Locally Measured Functional Traits Predict Species Registrability in Herbaceous Flora

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Wei et al. present their study on interesting idea about  the use of local field-measured trait data for the explanation of plant global disturbance. Authors extracted species occurrence data from the GBIF, using the total number of global records as a proxy for species’ distribution. Despite that this is a methodically competent work which describes important leaf traits, the authors should consider also the following comments to the presentation of this work:  

1. Results contain many parts which are more appropriate in Discussion (see below) and, in turn, the discussion is a retelling of the results. Discussion is too scarse and contains unsufficient references on recent studies.
2. The next definitions, namely their use by authors, are unclear: «physiological performance» (L.19, L.264) and «physiological strategies» (L.24) Any plant species well adapted to its environment possess morphological, physiological, biochemical and other performance. «physiological strategies» is unclear, ecologists use often «life strategies» or «ecological strategies».
3. Authors specify that «LWC, a trait related to resource storage and drought avoidance, may serve as a key physiological determinant of a species’ ability to establish in a broader range of environments». However, succulent plants a known to have the most maximal values of LWC, but they grow in very severe droght conditions and have local distribution.
4. I am not agree with the statement L. 282-283 «Structural attributes set the physical framework for a plant’s life history, but they do not guarantee physiological resilience to different climates or soil conditions.» It is well known that LMA and LT are related to climate, especially in a globall context LMA and LT are positively related to climate aridity. May be the variation in these traits but not absolute value in one local site could serve as indicator of distribution range?

Some other comments:

L.22 It is not necessary to specify in percents the decimals Replace «52.06%» by «52%»

L.178 I am not agree with the statement «plant height (H) showed a very weak positive correlation with OCC (R = 0.10, P = 0.22, Figure 1a)». There is NO correlation becouse of non significant R! The whole statement L. The same should be made with traits LT and LA L.186-187. It is not «weak psotive correlation», it is NO correlation! Becouse «none of which reached statistical significance» as authors mention on L.188

L.188-190 unclear sentence 

L.200 -202 should be replaced to discussion, and section 3.2. should be deleted (LWC should be combined with other traits in results)

L 228, L.230-232 should be replaced to discussion,

L.235-236 «mass-based stomatal conductance (gs) displayed only a weak and statistically non-significant relationship» is wrong. There is NO relation.

L.237 - 239 should be replaced to discussion,

L.254-259 should be replaced to discussion,

Figure 1 Should be specify what OCC is in signature to figure and why dots are of different colors. What means the line and R, P? The same for figure 2 and 3.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

I reviewed the article entitled “Locally Measured Functional Traits Predict Global Distribution Patterns in Herbaceous Plants”, which addresses how functional traits measured in 144 herbaceous urban species relate to the number of occurrences in GBIF. Overall I really enjoyed reading the manuscript and the rationale behind this work, but at the same time I feel that the data still needs a bit more of rethinking and focusing, since there are some elements that are not sufficiently explain as well as a conceptual mistake that needs to be addressed and would reshape the whole presented idea. After my detailed revision, I come with some suggestions for refocusing their work that are included in my comments. I would encourage the authors to rethink this and present an updated version since the idea is quite interesting and fitted for the journal’s readers.

Before going to my concerns and minor corrections, I would like to highlight some aspects that I find very good explained and very interesting about this work. Firstly, the authors settle their work in the fact that trait-distributed studies use solely a very limited number of traits that were collected from different conditions and maybe with different methodologies, which I totally agree. Traits in databases might not represent the full spectrum of trait values to compare species in a consistent way. The authors collected trait data for 144 species, which is an impressive amount of data. The authors frame this work within “functional biogeography”, a novel field that worth more attention. I agree that trait spectrums or syndromes (multivariate trait data) are definitely important to explain species fitness, and might condition species distribution. Overall, I agree on the frame of the article, the idea is novel and interesting.

 

1. GLOBAL REGISTERS DO NOT REPRESENT SPECIES EXTENT

One of my major concerns is that the article take home message is based on the assumption that the number of registered presences in GBIF represent “global distribution patterns”. This assumption is too simplistic since GBIF data must be considered carefully. To list a few examples:

• Beck et al 2014 Ecological Informatics: Spatial bias in the GBIF database and its effect on modeling species' geographic distributions
• Hughes et al 2021 Ecography: Sampling biases shape our view of the natural world
• Troudet et al 2017 Scientific Reports: Taxonomic bias in biodiversity data and societal preferences

The authors recognised some limitations in the discussion (L349-350) with this sentence: “First, the use of GBIF occurrence data as a proxy for global abundance or distribution breadth is subject to inherent sampling biases, including spatially uneven data coverage and taxonomic inconsistencies.”, which does not explain properly the misconception here.

GBIF is a very powerful dataset of species presence but data is strongly influenced by sampling bias. The authors recognise some limitations of this database: L108-111 “To reduce sampling bias and ensure data quality, we filtered the occurrence records by removing entries lacking geographic coordinates, duplicate records, and spatial outliers (e.g., marine or institutional coordinates). ”. However, the used methods do not correct sampling bias and ensure data quality. The used methods delete incomplete registers, duplicated data and some outliers (I assume based on specific areas). For example, it is very typical that GBIF registers are higher in areas with more human density, since people is more prone to register species that are seen. This could result in duplicated registers if the coordinates absolutely match, but usually it leads to “almost duplicated” registers, this is, a single individual is registered more than once with coordinates slightly different, depending on the GPS precision or the position of the observer. One way to solve this bias is to reduce aggregated occurrences based on a minimum distance. The authors should at least apply this type of correction as well as other sampling bias correction such as accounting for species visibility, etc.

The authors should apply measures to minimize this kind of bias, but most importantly, they should be aware that the number of total occurrences (whay they used as L111“proxy for global distribution extent”) only represents how much this species is registered. Species with more charisma, bigger and associated with humans (pets, ornamental plants) are much more likely to be registered.  Therefore the authors correlations reflect how likely a plant is to be registered.

I would say that this is a very interesting questioning as well. One might ask, do trait syndromes correlate with how likely are species to be registered in GBIF? Which kind of traits make species more “registrable”?

I would ask the authors to rethink their question since OCC is not a proxy for global distribution extend, for a second reason as well; they have not considered the geographic distance between occurrences. Perhaps a species has a lot of occurrences in a specific but very limited area, according to the authors methodology, this species would have a strong global distribution, which might be completely false. Not only geographic distance but also climatic differentiation on areas where species have been registered has been used as a proxy of species plasticity or realised niche breadth. I encourage the authors to rethink this and (1) adjust the proxy of global distribution to something that truly represents it, or (2) shift the question of the article to what makes some species more “registrable”.

I would be happy to discuss more these aspects if necessary.

 

2. MORE METHODOLOGICAL DETAIL IS NEEDED

I would like to ask the authors to include the following information:

• List of the 144 species and a summary of the species types: functional groups representation, invasive/native status, etc. I believe the species list is found in supplementary material of another publication but it should be added as well here.
• Coordinates of the locations where the species were sampled as well as some data showing that the environmental conditions were comparable, and how many days were needed.
• Number of replicates per species.
• Table of the traits defining them and methodological detail. For example TSW was acquired but not measured.
• Light conditions during measures such as stomatal conductance and range of light used for calculating Assimilation rates. I am assuming light curves were performed since the authors present Amax, but please confirm and detail it properly.
• Multivariate regression was performed, but I can’t find which type of functions were tested, only detail on variable’s selection.
• Field surveys were conducted between 2022 and 2023. Does that mean that not all species were measured in the same year?
• How diverse was OCC across species? Perhaps present an histogram and discuss if this truly represents enough variation. Also, I do not understand OCC scale in figures (0 to 6?). Explain.

 

3. Other aspects/minor suggestions

L17-20: “ The results indicate that traits associated with resource acquisition and physiological performance, such as stomatal conductance and leaf water content, are positively correlated with global occurrence, whereas traits related to nutrient storage and carbon content show negative associations.”: Please replace “positive/negative associations” by a more specific term. I don’t understand what the authors mean by this. Do they mean that it is or it isn’t a predictor?

L21-22: “Multivariate analysis explained 52.09% of the variance in global occurrence”. Variance in what? On my first read I thought that might be climatic conditions or latitude. Please explain.

Abstract and title: Consider changing it completely according to my comments above.

L34-36: “While traditional models of species distribution often assume natural conditions, recent studies highlight that species' distributions, particularly in human-modified landscapes, are influenced by factors such as land use, climate change, and ecosystem disruptions”. I would encourage authors to replace “assume natural conditions” by another term that represents better what the authors intended to communicate. Do authors mean perhaps “unaffected by human activities”?

L41-42: sentence “Given this shift in our understanding of species distributions, it is crucial to reconsider how functional traits— measurable characteristics of plants that influence growth, survival, and reproduction—interact with these human-influenced ecological factors”. After reading this sentence at this point of the introduction it sounds like the authors think that what would truly help understanding species distributions is considering human interaction, and how it affects species traits. However, this is not the focus of the article. Please rewrite the introduction to frame your study question.

L 59-60. Similarly to my last comment, after this sentence I thought that the article would be addressing trait variation with invasive and native species, but again is not the focus of the article. This is ok, but then I would suggest rewriting the introduction.

L 61-62 “Such species could therefore accumulate more occurrence records in global databases and attain broader ranges.”. Invasive species can accumulate more occurrences indeed, but not because they can attain broader ranges. Revise considering my comments above about linking geographic distribution with occurrences numbers.

L97: Were all the individual’s collected from an urban environment? How do the authors think this might impact the results? Are the selected species typical of urban areas? Is species selection linked to a question not considered in this work? Considering L112-113 it seems that the sampling of species was meant to represent urban environments. I feel that  changes the main question of the article, and that this information must be on the title/abstract, as well as contextualize it in the introduction, which kind of traits are expected to have urban species?

L122. This is the first time authors say TSW, define it here.

L 130. How was dry weight measured? I imagine 48h at 70 Celsius degrees? Detail.

L137-142. Provide more detail for photosynthesis measurements. Co2 level used, PAR level, Did you do light curves? Etc.. What was the VPD or humidity during sampling? How could that affect the results if measures were taken on different days? Was photosynthesis measured in detached or attached leaves?

Figure 1. The caption requires further detail Define R and P. Why is R and not R2? When significant, use bold for P-values. In my version I see that the points are coloured, what is the rationale behind it? Also, Some lines are not clear lines but cut at some point, did the authors edit the figures? Define OCC on the axis, and its units.

Figures 2 and 3. Same as in Figure 1. In addition, start axis at 0 values and use a break when necessary. Do not leave end-axis open (for example, panel e, x axis in Figure3).

At this point I was expecting a multivariate plot showing trait multidimensional variation and perhaps a third axis with OCC so we could identify if some trait syndromes are associated with higher total occurrences in GBIF. Consider adding this type of analysis (perhaps MDS, PCA, CA..).

 

Table 1. Consider also showing predicted vs real values of the multiple regression analysis. Why plural in “models”? Define B, SE B. Use superindex for the 2 in R2.

 

Discussion. Since I do not agree that OCC represents global distribution, I would suggest the authors to rethink and refocus their work. They have done an amazing work and their results are quite promising as well as their ideas. However, global distribution cannot be discussed based on the current analysis. Specifically expressions like some species with X characteristics “are often more locally specialized”, “ecological flexibility”, “wider geographic distributions”, “global establishment”, are not supported by how the data was acquired in the present version of the manuscript.

 

If useful, I encourage authors to consider these works related to niche breadth

• Higgins and Richardson 2014 PNAS Invasive plants have broader physiological niches
• Higgins, et al., 2012. J Biogeogr A physiological analogy of the niche for projecting the potential distribution of plants.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I thank the authors for their detailed and structured responses to my comments. The authors have done an amazing job at reshaping the manuscript (discussion has been fully rewritten as well as major edits in the title and abstract) and included extra analysis that support the main message. I really appreciate the PCA added by the authors to explore trait syndromes. The authors have added the requested methodological detail.

I do not have further comments.