Floristic Groups, and Changes in Diversity and Structure of Trees, in Tropical Montane Forests in the Southern Andes of Ecuador

Round 1

Reviewer 1 Report

The article has been almost completely rebuilt compared to the previous version, and most of the changes suggested by reviewers and editors have been incorporated. We suggest accepting the article for publication, of course after the Authors take into account at least some of the stylistic notes and correct the typos noticed in the final version of the text. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript “Changes in composition, diversity, and structure of trees in native forests in the southern Andes of Ecuador” is a very interesting paper that intends to answer two main questions

• Are there floristic groups of trees, differentiated by their floristic composition?
• Which environmental or predictor variables control the variation in tree diversity and structure?

Regardless of different detail issues, I sincerely believe that this article presents an important deficiency in the presentation of the results. In this sense, there are, in my view, three aspects to be corrected in a first review of the manuscript that refer to:

Figures and Tables must be a fundamental tool in order to facilitate both the reading of the results and the understanding of the related arguments

In this sense Figure 2 must include the 28 sampling points (currently only 23 can be counted), and also show their numbering (1-28) so that they can be associated with their location in Figure 1.

I think that is necessary to include a matrix of results with the local values ​​of the different variables for the 28 plots (replacing Table 1 or as a supplementary Table if it takes up too much space) including the values ​​of the spatial correlation index.

The authors do not to include in this manuscript, unlike the work published by Jadán et al. [29], the results of the other two indices of diversity calculable with abundance data (exponential of the Shannon index and inverse of the Simpson index). I think that their interpretation in the context of Hill numbers [total species (0H or richness), common species (1H or exponential of the Shannon index) and abundant species (2H or inverse of the Simpson index)] could be an argument for support the results both the floristic groups both the indicator species. (see by example Chao et al 2012 Proposing a resolution to debates on diversity partitioning. Ecology, 93(9), 2012, pp. 2037–2051

Finally I encourage you to give a suitable and readable format (same number of decimal places, etc) and reviewing the results of the correlations in Table S1 correcting if necessary (I guess that same values were displaced in rows and/or columns ​​). Please, also comment on why among the correlated variables the ones that have been chosen have been chosen (bear in mind that some of the selected climatic variables have been calculated using a model and are themselves subject to a high level of spatial autocorrelation).

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

First of all, thank the authors for their quick response, with the changes that have resulted in the original manuscript. I would not like to get too long in my comments, but I am afraid that some will be necessary in my opinion for the improvement of the manuscript.

As I am a second-round reviewer, I cannot stop at the initial trajectory of the manuscript and it is possible that some of my comments are contradictory with those of other reviewers or the scientific editor himself. However, and in view of the starting data in table S2, and the results shown in tables and figures throughout the manuscript, I wonder to what extent the systematic application of the different statistical procedures to the results of the field sampling, it clarifies, or on the contrary it masks, the ecological reality of the forests under study.

• With respect to the floristic groups found, the statistical procedure establishes an ideal number of groups for the number of starting samples and the relative similarity between them. In this case, three groups are established for 28 samples, although in a previous work (Jadán et al 2017) three groups were also established, in that case for 20 of the 28 samples analyzed in this work. Summarizing, broadly speaking, two of the three original groups have been merged in the recent analysis (Communities I and II) which becomes Group I, the old Community III is the new Group 2 and a new group has been delimited (Group III ) formed by 5 of the 8 samples added (those corresponding to mature forests). I suppose that by adding new samples (the cartography collects 100 forest patches in the study area according to the authors' criteria) we could reach either more groups or the same "ecologically" better defined groups.
• Regarding the indicator species of each group. The two species from Group 1 were already indicative of Communities I and II, three species from the former Community III are now representative of Group 2 (with another species joining Group 2), while one species from Community III It is now Group 3, which is also characterized by two new species. From the table of indicator species (Table 2) I have omitted all those represented in the sample by less than 10 individuals (despite their significant p).
• With respect to the Non-multidimensional Metric Scaling Analysis (NMDS) the first axis is related to two highly correlated variables in a negative way. When the annual mean temperature increases, the values ​​of the spatial correlation index (negative values) are lower. This axis orders groups 1 and 2, with group 1 corresponding to the areas of lower altitude and the most recent agricultural abandonment. In group 2 there are the highest values ​​of spatial correlation, corresponding to the other end of the axis, and are related to plots in areas with higher altitude, lower mean annual temperature and intermediate age of abandonment (around 100 years). The other orthogonal axis is characterized by increasing precipitation in areas with positive spatial correlation and where the concentration of manganese (¿?) In the soils is clearly higher than in group 2 and slightly higher than the mean of group 1. This group is It is also characterized by being that of the oldest forest plots (they are classified as mature forests).
• With respect to the GLM analyzes, the authors do not provide the data of the explained deviance. The adjustments for the heights, both total and dominant, present a very poor significance for the constants, which suggests that their predictivity is very low. The adjustment for richness is based on a spatial variable related to the geographical position of the plots in relation to the coordinates of the proximity matrices, which is difficult to interpret. Finally, the results for the basal area are the ones that seem most coherent, although it would be necessary to contrast with the behavior of the age variable (r = 0.83 in table S1).

Once these observations have been made, I would like to dwell on two aspects that have not been highlighted throughout the text. The variable age of the mass or time elapsed since abandonment (AGE) is fundamental to explain most of the dasometric variables, and even the richness of species. Since it is a variable that is not available in many cases, if the authors consider that its values ​​are adjusted, they should try to use it (although in the previous statistical analyzes it has been relegated by another correlated variable such as MAT - remember also that this variable has been obtained from a model and not by direct measurement-). Regarding the spatial correlation, it would be interesting to carry out the study for each response variable and see its relationship with the predictor variables.

This in relation to general considerations, with which the authors may or may not be in tune. Regarding more specific aspects, but that if they are not corrected they do not make the manuscript suitable for publication, I will list the following

Page 1 The abstract has a few inaccuracies and sentences that cannot be corroborated when reading the manuscript. I offer an alternative writing.

Abstract

Tropical Montane forests in Ecuador have a highly variable structure and floristic composition of trees in response to ecological gradients. At the local level, these traits are a result of ecological synergies between vegetation, environment, and spatial influences. In this context, this study identified floristic groups and evaluated how the structure and tree species composition correlate with environmental variables taking into account the spatial correlation. We also identified the indicator species within floristic groups. For this porpoise, we identified and measured diameter and height of all trees with a diameter (DBH)> 10 cm, in twenty-eight 500 m² plots, located at elevations between 2900 and 3500 m. We found that floristic composition was explained by climatic variables, indicating similarity to the niche assembly but also with the age of stand, which confirms the succession process. The richness and structure of the arboreal vegetation were influenced by complex associations between variables such as precipitation, temperature, and soil nutrients concentration. We concluded that in fragmented landscapes where there are particular environmental and spatial influences, it is possible to find different floristic groups that represent high variation in species composition within the mountainous region of southern Ecuador. The parameters of the floristic groups were explained mainly by environmental variables, but spatial correlation also has an influence on certain parameters.

 

Other comments

Pag 1 line 34. The citation 1 can not be found as is described in references section. Please complete.

Pag 1 line 42. 1800 ha is an extension out of context. Could you rewrite as percentage or other measurement relative to study area

 Pag 3 line 95. randomly distributed in an approximate area of 30 km2 (Figure 1) immersed in a strip of high montane forest. Eliminate. The sentence is not precise and redundant (see page 3 line 103. From the over 100 forest patches, we randomly installed 28 plots of 20 x 25 m (0.05 ha).

page 3 line 107. Replace 20 x 25 m by 20 m x 25 m or 20 x 25 m²

page 3 line 110 Here there…….. Eliminate the full sentence

page 3 line 121. Include “The total height was calculated through the average of all the trees in the plot”. The dominant …………

page 4 line 139. To have a general taxonomic vision of the study area, we first quantified the number of families, genera, and species. Also, we identified the families with the highest species richness and the abundance for each species. Delete the paragraph, is out of place and is irrelevant.

Pag 4 line 141. Table 1. Descriptive values of response variables and predictor variables recorded in 500 m2 plots in high Andean montane forests.

First there are not response variables in Table 1. Second, this is a results table and could be eliminated here and replaced by Table S2 in the results section.

Page 5, line 163. Diversity was evaluated through species richness. Vegetation parameters for richness and structure such as tree density (n/ha), basal area (m2/ha), total height (m), and dominant height (m) were considered response variables. Replace with Diversity was firstly evaluated through species richness. Parameters of plot structure such as tree density (n/ha), basal area (m2/ha), total height (m), and dominant height (m) were considered response variables.

Page 6, line 168 Adittionaly, we analysed differences in species richness and diversity (the above is redundant)

Pag 6 185-188 “For this procedure, the absolute values of the pair-wise correlations are considered. For two highly correlated variables, the function looks at the mean absolute correlation for each variable and removes the variable with the largest mean absolute correlation”. This is true but a pre-check of the ecological meaning of each variable and the main variables correlation groups, could help us to probe with different variables combinations.

Page 6, line 204 51 genera, 83 species and 34 families. Reorder please (family, genera, species)

Pag 6, line 191. I have already commented on the meaning of this correlation index.

Page 6, lines 225-233. I think that the paragraph could be modified in relation with the my comments about this topic.

Pag 7. Figure 2 captation. The codes of the indicator species are described in table  3. I can not find the codes on table 3. If you are referring to table 2 there are no codes too.

Pag 7 Table 2. Do not include the species with 10 or less individuals.

Pag 8 line 248. Include average and standard deviation of richness for each gropu

Pag 8 line 253. Fig 3 captation. Which line for total height and dominant height in figure 3D?

Page 8, line 255. Delete this line (it has been written in line 248).   The richness of tree species was higher in floristic groups two and three.

Pag8 lines 255-260. This paragraph, as the only one that explains the results shown in figure 4, should indicate the values (mean and standard deviation) for q = 0, 1 and 2 of the three floristic groups. From my point of view, it would be remarkable that for both common (q = 1) and abundant (q = 2) species the three groups reach an asymptote, while this does not occur for the richness of species in groups 1 and 2 due to to the presence of many species with very few individuals, which widens the extension of the confidence intervals. On the other hand, the incorporation of new plots of mature forests (currently 5 in the group with less than 250 total individuals) could vary not only their current richness but also the comparison with the other two groups.

Page 9, Figure 4 caption. Replace by Figure 4. Individuals-based abundance data rarefaction (solid lines) and extrapolation (dashed lines) of diversity based on the Hill numbers q = 0, q = 1, and q = 2 for three floristic group in Montane Forest. The 95% confidence intervals (gray-shaded regions) were obtained by a bootstrap method based on 50 replications. FG: floristic group.

Page 9 Lines 2567-273 and Table 3. See comments about NMDS above.

Page 10 Table 4. See comments about GLM above.

Pag 10-11 lines 295-302. Which is the new information/explanation of the discriminant analysis in relation with the floristic grouping. Dominant height is an indicative of succession stage, as is age of the stand or basal area ....

Page 11 lines 337-338. “In our plots the mean annual temperature varied minimally but, remarkably, explained the composition with greater significance”. This is the kind of sentences relatives to the statistical analysis but not to the ecological one. The variable values are obtained from a continuum model that take into account other variables. One such variable adjust to but not explain the composition. I think that the rest of the paragraph is more ecologically soundness that the sentence commented.

Page 12 lines 358-360. I think that the spatial correlation index used is not definitive to explain the variation in species richness (it could be tried with Moran's i to compare).

Page 12 lines 366-367. The highest richness values recorded in floristic groups two and three, located at lower altitudes,….  I think that must say higher .

Page 12 Lines 371-375. One proof more that temperature is a correlated variable and not and explanatory one of the forests patches development. Basal area is mainly related with stand age (0.83 correlation and 0.00 level of significance in table S1) as you commented at the end of the paragraph.

Page 12. Lines 389-399. This paragraph could be part of the last one, because the three dasometric variables (BA, AH, DH) are mainly correlated with stand age.

Page 13 lines 403-406. If you are agree replace by: The studied forests patches are heterogeneous. Three floristic groups can be recognized by the presence of indicator species and the influence of environmental factors. Structure variables and species richness of the patches are dependent of complex interactions of the climate and soil variables, associated with spatial effects. These patches could be considered as important species reservoir in the planning, ordering and management of the few areas that maintain their forest cover.

Page 13. Last paragraph. Delete.

Supplementary material. Please be sure that the table captions include the abbreviation (and full name) of every variable and this abbreviation (heading of each row and column) corresponds to the values below. Both Tables S1 and new S2 must be corrected. I think that old Table S2 should be the new Table S3.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

I think you have done a thorough and appropriate job of rewriting your manuscript, with results and discussion in line with the data sampled. I congratulate you for it. Only provide two small nuances regarding the writing of the abstract (I attach a new version with slight modifications) and the nomenclature of Table 2 (in order to unify with figure 3)

Abstract

 

Composition, diversity and structure of trees in tropical montane forests are responsive to ecological gradients and local succession.  Those parameters are a result of ecological interactions between vegetation, environment, and location. This study identified floristic groups on mainly secondary forests and evaluated how the composition, diversity and structure of trees correlate with climate, soil, and age since abandonment. We included in our models a measurement of spatial correlation, to explore the role of dispersion. For this purpose, we measured diameter and height of all trees with DBH ≥ 10 cm, in twenty-eight 500 m2 plots, in an elevation range between 2900 and 3500 m.  We found 14 indicator species in three floristic groups. Group composition was explained by age since abandonment, which showed strong succession effects. Mean monthly precipitation and Manganese, but not spatial correlation, explained plant composition in these montane forests, suggesting a minor role of dispersion. Species richness and structure of the arboreal vegetation were influenced by interactions between age, precipitation, and soil nutrients concentration. We concluded that in fragmented landscapes, within the rugged region of southern Ecuador, it is possible to find different floristic groups that encompass high variation in their composition.

Table 2. Species richness (q = 0), Shannon diversity (q = 1), and Simpson diversity (q = 2) based on the Hill numbers obtained by a bootstrap method based on 50 replications, for three floristic group in Montane Forest. It must be written;  Table 2. Species richness (q = 0), exponential of Shannon index (q = 1), and inverse of Simpson index (q = 2) based on the Hill numbers…….. In the table, in the diversity column it must be written q=0, q=1 and q=2 in each floristic group instead Species richness, Shannon and Simpson.

Other minor commentaries

Page 2 lin2 52 space, change by location

Page 2 line 58 (if the measurement is deforestation the sign is positive) change -3 % by 3 %

Page 2 lines 62-65.  I think that this sentence should be written in past

Page 2 line 71 Are there floristic groups of trees (add) in the remaining forest patches?

Page 2 line 71 Are these indicator species unique… Are there indicator……

Page 2 lines 86-86 The study area is located between 2900 and 3500 m in Azuay province, southern Ecuador. The area is located on the eastern slope of 87 the Western Andes Mountain range. Replace by   The study area is located on the eastern slope of the Western Andes Mountain range between 2900 to 3500 m of elevation.

Page 6 Table 2. It is necessary to be careful with the nomenclature.

 Page 10 lines 342. FG2 and 3. Replace by FG2 and FG3 here and elsewhere of the manuscript ahead.

Page 11 line 396. Floristic composition was not ¿poorly? influenced by spatial correlation

Supplementary files.

Table S1. Correct Richness in rows (it is written Richnnes). I think it could be helpful for reader to remark more significant correlations in bold

Table S3. I think that column 3 must say (SR or species richness) instead TD, column 4 TD column 5 BA, column 6 TH and column 7 DH. Finally SR must be described in the table header.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Review report

According to the authors, the aim of the study presented in this paper was defined by the following two questions:

1) How does the floristic composition, diversity and structure of the tropical montane forest community in Southern Ecuador (Azuay Province) change due to the influence of environmental variation and different successional stages?

2) How are the indicator species of the specific floristic groups related to the different ecological guilds?

The study relates to issues of conservation/regeneration of habitats that are valuable due to biodiversity conservation and ecological stability, i.e. important for the provision of long-term ecosystem goods and services. The topic chosen is relevant and important, and the results presented would certainly be of interest to Diversity readers. The structure of the article, the literature review, the fieldwork methods used and the data analysis are adequately and sufficiently described. The same is true for the discussion and the final conclusion. Unfortunately, the manuscript is very poorly written in terms of English usage and sentence structure and cannot be published in its current form. Significant revisions would be required to achieve publication-level quality. Peer review is not the stage at which this element should be addressed. Therefore, I recommend that the authors improve the language and resubmit the manuscript.

Author Response

Dear reviewer in the attached document we respond and describe in detail what corresponds to the observations requested by you in the first review.

Author Response File: Author Response.docx

Reviewer 2 Report

L37 The high Andean forest is a Tropical montane forest but not viceversa.

L70 " TMF are important for the provision of long-term ecosystem goods and services." The same was mentioned in line 42. It is repetitive.

L70-75 The ideas were mentioned in previous lines; try do not to be repetitive.

L76-79  The predictions should be improved. For the first question

" 1) How does the floristic composition, diversity, and structure of the forest community change due to the influence of environmental variation in different succession status?"

The authors should expose which variables they are expecting to be more relevant in defining the tree community and why. Moreover, they should expose how they are expecting according to the present knowledge, that the tree community will be regarding species richness, species composition or basal area in the environmental gradients where they are sampling.

L100 The plots in the image do not seem randomly distributed. Please, explain why the plots have an aggregate distribution.

L116-120 It is necessary to describe better the methods used to measure soil variables (e.g. Equipment model,  manufacturer, accuracy, etc).

L121-124 The species richness is biased by sampling. Maybe using rarefaction analysis could improve the analysis (See: Chao, A., Gotelli, N. G., Hsieh, T. C., Sander, E. L., Ma, K. H., Colwell, R. K. and Ellison, A. M. (2014). Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species biodiversity studies. Ecological Monographs 84, 45-67 ).

L124. Which guilds were used, and how the information of the guilds for each species was obtained?

L151-153 It is no clear how the correlation index was obtained for each plot.

L155-159 It is not necessary the introduction for the method; You only should indicate the method with the reference and explain why it is functional.

L161 Condition of what?

L171 "51 genera" instead of "51 generous"

L185 "Lower" altitude instead of "the less altitude"

L183-195 The three floristic groups are defined posterior to cluster analysis, which is not correct. It is necessary to test based on predictions, for instance, the groups in the analysis will be formed according to the AAL or altitude.  In the NMDS the groups will be perfectly separated in the floristic groups because it represents the same analysis (similitude or beta diversity).

L201 "Spatial correlation" instead of "special correlation."

L213 Spatial correlation" instead of "special correlation."

254-257 The relation of composition with Mn could not be causal. Instead, it could be correlated with elevation or soil formation.

L337-339 The statement needs a reference.

L347-350 Could the authors explain how to link ecological factors explaining tree community and forest fragments' environmental planning and management?

Author Response

Dear reviewer in the attached document we respond and describe in detail what corresponds to the observations requested by you in the first review.

Author Response File: Author Response.docx

Reviewer 3 Report

The paper by Jadán et al. entitled “Changes in diversity, composition, and structure along a successional and environmental gradient in endangered Andean montane forest. Local-scale analysis in the southern region of Ecuador” refers to patterns in composition and structure of secondary Andean montane forests shaped by different environmental conditions and the stage of succession.

Major comments

One of the most serious flaws, which disqualifies this paper, is the way it has been written. The text is full of grammatical mistakes, typos (even in latin names) and wrong vocabulary. The number of language errors makes the paper extremely difficult to read and thus follow the Authors’ argumentation. It cannot be published until it has been completely linguistically corrected and verified by a native speaker.

The style of the paper causes that its very important parts such as abstract, conclusions, and discussion must be extensively improved – they are full of very general statements and repetitions in the places where sentences, which are more precise and more related to the results are required. Although, the methods were correctly chosen and applied, some fragments of this part should be completed and/or further explained – the same refers to the results.

With reference to conclusions’ part: I did not find any fragment in the text which would link the presented findings with conservation status of studied forests mentioned in the conclusions. Please explain this. The Authors mention that “succession age associated with the geographical distance and climate variables can explain response variables.” Apart from the construction of this sentence, can you explain why the models which you used did not confirm this conclusion? According to the presented results, any of the response variables was correlated with age (Table 3). Conclusions should be re-written to make them compatible with the results and what is the most important with hypotheses presented at the beginning of the paper.

Minor comments:

Due to the large number of comments referring to the main text, I decided to include them directly in the pdf file.

Comments for author File: Comments.pdf

Author Response

Dear reviewer in the attached document we respond and describe in detail what corresponds to the observations requested by you in the first review.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

General note

The manuscript submitted for review has been linguistically corrected throughout the text. Unfortunately, it is still of very low quality in terms of English usage, which makes it difficult to understand and follow. I must uphold the recommendation from my previous review that significant linguistic improvement is needed to achieve publication-level quality. I encourage authors to ask a native speaker to assist them in this task.

Below is an example of the linguistic correction of the abstract and some comments on its content.

Linguistic correction of Abstract

Abstract: Montane forests have highly variable composition and structure in response to ecological gradients, but also to their successional stage. At the local scale level, these characteristics traits are a result of ecological synergies between vegetation, environment, and succession age. In this context, in this study we identified floristic groups of arboreal vegetation and investigated evaluated how their composition, diversity and species structure correlate with environmental variables and forest the age of successional age. of the forests. We also identified the association of species into floristic groups, and assessed evaluated the presence of indicator species in ecological associations in relation with the stated predictors. variables. To do For this, we identified all trees with a diameter greater than 10 cm and measured their DBH and height within twenty-eight plots located at elevations between 2900 and 3500 m a.s.l. and with a specific chronosequence between 14 and more than 100 years after abandonment. measured (DBH, height) and identified all trees with a diameter (DBH)> 10 cm, within twenty-eight plots, located at elevations between 2900 and 3500 m, and with a determined chrono sequence between 14 and > 100 years from their abandonment. We found that structural and compositional parameters of the arboreal vegetation were influenced by the climatic variables and the spatial correlation. Two floristic groups were the product of "late succession", and they emphasized long-lived heliophytes and sciophytes as indicator and exclusive species long-lived heliophytes and sciophytes. We conclude that climatic predictors and spatial correlation, in addition to altitude and the age of succession, largely explained the variation in the parameters of the arboreal vegetation parameters in the mountainous region of southern Ecuador. Indicator species were not reliable solid indicators of succession, but the spatial correlation in combination associated with different environmental and edaphic variables played a very important role in the establishment assembly of forest communities.

 

Comments to corrected Abstract

Abstract: Montane forests have highly variable composition and structure in response to ecological gradients, but also to their successional stage. At the local scale, these characteristics are a result of synergies between vegetation, environment, and succession age. In this context, we identified floristic groups of arboreal vegetation and investigated how their composition, diversity and species structure correlate with environmental variables and forest successional age. We also identified the association of species into floristic groups, and assessed the presence of indicator species in ecological associations in relation with the  predictors. (Comment: what predictors? It should be explained in more detail which variables were examined as predictors) To do this, we identified all trees with a diameter greater than 10 cm and measured their DBH and height within twenty-eight plots (500 m² each) located at elevations between 2900 and 3500 m a.s.l. and with a specific chronosequence between 14 and more than 100 years after abandonment. (Comment: it is not clear here what you mean by "abandonment". Please clarify what the term is about) We found that structural and compositional parameters of arboreal vegetation were influenced by climatic variables and the spatial correlation. Two floristic groups were the product of "late succession", and they emphasized long-lived heliophytes and sciophytes as indicator and exclusive species (Comment: sentence is not clear). We conclude that climatic predictors and spatial correlation, in addition to altitude and of succession, largely explained the variation in arboreal vegetation parameters in the mountainous region of southern Ecuador. Indicator species were not reliable indicators of succession, but the spatial correlation in combination with different environmental and edaphic variables played a very important role in the establishment of forest communities. (Comment: the last two sentences are repetitions or include repetition of a meaning of the previous sentence. Instead, please explain which predictors influenced the vegetation parameters and to what extent)

Reviewer 2 Report

The authors followed the previous comments and observations and they improved the paper.

Reviewer 3 Report

Although the authors accepted most of my comments concerning language they did not answer for any of my question or doubt expressed in the review or in the form of comments attached to the manuscript. In fact, their response letter was brief, rather in the negative sense. However, it encouraged me to review the current version of the manuscript even more thoroughly.

Firstly, although the authors improve wording in some places, the paper still needs the extensive English improvement. I strongly insist on competent correction of the language by a native speaker with an experience in academic English. The latter condition is the more important, because of using by the authors the vocabulary typical of spoken English or phrases which were probably literally translated from their mother tongue. Apparently I didn't make it entirely clear that because of the numerous linguistic mistakes, the article should be rewritten in correct English. It probably cannot be corrected sentence by sentence as the Authors attempt to do.

Secondly, the substantive part of the introduction and methods were supplemented with some important paragraphs for the understanding of the article, but the following issues still remain:

• The constant use of terms such as "floristic composition", "floristic groups", etc., erroneously implying that the entire vascular flora (including herb species) was studied, while only trees larger than 10 cm in diameter were examined. This misleads the reader in a factual and not only linguistic way. Therefore, even research questions (hypotheses) should be formulated from a scratch to better express the real purpose of this study.
• It is still unclear what the authors mean by abundance - one can only guess that they mean the number of individuals of given species.
• I do not understand why authors added the two next graphs in Figure 2 which are practically identical as the previous one. Choose one type of them. I did not receive the answer why continuous variables such as temperatures and precipitations were presented on the NMDS graph as dummy variables not vectors.
• Figure 3 is not related to the topic of the article, and does not contribute anything to it.
• Although the spatial correlation was indicates as the main factor shaping tree composition of studied forests it was not included in Table 1 as explanatory variable.

Third, the paper de facto describes the diversity of secondary forest in a 30 km2 area in the Azuay province and shows the relationship of this diversity with habitat factors. The authors performed a classical ecological-phytosociological study, although on a small number of samples. Aa such this study can be an interesting contribution to vegetation ecology of still poorly known Andean region. However, the objectives that the authors set for themselves - determination of succession changes in the studied area, simply could not be achieved. The paper shows that the main factors influencing the variability of the communities studied are spatial correlation, mean annual temperature and mean monthly precipitation. The additional influence of manganese or copper on the interchangeability of the studied communities may be of a completely random character, connected with a small number of samples, and their analysis in the Discussion is without basis. Moreover, the presented stand age (AAL) ranges for Group I and II partially overlap, whereas the age of Group III, comprising the oldest stands, was estimated using interviews with local people, which does not allow to treat these data as allowing to draw far-reaching conclusions. The influence of forest age on its species composition has not been proved in any way, and all conclusions on this subject, both in the Results section and in the Discussion, are purely speculative. What is more, we are talking about succession when we consider a sequence of organisms (species) following one another over time in the same locality. In no case can we consider succession as a sequence of unrelated objects, in addition showing statistically significant differences in habitat conditions. Similarly, we cannot talk about changes without a documented initial condition of a specific area.

Therefore, I suggest changing the purpose and consequently the title of this study to "Tree composition and structure of secondary Ecuadorian montane forest along environmental gradients", and resigning from any consideration of the influence of forest age on its species composition, including separation of helio- and sciophytes’ groups, as well as chapter 4.2 of the discussion and Table 5.

In other words this paper cannot be published in the current form and must be written from a scratch taking into account new research goals and hypotheses.