Land Use Effects on the Space Use and Dispersal of an Apex Predator in an Ecotone Between Tropical Biodiversity Hotspots

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General comments

This study represents a valuable contribution to the field, and I believe it will be of interest to researchers in movement ecology, conservation biology, and landscape planning.

This is a well-designed and timely study that provides valuable insights into puma space use and dispersal in a highly fragmented ecotone between the Atlantic Forest and the Cerrado. The integration of GPS movement data with behavioral modeling and landscape variables is a strong methodological contribution, particularly given the scarcity of such studies in South America. The manuscript is generally well-written and logically structured, and the research questions are clearly stated and addressed. I commend the authors for the use of advanced movement models and for placing their results in a broader conservation context. However, several points could be clarified to improve transparency and replicability:

• Please ensure consistent use and definition of the three behavioral phases (pre-dispersal, dispersal, and post-dispersal, and residency) and other terms (home range, movement, ranging, space use) throughout the manuscript. Please revise to ensure that key terms are used consistently and not interchangeably unless justified.

• Clarify whether all 14 individuals, including five translocated pumas (one after captivity), were included in all analyses, and discuss how translocation may have influenced results.

• Address the implications of the strong sex bias in the sample (only three females, none of which dispersed), and discuss how this may affect interpretations related to connectivity and dispersal ecology.

• Provide additional clarity in the Methods regarding filtering of GPS data, model selection (e.g., OUF justification), and treatment of multicollinearity in landscape analyses.

• Consider refining figure legends and improving transparency in how movement data and landscape interactions are presented.

Detailed suggestions

Below I provide specific comments organized by manuscript section, with references to line numbers when applicable.

Abstract

The abstract presents the study’s objectives, methods, and key findings in a logical and coherent structure. It successfully emphasizes the novelty of documenting puma dispersal using GPS data in Brazil.

Suggestions:

L18 to L24: “We present empirical data on space use and dispersal patterns of GPS-collared pumas (Puma concolor) and evaluate how different land uses affect their movement behavior in a fragmented ecotone...”

L21: Italics in Puma concolor

L29-32: "Pumas moved faster and more linearly during dispersal than residency. Their movement was slower and home ranges smaller in more forested areas, underscoring forest importance as habitat. In contrast, movement increased in open pastures, particularly at night and during dispersal."

L33-38: The conclusion does well to highlight the research gap in South America and the need for integrated movement and population monitoring. However, adding a clearer link to conservation policy implications—e.g., corridor planning, land-use management—would enhance the applied relevance.

Introduction

The introduction is well-structured, informative, and clearly articulates the relevance of studying space use and dispersal of apex predators in fragmented tropical landscapes.

Suggestions:

L105-106: Given that the study area lies within the interface of the Atlantic Forest and the Cerrado, two highly threatened biodiversity hotspots, the introduction could benefit from a clearer emphasis on the global conservation relevance of this setting.

L49 to 121: Terms such as “space use,” “residency,” “home range,” “movement,” and “ranging” are used somewhat interchangeably. It would be helpful to briefly define these key concepts to aid clarity, especially for interdisciplinary readers. A few words briefly defining the terms would be enough to clarify them for later sections. The Methods section can then emphasize and detail how these behaviors were analyzed.

Methods

The description of how puma movements were classified into residency and dispersal phases is methodologically sound and well-referenced. However, I recommend that the authors explicitly define the phases — pre-dispersal, dispersal, and post-dispersal and residency — within this section, ideally with clear, concise criteria for each. Although these phases are implied and used throughout the manuscript, clearly stating their definitions early in the Methods would improve transparency and reproducibility. This is especially important for researchers aiming to compare findings or replicate the approach in other regions or species. A visual summary or table listing the criteria used and the number of individuals assigned to each phase would also be helpful (within the manuscript or in the supplementary material). Selected models capture both spatial and temporal autocorrelation, which is crucial in animal movement data. They allow for an objective distinction between constrained (residency) and unconstrained (dispersal) movement patterns. The choice to resample the GPS data to one location per day is appropriate for identifying broad-scale behavioral phases such as dispersal, rather than focusing on fine-scale hourly movements.

Suggestions:

L134: Most, but not all, pumas were captured along the Tietê River basin. Could the authors clarify where the remaining individuals were captured and whether their movement data were included in the same analyses? Briefly clarify whether all individuals were included in the same analyses, or whether the coastal individual was treated differently due to landscape context.

L152-176: The collar recorded positions hourly, but later analyses resampled to one location per day (line 205). It is not clear whether or how data quality or outlier filtering was applied before model fitting. Briefly mention any filtering protocols or quality control procedures applied to the GPS data.

L180: As mentioned in the Introduction, it would be important to define key concepts such as “dispersal” and “residency.” Since these definitions appear in the following section, the authors could either indicate this by adding “(see next section)” or consider moving the definitions earlier in the text.

L215: Define pre- and post-dispersal phases.

L247: I emphasize the need to explicitly define the three behavioral phases — pre-dispersal, dispersal, and post-dispersal ranging and residency— so that they are clearly stated. This would make the study more practical and accessible for anyone aiming to compare results or replicate the analysis. I recommend including concise definitions of each phase (e.g., time frame, movement characteristics, biological meaning) to make the classification criteria transparent.

L239-257: Although the manuscript mentions covariation between land use types (line 324), there is no mention of steps taken to address multicollinearity in the GLMMs. Authors may add a sentence clarifying whether multicollinearity was tested (e.g., VIF) and how it was handled in model selection.

Results

The section is clearly structured and presents rich, well-analyzed data on puma movement, home range, and dispersal behavior. The figures are informative and support the main findings, and the integration of movement models with landscape features is a notable strength. However, several clarifications and contextual additions would improve the rigor and interpretation of the results:

The results include all 14 individuals, including five that were translocated (lines 282–284), one of which was held in captivity for 8 months prior to release. However, the potential behavioral effects of translocation — particularly for the individual kept in captivity — are not discussed in relation to dispersal timing, movement patterns, or home range size. I suggest that the authors clarify whether all individuals were included in the same analyses, and if so, discuss the potential influence of translocation on their behavior. If any individuals were excluded from specific analyses, this should be clearly stated.

About males and females, while the results report that only males dispersed, there is no discussion of whether this reflects a biological pattern (e.g., female philopatry) or simply a result of the small female sample size. I recommend that the authors address this explicitly — either by referencing known sex-biased dispersal patterns in pumas or by acknowledging the limited inference due to low female representation. Additionally, since many dispersal-related analyses only include males, the potential for sex bias should be clearly stated in the interpretation of results.

Throughout the manuscript, there is occasional ambiguity regarding whether analyses distinguish among the three behavioral phases (pre-dispersal, dispersal, post-dispersal) or group them into two categories (dispersal vs. residency). While the classification is clearly defined in the Methods (line 215), the terminology is inconsistently applied in later sections. For instance, in the analyses of landscape effects on movement and in Figures 4 and 5, it is unclear whether pre- and post-dispersal were analyzed separately or merged as a single residency category. Clarifying this throughout the text and figures would improve transparency and allow for more accurate interpretation and replication.

Suggestions:

L276-281: It is noted that six individuals were residents and eight dispersed. However, it would be helpful to explicitly state whether all 14 individuals were included in the same analyses, or if any were excluded from specific models (e.g., the one individual with incomplete dispersal data, or the translocated animals). Add a clarifying sentence specifying which individuals were used in each analysis and whether translocation status affected inclusion.

L303–309, 331–334: The authors correctly acknowledge variability among individuals, which is an important point. Consider adding a brief summary of how consistent or variable the main movement patterns were across the sample (e.g., how many individuals deviated from the general pattern of faster, more directed movement during dispersal).

L315-317: The text states that all home ranges were modeled using OUF, but it’s not clear if this was because OUF was always the best model, or if other models were rejected based on AICc. Clarify whether OUF was the best-fitting model for all individuals (as per AICc results) or if it was chosen for consistency.

L320-326: While forest and road effects are highlighted, interactions or potential collinearity among land use variables are only briefly mentioned. Consider presenting a simplified summary (or referring to Appendix) that helps interpret these results in light of correlated predictors (e.g., forest vs. sugarcane).

L352-391: The comparison is well executed and valuable. However, the distinction between natal dispersal vs. post-translocation dispersal in the current study should be made clearer. A summary table (in main text or appendix) comparing key metrics (age, distance, method) across studies would be helpful.

Discussion

The discussion provides a strong synthesis of the study’s findings and their relevance in a broader ecological and conservation context. It successfully compares results with the literature and reflects on the implications of land use and behavioral plasticity in Puma concolor.

Suggestions:

L409-414: The authors note the elevated dispersal ages in their sample and suggest possible explanations, including the lack of natal range monitoring and individual translocation. However, these explanations remain speculative and are not discussed in sufficient depth. Expand on how translocation — especially after captivity — may alter dispersal behavior and compromise comparability with natural dispersal events. If possible, state whether translocated individuals displayed systematically different movement patterns.

L415-417: The study includes only three females, none of which dispersed. Although this is acknowledged, the potential implications for understanding sex-biased dispersal, landscape permeability, and population connectivity are underdeveloped. Discuss how this sex bias may affect conclusions and stress the need for future research incorporating a more balanced sex ratio.

L469-471: The use of statistical models (OUF, CVM, AKDE) to delineate behavioral phases is a methodological strength and still rare in the literature. This is mentioned briefly but deserves more emphasis. Highlight more clearly how this methodological framework improves upon the common visual or subjective classification used in most previous studies (as noted in the literature review).

L450-499: The discussion touches on conservation applications, such as corridor design and population monitoring. These points are highly relevant but could be developed further. Consider suggesting how specific movement patterns (e.g., use of pastures at night, forest dependency) could inform conservation strategies — for example, identifying priority zones for connectivity or avoiding road expansion in sensitive corridors.

Author Response

Please see my responses to the comments in the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript is very interesting, well written and definetly fills a void of knowledge both for pumas and landscape use in a fragmented environment.

Although I understand that clustering of data feels necessary fo have a sound sample size, it creates an artificial situation regarding dispersal individuals, in addition to include translocated individuals in a dispersal sample. I would recommend to separate residents, dispersal, and translocated individuals in order to see the effect of the translocation. From table D2, five of the eight dispersal individuals were the result of translocations, reinforcing my coment about considering this a treatment on its own. Older dispersal individuals as stated by the authors may be an artifact of capturing animals that already dispersed and were settled, there is no explanation as to why the animals were captured and then translocated. Two of five individuals, were likely natural dispersals the other three appear to have settled down.

This coupled with the clustering of data for multiple years, dispersal in puma populations usually takes multiple years to gather a large enough sample size and also reflects the opportunistic capture of those individuals that may disperse, which are usually low numbers. The landscape presented by the authors may be a sink for the overall population of pumas, as reflected by the low survival of most individuals. There were no pumas surviving or continued monitoring after a year (with the exception of 1 male), therefore its hard to say if the animals are indeed surviving, particularly with no follow up. Puma populations without harvest tend to have high survival rates, which is not the case for the authors study area.

Model selection may be artifically influenced by the use of a one time landscape image instead of multiple years, or a representation of landscape change, and may need to be redone incluiding time as a covariable. 

The authors need to be consistent on the presentation of data, for example in figure D7, they label female Euclidian dispersal distance, this is not accurate as they said females were resident. 

I would recommend to make a supplementary figure per year (or years in the case of monitoring encompassing a 2 year period), this should reflect the differences not only on time but likely in habitat use. A five year period particularly in a fragmented landscape should help explain changes during such period. 

I would recommend the exploration of the data in those separate categories. And although the information is very valuable it may be highlighting a non-existing pattern. On the review section, I would separate the temperate studies of South America as they will clearly resemble those in North America, as there as similirtaties in habitat estructure. And may not be comparable to those of the tropical Americas.        

I think the authors need to make a clear separation of analysis, another example is figure 3, with the predicted Home Range size in relation to forest proportion and roads. As the authors are including all data in the image, but are not labeling the individuals according to their dispersal or resident status. This would potentially reinforce the hypothesis.

In addition, the discussion and conclusions sections should be shortened.

As I recommend to spli residents, dispersers and translocations, I wont attach a document as the results will likely change. It is information that is necessary to be publish, but looking at the overall batch of results, it needs to have some sort of reanalysis before acceptance. 

 

Author Response

Please see my responses to the comments in the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All reviewers’ comments were addressed thoroughly, clearly, and with strong technical accuracy. The authors responded to each point with care, incorporating relevant suggestions and expanding on key aspects where necessary. These revisions have resulted in a significantly improved manuscript. I recommend the publication of this work in its current form.