Indirect Effects of Cattle Trampling on the Structure of Fruit-Feeding Butterfly Assemblages Inhabiting Restinga Forests in Southern Brazil

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I very much welcomed this well-conducted study, which has no flaws in execution, only minor flaws in description of the methods, and brings highly interesting and well presented results. Concurrently, however, it has serious shortcomings regarding the theoretical background, and hence needs general re-framing, which should reflect current state of the scientific debates (Introduction), and shall be more fully reflected in Discussion.

Ti start from the basics, domestic cattle is a farm animal, and neither it, nor its feral or natural forms (the now extinct “auroch”), is native for the Neotropics. Hence, it might be considered an exotic element, and disturbance factor. However, the Neotropics had hosted wide spectrum of its native megafauna, starting from the unique pre-interchange taxa (the Megatheridae giant ground sloths, through glyptodonts, Pampatheridae mega-armadillos, many other sloth weighing around 1 ton per individuum, the SANU ungulates, most notably Toxodon and Macrauchenia) … all the way to Laurasian elements, which included proboscideans, horses, and many others. There is rich and growing literature on this topic (I recommend Svenning et al., PNAS 113, 898-906 (2016), to begin with), which agrees that the Neotropics have been uniquely impoverished, in terms of mammalian megafauna, when European settlers first entered the Continent.

Crucially, all these megafaunal beasts had been present there for millions of years, the insects fauna, as well as vegetation, long co-existed and co-evolved with them, and it is hence fully expectable, that some of the currently existing butterflies exploit, or even need, the disturbances provided in the past by native large mammals. Next, with native large mammals generally gone, such species would suffer in the “undisturbed forests” habitats; and some of them can be expected to profit from presence and activities of non-native mammals, including the cattle. Which is precisely what you discovered, and which can make your study fascinating, novel and unique – if adequately framed. (To foster this argument, refer to the growing body of literature on “ecological replacements”, “ecological analogues”, etc.).  

Further, any effect of big herbivore, which the cattle of course is, will necessarily depend on such circumstances as cattle density, mode and intensity of grazing, etc. Your study setting (pampa grasslands interspersed with short-trees woodlots on rather unfertile soils), and the fact that both the grassy and woody patches are grazed, seems quite “naturalistic” grazing mode to me, so  that the cattle are likely replacing the native herbivores, whose densities would likely be quite high in such a habitat. The aspects of grazing mode also need to be involved, presumably in Discussion.

To conclude this long explanation: Your study needs re-framing in the way, that some big herbivores are “natural” for your system, although the humans had mostly killed them some millennia ago. Hence, the simple notion that “undisturbed forest is good, disturbed forests is bad” is clearly false, both types of “forest” are necessary for diversity, and the cattle, if managed in a nature friendly way, may serve as replacement of native, but extinct, fauna.

 

Minor remarks:

 

Introduction

Framing the manuscript --

Line 33: „One of the main activities resulting from human disturbances is livestock farming…” – unhappy expression, livestock farming does not result from human activities. It is (quite vital) human activity per se. Re-word.

Around 49-56: The text, and selection of citations, seems to be one-sided, as it is presenting only studies reporting negative impacts of cattle grazing. I also noted that the newest citations are almost decade old (2015). The selection should be more balanced, with some representation of newer literature.

73: “the impacts of real estate speculation” – is not this just “real estate development”?

91: “which primarily consume” => which, as adults, primarily consume

106-115: Your predictions basically are that lower cattle trampling will lead to somehow better conditions to butterflies. This will need re-thinking, in line with reframing the introductory ideas of your article. Also, the current ordering of ideas (i) – (iii) may require reordering, starting that the assemblages shall differ (ii), and continuing towards vegetation structure, light, etc.

 

Materials and Methods

133-4: “The sites were chosen based on some criteria, such as isolation…” – always avoid vague expressions, such as “some criteria”. Here, it is easily remediable, e.g., “The sites were chosen based on such criteria as isolation…”

138-9: “…in which each SU was distant, at least, 300 m apart to ensure the 138 independence of the replicates…” => “…in which each SU was separated by at least 300 m from other SUs to ensure the independence of the replicates…”

142: although you provide a citation, some brief specification of the traps, such as diameter, or sack height, would be appropriate here.

145-7, plus Table 1:  be more specific about the “qualitative measurements of faeces density across all seasons during the first year of the study”. E.g., for the faeces,  let reader know from how large area this was estimated, and how.

149: the word “butterflies” should occur here, so that reader knows, that you are not talking about the faeces any more…

174: “eight meters before each trap” => “eight meters in front of each trap” (likely, but I am not sure about it)

somewhere between  141 and 191: For the sake of logic, I recommend writing about all environmental measurements, i.e., faeces, temperature, humidity, vegetation structure… and then about sampling the butterflies. Small reorganising the paragraphs will do this.

Table 1 (once more). All average values should be accompanied by a measure of dispersion (standard errors and/or deviations).

Results

No problems here.

 

Discussion

This section needs deep re-framing in the light of current naturalistic grazing / rewilding / ecological analogues debates.

 

 

 

Comments on the Quality of English Language

It was generally good, but further check would help.

Author Response

3. Point-by-point response to Comments and Suggestions for Authors   Comments 1: I very much welcomed this well-conducted study, which has no flaws in execution, only minor flaws in description of the methods, and brings highly interesting and well presented results. Concurrently, however, it has serious shortcomings regarding the theoretical background, and hence needs general re-framing, which should reflect current state of the scientific debates (Introduction), and shall be more fully reflected in Discussion.   Ti start from the basics, domestic cattle is a farm animal, and neither it, nor its feral or natural forms (the now extinct “auroch”), is native for the Neotropics. Hence, it might be considered an exotic element, and disturbance factor. However, the Neotropics had hosted wide spectrum of its native megafauna, starting from the unique pre-interchange taxa (the Megatheridae giant ground sloths, through glyptodonts, Pampatheridae mega-armadillos, many other sloth weighing around 1 ton per individuum, the SANU ungulates, most notably Toxodon and Macrauchenia) … all the way to Laurasian elements, which included proboscideans, horses, and many others. There is rich and growing literature on this topic (I recommend Svenning et al., PNAS 113, 898-906 (2016), to begin with), which agrees that the Neotropics have been uniquely impoverished, in terms of mammalian megafauna, when European settlers first entered the Continent.   Crucially, all these megafaunal beasts had been present there for millions of years, the insects fauna, as well as vegetation, long co-existed and co-evolved with them, and it is hence fully expectable, that some of the currently existing butterflies exploit, or even need, the disturbances provided in the past by native large mammals. Next, with native large mammals generally gone, such species would suffer in the “undisturbed forests” habitats; and some of them can be expected to profit from presence and activities of non-native mammals, including the cattle. Which is precisely what you discovered, and which can make your study fascinating, novel and unique – if adequately framed. (To foster this argument, refer to the growing body of literature on “ecological replacements”, “ecological analogues”, etc.).    Further, any effect of big herbivore, which the cattle of course is, will necessarily depend on such circumstances as cattle density, mode and intensity of grazing, etc. Your study setting (pampa grasslands interspersed with short-trees woodlots on rather unfertile soils), and the fact that both the grassy and woody patches are grazed, seems quite “naturalistic” grazing mode to me, so that the cattle are likely replacing the native herbivores, whose densities would likely be quite high in such a habitat. The aspects of grazing mode also need to be involved, presumably in Discussion. To conclude this long explanation: Your study needs re-framing in the way, that some big herbivores are “natural” for your system, although the humans had mostly killed them some millennia ago. Hence, the simple notion that “undisturbed forest is good, disturbed forests is bad” is clearly false, both types of “forest” are necessary for diversity, and the cattle, if managed in a nature friendly way, may serve as replacement of native, but extinct, fauna. Response 1: Thank you for pointing this out. We have made some adjustments throughout the manuscript to attend this suggestion. Please check the main file, where our revisions are highlighted in red. In summary, we pointed that the presence of grazing animals is potentially beneficial to grassland ecosystems (Baggio et al 2021), as it increases the richness and abundance of arthropods under deferred grazing systems (Ferreira et al 2020). However, we politely disagree with the reviewer regarding the matter that the cattle presence would act as a rewilding in forest ecosystems, such as the studied Restinga forest. In forests, the cattle presence often causes a filter to shade-tolerant tree species, benefiting mainly light-demanding trees (Vefago et al 2019). The intensive use of cattle raising in an unmanaged way occupying several hectares of the mosaic of grassland-forest is not according to the presence of the large herbivores of the megafauna in the past. Moreover, there was no restricted use of the ecosystems by the megafauna, thus the disturbance made was not too severe and without the influence of human pressure (cattle confinement, the presence of fences to avoid cattle escape, etc.). Thus, nowadays the disruption of forested ecosystems worldwide and, mainly, in forested ecosystems like the transition between Pampa and Atlantic Forest in southern Brazil and Amazonia in northern Brazil, causes irreversible and severe damage to the structure of forest sites letting them hollow and without understory, changing dramatically the native environments and microclimate inside these native forests. These may negatively influence the insect fauna, but further tests are required. Therefore, our study justifies itself as it addresses both the positive effect of cattle presence found on arthropods in grasslands and the negative effect of cattle trampling in forests aiming to better understand the overall effects of cattle trampling on butterfly communities in forest ecosystems. Please see lines 33-35; 59-62; 71-75.   Comments 2: Line 33: „One of the main activities resulting from human disturbances is livestock farming…” – unhappy expression, livestock farming does not result from human activities. It is (quite vital) human activity per se. Re-word. Response 2: Done. Lines 33-35.   Comments 3: Around 49-56: The text, and selection of citations, seems to be one-sided, as it is presenting only studies reporting negative impacts of cattle grazing. I also noted that the newest citations are almost decade old (2015). The selection should be more balanced, with some representation of newer literature. Response 3: We agree with the reviewer. We added the positive point-of-view regarding the cattle presence, along with more recent research across the text. Please see lines 59-62; 70-78; 396-402.   Comments 4: 73: “the impacts of real estate speculation” – is not this just “real estate development”? Response 4: Done.   Comments 5: 91: “which primarily consume” => which, as adults, primarily consume Response 5: Done.   Comments 6: 106-115: Your predictions basically are that lower cattle trampling will lead to somehow better conditions to butterflies. This will need re-thinking, in line with reframing the introductory ideas of your article. Also, the current ordering of ideas (i) – (iii) may require reordering, starting that the assemblages shall differ (ii), and continuing towards vegetation structure, light, etc. Response 6: As described above, we included a positive perspective regarding the effect of cattle on grassland ecosystems. However, our literature research on the effect of cattle trampling in forests shows that their effects are mainly negative for vegetation, which would translate into a negative effect on the butterfly fauna as well (remembering that there is a strong association of butterflies, in the immature and adult stages, with the richness and composition of plant species). Indeed, butterflies are great environmental indicators because of this, responding very fast to alterations in the structure of plant communities and the landscape. Therefore, our predictions remain unaltered. In addition, we followed your suggestion and re-ordered the aims as follows: (i) Is there a segregation in butterfly species composition among these habitats, including identifying indicator species for each habitat? (ii) Are there differences in butterfly diversity parameters among Restinga forests with low levels of cattle trampling compared with medium and high levels of disturbance? (iii) How do environmental variables and vegetation structure of Restinga forests affect the richness, abundance, and dominance of fruit-feeding butterfly assemblages due to indirect disturbances caused by cattle presence?   Comments 7: 133-4: “The sites were chosen based on some criteria, such as isolation…” – always avoid vague expressions, such as “some criteria”. Here, it is easily remediable, e.g., “The sites were chosen based on such criteria as isolation…” Response 7: Done.   Comments 8: 138-9: “…in which each SU was distant, at least, 300 m apart to ensure the 138 independence of the replicates…” => “…in which each SU was separated by at least 300 m from other SUs to ensure the independence of the replicates…” Response 8: Done.   Comments 9: 142: although you provide a citation, some brief specification of the traps, such as diameter, or sack height, would be appropriate here. Response 9: Done.   Comments 10: 145-7, plus Table 1:  be more specific about the “qualitative measurements of faeces density across all seasons during the first year of the study”. E.g., for the faeces, let reader know from how large area this was estimated, and how. Response 10: Done.   Comments 11: 149: the word “butterflies” should occur here, so that reader knows, that you are not talking about the faeces any more… Response 11: Done.   Comments 12: 174: “eight meters before each trap” => “eight meters in front of each trap” (likely, but I am not sure about it) Response 12: Done.   Comments 13: somewhere between 141 and 191: For the sake of logic, I recommend writing about all environmental measurements, i.e., faeces, temperature, humidity, vegetation structure… and then about sampling the butterflies. Small reorganising the paragraphs will do this. Response 13: Done.   Comments 14: Table 1 (once more). All average values should be accompanied by a measure of dispersion (standard errors and/or deviations). Response 14: Done.   Comments 15: Discussion: This section needs deep re-framing in the light of current naturalistic grazing / rewilding / ecological analogues debates. Response 15: We included a discussion in this regard at the last paragraphs seeking to highlight the importance of large herbivores in grassy ecosystems and their potential impacts on forests. Please see lines 392-405; 416-420.

4. Response to Comments on the Quality of English Language

Point 1: It was generally good, but further check would help.

Response 1: We conducted a careful English language revision as requested to improve the quality of the manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This is a well produced paper showing the effects of cattle farming in the endemic  ecosystem of coastal plain region in Rio Grande do Sul State, south-eastern Brazil. They present good evidence how various intensities of cattle rampling influence the forest composition using fruit-feeder nymphalid butterfly species as indicators. The methods are well chosen resulting good results. I recommend the paper to be published in the present form. Below I give some remarks or comments according to line numbers of the manuscript.

59: "North Hemisphere" should be "Northern Hemisphere";

60: There are many papers on forest butterflies dwelling in Europe (as well a sin USA) and their habitat focussing to changed land-use and popualtion dynamics, and often tackle livestock grazing in forests. There are three flaghsip species in Europe: Coenonympha hero (satryrid), Lopinga achine (satryid), and Apatura iris (nymphalid). A brief search on the internet could yield some papers worth to cite. One brand new example for Lopinga achine: https://fartmann.net/downloads/articles/Streitberger_et_al_Lopinga_achine_Forest_management_ForestEcolManag_2012.pdf.

67: The opening word "However" has no syntax here, should be deleted, so the sentence should start "Several studies suggest that...";

75-76: that develop... should be better like this "that develop on sand dunes formed along the Brazilian Atlantic coast during the not so distant past ";

87: "luminosity", would be better "light";

92: probably one or two sentences would be necessary which kind of trees offer fruits for the butterfly community examined; and for the uninformed readers it also should be mentioned that cattle faces here do not attract the butterflies in any conditions;

154: I wonder whether there is a list of recorded species; it would be worth to publish that list showing (as an appndix) the full composition of the butterfly fauna recorded during the study;

160: Fig. 1B would be much more informative with showing some kind of geography or forest silhouettes than having a pure blank background;

161: Rio Grande do Sul State in south-eastern Brazil;

191: I recommend to put the two figures of S1 into the main text, there is no sense to hide them;

308: please compose a plate showing the species either in situ or via specimens.

Author Response

Comments 1: This is a well produced paper showing the effects of cattle farming in the endemic ecosystem of coastal plain region in Rio Grande do Sul State, south-eastern Brazil. They present good evidence how various intensities of cattle trampling influence the forest composition using fruit-feeder nymphalid butterfly species as indicators. The methods are well chosen resulting good results. I recommend the paper to be published in the present form. Below I give some remarks or comments according to line numbers of the manuscript. Response 1: Thank you for the careful reading and important suggestions to improve our manuscript. All changes in the manuscript were made in red.   Comments 2: 59: "North Hemisphere" should be "Northern Hemisphere"; Response 2: Done.   Comments 3: 60: There are many papers on forest butterflies dwelling in Europe (as well a sin USA) and their habitat focussing to changed land-use and popualtion dynamics, and often tackle livestock grazing in forests. There are three flaghsip species in Europe: Coenonympha hero (satryrid), Lopinga achine (satryid), and Apatura iris (nymphalid). A brief search on the internet could yield some papers worth to cite. One brand new example for Lopinga achine: https://fartmann.net/downloads/articles/Streitberger_et_al_Lopinga_achine_Forest_management_ForestEcolManag_2012.pdf. Response 3: Thanks for the advance and for the suggestion, but we choose to maintain without the citation of works in Europe and USA because is a distinct kind of habitat type, generally related to temperate areas, thus in another perspective and approach. The study was conducted in a subtropical region of the Neotropics and the species indication occurs differently, much more related to some specific groups of fruit-feeding butterflies. Thus, to avoid confusion by the readers we opted to maintain its original form.    Comments 4: 67: The opening word "However" has no syntax here, should be deleted, so the sentence should start "Several studies suggest that..."; Response 4: Done.   Comments 5: 75-76: that develop... should be better like this "that develop on sand dunes formed along the Brazilian Atlantic coast during the not so distant past "; Response 5: Done.   Comments 6: 87: "luminosity", would be better "light"; Response 6: Done.   Comments 7: 92: probably one or two sentences would be necessary which kind of trees offer fruits for the butterfly community examined; and for the uninformed readers it also should be mentioned that cattle faces here do not attract the butterflies in any conditions; Response 7: Done.   Comments 8: 154: I wonder whether there is a list of recorded species; it would be worth to publish that list showing (as an appndix) the full composition of the butterfly fauna recorded during the study; Response 8: Thanks for the suggestion. We included the species list in the repository of the manuscript associated with a Zenodo file.   Comments 9: 160: Fig. 1B would be much more informative with showing some kind of geography or forest silhouettes than having a pure blank background; Response 9: Thanks for the suggestion. But, the region is near the ocean in a Coastal Plain without great variation in altitude and vegetation. We choose to maintain the Figure in its original form for a better map fit. We think that the images of the distinct levels of disturbance of the Restinga forests in the same figure is enough to well represent the region, thus complementing the map.   Comments 10: 161: Rio Grande do Sul State in south-eastern Brazil; Response 10: Done.   Comments 11: 191: I recommend to put the two figures of S1 into the main text, there is no sense to hide them; Response 11: We adjusted the text to include the Figure S2 in the main file, as suggested. However, we decided to maintain the Figure S1 in the supplementary information file, as their results are already described in the main text (see below). “All environmental variables were used as predictors in the statistical analysis, assessed through average values, and showed no correlation according to Pearson’s correlation test (Fig. S1).”   Comments 12: 308: please compose a plate showing the species either in situ or via specimens.

Response 12: Done, we agree with the suggestion, improving a lot the manuscript. Thanks for the suggestion!

4. Response to Comments on the Quality of English Language

Point 1: English language fine. No issues detected

Response 1: Thanks.

5. Additional clarifications

None.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Dear colleagues, 

Your paper considerably improved (it was quite good already in the first version!), but my overall impression is, that you still did not fully grasp its potential, which would be, to frame it within the current Pleistocene defaunation -  incomplete ecosystems - rewilding potential - ungulates as ecosystem engineers (etc.) debates. As the debates rage quite hot lately, it is also not easy to remain impartial in them, so I am trying to respect that you are trying not to get too immersed into them. Still, by carefull reading your paper, I cannot help the impression that you are too defensive about your results, or, rather, trying to stay within a traditional views of your system so hard, that you are neglecting quite contradictory evidence. 

Let me reiterate. 

1. you claim that your study sites is situated "within the pampa biome". Pampa (grassland) with patches of forest is, of course, savanna (neglecting the details), and savannas across the World have evolved under heavy influence of grazing. Hence, your argument (from Cover letter) that cattle is "not native species" is of little validity in a system, which had lost almost all native herbiveres ~10 000 y ago, and in which, even in the pre-exinction period, a substantial proportion of herbivores (mainly elephants and horses, but also still extant deer and tapir) were relative newcomers from the Nearctic. If the cattle is fulfilling some role of native herbivores, it may be non-native, but can be "useful" - of course only under some circumstances, not if overstocked/overgrazed, etc. 

2. critically, all the extant neotropical butterfllies (and the rest...) have evolved in megafaunal world. You are actually stating, in Discussion, that "these species are common in various habitat types due to the more suitable conditions in open and sunny areas. Thus, the studied disturbed Restinga forests presented numerous clearings allowing the edge environmental conditions to penetrate harshly the core of forest fragments...". The description fits quite well to an ecosystem under megafauna influence. Have you considered the option that such conditions are more "natural", in evolutionary time scale, than the post-megafaunal dense forests? Allthough you are claiming that your disturbed forests were "too much" disturbed, we have only little idea, how would the ecosystem look with all the giant sloths, glyptodons and mastodons present. 

3. I woud not mind, if your data contained some evidence for the benefits of "undisturbed forest" for butterflies. But checking your results: The undisturbed conditions hosted much fewer butterflies, plus, had no indicator species, implying that the undisturbed forest is novel or impoverished ecosystem, if compared with the disturbed one!

4. Although it is really up-to-you how you will deal with the comments above (I would personally reframe it all, but it is your right not to!), the text in Discussion, 396-402, must be reconsidered. You are making a distinction grassland x forest regarding the impact of grazing, but this distinction is very artificial, and forest-centric. In a megafaunal past, most of forest would experience disturbance by animals, and many "grasslands" would contain some trees, generally resulting in less clear forest-grassland distinction. There is enough evidence that grazing in forests (not only grasslands) may be beneficial for small animals including insects - so change different exapmples, or be more nuanced. 

Here is some reading to illustrate my point. 

 

Pearce, Elena A., et al. "Substantial light woodland and open vegetation characterized the temperate forest biome before Homo sapiens." Science advances 9.45 (2023): eadi9135.

Storch, David, et al. "Biodiversity dynamics in the Anthropocene: how human activities change equilibria of species richness." Ecography 2022.4 (2022).

Maicher, Vincent, et al. "Effects of disturbances by forest elephants on diversity of trees and insects in tropical rainforests on Mount Cameroon." Scientific Reports 10.1 (2020): 21618.

 

 TECHNICAL COMMENT (must be dealt with) 
275-7: "the species composition across the low-level disturbed area is mainly determined by high mean humidity, understory height, and a higher percentage of canopy cover" - it may well be so, but this claim is based on observational evidence (comparison with values in the table). I would tone down the claim ("seems mainly determined"). 

313-16 (and Table 4). Here, I see that whatever I wrote above above lines 275-7 is invalid, as you, in fact, performed the relevant analysis. Given this, your mention about "causes" of the pattern (275-7) were preliminary, you should combine these two arguments (and not repeat them twice). 

Wishin you good luck with the revision

Martin Konvicka

 

 

Author Response

Thank you for reviewing the new version of our manuscript and the advances you have made. We have addressed all your comments and queries in the following text. Despite our differing views on cattle raising and livestock farming in South America, we have included some parts of your argument, particularly in discussing the significance of non-native cattle compared to extinct herbivore megafauna.

Comments 1: Dear colleagues,

 

Your paper considerably improved (it was quite good already in the first version!), but my overall impression is, that you still did not fully grasp its potential, which would be, to frame it within the current Pleistocene defaunation -  incomplete ecosystems - rewilding potential - ungulates as ecosystem engineers (etc.) debates. As the debates rage quite hot lately, it is also not easy to remain impartial in them, so I am trying to respect that you are trying not to get too immersed into them. Still, by carefull reading your paper, I cannot help the impression that you are too defensive about your results, or, rather, trying to stay within a traditional views of your system so hard, that you are neglecting quite contradictory evidence.

 

Let me reiterate.

 

1. you claim that your study sites is situated "within the pampa biome". Pampa (grassland) with patches of forest is, of course, savanna (neglecting the details), and savannas across the World have evolved under heavy influence of grazing. Hence, your argument (from Cover letter) that cattle is "not native species" is of little validity in a system, which had lost almost all native herbiveres ~10 000 y ago, and in which, even in the pre-exinction period, a substantial proportion of herbivores (mainly elephants and horses, but also still extant deer and tapir) were relative newcomers from the Nearctic. If the cattle is fulfilling some role of native herbivores, it may be non-native, but can be "useful" - of course only under some circumstances, not if overstocked/overgrazed, etc.

 

2. critically, all the extant neotropical butterfllies (and the rest...) have evolved in megafaunal world. You are actually stating, in Discussion, that "these species are common in various habitat types due to the more suitable conditions in open and sunny areas. Thus, the studied disturbed Restinga forests presented numerous clearings allowing the edge environmental conditions to penetrate harshly the core of forest fragments...". The description fits quite well to an ecosystem under megafauna influence. Have you considered the option that such conditions are more "natural", in evolutionary time scale, than the post-megafaunal dense forests? Allthough you are claiming that your disturbed forests were "too much" disturbed, we have only little idea, how would the ecosystem look with all the giant sloths, glyptodons and mastodons present.

 

3. I woud not mind, if your data contained some evidence for the benefits of "undisturbed forest" for butterflies. But checking your results: The undisturbed conditions hosted much fewer butterflies, plus, had no indicator species, implying that the undisturbed forest is novel or impoverished ecosystem, if compared with the disturbed one!
4. Although it is really up-to-you how you will deal with the comments above (I would personally reframe it all, but it is your right not to!), the text in Discussion, 396-402, must be reconsidered. You are making a distinction grassland x forest regarding the impact of grazing, but this distinction is very artificial, and forest-centric. In a megafaunal past, most of forest would experience disturbance by animals, and many "grasslands" would contain some trees, generally resulting in less clear forest-grassland distinction. There is enough evidence that grazing in forests (not only grasslands) may be beneficial for small animals including insects - so change different exapmples, or be more nuanced.

Response 1: Thank you for pointing this out again. We have made some adjustments throughout the manuscript to support this suggestion. Despite our differing views on cattle raising and livestock farming in South America, we have included some parts of your statement, particularly in discussing the significance of non-native cattle compared to extinct herbivore megafauna. However, we maintain our position and statements regarding the problematic use of extensive cattle farming practices and their influence on biodiversity.

   There is a misunderstanding in the interpretation of the ecosystems of southern Brazil. The Savannah habitats are globally distributed in the tropical zone and in Brazil including the Cerrado biome, which is entirely different from the grasslands and forest areas of Southern Brazil. The grasslands (Pampa biome) interspersed with fragments of the Atlantic Forest biome (Restinga forest in our study) belong to two main biomes in Brazil. The region of our study is unique worldwide and is considered an ecotone between these two biomes within a classification of Pioneer Formations. Thus, the evolution of the Pampa and Atlantic Forest is not the same, including distinct pressures from grazing and cattle raising.   According to several studies, cattle use in these complex forests is quite harmful to insects, and these patterns may extend to the Atlantic Forest as a whole. Therefore, the importance of grazing in the Pampa is entirely distinct from that in the forests. Moreover, the balance of the mosaic grassland-forest in southern Brazil is driven by fire regimes as well. The maintenance of this mosaic, with well-preserved and undisturbed forests, is crucial for the balance and dynamics of these ecosystems. Excluding cattle grazing increases the richness and diversity of arthropods, and the same occurs in the Atlantic Forest. The fragmentation and loss of native vegetation driven by large-scale agricultural and monoculture activities are among the main threats to biodiversity in terrestrial ecosystems. In Brazil, these problems are even more severe due to significant economic and political pressure to weaken environmental regulation in the last decade (Soares-Filho et al., 2014; Vale et al., 2021).

   The mosaic between Pampa and Atlantic Forest forms a biocultural landscape (Overbeck et al., 2007; Luza et al., 2014). This mosaic has been maintained in bi-stability (spatial co-occurrence of both physiognomies) by a combination of biotic and abiotic factors (Innes et al., 2013; Luza et al., 2014; Cruz et al., 2020). Therefore, it is essential to preserve the natural mosaic in which the Atlantic Forest plays a crucial role in maintaining biodiversity. These forest habitats serve as a reference vegetation type for the succession process and the dynamics of grassland-forest mosaics (Luza et al., 2014), as it is poorly understood what the effects and consequences of biodiversity changes will be given the biotic homogenization of ecosystems on a broader spatial scale (Wang et al., 2021).

   Changes in the environment can be detected by assessing sensitive bioindicator taxonomic groups. In tropical forests, fruit-feeding butterflies have been widely used as bioindicators for their ease of standardized sampling and their response to subtle vegetation changes, such as sites undergoing a forest succession process at different stages (Wood and Gillman, 1998; Barlow et al., 2007, 2008; Ribeiro and Freitas, 2012; Sant’Anna et al., 2014; Spaniol et al., 2019). These environmental changes can affect the relationship between host plants, butterfly larvae, and food source availability for adults, reflecting the dominance of a few species in environments with biotic homogenization. For butterflies, dramatic changes in forested habitats modify the species composition of these insects, and the deleterious effects of deforestation and modification on biotic and abiotic factors on the structure of taxonomic and functional diversities of butterflies are widely recognized.

   Therefore, it is difficult to consider cattle trampling and the mischaracterization of the complex and well-structured Restinga forest as having a positive effect. Butterflies using the disturbed Restinga Forest are quite distinct and commonly associated with perturbed areas across the entire Atlantic Forest domain. The novel ecosystems formed are very different from those found in preserved forests, even though the richness and abundance of butterflies are lower in preserved Restinga forests (and lower richness and abundance are not indications of decreasing diversity; it is important to show the dominance patterns, which tend to be less in preserved Restinga Forests). Time for succession to occur naturally is not the only parameter for measuring the capacity of an ecosystem to recover and retain local species. Dramatically altered ecosystems will hardly return to their original conditions, even under the regeneration process (Ferraro et al., 2020; Casas-Pinilla et al. 2022).

   The interior of well-preserved Restinga Forests contains a subset of forest-dwelling butterfly groups that were quite distinct from the groups found in disturbed Restinga forests, consisting mainly of butterflies associated with open areas or edge environments. Habitat use by fruit-feeding butterfly groups may be highly variable, with individual species responding differently according to their biology. The subfamily Satyrinae was the most representative in relation to total abundance, with Satyrini having more species and much greater abundance. For the most part, the disturbed Restinga forests had generalist and common species that are associated with clearings, sunny, and perturbed areas (Brown, 1997; Uehara-Prado et al., 2007). Immature stages usually feed on grasses (Poaceae), such as those in the component matrix with cattle grazing associated with fragments of the Atlantic Forest in this study. Therefore, butterflies that fly in extreme conditions have eurytopic habits with great vagility to disperse through the matrix due to it being easier to take flight given the high luminosity and temperature, besides using pioneer or edge host plants (Uehara-Prado et al., 2007). For forest-dwelling butterflies, the more stable the abiotic variables, the easier it is for them to use the interior of the forests (Spaniol et al., 2019). The specific microhabitats and, possibly, the availability of host plants and decomposing fruits within fragments of the Atlantic Forest (Uehara-Prado et al., 2007) allow butterflies to survive in this habitat type.

   Environmental heterogeneity and different landscape scales, such as in the Atlantic Forest, are important components for both local and regional biodiversity (Ribeiro et al., 2012) and may be modified by natural or human-made disturbances. It is important to maintain a heterogeneous matrix with distinct land-use systems, following recent trends in the recovery of degraded habitats by landscape management in subtropical areas (Bellaver et al., 2022). This heterogeneity includes agroforestry and the preservation of several small fragments (Lorandi et al., 2021), which may effectively support more species than larger fragments (Fahrig, 2020), both composed of secondary vegetation or high-quality habitat serving as steppingstones (Filgueiras et al., 2016; Iserhard et al., 2019; Melo et al., 2019; Schulze et al., 2004). Finally, the changes adopted for the Brazilian Forest Code (Freitas, 2010) regarding deforestation and the reduction in forested areas might put severely fragmented regions, such as most remaining patches of the Atlantic Forest, at severe risk and cause them irreversible damage (Ribeiro et al., 2009).

   Thank you for your valuable contribution to our manuscript, which has generated an interesting and exciting discussion about these issues and the importance of evaluating this kind of study system.

 

References:

- Freitas, A.V.L., 2010. Potential impacts of the proposed Brazilian Forest Act on native butterflies. Biota Neotrop. 10, 53–57. https://doi.org/10.1590/S1676- 06032010000400007

- Soares-Filho, B., Raj˜ao, R., Macedo, M., Carneiro, A., Costa, W., Coe, M., Rodrigues, H., Alencar, A., 2014. Cracking Brazil’s Forest Code. Science 344 (6182), 363–364.

- Vale, M.M., Berenguer, E., Argollo de Menezes, M., Viveiros de Castro, E.B., Pugliese de Siqueira, L., Portela, R.d.C.Q., 2021. The COVID-19 pandemic as an opportunity to weaken environmental protection in Brazil. Biol. Conserv. 255, 108994. https://doi. org/10.1016/j.biocon.2021.108994

- Overbeck, G., Muller, S., Fidelis, A., Pfadenhauer, J., Pillar, V., Blanco, C., Boldrini, I., Both, R., Forneck, E., 2007. Brazil’s neglected biome: The South Brazilian Campos. Perspect. Plant Ecol. Evol. Syst. 9 (2), 101–116. https://doi.org/10.1016/j. ppees.2007.07.005.

- Luza, A.L., Carlucci, M.B., Hartz, S.M., Duarte, L.S., 2014. Moving from forest vs. grassland perspectives to an integrated view towards the conservation of forest–grassland mosaics. Nat. Conserv. 2, 166–169. https://doi.org/10.1016/j. ncon.2016.05.002.

- Innes, C., Anand, M., Bauch, C.T., 2013. The impact of human-environment interactions on the stability of forest-grassland mosaic ecosystems. Sci. Rep. 3, 2689. https://doi. org/10.1038/srep02689.

- Cruz, A.P., Giehl, E.L.H., Levis, C., Machado, J.S., Bueno, L., Peroni, N., 2020. Pre-colonial Amerindian legacies in forest composition of southern Brazil. PLoS ONE 15. https://doi.org/10.1371/journal.pone.0235819.

- Wang, S., Loreau, M., Mazancourt, C., Isbell, F., Beierkuhnlein, C., Connolly, J., Deutschman, D.H., Dolezal, J., Eisemhauer, N., Hector, A., Jentsch, A., Kreyling, J., Lanta, V., Leps, J., Polley, H.W., Reich, P.B., Van Ruijven, J., Schmid, B., Tilman, D., Wilsey, B., Craven, D., 2021. Biotic homogenization destabilizes ecosystem functioning by decreasing spatial asynchrony. e03332 Ecology 102 (6). https://doi. org/10.1002/ecy.3332.

- Wood, B., Gillman, M.P., 1998. The effects of disturbance on forest butterflies using two methods of sampling in Trinidad. Biodiv. Conserv. 7, 597–616.

- Barlow, J., Overal, W.L., Araujo, I.S., Gardner, T.A., Peres, C.A., 2007. The value of primary, secondary and plantation forests for fruit-feeding butterflies in the Brazilian Amazon. J. Appl. Ecol. 44, 1001–1012. https://doi.org/10.1111/j.1365- 2664.2007.01347.x.

- Barlow, J., Araujo, I.S., Overal, W.L., Gardner, T.A., da Silva Mendes, F., Lake, I.R., Peres, C.A., 2008. Diversity and composition of fruit-feeding butterflies in tropical Eucalyptus plantations. Biodiv. Conserv. 17 (5), 1089–1104. https://doi.org/ 10.1007/s10531-007-9240-0.

- Ribeiro, D.B., Freitas, A.V.L., 2012. The effect of reduced-impact logging on fruit-feeding butterflies in Central Amazon, Brazil. J. Insect Conserv. 16 (5), 733–744. https://doi. org/10.1007/s10841-012-9458-3.

- Sant’Anna, C.L.B., Ribeiro, D.B., Garcia, L.C., Freitas, A.V.L., 2014. Fruit-feeding butterfly communities are influenced by restoration age in tropical forests. Rest. Ecol. 22 (4), 480–485. https://doi.org/10.1111/rec.12091.

- Spaniol, R.L., Duarte, L.S., Mendonça Jr., M.S., Iserhard, C.A., 2019. Combining functional traits and phylogeny to disentangling Amazonian butterfly assemblages on anthropogenic gradients. Ecosphere 10. https://doi.org/10.1002/ecs2.2837.

- Ferraro, A., Fidelis, A., Silva, G.S.d., Martins, A.R., Piedade, S.M.D.S., Appezzato-da- Gl´oria, B., Rosario Acosta, A.T., 2020. Long-term Pinus plantations reduce the bud bank in Cerrado areas. Appl. Veget. Sc. 24 (1) https://doi.org/10.1111/avsc. v24.110.1111/avsc.12537.

- Casas-Pinilla LC, Iserhard CA, Richter A, Gawlinski K, Cavalheiro LBD, Romanowski HP, Kaminski LA. 2022. Different-aged Pinus afforestation does not support typical Atlantic Forest fruit-feeding butterfly assemblages. For Ecol Manag 518. https:// doi. org/ 10. 1016/j. foreco. 2022. 120279.

- Brown, K.S. 1997. Diversity, disturbance, and sustainable use of neotropical forests: insects as indicators for conservation monitoring. Journal of Insect Conservation, 1, 25–42.

- Uehara-Prado, M., Brown, K.S. & Freitas, A.V.L. 2007. Species richness, composition and abundance of fruit-feeding butterflies in the Brazilian Atlantic Forest: comparison between a fragmented and a continuous landscape. Global Ecology and Biogeography, 16, 43–54.

- Ribeiro, D.B., Batista, R., Prado, I.P., Brown, K.S. & Freitas, A.V.L. 2012. The importance of small scales to the fruit-feeding butterfly assemblages in a fragmented landscape. Biodiversity and Conservation, 21, 811–827.

- Bellaver J, Romanowski HP, Richter A, Iserhard CA. 2022. Living on the edge: the use of fruit-feeding butterflies to evaluate edge effect on subtropical assemblages. Austral Ecol. https:// doi. org/ 10. 1111/ aec. 13261.

- Lorandi, S., Halinski, R., Mustin, K. & Iserhard, C.A. 2021. Are there differences in the diversity of bees between organic and conventional agroecosystems in the Pampa biome? Journal of Apicultural Research, 60, 1–13. Available from: https://doi.org/10.1080/00218839.2021.1888524

- Fahrig, L. 2020. Why do several small patches hold more species than few large patches? Global Ecology and Biogeography, 29, 615–628.

- Filgueiras, B.K.C., Melo, D.H.A., Leal, I.R., Tabarelli, M., Freitas, A.V.L. & Iannuzzi, L. 2016. Fruit-feeding butterflies in edge-dominated habitats: community structure, species persistence and cascade effect. Journal of Insect Conservation, 20, 539–548.

- Iserhard, C.A., Duarte, L., Seraphim, N. & Freitas, A.V.L. 2019. How urbanization affects multiple dimensions of biodiversity in tropical butterfly assemblages. Biodiversity and Conservation, 28, 621–638.

- Melo, D.H.A., Filgueiras, B.K.C., Iserhard, C.A., Iannuzzi, L., Freitas, A.V.L. & Leal, I.R. 2019. Effect of habitat loss and fragmentation on fruit-feeding butterflies in the Brazilian Atlantic Forest. Canadian Journal of Zoology, 97, 588–596.

- Schulze, C.H., Waltert, M., Kessler, P.J.A., Pitopang, R., Shahabuddin, G., Veddeler, D. et al. 2004. Biodiversity indicator groups of tropical land-use systems: comparing plants, birds, and insects. Ecological Applications, 14, 1321–1333.

- Ribeiro, M.C., Metzger, J.P., Martensen, A.C., Ponzoni, F.J. & Hirota, M.M. 2009. The Brazilian Atlantic Forest: how much is left, and how is the remaining forest distributed? Implications for conservation. Biological Conservation, 142, 1141–1153.

 Comments 2: 275-7: "the species composition across the low-level disturbed area is mainly determined by high mean humidity, understory height, and a higher percentage of canopy cover" - it may well be so, but this claim is based on observational evidence (comparison with values in the table). I would tone down the claim ("seems mainly determined").

Response 2: Done, we agree! We included the wrong number of Tables in the text. Now it is according to the expectation.

Comments 3: 313-16 (and Table 4). Here, I see that whatever I wrote above above lines 275-7 is invalid, as you, in fact, performed the relevant analysis. Given this, your mention about "causes" of the pattern (275-7) were preliminary, you should combine these two arguments (and not repeat them twice).

Response 3: Done. We corrected the number of Tables in the manuscript text to well-address the results. Thanks for the crucial observation.

Author Response File: Author Response.pdf