RESULTS

Some sections of the results are still very messy (e.g. lines 320-324). I strongly recommend structuring them in sections / paragraphs corresponding to any of the hypotheses (and any of the analyses!). In each of these sections, it will be also necessary to start commenting the most important results (those in the figures that will be included in the main text) and just then the results which appear in the supporting information. Some paragraphs are structured the other way round, which makes very difficult for the reader to understand what did you do first, and which results are the most important ones.

The use of terms is not consistent. For instance, Soil Cationic Capacity (SCC) is mentioned as “Soil fertility” in Table 1 and Fig. 2, but as SCC when referring to these results in the main text (line 319). This variable also appears as “Soil fertility-SCC” in Figure S2. A similar confusion can be seen in lines 331-332. I suggest using SCC all over the main text and in all figures and tables, indicating in their captions, that these abbreviations refer to Soil Cation Concentration (SCC). In the Methods, you have appropriately indicated that soil fertility was taking into account by measuring Soil Cation Concentration (SCC; lines 183-184).

More importantly, there is one analysis (the Spearman correlations between tree species and composition) which is neither justified in the Introduction nor mentioned in the Methods, and another one (the PCoA) which is mentioned in the Methods but its results not given!. The authors should explain why they have included these two analysis, to which of their hypotheses do they respond. On the contrary, if they are not relevant at all, they should be removed from the main text.

DISCUSSION

In contrast to the Methods and Results section, the Discussion is relatively well structured, although there are some statements (particularly those referring to species sorting and trait selection), that could be somehow speculative. The section on the most important variables determining changes in wood density and species composition (lines 421-474) is not clear at all. I recommend the authors to split it in two separate sections, explaining the results obtained for each of these response variables.
Moreover, according to the Results (Table 3), Soil fertility (Soil Cationic Capacity; SCC) is apparently the most important variable to determine changes in wood density (CWN-wd), but this variable is not even mentioned in this section, which mostly refers to hydro-topographic conditions. I think the authors should give more importance to Soil fertility as a major determinant of Community Weighted Mean of wood density.

MINOR COMMENTS

Introduction

Lines 30-33: Thanks for the changes, now I understand what you mean with this sentence. Anyway, I would re-structuring it in order to make clearer what you meant with the term “local hydrological conditions”, which is very important for understanding what you are doing in the manuscript. To re-structure this sentence I recommend a zoom-in perspective (from the largest to the smallest scale). In the first sentence you have referred to climate, don’t you?. Then, start this one with “Nevertheless, within any given climate…” and afterwards “at local scale, water availability for plants is modulated by local topographic and edaphic conditions (i.e. local hydrological conditions), which in turn affect large-scale variations in plant species diversity”. I would delete “hydrological patterns”, because they are mentioned just before, and “distribution of plants”, because it is a very vague term. In contrast, “species diversity” clearly refers to richness of species and traits, and species composition.

Lines 44-62: This paragraph is still poorly structured. I guess it refers to the importance of local hydrological conditions to determine species composition, which is mentioned three times (lines 45, 50 and 58). Then, why do you suddenly mention neutral processes, dispersal limitation and biotic interactions?... I strongly suggest deleting the reference to all these processes in this paragraph. Optionally you can mention them at the beginning of the former paragraph, when you refer to other drivers of species richness and composition in tropical forests. But, as soon as you start focusing on local hydrological conditions as an important determinant of species richness and composition, please, keep the focus on. After its reading, it should be clear how local hydrological conditions can determine species composition all over the Amazon.

Lines 63-80: Again, this paragraph is not focused. In my opinion, here you should clearly state that local hydrological conditions determine species richness, providing support for hypothesis 2. However, in this paragraph you mainly explain how topography and soil texture determine local hydrological conditions (lines 63-78), and just at the end (lines 78-80) that local-hydro-topographic conditions and precipitation affects the community assemblage. Does it mean that local hydrological conditions affect species richness?... Then, write it explicitly, it will help the readers to link this paragraph with hypothesis 1. I would also add here the information given in lines 113-116 (see comment on them).

Lines 84-87: This sentence on the influence of local hydrological conditions on wood density is, in my opinion, very important to justify the study of this trait. However, I would locate it at the end of the paragraph (line 96), after the (more general) sentence referring to the influence of local hydrological conditions on all traits. See also comment in lines 101-104.

Lines 96-97: I suggest removing this sentence, because you will study whether local hydrological conditions affect plant traits variation (i.e. variations in the Community Weighted mean of wood density) or not. The sentence is not related with this objective.

Lines 98-99: You are not studying the influence and importance of local soil hydrological conditions “on tropical forests” but on patterns of variation in species richness, species composition and plant functional traits in tropical forests (i.e. over the Amazon basin). I suggest removing and starting directly with “The aim of this study”.

Lines 101-104: I would place this sentence just at the end of the paragraph on how local hydrological conditions filter plant traits (line 96). It is important to justify (in the corresponding paragraph) why you just finally selected wood density. See also comment in lines 84-87.

Lines 104-107: I would remove these questions. In them, you are basically saying the same as in lines 99-101.

Lines 109: Better, “specifically, we predict that”.

Lines 113-116: This information would be better placed in the paragraph referring to the influence of local hydrological conditions on species composition (see also comments in lines 63-80 and 191-192).

Methods

Lines 129-131: Are the mean, minimum and maximum values mentioned in line 129 the mean, minimum and maximum values of the number of species per plot?. If affirmative, please, indicate it. Otherwise, I am not sure if the statement in lines 129-131 is necessary. The one given in lines 131-133 is enough to justify the inclusion of just those individuals identified to species.

Lines 141-142: Please, remove the terms in brackets; “(Height Above Nearest Drainage and soil texture)” and “(the climatic water deficit)”, as they are correctly specified in the two paragraphs below

Line 146: Please, write “Height Above Nearest Drainage” and then, HAND, in order to identify the name to which those initials are referring to.

Line 152: It is not necessary to include these citations here, as you had just mentioned them in line 146.

Line 158-159: I suggest joining this sentence with the latest, writing “bilinear interpolation, which assigns”.

Line 161: Include the reference to Banon & Novo (2018) here, just after “drainages”.

Lines 164-166: Here, as well as in lines 257-258, you mention the proportions of sand and clay, but they are not mentioned anymore in the text. In Table S1, when presenting the models, you just refer to “Texture”. Why you did not considered the proportion of silt?, second, which means “Texture” in the models? Proportion of sand or of clay?... Please, clarify this issues, because otherwise it will be very difficult to understand your results.

Lines 191-192: I strongly suggest including this sentence, together with the references to Feldpausch et al. (2011) and ter Steege et al. (2013) in the paragraph of the Introduction presenting the importance of local hydrological conditions to determine species composition (i.e. the one that gives support to hypothesis 2; 44-62). In the Introduction you refer to the importance of zones (in general) to determine species composition, in the Methods you just refer again to Feldpausch et al. (2011; lines 190-191) and describe the zones (lines 192-202). See also comments in lines 63-80 and 113-116.

Lines 218-220: I think it should be more appropriate to refer to the Community Weighted Mean of wood density (CWM-wd) in each plot.

Line 224: Which “regression analyses”? To refer to the PERMANOVA I would use the term permutational (or non-parametric) analysis of variance.

Line 224-227: You could remove this sentence, as you have already referred to this fact in lines 191-192 (the sentence that I suggest you to place in the Introduction!).

Line 228: Which was the dependent variable in PERMANOVA? The region to which 1-ha plot belonged to?, and which were the independent variables?... The presence / absence of each species in each plot? Abundance of each species in each plot?. Please, clarify.

Line 232: By “one ordination” you refer to one NMDS, don’t you? If affirmative, please, specify.

Line 239: “occurring in less than two plots”, maybe?

Lines 242-245: Do you regressed one matrix against another one?. Sorry, but I cannot understand this sentence. Please, explain it.

Lines 246-249: If you quantify the variability in species composition within each region, why you applied the PCoA over all plots?. Please, clarify. Moreover, all over the manuscript, this analysis is just mentioned here. In ECOG-06125 you just mentioned the PCoA in the Supplementary materials (Table S4 and Figure S2). When responding to my question on Table S4 (ECOG-06125) you affirmed that "the results of these analyses have been included on the main text”. However I cannot see the results of the PCoA on ECOG-06125_R1. Please, clarify why you included this analysis, where are its results and what do they add to those of the NMDS ordinations performed in each region (lines 232-233).

Lines 255-257: If you have a reference for this statement, please, include it.
Line 275: Please, substitute “it was also necessary to include” by “we included”. It is much simpler.

Lines 277-279: I suggest joining this sentence with the previous one substituting “In other words, this variance stricture allows” by “as it provokes”.

Lines 287-289: I don’t understand this sentence. What do you mean with “to interpret” the species composition axes”? and why is it necessary to examine the correlations between each axis (I guess NMDS-1 and NMDS-2), and the Community Weighted Mean of wood density (CWM-wd)?. Please, clarify, and if this analysis does not give any relevant information, please, remove it.

Results

Lines 297-298: I suggest removing this sentence.

Lines 300-302: Please, join these two sentences substituting “Fisher’s alpha at family level” (line 301) by “whereas at family level it”. Give the reference to Table 1 after “SCC” (line 302), it makes easier for the reader to identify the specific result you mention here.

Line 303: Please, replace the reference to Table 1 and Figs. 2a-c after “levels”.

Line 303: This is the first time in the manuscript you mention these “partial effects”. You should indicate that you will analyse them much before, in the Methods, when commenting on the GLS models.

Line 304: Please, include a reference to Fig. 3 after “MCWD and HAND”.

Line 316: Here, at the beginning of this section, you should start indicating that Wood density (CWM) was mainly (and negatively!) determined by Soil fertility and Region (Table 1). This is much more important than referring to the supporting information. I already indicated that in the former version of the manuscript (see comment on line 338 of ECOG-06125), and you have not accomplished this change. Please, do it or justify why you are not doing that.

Line 319: I will include this reference to Fig. 2d at the beginning of the paragraph, together with that of Table 1 (see comment just above). Please, start commenting the most important results (those in figures included in the main text) and then the results in those figures given in the supporting information.

Lines 320-222: I will include this sentence, referring to Fig. 3, just after the one referring to Table 1 and Fig. 2. Afterwards, you can refer to the less important and/or non-significant results included as supporting information.

Line 328: Please, substitute “at least half” by “> 50 %”.

Line 331: In this sentence you mention SCC but this variable is written as “Soil Fertility” in Table 3. Please, use the abbreviation (SCC) or “Soil fertility”, but just one term. I suggest using SCC all over the main text and in all figures and tables, indicating in their captions, that these abbreviations refer to Soil Cation Concentration (SCC). In the Methods, you have appropriately indicated that indicating soil fertility was taking into account by measuring Soil Cation Concentration (SCC; lines 183-184).

Line 332: Again, use SCC, Soil fertility or Soil Cation Concentration, but just one (and the same) term, always.

Line 333: I guess you mean a significant interaction between HAND and MCWD. Please, correct it, and refer to Table 3 “(Table 3)” just after “axis”.

Lines 338-341: Why did you perform these correlations between the NMDS axes and CWM-wd? You did not explain that in the Methods, but in the Discussion (see comment in line 457). I suggest removing this analysis and it results. Anyway, if you consider them important, you should; (1) include a couple of sentences to justify the analysis of these correlations in the Introduction, and (2) write that you will analyse them in the Methods.

Lines 341-345: See comment in lines 246-249.

Discussion

Lines 354-355: This sentence is very vague. How does species composition was affected by local hydrological conditions? According to Table 3 it seems that it varies depending on Soil Cationic Capacity (SCC).

Lines 360-361: Are you sure that there are no previous studies on the effects of hydro-topographic conditions (HAND) and climatic water deficit (MCWD) on diversity at large-scale, in wet tropical forests, and particularly, in the Amazon basin?

Lines 368-373: I recommend joining these two sentences removing “that can be stressful to plants. These effects can decrease” and substituting this sentence by “decreasing”.

Lines 385-387: If the differences between the upper (and well-drained) and lower areas in terms of species diversity were not significant at all, I suggest; (1) removing this sentence, (2) adding one sentence more explaining such result.

Line 401: What do you mean with “higher dynamics”? According to the mention to disturbances in line 405, I guess you mean that higher (upper?) and well-drained areas are more prone to human-induced disturbances and therefore, that forests on them are more dynamic? Please, clarify.

Lines 411-415: This sentence is awkward. The first part (lines 411-413) is not apparently related with the second one (413-415). Moreover, according to what you say in the next sentence (lines 415-416) I guess that instead of “the effects of soil fertility” here you are just referring to “soil fertility”. Don’t you?. That is, to studies indicating that disturbances positively affect tree diversity. Please, clarify and re-write accordingly.

Lines 417-419: Please, include this explanation earlier in the paragraph, when you first associate disturbance to soil fertility (line 413). It would help to make the paragraph in lines 411-419 more clear and structured.

Lines 422-425: I am not sure that your results indicate that. More importantly, according to Table 1, Soil fertility (Soil Cationic Capacity; SCC) is the most important variable to determine changes in wood density (CWN-wd). However, Soil fertility is not mentioned in all this section (lines 421-451)!!. Please, clarify and correct.

Lines 426-427: You mention “Amazon” twice in the same sentence, please, simplify this sentence deleting one of them.

Lines 429-433: This is a very long and complex sentence. Please, split in two starting with the agreement (lines 430-433) and continuing with the disagreement (line 430) with your results.

Lines 433-435: If you did not detect the same effect as that referred in Blanchard et al. (2019); (1) do not cite this article at all or (2) cite it, including a short sentence to explain why you could have not detected this effect.

Lines 446: Please substitute “in addition to” by “as well as”.

Lines 446-449: Do you mean that species common in upper, well-drained areas do usually have a greater resistance to water stress and drought (refs) and root depth (ref)? If affirmative, please, write it properly.

Lines 449-451: Maybe I am losing something, but it seems that soil texture is not important as a driving factor of changes in CWM-wd (Table 1). Then, I suggest removing this sentence.

Lines 452-454: Please, may you provide a reference for this sentence?. Moreover, I think it will be better placed in the Introduction, in order to justify the analysis of the changes in species composition as a consequence of local hydro-topographical conditions.

Line 455: What does hydro-topography mean? Please, mention the specific variables which appear as more significant to determine changes in species composition, within each region, in Table 3. Namely, Soil fertility for CA-EA, and MCWD (hydro-topography?) for SA and WA. Note also that soil fertility is the second most important variable for SA and WA.

Line 457: Here is the explanation of the correlation!! (lines 338-341). But I should have explained in the Introduction. Otherwise, I am not sure this correlation demonstrates that the filtering of species traits such as wood density is the mechanism of species sorting... I honestly suggest removing these analyses from the manuscript.

Lines 459-464: I don’t understand these sentences. In lines 459-462 you refer to the fact that in a specific region, the species composition changes along the hydro-topographic gradient, and then in lines 463-464 you refer to a particular effect… on wood density. To which of your variables of interest are you referring on this paragraph, to species composition or to wood density? Please, clarify and re-write accordingly.

Line 465: With “hydrology” you mean “hydro-topographic conditions”? If affirmative, please, correct. And be consistent in the use of terms.

Line 467: Do you mean “the larger variability in tree composition occurring in this comparing to other regions”? If affirmative, please, correct.

Lines 472-474: Yes I agree. But based on this sentence some of the statements made above could be very speculative!. I strongly recommend re-structuring the full text between lines 421-474 explaining what are the most important variables determining changes in wood density and species composition all over the Amazon (and in the regions considered), separately.

Tables and Figures

Lines 874-885 and 887-889: I have wondered on this issue all the time while revising the manuscript. This information should be given in the Methods, not in the caption of Table 1.

Table 1: Please, in the first column of the table you should indicate that “Species”, “Genus” and “Family” refers to Fisher’s Alpha (e.g. FA-species, FA-genus, FA-families). Instead of “Wood density”, use CWM, which is the variable considered in the analysis. I already indicated when reviewing the first version of this manuscript (see comment on lines 390-394 of ECOG-06-125), but you neither did it nor provided a justification.
Table 1 and Table 3: Please, in the tables, include the names of the specific variables you have used (e.g. Soil Cationic Capacity; SCC), not the names of the general variables they represent (e.g. Soil fertility).

Table 2: I strongly suggest placing this table as supporting information, because it is just mentioned once in the manuscript and it is, in my opinion, not important at all.

Table 4: I would remove this analysis from the manuscript. If you consider it is important, please, provide a justification for it in the Introduction, mention it in the Methods, and include this table as Supporting information.