the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Mar 2025
Comparative analysis of nutrient concentrations in generalist and specialist tree species on clay and sandy soils in the Central Amazon
Abstract. Tropical forest soils generally have low nutrient availability. Some species exhibit specialized behavior, occurring exclusively in a single soil type, while others are generalists, thriving across different soils and water table depths. This study assessed the influence of topographic variation on leaf and trunk macronutrient and carbon amounts of tree species occurring only in one topographic position and species occurring across topographic positions and their relationship with soil macronutrient and carbon stocks. We selected nine species occurring in different topographic positions: three plateau specialists, three valley specialists, and three generalists (with four replicates each, totalling 35 individuals), where leaf and trunk samples were collected from each individual, and soil samples for carbon and nutrient analysis and quantification. Leaf and trunk nutrient concentrations varied across specialist and generalist functional groups, with valley specialists showing the highest concentrations of leaf and trunk nutrients and carbon. Nutrient concentrations within generalists remained consistent across topographic positions, underscoring their adaptive strategy to sustain productivity across environments. The concentrations of certain trunk nutrients of plateau and valley specialists and generalists mirrored those found in leaves, albeit at lower relative concentrations. Trunk carbon concentrations did not vary significantly compared to leaves, suggesting that other biological or environmental factors influenced tree nutritional status. We found evidence of variations in plant carbon and nutrient concentrations between generalist and specialist species inhabiting plateau and valley habitats in Central Amazonia, and a weak correlation between the stocks of some soil nutrients and leaf and trunk nutrient amounts.
- Preprint
(2241 KB) - Metadata XML
-
Supplement
(728 KB) - BibTeX
- EndNote
Status: open (until 31 May 2025)
-
RC1: 'Comment on egusphere-2025-391', Anonymous Referee #1, 02 Apr 2025
reply
I thank the authors for their detailed and relevant look into nutrient cycling in tropical forest and appreciate the hard work that has undoubtedly been done to achieve the results presented. However, I have many remarks throughout the text on phrasing, word choice, nuance, and more importantly, on the quality of the scientific analysis, discussion, and the data analyses. Nonetheless, I believe the expected changes/clarifications are feasible, which is why I would reconsider the manuscript after major revisions. Please find my general remarks below; for the detailed and textual remarks, I refer to the attached annotated pdf (please do not mind what is written on the first page - I noted those comments more clearly in the 'major general comments' below). The reason for providing the annotated pdf is that there were many textual issues that would simply take too much time to write in this reviewer comment.
Major general comments:
- The title does not represent the core of the study: at first, it seemed the study was about the difference between specifically sand- or clay soils, whereas the focus was actually on topography.
- Did you use pith-to-bark samples for basic wood density? You always mentioned the diameter of the wood samples you collected, but not their length (thus also not if they were pith-to-bark cores). Additionally, these samples are usually collected with 2 replicates per stem, to account for asymmetry in the stem's shape. Note that you can correct for not using pith-to-bark sensu https://bsapubs.onlinelibrary.wiley.com/doi/full/10.3732/ajb.0900243).
- Additionally, please explicitly mention how you treated the bark - was it included in measurements or not? This is also relevant for the wood nutrient analyses.
- The methodology of submerging the cores in water for 20 days has certain limitations: you probably over-estimate the fresh volume, so mention that one can also use conversion factors sensu https://bsapubs.onlinelibrary.wiley.com/doi/full/10.1002/ajb2.1175.
- Add more info in the methods section. The authors often refer to a protocol used by the lab, but the reader would have to see which steps were taken in the sample preparation, extraction methods, which machines were used to analyze the contents, ....
- The statistical analyses were carried out or discussed (or both) to a limited extent. The paragraph in the M&M section discusses the check of normality, but not of homoscedasticity. Additionally, and more importantly, the authors mention ANOVA as the method to test "whether leaf and trunk nutrient- and C concentrations differ between the generalist and specialist species," or Kruskal-Wallis in case the normality assumption was violated. The authors need to acknowledge that they have a categorical variable 'tree functional type' with 3 levels: valley spec., plateau spec., and generalist. ANOVA can only tell you if at least 1 of these levels differs significantly from the others, not if which levels differ from each other. For the latter comparison, you would need a suitable post-hoc test. Nonetheless, e.g. Fig. 4 shows the 3 factor levels with specific indications of significance marked as '*', which would have to mean post-hoc testing has been conducted. If this is the case, please mention in the M&M section. Additionally, Fig. 3 reports t-test results, while the method was not mentioned in the M&M, nor the assumptions to be evaluated. Finally, Fig. 6 shows Spearman correlations with p-values. Which statistical test was carried out? And why was a non-patametrical correlation coefficient used?
- Further, for now, the discussion of the results is quite limited and a bit 'shallow'. I believe the results are valuable, but require better understanding and more consistent and robust analysis. It is, e.g., important to acknowledge that you are not looking into soil nutrient availability, only to total nutrient stocks. Hence, is it surprising you don't find strong correlations with tissue nutrient contents?
- I'm still a bit unclear to what the actual impactful result is you found. As mentioned before, the discussion contains many general statements that are not necessarily informative. Following is an example from the abstract: "Trunk carbon concentrations did not vary significantly compared to leaves, suggesting that other biological or environmental factors influenced tree nutritional status." --> What do we learn from this? Is it maybe simply to be expected that wood vs. foliar carbon contents vary to a different extent? How is this linked to the 'tree nutritional status'?
- Additionally, in the discussion, always start with your results, and then put them in a context using literature.
Additional comments were written on the attached document, close to the corresponding sentence in the MS. Please also formulate a reply to these comments/questions in case of resubmission.
Minor general comments:
- Use 'wood' instead of 'trunk' or 'trunk woody material'
- Use 'content' instead of 'concentration' for mass-based measures (µg /g etc)
- Add a soil properties description table to the site description or results.
- The colors in Figures 3, 4, 5 are not particularly meaningful
- Make a clear distinction in terminology between plant functional types, often defined in the context of pioneer trees vs. shade-tolerants etc. Just to avoid confusion.
- Make sure to use abbreviations in full only at first mention: e.g., 'carbon (C)' --> Later only 'C'
- Fig. 3: plot means with diamond shapes, make y-axis labels bigger and simpler: 'Soil C (g C kg-1)', make sure to add the specific element after 'g' as in this example, why are they called 'average' concentrations? (similar remarks for Fig. 4 and 5)
- Fig. 6: what is the shaded area?
- Fig. 7: improve names: "Ca_Soil_Mg.ha" --> "Soil_Ca", put units in caption + What is the logic behind the colors?
- Why do you note down doi links for a few references only?
- Fig. S1: use different colors for different tree categories (--> how spatially separated are the classes actually?) + add classification what you define as valley and what as plateau. + relative height is relative to what?
- Fig. S2: I count 13 trees for valley specialists
- Fig. S3: Give info only for relevant species
- Generally: what do the confidence intervals indicate? How were they constructed?
- Is there no suppl. fig. 6?
- see comments in attached file for Suppl. Fig. 7 & 8.
-
RC2: 'Comment on egusphere-2025-391', Anonymous Referee #2, 28 Apr 2025
reply
In this study, the authors ask how environmental variation drives specialist and generalist species distributions in a terra-firme tropical forest in the central Amazon. They examine how tree nutrient tissue concentrations across specialist and generalist species are distributed along a topographic gradient (valley vs. plateau) that varies in soil nutrients. The authors hypothesize that specialists and generalists differ in their tissue nutrient content which could explain their distributions. In addition, they hypothesize that wood density is a strong predictor of wood nutrients, and that wood nutrients correlate with soil nutrient availability. They selected three plateau specialists, three valley specialists, and three generalists, replicating four individuals per treatment for a total of 35 individual trees, collecting samples for leaf and trunk measure carbon, nitrogen, phosphorus, calcium, potassium, and magnesium in addition to soil measurements.
In summary, they found that specialist species restricted to the valley had the highest leaf and wood nutrient content, reflecting the soil nutrient availability, and that nutrient concentrations within generalists remained relatively consistent across topography, suggesting adaptability across nutrient gradients. They also found coordinating between leaf and wood nutrients, but no relationships between wood density and nutrient content.
These are important questions as we know very little about nutrient content in wood, despite wood pools being a very large sink of nutrients in highly weathered tropical forests (e.g. Bauters et al., 2022 and Dalling et al., 2024). In addition, we know very little as to how leaf and wood tissues are coordinated, and how well tissue nutrient content reflects soil nutrient availability and bioavailability of soil nutrients to plants.
Overall, I have some major comments to improve the interpretation of the findings.
- Contextualizing the importance of the study: I think that one of the main arguments for this study is that we understand that soil nutrient availability can drive species distributions, but the mechanisms are relatively unknown. In addition, nutrient stocks in leaves and wood are often understudied, but could help explain species distributions and how species are able to persist on soils with low nutrient availability via changes in nutrient-use efficiency. This background could be set up a bit better; currently the first paragraph doesn’t introduce what the study is focusing on, which is on tissue nutrient concentrations.
- Data presentation: I understand the value of focusing on stocks, but there are a lot of extrapolations made to estimate these stocks (lines 214-230). To understand how soils and tissues are coordinated, it would be very helpful to see scatterplots such as in Heineman et al. 2016 with tissue concentrations, similar to Figure 6. You could make side-by-side panels with concentration data and stock data, for example. In addition, for the correlation matrix, the same could be done with concentrations rather than stock data; for example a correlation matrix between leaf and wood nutrient concentrations, soil and leaf nutrient concentrations, and wood and soil nutrient concentrations. Currently, it is not super clear based on the labels in Figures 7 and 8 what properties are being compared/correlated.
- Data presentation: It might be easier to interpret the differences between valley vs. plateau specialists compared to generalists if information from Figures 3, 4, and 5 were integrated more. For example, since there are no differences between C and Ca between the valley and plateau soils, this is important to remember when looking at Figures 4 and 5 when comparing valley and plateau specialists vs. the generalists. I might try a multipaneled figure where the soil, wood, and leaf data are presented across columns and the different nutrients across rows or something like that. Similarly, the results can also integrate the soil and tissue concentration data; for example, “Consistent with the higher Mg concentrations in the valley, wood and leaf Mg concentrations were also higher for valley specialists”
- Explaining species selection: why were these nine species selected, and what is the relative abundance of the selected species at the site? Why were the species selected ranging in diameter classes? More context as to how specialists vs. generalists are defined would help; for example, if they are common and abundant, these characteristics may be more reflective of generalist species.
- Nutrient availability at the site: First, it would help to contextualize general nutrient limitation at the site: are these soils low in base cations or phosphorus? This could help understand which nutrients may be more critical for structuring community composition. The setup for distinguishing valley vs. plateau soils could also be structured better, and topography as a driver of soil nutrient availability should be introduced in the introduction. In addition, the authors need to explain how the soils were measured and what extractants were used; total or “available” fractions for example, as well as the tissue concentrations: were they ashed at 500C and digested in aqua regia? Reporting the concentration rather than the stock data would also help contextualize how nutrient concentrations in soils and tissues vary across sites. For tissues, C:N ratios might also be more helpful in textualizing N rather than C and N alone, for example. Another aspect could be to examine tissue relationships with the topsoil only, and to show depth distributions of soil in the supplement to see how much the soils vary across depth rather than integrating the entire top 50 cm.
- Why was the growth rate (periodic annual increment) measured, and could an analysis be done comparing nutrient content with growth rates?
- Spend the discussion more on interpreting the results. I would recommend restructuring the discussion around the main findings, discussing whether or not the findings supported your hypotheses (and if they were surprising or not), contextualizing them with previous work, and discussing the generality of these results and what they might tell us about tropical forest nutrient cycling or species distributions more broadly.
Line by line comments:
Title: the title should include topography rather than soil texture
Abstract: The abstract could benefit from a hypothesis: “we hypothesized that across a nutrient gradient, generalists might differ from specialists in their stoichiometric demands for nutrients and carbon.” The abstract could also benefit from explaining the role of topography upfront: how do you expect nutrient availability to change across topography; higher nutrient availability in the valley and lower nutrient availability in the plateaus?
Line 15: Some “tree” or “plant” species?
Line 16: “We” instead of “This study”
Line 20: I believe analysis and quantification are the same thing; you can just write “for total carbon and nutrient analysis”
Line 34: perhaps “lowland” Amazon? The western Amazon is thought to be much higher in soil fertility?
Line 46: “The primary reservoir of essential nutrients for most forest ecosystems is from the soil”
Line 51: another citation could be Dalling et al., 2024
Dalling, J.W., Flores III, M.R. and Heineman, K.D., 2024. Wood nutrients: Underexplored traits with functional and biogeochemical consequences. New Phytologist, 244(5), pp.1694-1708.
Line 56: do you mean, serving as a long-term nutrient reserve?
Line 60: perhaps “experiments testing tree responses to P limitation”?
Lines 60-62: this sentence is very confusing; it is not clear how maximum electron transport rates are related to foliar nutrient concentrations and wood density
Line 92: “trunks”
Line 100: perhaps describe the site in a more general pattern: “the study was conducted in tropical forests in the central Amazon Basin…”
Line 106: describe the elevational changes, 47-114 m?
Line 133: list the ranges of sizes of the trees so we know can estimate their canopy position. Also explain why a variety of sizes were sampled?
Line 138: Brosimum rubescens should be italicized
Line 141: just “associated” unless there are N fixers that are not in the Fabaceae family at your sites
Line 143: maybe just diameter growth instead of “PAI”? you can summarize it as DBH t2- DBH t1/ (t2-t1) instead of “current”
Table 1: might help to list the initial size or report the relative growth rate since interpreting a change in the DBH increment also is relative to the size of the tree
Lines 170-178: how were the samples extracted or digested? Are the concentrations estimated as total nutrient content?
Line 182: “sampled” instead of “opened”
Line 193: “Soil bulk density”
Line 219: “o” is a typo?
Line 256: “were significantly higher” rather than “presented significantly higher”
Line 262: Referring to Figure 4 here?
Lines 331-333: nutrients also accumulate in valleys because of gravity and weathering. I would be hesitant to argue that valley soils typically have higher nutrient availability because of litterfall decomposition (also lines 418-425), as the landscape erodes into the valley.
Berhe, A.A., Barnes, R.T., Six, J. and Marín-Spiotta, E., 2018. Role of soil erosion in biogeochemical cycling of essential elements: carbon, nitrogen, and phosphorus. Annual Review of Earth and Planetary Sciences, 46(1), pp.521-548.
Line 341-342: Soong et al. (2021) did not attribute the higher growth rates in the clayey soils solely due to the greater prevalence of AMF?
Lines 351-354: the growth rates might need to be presented as relative growth rates since it’s hard to interpret the growth without knowing the size of the trees
Lines 363 to 367: why not instead that the nutrient content mirrors the environmental variables to some extent?
Line 413: Bauters et al. (2022) likely found correlation between nutrients in tissues probably because of the soils, not because of the internal demand
Citation: https://doi.org/10.5194/egusphere-2025-391-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 98 | 32 | 7 | 137 | 14 | 6 | 6 |
- HTML: 98
- PDF: 32
- XML: 7
- Total: 137
- Supplement: 14
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1