Morphometric Analysis of a Trapdoor Spider (Araneae, Idiopidae) across Different Brazilian Biomes Reveals the Geographic Variation of Spiders from the Caatinga Biome

Round 1

Reviewer 1 Report

This paper inquiries into the existence of morphological differences among several populations of Idiops pirassununguensis Fukami & Lucas 2008, a trapdoor spider known to occur across the main Brazil biomes (Amazon, Caatinga and Cerrado). The question is addressed through 3 different analytical approaches, based on either linear or geometric measurements (including outlines or landmarks, a technique known as Geometric Morphometrics -GM-) of 64 specimens collected at almost the whole range of the species distribution area. Results show size differences between the Caatinga and both the Amazon and the Cerrado biomes, affecting the whole size, the legs and pedipalps length, the sternum shape and the position of the ALE (anterior lateral eyes). These results lead authors to speculate that differences may reflect either morphological and physiological adaptations to the availability of food and water resources, or the evolutionary and biogeographical history of the species.

GM tools have proved to be useful at revealing otherwise undetectable morphological differentiation in several taxa, particularly spiders. The application of these techniques has been so far relatively uncommon in the analysis of geographic differentiation, and this paper combines this tool with two other techniques (analyses ratios and outlines), likely helping other authors to know any apply the more suitable technique. This is an interesting approach but, at this point, it suffers from important shortcomings.

BACKGROUND, METHODOLOGY AND RESULTS:

1.       The authors develop no hypothesis about which and why do they expect the morphological differences found, nor do they provide detailed information that could sustain their aims, methodology and conclusions.

a.       Environmental information is absent. The Caatinga biome is simply described as arid and opened, and there is no information about the habitat characteristics and diversity within this and any of the other two biomes. Figure 1 shows that biomes are differentially fragmented, and species distribution continuity could vary depending on habitat fragmentation. Together with limited dispersion, habitat fragmentation may greatly affect genetic isolation, and morphological differences may not relate to adaptations or genetic drift, but to phenotypic plasticity (PP), in case populations are not genetically isolated. PP could be a rather likely expectation in case that size or shape variation relate to resources or water availability, but authors do not even discuss this possibility.

b.       The species lifestyle is neither mentioned. This information is particularly relevant to substantiate the exclusive use of males in this study and the likely genetic isolation. Spider males are known to wander in search for females, becoming relatively less sedentary, and the fact that 90% of the specimens collected were males gives room to consider this possibility. Provided that males usually die earlier than females and scarcely never feed, a study addressed at detecting morphological variation among populations related to foodd availability should include both males and females.

2.       The methodology is difficult to understand.

a.       Authors use different measurements and techniques, but it’s not clear why and how do they complement each other. Each technique allows measuring different body parts (the carapace and legs are linearly measured, whereas the sternum and the bulb are not). However, it’s always shape analysis, as linear measurements are used to calculate ratios. Information about size is extracted from GM (centroid size), but this is not clear in the paper. However, it’s only size that varies among populations.

b.       There is no summary table or figure that shows how specimens are distributed across each biome and the distance among the study sites. This is a very important point, provided the different extension and fragmentation of the three biomes, likely reflecting differences in habitat diversity. This information could be calculated from the spatial coordinates at which each specimen was caught that is included as supplementary material, but this is not enough. Moreover, Figure 1 shows that several study sites (almost half) are located at or near the border between two biomes; because analyses are based on averaged population data, it’s necessary to discuss this information.

c.       Specimens came from different collections and were likely preserved. Preservation methods generally dehydrate spiders, and there is room for morphological variation resulting from time and the preservation method used.

3.       Results are poorly presented.

a.       Table 1, which summarizes the averaged linear measurements, does not give  any detailed information on the three particular biomes, nor the sample sizes. The averaged data dispersion is not informative.

b.    Summary data on ratios or the results of PCA analysis on which allows to reduce all linear measurements to a single variable are never provided.

c.       In a study covering such a great geographical area, that only includes 64 males specimens, collected at a rather small number of study sites, it becomes necessary to identify the plots and summarize both biome and plot variation.

4.       Conclusions are not drawn from any hypotheses, nor based on results, it’s just speculation. The main conclusion reflects size variation (smaller, shorter legs and pedipalps….), but information on size is never provided.  

 

MINOR POINTS:   

1.       The title does not reflect your approach and results, as it’s not geographical variation that reveals morphometric differentiation, but just the opposite, morphometric variation could reflect geographical differentiation.

2.       The abstract has to be completely rewritten: (1) a sedentary lifestyle does not necessarily relate to a more or less restricted distribution, which mainly depends on habitat availability and species strict habitat specificity; (2) at this point, the abstract does not include any mention of the particular methods used; (3) your analysis of linear measurements is also addressed at describing the animal shape, through a different techniques, and (4) your conclusions are not drawn from your results.

3.       L33: what “spatially distributed populations”. It’s an awkward expression, as any population is necessarily distributed in the space.

4.       L37: perhaps "body shape and/or size"

5.       L34: you do not analyze any spatial or environmental gradient, I guess. Perhaps this sentence must be rewritten or erased.

6.       L38: not necessarily, they may also result from phenotypic plasticity.

7.       L47: you don’t use two different types of analytical methods, but 3, and also 3 types of measures: linear, outlines and landmarks. Landmarks also give information about outlines, but you use Fourier analysis in this case.

8.       L49: you don’t apply the traditional linear morphometry, which generally does not include ratios but a single linear measure (CW) in the case of spiders.

9.       L50: The term GM is usually spelled as “Geometric morphometrics”, not “morphometry”.

10.   L53: the term “shape” is preferred, instead of “form”.

11.   L61: “among” instead of “between”, as intraspecific analyzes should include more than 2 spots (again L98, L166, L227…….).

12.   L71: be careful, it’s not clear that you are exploring variation along environmental gradients.

13.   L106: Information from Table S1 should be included in the text, indicating the number of specimens per area, as well as the distance and some identification of the study plots, which allows to know where the plots located at the border of two different biomes (Figure 1) were classified.

14.   L110: general information about the biomes or the simple reference to the map is not enough, information about the study sites is necessary if you want to relate morphological variation to environmental conditions.

15.   Figure 2: pedipalp length is not strictly linear, why did you decide not to use any alternative technique? Tarsus shape looks as important.

16.   L147: using ratios instead of simple measurements means that you are analyzing shape. Why could ratios better reflect adaptation of the specimens? Perhaps shape.

17.   L232: the total variation is not informative, one needs to know biome variation, as the analyzes always include the biome as a grouping factor. Also, provided that you use ratios, the summary table should include information about ratios.

18.   Figure 2 is extremely confusing, as it provides information about the whole techniques in a section entitled “linear morphology”, before any information is given about the results of the Principal Coordinate Analysis or the Gower distance of similarity, but just the Canonical Variate Analysis. Moreover, the figure shows a remarkable difference in data overlap depending on the technique (I guess carapace and sternum shape should correlate).

19.   Table 2: why bulb copulatory data are presented this way?

20.   L258: please review “related to mainly with the angulation”  

21.   Figures: I do not understand the need to include figures both at the body of the paper and as supplementary material.

22.   L312: please, explain why results based on different techniques are congruent and complementary. You should consider the need of starting this section with a brief summary of results. For example, provided that you didn’t found differences in bulb morphology, one could presume that specimens belong to the same species.... Eye arrangement was only studied through GM, and you have no information based on any other technique.

 

Author Response

Dear reviewer, thank you for your contributions.

Below are responses to the main points raised, as well as some considerations.

 

Best regards.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors, congratulations for the complex and important work. I recommend the acceptance of this paper after minor English checks, and after reorganizing the Introduction and discussion. My suggestions are attached.

Sincerely

 

Comments for author File: Comments.pdf

Author Response

Dear reviewer, thank you for your contributions.

Your suggestion was accepted and included in the text.

I also emphasize that the manuscript was partially rewritten as a way to clearly expose the hypotheses, arguments, methods, results, and discussion of the data.

 

Reviewer 3 Report

This is a great mss which thoroughly examines the question of variability of Idiops pirassununguensis both in morphometry and morphology. The data are generally excellent. The conclusions really crush the good  work and the referencing fails drive in the final stake. Also, there are no conclusions: the entire direction of the mss is about the different groups but no conclusion is given. Why? Logically, there is ample evidence that there are 3 species and DNA is NOT needed to support that conclusion.

Major revision because of the conjectures and referencing fails.

 

Expression fails

Loaded statements

“, have a continental distribution,” implies a continent-wide distribution but it is not. A better choice of words is needed.

 

But : indirect writing

l. 318: According to our results, individuals from the Caatinga tend to be smaller when compared to specimens from the Amazon and Cerrado…

No. You need to state the facts, these are your results so don’t repeat that be more direct:

Individuals from the Caatinga were smaller compared to specimens from the Amazon and Cerrado.

The Discussion has a severe overload of conjecture, sentences including may, e.g.,

“This differentiation of Caatinga specimens may be associated with morphological and physiological adaptations related to thermoregulation and water balance of these spiders to the restrictions imposed by the arid environment [52-54], such as that found in the semi-arid Caatinga, which…”

Referencing fails

l. 74 “few studies have explored the morphometric” …? The papers cited are trivial. Look at all the work of Coyle and Gray 2010. This is a severe failure.

 

“According to Main [52] and Cloudsley-Thompson [53]…”

Main, B. Y. 1982. Adaptations to arid habitats by mygalomorph spiders.

Note: the species concepts on which this work are based have been vastly overturned: what were considered “races” of one species are now full species: check Rix’s work. So the conclusions formed are incorrect.

Likewise, Cloudsley-Thompson, J. L. 1983. Desert adaptations in spiders. However, both papers are about desert or arid habitats which I understand are not the habitats considered in this mss. The pressures in arid environments are considerably more unforgiving that most of those found in Brasil. However, if I am wrong and the authors consider the environments are comparable then you have assumed the reader understands the complexity of the Caatinga; that needs to be added.

 

Author Response

Dear reviewer, thank you for your suggestions.

The manuscript has been partially rewritten to present the hypotheses and predictions, as well as arguments, methods, results and data discussion more clearly.

We prefer to be conservative and keep all specimens of I. pirassununguensis as belonging to the same species. Neotropical species of Idiops have been recently revised and have a number of diagnostic characters that support this position. 

Regarding conjectures and indirect writing, although we follow a more conservative line, we changed some phrases as a way to improve reading.

Here are some specific points:

reference failure
I. 74 “few studies have explored morphometry” …? The cited articles are trivial. Look at all of Coyle and Gray's 2010 work. This is a serious flaw.

Answer: I did not find the work cited above (Coyle and Gray 2010) in any indexed database. Despite this, I disagree that the cited works are trivial. A wide review of morphometry works involving spiders was carried out, most of which were cited in the manuscript.

“According to Main [52] and Cloudsley-Thompson [53]…”
Main, B. Y. 1982. Adaptations to arid habitats by mygalomorph spiders.
Note: the species concepts on which this work is based have been largely reversed: what were considered "races" of a species are now full species: check out Rix's work. Therefore, the conclusions formed are incorrect.

Answer: Although the work of Barbara Main (1982) actually discusses data based on species divided into races, a well outdated concept, in my view, information about the adaptation of spiders to arid environments (morphological modifications, foraging strategies) continues to be valid and therefore the citation deserves to be maintained.

Likewise, Cloudsley-Thompson, J. L. 1983. Desert adaptations in spiders. However, both documents are about desert or arid habitats that I understand are not the habitats considered in this manuscript. Pressures in arid environments are considerably more relentless than most found in Brazil. However, if I am wrong and the authors consider the environments to be comparable, you have assumed that the reader understands the complexity of the Caatinga; that needs to be added.

Answer: We deleted the sentence that compares our results with data obtained for species from desert and arid regions. In addition, more information about the Caatinga biome was included in the manuscript.

See the excerpt below, taken from the Book Caatinga - The Largest Tropical Dry Forest Region in South America (DA SILVA, José Maria Cardoso; LEAL, Inara R.; TABARELLI, Marcelo (Ed.. Springer, 2018): "Most of the rainfall is concentrated in three consecutive months, although wide annual variations and recurrent droughts are frequent (Nimer 1972). The number of dry months increases from the edges to the core of the region, with some areas experiencing periods of 7–10 months without water availability for plants (Prado 2003). An important feature of the Caatinga is the high inter-annual variability in rainfall, with droughts that can last for years."

Reviewer 4 Report

This study presents a morphometric analysis of a trapdoor spider species widely distributed in Brazil, and search for correlations of morphological variation with Brazilian biomes and their climate. The morphometric and statistical analyses seem solid (although I am not specialist).

 

The authors found confirmation of previous theories and congruence with previous empirical results about associations of intraspecific variation with climatic variables. This seems fine up to this point.

 

Afterwards the discussion goes on other possible scenarios of genetic structuring and cryptic species, which seemed to me as a big leap in the argument. For example:

 

“The association of these characteristics suggests a possible morphological stasis [40, 78] and the existence of a cryptic species complex”

“The observed differences in non-genital, but not in genital traits among populations from different biomes, suggests that if there are cryptic species within the current concept of this species, ...”

.. This seems speculative to me. I would expect first to present the evidence for cryptic species or genetic structure, and then make a discussion of its possible causes.

One of the referred studies presented as theoretical context in support of the warm-small association (Hoerne et al. 2015) is about intraspecific patterns of adult size variation. The authors also find to corroborate Lira et al.'s (2021) finding of intraspecific variation of smaller sized scorpions in the Caatinga. I feel that the association of intraspecific size variation with warmer climate goes in the opposite direction of explaining size variation due to incipient speciation, genetic structure, or cryptic species. Perhaps the connection is obvious for the authors and a better explanation may help understanding the argument?

 

“possible adaptive processes associated with the evolution of these spiders in arid environments (e.g., reinforced trapdoor tubes)”

.. this comes out of the blue. What is a reinforced trapdoor tube? Is the reinforced tube specific of the arid populations?

 

I think that this manuscript can be reviewed with a better connection of the theoretical arguments with the patterns found, plus perhaps shortening or eliminating the most speculative interpretations.

 

(I wonder if I saw the proper manuscript file. The abstract in the PDF is different from that in the system, and the PDF has many MS-Word tracked changes.)

 

An additional thought that may be useful: There is a hypothesis that the size of male genitalia is intra-specifically less variable than the body size (Eberhard et al. 1998, https://doi.org/10.1111/j.1558-5646.1998.tb01642.x). I wonder if the morphometric data of the copulatory bulb can be used here to suggest the existence of intra- (uniform size) vs. inter-specific (bulb size correlated with body size) variation. The finding that “Regarding the results of the morphogeometric analysis of the copulatory bulb, there was no separation of the specimens in relation to the biomes” may indicate the first case.

 

 

Small edits:

 

“specific specimens”

.. conspecifics

 

“I. pirassununguensis can, therefore, help us understand the processes which led to the current Neotropical biodiversity.”

.. I think this claim is too vague.

 

“previously defined landmarks”

.. provide reference?

 

 

Author Response

Dear reviewer. Thank you for your contributions.

Your suggestion was accepted and included in the text.

After evaluating your comments, we agreed that the final part of the discussion was speculative. As we did not perform molecular analyses, we decided to remove the excerpts that discussed ecological speciation and cryptic diversity.

Best regards,

Rafael

Round 2

Reviewer 3 Report

The authors response even more increased my concern. I may have confused them. Gray (2010) is a major recent work using morphometric. Coyle's work in monumental, the elephant in the room: search WSC for Coyle for the years 1984 to 1995. Their claims must be compared to the paper: with it the mss is rejected, outright for very poors scientific research.

Author Response

Dear reviewer 3

We rewrote the paragraph on morphometric studies in spiders in the introduction, in order to highlight studies that discuss geographic variation and those performed on mygalomorph spiders, which include multivariate and/or morphogeometric analyses.

We emphasize that we have included some studies by Fred Coyle that discussed geographic variation (including scatterplots) in species of Aliatypus (Antrodiaetidae) and Ischnothelinae (Ischnothelidae). Despite this, we do not cite the study by Gray (2010) on Hexathelidae, which, despite being relevant, presents only morphological variation data, not presenting the analyzes mentioned above.

Best regards

Round 3

Reviewer 3 Report

The issue is simple. Scientific papers that may directly be relevant or marginally should be mentioned; they are not. The authors have inserted two references to Coyle in the lit cited but nothing about them in the manuscript: that then appears to be a deliberate attempt to avoid the issue. I doubt that it is and wonder that they have not included the wrong version. 

In the core of it , the mss is good but this action is fatal. Rejected.

Author Response

Dear reviewer, we continue with the same arguments we presented before.

Part of the discussion has been reworded to be less speculative.

Best regards,

Rafael Fonseca Ferreira