Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

Behavioural Time Allocation and Responses to Environmental Enrichment in Zoo-Housed Yellow-Breasted Capuchin Monkeys (Sapajus xanthosternos)

J. Zool. Bot. Gard. 2026, 7(2), 17; https://doi.org/10.3390/jzbg7020017

by Djalma da Nobrega Ferreira¹, Sérgio L. G. Nogueira-Filho^1,2,†

, Guillermina Hernández-Cruz³

, Stella G. C. Lima¹

, Mike Mendl³ and Selene S. C. Nogueira^1,2,*

Reviewer 1:

Guoan Yin

Reviewer 2: Anonymous

Reviewer 3:

Abd Rahman Mohd-Ridwan

Reviewer 4: Anonymous

J. Zool. Bot. Gard. 2026, 7(2), 17; https://doi.org/10.3390/jzbg7020017

Submission received: 7 July 2025 / Revised: 23 March 2026 / Accepted: 26 March 2026 / Published: 2 April 2026

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The large variation range in behavioural distinctiveness among the observed capuchins (from -0.86 to 3.11 for neuroticism) and the limited sample size result in insufficient reliability of the results. And, interfering factors such as sex and age could not be effectively eliminated. Consequently, the conclusions should be more cautiously stated. Besides, the discussion needs further improvement.
However, the study exhibits strong novelty, and the results hold certain value.

Author Response

Comments and Suggestions for Authors

Ferreira et al.: We appreciate the reviewer’s insightful comments and fully agree that the considerable variation in behavioural distinctiveness and the limited sample size may constrain the reliability and generalizability of our findings. We have now explicitly acknowledged these limitations in both the Discussion and Conclusion sections and have rephrased our conclusions more cautiously. Additionally, we expanded the Discussion to better contextualize the influence of sex and age, emphasizing that the observed patterns should be interpreted with caution and viewed as preliminary but informative for future investigations.

The revised sections and the conclusion session now read as follows:

“Overall, our findings demonstrate that environmental enrichment effectively improved welfare in captive yellow-breasted capuchin monkeys by reducing aggression and promoting exploratory behaviours across individuals, regardless of their personality scores. These results reinforce the central role of enrichment in fostering naturalistic behaviours that align with the conservation and educational missions of zoological institutions. While personality traits did not directly predict behavioural responses, individual differences remain crucial for tailoring management strategies that optimise welfare outcomes. Future studies incorporating larger samples and long-term monitoring are needed to assess the stability of these effects and to explore how different enrichment types, social structures, and visitor presence may further enhance welfare and conservation value.”[lines 558-567]

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript describes a study of the behavioural responses of yellow-breasted capuchin monkeys to an environmental enrichment intervention. The study also seeks to assess the influence of individual personality traits on the behavioural responses of individuals to the enrichment intervention. The study is well-written with few formatting problems and the language use is appropriate for the journal. The reference style is incorrect for the journal but this is a minor fix. There are also issues with the definitions of the behavioural categories used however, more concerningly, there are fundamental issues with the study which compromise its scientific validity.

The biggest issue that I find with the study is the inclusion of the personality dimension. The approach to assess personality using the ethogram and observations of the animals during the study is logically flawed. By scoring the behaviour of the animals, using the scored behaviours to derive metrics about personality traits and then correlating these metrics with the scored behaviours, the logic here is circular and confounds any interpretation of the results. An appropriate approach to assessing the impact of personality traits on enrichment intervention efficacy would be to assess personality independently of the experimental context and then link the independently assessed traits to the responses in the experimental context. This is the standard approach to personality measurement and underpins the establishment of standard tests of personality (e.g. novel object test). However, this was not done in this study and the method selected in this study is logically unsound. I would recommend that the authors consider removing the personality aspect of the study entirely or, alternatively, conduct independent personality tests on the subjects and use the scores from these tests to assess the role of personality traits in driving responses to enrichment. There is also the exclusion of a seemingly obvious personality trait (i.e. openness) which requires explanation.

The second issue lies in the analysis of the data. As pointed out in the specific feedback, the authors have run multiple independent similar tests on behavioural category responses; this assumes that each behaviour category is independent of each other when, in actuality, they are inter-dependent and cannot be considered in isolation. This increases risk of Type I error and should be addressed through a re-analysis of the data using a single test instead of multiple tests. It appears that this was done in order to investigate the impact of personality as a modulator of behavioural responses to enrichment but this is statistically problematic.

In light of the above, I am recommending that this study be reconsidered following major revisions. I urge the authors to reassess the study as a whole and reconsider whether the personality aspect is a constructive inclusion in the study. Given the species, a study of enrichment alone would be a valuable inclusion to the literature.

Comments for author File: Comments.pdf

Author Response

Comments and Suggestions for Authors

Ferreira et al: We thank the reviewer for raising these concerns and appreciate the opportunity to clarify our methodology. We agree that personality assessment must be conducted independently of the experimental context to avoid conceptual circularity. We apologise that the temporal separation between the personality assessment and the enrichment experiment was not sufficiently clear in the original submission. Personality Data Were Collected Before the Experimental Manipulation. In the revised manuscript, we explicitly clarify that all behavioural observations used to generate personality scores were conducted 10 days before the onset of the ABA (A1–B–A2) enrichment phases. Thus, personality scoring did not include, overlap with, or derive from any behaviours recorded during the enrichment intervention. To avoid any ambiguity, the revised Methods section now states:

“Behavioural observations used to quantify personality traits were carried out during a 10-day baseline period after a 20-day habituation phase and before the beginning of phase A1 of the ABA design. No enrichment was provided during this period, and these observations were used exclusively for the derivation of personality scores. Behavioural data were gathered through continuous focal animal sampling (Altmann, 1974), conducted twice daily—from 8:00 am to 10:00 am and from 1:00 pm to 3:00 pm.” [lines 150-156]. Thus, the behaviour used to derive personality traits was recorded in a different temporal window, different context, and independent of the experimental treatment.

Moreover, our personality coding closely follows the recommendations of Réale et al. (2007) for deriving trait-level measures from repeated observations in naturalistic contexts when formal test batteries cannot be applied (as is common in primatology, e.g., Freeman et al., 2016; Carter et al., 2012). We further clarify this by explicitly describing that: Aggressiveness/Neuroticism was derived from the summed proportion of time in threats, chases, and agonistic vocalisations; Sociability/Extraversion was derived from social grooming (given/received) and social play.

These behaviours are considered trait-like indicators of the corresponding personality dimensions and were measured across 100 minutes per individual, collected across the baseline period.

Therefore, there is no analytical circularity, because:personality scores were created only from the baseline data, and enrichment responses were analysed only from the A1–B–A2 data. Thus , there is no circularity: the predictor (personality) and outcome variables (behavioural responses to enrichment) come from non-overlapping datasets. We have revised the manuscript text to make this distinction unambiguous. We also added a short justification explaining why classical test batteries (e.g., novel object tests) could not be implemented due to:

restricted access imposed by the species’ conservation status,
strict zoo management protocols limiting animal manipulation,
group-living conditions, which prevent isolating individuals.

We now explicitly state that our approach aligns with validated ethological approaches for assessing primate personality using observational data in naturalistic settings.

On the Number of Behavioural Models (Multiple GLMMs). The reviewer is concerned about Type I error inflation. We appreciate this point and have now:

clarified that the behavioural response variables represent distinct behavioural domains, supported by ethological literature,
acknowledged potential interdependence,
and added a rationale for analysing each behavioural category separately to maintain biological interpretability (rather than collapsing behaviours into a composite index, which would obscure domain-specific effects).

Because we analysed six biologically meaningful response variables and we used:

GLMMs with random effects controlling for repeated measures,
models designed to test specific hypotheses for each behavioural domain,
and significance threshold (P < 0.05),

we argue that the analysis strategy remains sound and interpretable.

We added a paragraph explicitly stating this reasoning in the revised Data Analysis section.

Regarding, the statistical Analyses (Clarified and Strengthened)

We added the following sentence near the beginning of the Data Analyses section:

“Because the personality scores and enrichment-response behaviours were derived from non-overlapping datasets collected in different phases, personality acted as an independent predictor rather than a variable confounded with the behavioural response.” [lines 213-215]

And added the following rationale regarding multiple models:

“Each behavioural response variable corresponded to a distinct ethological domain (aggression, exploration, inactivity, abnormal behaviour, affiliative behaviour, and behavioural diversity). These behaviours are known to reflect different motivational systems and are not interchangeable; thus, separate GLMMs were used to preserve biological interpretability (Martin & Bateson, 2007). While some degree of interdependence among behaviours is possible, the set of six models represents a modest number of analyses and each addressed a clearly articulated hypothesis.”[lines 237 – 241]

Reviewer 3 Report

Comments and Suggestions for Authors

This is a strong paper with important findings. The reduction in aggression and increase in exploration during enrichment are very valuable for improving zoo management and supporting conservation goals of capuchin in captivity

Comments for author File: Comments.pdf

Author Response

Comments and Suggestions for Authors

Ferreira et al. :Thank you very much for your comment.

Reviewer 4 Report

Comments and Suggestions for Authors

General Comments:

The manuscript raises a significant question regarding behavioural distinctiveness in responses to environmental enrichment in Sapajus xanthosternos. I found that the study is conceptually fine with appropriate methods and contributes to the broader field of zoo animal welfare and conservation management. Key strengths include sample size, application of a nice enrichment design, and use of behavioural coding. However, I see a need for clarification of the hypotheses, streamlining the methodological description, and sharpening both the presentation of results and the discussion. Providing stronger connections to practical welfare and conservation management would increase the impact of the paper.

Specific comments:

The abstract is overly long.

Lines 21–27: The research aim could be phrased more directly like “We tested whether behavioural distinctiveness (neuroticism, extraversion) predicted individual variation in responses to environmental enrichment.”

Lines 34–39: Condense the reporting of results to only the major outcomes (aggression reduction, exploration increase, and no predictive effect of traits). Consider moving statistical details to the Results section.

Lines 45–49: Simplify to highlight the broader conclusion: enrichment benefits were consistent across individuals regardless of personality scores.

Suggestion: End the abstract with a stronger applied message, such as how the results could guide enrichment planning in zoos or inform reintroduction programmes.

Lines 54–66: The literature review is comprehensive but heavy with references.

Lines 80–83: Remove redundant use of “Therefore.”

Lines 102–121: Summarize hypotheses more clearly at the end of the section.

Highlight the novelty of the study and emphasize its potential contribution to both welfare and conservation outcomes.

Lines 128–169: I would consider summarizing the enclosure features rather than lengthy prose and include size, substrate, visitor distance, and existing enrichment.

Line 131–132: Giving some justification of the age categories, given their extended developmental period.

Lines 180–190: Move the explanation of how neuroticism and extraversion (Lines 232–237) to this section for clarity.

Lines 191–204: I would provide a rationale why 10 days per phase were chosen and address whether enrichment effects could persist into A2.

Lines 205–210: Clarify the sampling design, were the 10-min focal samples for trait assessment conducted separately from the 5-min samples?

Lines 232–237: I suggest moving trait calculation methods earlier.

I suggest explicitly reporting the intra- or inter-observer reliability.

Affiliative behaviours are limited to grooming and play. Either justify this restriction or expand to include proximity/approach behaviours. For clarity, mark which behaviours contributed to each personality dimension (neuroticism, extraversion) directly in the ethogram. Clarify whether exploratory and affiliative behaviours were weighted equally when calculating diversity. However, the presentation of the behavioural diversity results is still poor.

Lines 286–290: Discuss potential group-level drivers (social hierarchy, enclosure features) that may explain correlation differences.

Lines 292–299: Emphasize aggression reduction as it is the key enrichment outcome.

Lines 310–315: Present exploration increases graphically for clarity.

Lines 323–327: Highlight age-related inactivity in the discussion as a welfare implication.

Lines 338–341: Interpret low abnormal behaviour carefully because it may indicate strong baseline welfare conditions.

Lines 343–357: Simplify affiliative behaviour findings, emphasize adult–juvenile differences, not complex regression outputs.

Lines 361–373: focus only on methodological differences.

Lines 374–379: I would highlight the absence of sex differences in aggression, it is highly important

Lines 381–391: I suggest expanding the discussion of the positive correlation between extraversion and neuroticism; consider whether dominance or group stress could explain this.

Lines 401–423: Rethink practical recommendations but highlight applied relevance (e.g., enrichment planning, reintroduction suitability).

Lines 424–443: I suggest emphasising on broader conservation and welfare implications.

Lines 445–456: Interpret the low stereotypy levels as evidence of effective baseline management.

Lines 457–477: I suggest shortening the message on behavioural diversity.

Lines 478–489: Connect age-related affiliative behaviour to practical zoo management (e.g., social group design).

Suggestion: Conclude the discussion with a short paragraph outlining specific options for future research (e.g., testing other enrichment types, long-term trait stability, effects of visitor presence).

The conclusion could be a single paragraph emphasising that enrichment improved welfare across individuals regardless of personality scores, while individual differences remain relevant for management strategies.

Author Response

R4:

Comments and Suggestions for Authors

General Comments:

Ferreira et al.: Thank you for your time and suggestions that we accepted and reworded in this new version.

Specific comments:

R4:The abstract is overly long.

Ferreira et al.: Thank you, we accepted your suggestion and reworded as follows:

“We tested whether behavioural distinctiveness (neuroticism, extraversion) predicted individual variation in responses to environmental enrichment (EE) components, with the aim of refining husbandry practices for yellow-breasted capuchin monkeys (Sapajus xanthosternos) in zoological settings. Employing the standard ethological approach of behavioural coding, we observed 20 capuchins housed in three groups comprising adult and juvenile males and females. Behavioural profiles were categorised into two distinctiveness dimensions: neuroticism and extraversion. To evaluate individual engagement with EE, we applied the ABA paradigm, wherein phases A1 and A2 (controls) represented standard zoo routines, while phase B corresponded to the implementation of an EE programme. Each phase spanned 10 days, and behavioural data were collected via focal animal sampling (10 minutes per animal per day), resulting in a total of 1,200 focal observations. During the environmental enrichment phase, capuchin monkeys exhibited a clear reduction in aggressive interactions and a concurrent increase in exploratory activity, indicating a positive welfare response to enrichment. However, individual personality traits did not predict behavioural change: neither neuroticism nor extraversion scores were significantly associated with variations in aggression or exploration. These findings suggest that, while enrichment effectively promoted more positive and naturalistic behaviours, individual behavioural profiles may not reliably predict responsiveness to such interventions. From an applied perspective, our findings can guide zoo management strategies by identifying enrichment elements that foster naturalistic behaviours and social stability, ultimately enhancing welfare standards and informing pre-release preparation protocols in reintroduction programmes.” [lines 25-45]

R4:Lines 34–39: Condense the reporting of results to only the major outcomes (aggression reduction, exploration increase, and no predictive effect of traits). Consider moving statistical details to the Results section.

Ferreira et al.: We thank the reviewer for this helpful suggestion. In accordance with the recommendation, we have condensed this section of the Discussion to highlight only the main behavioural outcomes—namely, the reduction in aggression, the increase in exploratory behaviour, and the absence of a predictive effect of personality traits. Statistical details and secondary behavioural measures have been moved to the Results section for clarity.

The revised paragraph now reads as follows: “…However, individual personality traits did not predict behavioural change: neither neuroticism nor extraversion scores were significantly associated with variations in aggression or exploration. These findings suggest that, while enrichment effectively promoted more positive and naturalistic behaviours, individual behavioural profiles may not reliably predict responsiveness to such interventions. From an applied perspective, our findings can guide zoo management strategies by identifying enrichment elements that foster naturalistic behaviours and social stability, ultimately enhancing welfare standards and informing pre-release preparation protocols in reintroduction programmes.” [lines 37-45]

R4:Lines 45–49: Simplify to highlight the broader conclusion: enrichment benefits were consistent across individuals regardless of personality scores.

Suggestion: End the abstract with a stronger applied message, such as how the results could guide enrichment planning in zoos or inform reintroduction programmes.

Ferreira et al.: Yes, thank you very much. We reworded as follows: “…These findings suggest that, while enrichment effectively promoted more positive and naturalistic behaviours, individual behavioural profiles may not reliably predict responsiveness to such interventions. From an applied perspective, our findings can guide zoo management strategies by identifying enrichment elements that foster naturalistic behaviours and social stability, ultimately enhancing welfare standards and informing pre-release preparation protocols in reintroduction programmes.” [lines 39- 45]

R4:Lines 54–66: The literature review is comprehensive but heavy with references.

Ferreira et al.: We appreciate the reviewer’s observation regarding the density of references in the literature review. We agree that the section includes a substantial number of citations; however, this approach was intentional to provide a comprehensive overview of the existing research in this interdisciplinary field. Given the limited number of studies addressing personality and enrichment interactions in Sapajus species, we considered it essential to contextualize our study within broader primate welfare and behavioural frameworks. Nevertheless, we have carefully reviewed the section to ensure that all references included are directly relevant and contribute meaningfully to the background and rationale of the study.

R4:Lines 80–83: Remove redundant use of “Therefore.”

Ferreira et al.: Thank you, done.

R4:Lines 102–121: Summarize hypotheses more clearly at the end of the section.

Highlight the novelty of the study and emphasize its potential contribution to both welfare and conservation outcomes.

Ferreira et al.: We appreciate this valuable suggestion. We revised the end of the section to summarise our main hypotheses more clearly and to highlight both the novelty and applied significance of our study. The revised paragraph now explicitly outlines the predicted effects of environmental enrichment and clarifies how our findings may contribute to welfare improvement and conservation management for endangered yellow-breasted capuchins.

The new version was reworded as follows: “…we hypothesised that responses to environmental enrichment would vary according to individual behavioural traits. Using a standard ethological approach (behavioural coding, sensu Vazire et al., 2007), we described behavioural distinctiveness in two main dimensions: neuroticism (reflecting aggressiveness) and extraversion (reflecting sociability and playfulness). Given the pronounced sexual dimorphism in Sapajus spp. (Fragaszy et al., 2016), we expected males and females to differ in neuroticism scores, but not in relation to age class, as behavioural distinctiveness is typically stable across time and contexts (Réale et al., 2007). Specifically, we predicted that: 1. Environmental enrichment would reduce aggression, particularly among individuals with higher neuroticism scores. 2. Individuals scoring high in extraversion would show increased exploratory behaviour and behavioural diversity, and reduced abnormal or inactive behaviour, following enrichment. 3. Highly extraverted individuals would also exhibit more affiliative interactions during the enrichment phase. This study represents the first detailed assessment of personality–enrichment interactions in S. xanthosternos, an endangered primate species. By linking personality traits to enrichment responses, our findings are expected to advance the design of welfare-oriented management practices and inform conservation strategies, including behavioural preparation for reintroduction programmes.” [lines 85-101]

R4:Lines 128–169: I would consider summarizing the enclosure features rather than lengthy prose and include size, substrate, visitor distance, and existing enrichment.

Ferreira et al.: Yes, thank you. We summarize as follows:

“G1 was housed in a 47 m² (8.4 m length x 5.6 m width x 4.7 m height) enclosure with a dirt floor and surrounded and covered by a wired-mesh fence supported by concrete and iron bars. This enclosure was in a non-visited area of the zoo because this group of animals were being rehabilitated for a reintroduction programme. G2 occupied a 72m² enclosure divided into two areas: backstage area of 55.3m² (15.3 m length x 6.1 m width x 4.6 m high) and the exhibition area of 16.7 m² (4.9 m length x 3.5 m width x 2.6 m high). The exhibition area had a cement floor and concrete walls, tile roofing and a reinforced glass wall for exhibiting animals to zoo visitors. In this enclosure, the minimum distance between animals and visitors was 3.0 m. The backstage area was connected to the exhibition area by a door (0.8 m width x 2.1 m high), and the animals could choose if they wanted or not to have contact with the public. G3 was kept on an artificial island of approximately 120m² (12.0 m length x 10.0 m width), surrounded by an artificial stream (3 m width x 0.5 m depth). Outside the island area, there were four jackfruit trees (Artocarpus heterophyllus) and two cashew trees (Spondias mombin). In this enclosure, the minimum distance between animals and visitors was 6.0 m. “[lines 123-136]

R4:Line 131–132: Giving some justification of the age categories, given their extended developmental period.

Ferreira et al.: We appreciate the reviewer’s insightful comment. We have now included a justification for the adopted age categories. Specifically, we clarified that the classification followed Silva et al. (2016) but was also based on the developmental milestones and behavioural maturity described for Sapajus species, considering both physical development and independence from maternal care. The revised text now provides this rationale. We added a justification in this new version as follows: “This classification was adopted because Sapajus species exhibit an extended developmental period, but individuals around five years old already display adult body size, secondary sexual characteristics, and stable participation in adult social interactions, indicating functional maturity (Fragaszy et al., 2004; Izar et al., 2012). In contrast, individuals aged 9–12 months are fully weaned but not yet independent, corresponding to the early juvenile stage when exploratory and social play behaviours are most prominent.” [lines 112-117]

R4:Lines 180–190: Move the explanation of how neuroticism and extraversion (Lines 232–237) to this section for clarity.

Ferreira et al.: We thank the reviewer for this helpful suggestion. We have now added a brief explanation in the earlier section where the behavioural states are first introduced (former line 190). In this revised paragraph, we describe the conceptual framework used to classify behaviours into the two personality-related dimensions (aggressiveness/neuroticism and sociability/extraversion), and we now refer to Table 1 for clarity.

However, following standard practice and to maintain methodological transparency, we have retained the detailed explanation originally provided in lines 232–237 in the Data Handling and Analysis subsection. These lines contain essential technical information about how the framework was implemented during data processing (e.g., computation of proportion scores and z-transformation). This separation ensures that conceptual definitions are presented alongside the behavioural ethogram, while methodological procedures remain in the dedicated analysis section. We believe this restructuring improves clarity without compromising methodological transparency.

In this new version we added as follows: “These states were categorised into two behavioural distinctiveness dimensions following Réale et al. (2007): (1) aggressiveness/neuroticism, comprising threat, chase, and agonistic vocalisations; and (2) sociability/extraversion, encompassing the performance or reception of social grooming and engagement in social play. These categories provided the conceptual framework guiding our subsequent personality scoring procedures (see Table 1), while the detailed analytical implementation is described in the Data Analysis section.”[lines 220-226]-

R4:Lines 191–204: I would provide a rationale why 10 days per phase were chosen and address whether enrichment effects could persist into A2.

Ferreira et al.: We appreciate the reviewer’s observation. The 10-day duration per phase was determined by logistical and ethical constraints set by the hosting zoo, which limited the time animals could be exposed to experimental manipulations. Nevertheless, previous studies have shown that short-term enrichment periods of similar duration are sufficient to elicit measurable behavioural changes in captive primates. In addition, the inclusion of a second control phase (A2) allowed us to verify whether the behavioural effects observed during the enrichment phase persisted after the removal of enrichment stimuli.

In this new version we reworded as follows: “All enrichment items were introduced simultaneously into the enclosure on the first day of phase B, remaining for 10 days, being replaced when necessary and removed by the keeper on the last day of this phase. The duration of 10 days per phase (A1, B, A2) was defined according to the logistical constraints and animal management schedules established by the zoo. This period was considered sufficient to detect short-term behavioural adjustments to enrichment interventions, as previously reported in similar studies with captive primates [Albanese et al. 2021]. The subsequent control phase (A2) was designed to evaluate whether behavioural changes observed during enrichment persisted once the stimuli were removed, allowing us to assess potential carry-over effects from phase B. Throughout all phases, the feeding routine was maintained according to the times and locations already defined by the zoo and described above.” [lines 178 - 188 ]

R4:Lines 205–210: Clarify the sampling design, were the 10-min focal samples for trait assessment conducted separately from the 5-min samples?

Ferreira et al.: We thank the reviewer for the opportunity to clarify this point. The 10-minute focal samples were conducted exclusively during the personality assessment phase, prior to the onset of the environmental enrichment experiment. The 5-minute focal samples, in turn, were part of the enrichment study (phases A1, B, and A2) and followed the same sampling principles but were shorter in duration and collected in different experimental contexts. Therefore, these two datasets were collected independently and correspond to distinct phases and objectives within the study.

We clarified this point in this new version as follows: “The 10-minute focal samples described above were conducted exclusively for the assessment of individual behavioural traits before the enrichment experiment. Subsequently, during the environmental enrichment phases (A1, B, A2), independent focal samples of 5 minutes were collected following the same sampling principles but with distinct experimental aims (see below).” [lines 161 -165]

R4:Lines 232–237: I suggest moving trait calculation methods earlier.

Ferreira et al.: We appreciate the reviewer’s suggestion to move the description of the trait calculation methods earlier in the text. However, we prefer to maintain the current order, as it follows a logical sequence of procedures—from behavioural recording and coding to the subsequent computation of trait scores. This structure ensures that readers can clearly follow how the behavioural data were obtained and processed before understanding how the traits were calculated. We believe this organization improves the clarity and coherence of the methodological narrative.

R4:: I suggest explicitly reporting the intra- or inter-observer reliability.

Ferreira et al.: We appreciate the reviewer’s suggestion. In our study, all behavioural data were coded by a single experienced observer who was blind to the study objectives, in order to minimise observer bias and ensure internal consistency. Therefore, no formal intra- or inter-observer reliability test was conducted. Nevertheless, the use of a single trained observer and continuous focal sampling protocol has been shown to provide high reliability in behavioural studies of non-human primates (Altmann, 1974; Martin & Bateson, 2007).

We clarified in this new version adding a sentence as follows: “All behavioural data were coded by a single experienced observer blind to the study objectives, ensuring internal consistency of the dataset (Altmann, 1974; Martin & Bateson, 2007).” [lines 266-268]

R4: Affiliative behaviours are limited to grooming and play. Either justify this restriction or expand to include proximity/approach behaviours. For clarity, mark which behaviours contributed to each personality dimension (neuroticism, extraversion) directly in the ethogram. Clarify whether exploratory and affiliative behaviours were weighted equally when calculating diversity. However, the presentation of the behavioural diversity results is still poor.

Ferreira et al.: We appreciate the reviewer’s comment regarding behavioural diversity. Behavioural diversity indices (i.e., Shannon H indices) were calculated using Past version 4.07. The Shannon H index is calculated using the formula H = - ∑ (p_iln (p_i), where p_iis the proportion of each behavioural category in the sample (e.g., foraging) and ln is the natural logarithm. Hence, all behavioural categories included in the analysis are weighted equally. We excluded the behavioural categories inactivity and abnormal behaviour from the analysis as the calculation should include species-typical behaviours exclusively and omit inactivity and repetitive or abnormal behaviours (Miller at al., 2020).

We have modified Section 2.4 in the Material and Methods to reflect this as follows: “We calculated the Shannon-Weaver Diversity Index (SWI) or H-index using Past software version 4.07 for the enriched and unenriched subjects per observation day according to the observation phases (A1, B, A2). When calculating behavioural diversity (H-index), only species-typical behaviours were included (Miller et al., 2020); therefore, abnormal repetitive behaviour (pacing) and inactivity were excluded from the behavioural diversity analysis. [lines 206-211]

R4:Lines 286–290: Discuss potential group-level drivers (social hierarchy, enclosure features) that may explain correlation differences.

Ferreira et al.: We appreciate the reviewer’s valuable comment. We agree that group-level factors such as social hierarchy and enclosure characteristics may have contributed to the behavioural variation observed. We have now addressed this point in the Discussion, noting that dominance relationships and enclosure layout could have influenced both individual aggression levels and their correlations with personality traits, particularly neuroticism.

We added in discussion as follows: “It is important to note that group-level variables may also underlie part of the behavioural variation observed. In primate groups, dominance rank and social stability can modulate the expression of aggressive and affiliative behaviours (de Waal, 1986; Sapolsky, 2005). In our study, although the groups were stable and the sample size limited formal analysis of hierarchy effects, differences in dominance relationships could have influenced the expression of aggression and, consequently, the strength of its correlation with neuroticism. Additionally, enclosure-related features—such as the availability of refuges, visual barriers, and spatial distribution of resources—may have affected opportunities for avoidance or competition, thereby shaping group-level behavioural dynamics (Hosey et al., 2009; Young et al., 2012). These contextual effects should be explored in future studies to disentangle individual personality influences from group-level ecological and social factors.” [lines 411-420]

R4:Lines 292–299: Emphasize aggression reduction as it is the key enrichment outcome.

Ferreira et al.: We thank the reviewer for this insightful suggestion. We have now revised the Discussion to more clearly highlight the reduction in aggressive behaviour as the key outcome of the enrichment intervention, emphasizing its relevance as an indicator of improved welfare and social stability among the yellow-breasted capuchins.

We added in the discussion of this new version as follows: “The reduction in aggressive behaviour observed during and after the enrichment phase represents the central outcome of our study, providing clear evidence of the positive effects of environmental enrichment on the welfare of yellow-breasted capuchins. Lower aggression levels are widely recognised as indicators of reduced social tension and improved psychological well-being in captive primates (Boinski et al., 1999; Baker, 2016). Notably, the persistence of this reduction into the subsequent control phase (A2) suggests that the enrichment effects extended beyond the immediate exposure period, potentially promoting more stable social dynamics and decreased stress within the groups.” [lines 384-391]

R4:Lines 310–315: Present exploration increases graphically for clarity.

Ferreira et al.: We thank the reviewer for this suggestion. The exploration increases are already presented in the manuscript in Figure 4, which illustrates the changes graphically for clarity. We hope this figure adequately addresses the reviewer’s point.

R4:Lines 323–327: Highlight age-related inactivity in the discussion as a welfare implication.

Ferreira et al.: We thank the reviewer for this insightful comment. We have now highlighted the age-related differences in inactivity in the Discussion section and emphasized its potential implications for welfare.

We added in our new discussion as follows: “Our finding that adult yellow‑breasted capuchin monkeys (Sapajus xanthosternos) spent a significantly greater proportion of time inactive (16.9 % ± 1.2) compared to juveniles (10.4 % ± 2.0) suggests an age‑related shift in activity budget that may have important welfare implications. In captivity, elevated inactivity has been associated with reduced behavioural diversity and possibly compromised welfare states (e.g., increased anxiety indicators, decreased affiliative behaviours) in ageing primates. For example, elderly captive chimpanzees exhibited more inactivity and less behavioural diversity compared to younger individuals (Webb et al. 2019)

Moreover, inactivity in captive animals has been proposed as a potential indicator of affective state (positive or negative) and welfare risk, especially when it represents a deviation from species‑typical active behaviours (Fureix and Meagher 2015).

From a practical zoo‑welfare perspective, this age‑related increase in inactivity suggests that enrichment and management programmes should explicitly consider the needs of older individuals. Enrichment devices, foraging opportunities and structural complexity might need to be adapted or intensified to engage older capuchins, thereby maintaining activity levels, promoting behavioural diversity and reducing the potential welfare risk associated with inactivity. Such an approach is consistent with geriatric care recommendations for captive primates that emphasise increasing behavioural opportunities for older individuals. Novak & Suomi 1988]. In summary, the age class effect we observed is not simply a demographic footnote, but a key indicator of welfare‑relevant differences in how zoo‑housed capuchins engage with their environment. Future work should explore how these inactivity differences relate to physiological health, enrichment uptake and individual welfare outcomes in older versus younger animals. [lines 504 - 524]

R4:Lines 338–341: Interpret low abnormal behaviour carefully because it may indicate strong baseline welfare conditions.

Ferreira et al.: We thank the reviewer for this valuable comment. We agree that the low frequency of abnormal behaviour observed in our subjects should be interpreted carefully, as it may reflect strong baseline welfare conditions rather than an absence of welfare challenges. We have now revised the Discussion to clarify this interpretation and to acknowledge that low levels of abnormal behaviour are consistent with adequate environmental and social conditions for the animals.

We discussed better in this new version as follows: “The low and stable frequency of abnormal or stereotypical behaviours observed across the three study phases (below 2%) should be interpreted cautiously. Rather than indicating the ineffectiveness of enrichment, such low rates may reflect a generally high baseline welfare state within the group. Abnormal behaviours in captive primates often emerge under conditions of chronic stress, social instability, or impoverished environments (Mason et al., 2007; Hosey et al., 2020). Their scarcity in this population therefore suggests that the housing conditions and management practices already supported behavioural needs effectively. Similar findings have been reported in other well-managed zoo populations, where low levels of stereotypy were considered an indicator of good welfare rather than a lack of behavioural responsiveness (Brent & Eichberg, 1991; Kuhar, 2008). Nonetheless, caution is warranted when interpreting low frequencies of abnormal behaviour, as such patterns might also arise when opportunities for their expression are limited by enclosure design or when observation periods are too short to detect subtle forms of behavioural discontent (Mason & Latham, 2004). Future studies combining behavioural and physiological indicators would help confirm whether the low occurrence of stereotypies indeed corresponds to high welfare states in Sapajus xanthosternos.” [lines 457-470]

R4:Lines 343–357: Simplify affiliative behaviour findings, emphasize adult–juvenile differences, not complex regression outputs.

Ferreira et al.: We thank the reviewer for this helpful suggestion. We have simplified the presentation of affiliative behaviour findings to emphasize the clear differences between adults and juveniles, avoiding detailed regression outputs. The revised text now highlights the biological relevance of age-related patterns rather than statistical complexity.

We reworded the results in this new version as follows: “Age class influenced the percentage of time yellow-breasted capuchin monkeys engaged in affiliative behaviour (F₁,₄₀.₁₅ = 7.51, P = 0.009). Adults spent a significantly greater proportion of time in affiliative interactions (6.4 % ± 0.6) compared to juveniles (4.1 % ± 1.1). This suggests that social bonding and grooming behaviours are more prominent among adults, possibly reflecting their more established social relationships within the group. No other factors significantly affected affiliative behaviour (see Supplementary Materials for full statistical outputs).” [lines 334-339]

We also add some discussion as follows: “The higher levels of affiliative behaviour observed in adults compared to juveniles are consistent with patterns reported for other Sapajus species, where mature individuals invest more in social grooming and affiliative exchanges to maintain group cohesion (Izar et al., 2012; Tiddi et al., 2015). This age-related difference may also indicate that social integration strengthens with maturity, a factor that can influence how individuals engage with enrichment opportunities and respond to environmental change.” [lines 451-456]

R4:Lines 361–373: focus only on methodological differences.

Ferreira et al.: We reworded, basing our discussion on methods used as follows: “However, Delval et al. (2020) find a decrease in sociability with age. The difference between studies may be due to the methods used in both studies. Delval (2019) included behavioural categories mostly related to social interactions (i.e., grooming, lip smack, proximity, touch and nurse) for the individual behavioural distinctiveness category named as sociability-attractiveness, but also included the patterns do not move, forage, and vigilant. In contrast, we just considered the behavioural patterns of social grooming and social play in the sociability/extraversion behavioural distinctiveness dimension.” [lines 346-352]

R4:Lines 374–379: I would highlight the absence of sex differences in aggression, it is highly important

Ferreira et al.: Yes, thank you. We added this discussion as follows: “Contrary to expectations, no sex-based differences were observed, as aggressiveness/neuroticism scores did not differ between males and females. This outcome was unexpected, given that previous research (Fragaszy et al., 2016) suggests that the greater size and body mass of adult males, compared to adult females, may influence aggressive behavioural expression. The findings reported here may be attributable to social dynamics and resource distribution within the enclosures, which may have mitigated the anticipated variation in aggressive behaviour between sexes. However, this explanation warrants further empirical investigation to confirm its validity.” [lines 353-359]

R4:Lines 381–391: I suggest expanding the discussion of the positive correlation between extraversion and neuroticism; consider whether dominance or group stress could explain this.

Ferreira et al.: We thank the reviewer for this insightful suggestion. We have now expanded the Discussion to consider whether dominance relationships and group-level stress dynamics could underlie the positive correlation observed between extraversion and neuroticism. The revised paragraph also integrates literature on the behavioural and social correlates of these personality dimensions in primates.

We reworded as follows:

“…On the other hand, we identified a positive correlation between the behavioural dimensions of neuroticism and extraversion. This pattern aligns with previous studies in primates, where individuals that are highly sociable and active also tend to show greater emotional reactivity or tension (Weiss et al., 2017; Morton et al., 2013). One possible explanation is that more extraverted individuals, due to their frequent social engagement and assertive behaviours, may also occupy more dominant or central positions within their social networks. Dominance often entails greater responsibility for maintaining access to resources and social stability, which can be accompanied by elevated vigilance and stress levels (Sapolsky, 2005; Massen et al., 2013). Thus, individuals who are more socially interactive may also exhibit heightened neuroticism as a consequence of their increased involvement in social conflicts or competitive interactions.

Alternatively, group-level factors such as overall social tension or environmental unpredictability may accentuate both extraverted and neurotic tendencies within the same individuals. Under stressful or competitive group conditions, socially active animals might display greater reactivity or aggression when managing social relationships (Sussman et al., 2013; Higham & Maestripieri, 2014). Therefore, the observed positive correlation between extraversion and neuroticism in Sapajus xanthosternos could reflect an adaptive balance between sociability and emotional responsiveness, shaped by the social structure and stress dynamics of the captive groups.” [lines 360-376]

R4:Lines 401–423: Rethink practical recommendations but highlight applied relevance (e.g., enrichment planning, reintroduction suitability).

Ferreira et al.: Thank you for your suggestion, we added a new paragraph as follows:

“From a management standpoint, promoting exploratory and social behaviours aligns with broader conservation and educational objectives, as animals expressing positive welfare states are more likely to thrive in breeding programmes, participate successfully in reintroduction efforts, and elicit positive public perceptions [Verbeke, 2015; Packer et al., 2018; Paulino et al., 2018]. Thus, integrating personality-informed enrichment design and behavioural monitoring provides a practical framework for both improving animal welfare and enhancing the applied conservation value of captive populations. Moreover, our results support the development of enrichment protocols that vary in complexity and predictability according to individual behavioural profiles, alongside regular behavioural assessments to guide social grouping and pre-release conditioning. Implementing rotational enrichment schedules, offering socially interactive devices, and gradually increasing environmental challenges may enhance adaptability and stress resilience—key traits for reintroduction success. Therefore, personality-based management offers a feasible and evidence-driven strategy to align animal welfare practices with long-term conservation outcomes [Melfi, 2013; Carter et al., 2013; Watters, 2014].

Future research should aim to assess the long-term stability of individual behavioural traits in relation to environmental enrichment, as well as to test a wider range of enrichment types—including cognitive, sensory, and social stimuli—to determine their differential effects on welfare and behavioural flexibility. Additionally, examining how visitor presence and enclosure design interact with personality and enrichment responses would provide valuable insights for evidence-based management in zoological settings. Such studies would contribute to developing more refined, personality-informed welfare strategies that enhance both individual well-being and the broader conservation value of captive primate populations [lines 537-557]

R4:Lines 424–443: I suggest emphasising on broader conservation and welfare implications.

Ferreira et al.: Thank you very much, we add these aspects as follows: “From a management standpoint, promoting exploratory and social behaviours aligns with broader conservation and educational objectives, as animals expressing positive welfare states are more likely to thrive in breeding programmes, participate successfully in reintroduction efforts, and elicit positive public perceptions (Verbeke, 2015; Packer et al., 2018; Paulino et al., 2018). Thus, integrating personality-informed enrichment design and behavioural monitoring provides a practical framework for both improving animal welfare and enhancing the applied conservation value of captive populations. Moreover, our results support the development of enrichment protocols that vary in complexity and predictability according to individual behavioural profiles, alongside regular behavioural assessments to guide social grouping and pre-release conditioning. Implementing rotational enrichment schedules, offering socially interactive devices, and gradually increasing environmental challenges may enhance adaptability and stress resilience—key traits for reintroduction success. Therefore, personality-based management offers a feasible and evidence-driven strategy to align animal welfare practices with long-term conservation outcomes (Melfi, 2013; Carter et al., 2013; Watters, 2014).” [lines 537-549]

R4:Lines 445–456: Interpret the low stereotypy levels as evidence of effective baseline management.

Ferreira et al.: Yes, thank you, following you and other reviewers, we added this discussion as follows: “The low and stable frequency of abnormal or stereotypical behaviours observed across the three study phases (below 2%) should be interpreted cautiously. Rather than indicating the ineffectiveness of enrichment, such low rates may reflect a generally high baseline welfare state within the group. Abnormal behaviours in captive primates often emerge under conditions of chronic stress, social instability, or impoverished environments (Mason et al., 2007; Hosey et al., 2020). Their scarcity in this population, therefore, suggests that the housing conditions and management practices already supported behavioural needs effectively. Similar findings have been reported in other well-managed zoo populations, where low levels of stereotypy were considered an indicator of good welfare rather than a lack of behavioural responsiveness (Brent & Eichberg, 1991; Kuhar, 2008). Nonetheless, caution is warranted when interpreting low frequencies of abnormal behaviour, as such patterns might also arise when opportunities for their expression are limited by enclosure design or when observation periods are too short to detect subtle forms of behavioural discontent (Mason & Latham, 2004). Future studies combining behavioural and physiological indicators would help confirm whether the low occurrence of stereotypies indeed corresponds to high welfare states in Sapajus xanthosternos.” [lines 457-471]

R4:Lines 457–477: I suggest shortening the message on behavioural diversity.

Ferreira et al.: Yes, thank you for your suggestion, we shortened the message cutting the sentence “We calculated the Shannon-Weaver Diversity Index [SWI] or H-index using Past software version 4.07 for the enriched and unenriched subjects per observation day according to the observation phases [A1, B, A2]. When calculating behavioural diversity [H-index], only species-typical behaviours were included [Miller et al., 2020]; therefore, abnormal repetitive behaviour [pacing] and inactivity were excluded from the behavioural diversity analysis.” [lines 206-211]

R4:Lines 478–489: Connect age-related affiliative behaviour to practical zoo management (e.g., social group design).

Ferreira et al.: Thank you for your suggestion; we added in this new version as follows: “From a practical zoo‑welfare perspective, this age-related increase in inactivity suggests that enrichment and management programmes should explicitly consider the needs of older individuals. Enrichment devices, foraging opportunities and structural complexity might need to be adapted or intensified to engage older capuchins, thereby maintaining activity levels, promoting behavioural diversity and reducing the potential welfare risk associated with inactivity. Such an approach is consistent with geriatric care recommendations for captive primates that emphasise increasing behavioural opportunities for older individuals. (Suedmeyer et al 2024. In summary, the age class effect we observed is not simply a demographic footnote, but a key indicator of welfare‑relevant differences in how zoo‑housed capuchins engage with their environment. Future work should explore how these inactivity differences relate to physiological health, enrichment uptake and individual welfare outcomes in older versus younger animals.” [lines 514-524]

R4:Suggestion: Conclude the discussion with a short paragraph outlining specific options for future research (e.g., testing other enrichment types, long-term trait stability, effects of visitor presence).

Ferreira et al.: Thank you for your suggestion. We added options for future research as follows: “Future research should aim to assess the long-term stability of individual behavioural traits in relation to environmental enrichment, as well as to test a wider range of enrichment types—including cognitive, sensory, and social stimuli—to determine their differential effects on welfare and behavioural flexibility. Additionally, examining how visitor presence and enclosure design interact with personality and enrichment responses would provide valuable insights for evidence-based management in zoological settings. Such studies would contribute to developing more refined, personality-informed welfare strategies that enhance both individual well-being and the broader conservation value of captive primate populations.” [lines 550-557]

R4: The conclusion could be a single paragraph emphasising that enrichment improved welfare across individuals regardless of personality scores, while individual differences remain relevant for management strategies.

Ferreira et al.: Thank you for your suggestion that we agree, the text in this new version was modified as follows: “Overall, our findings demonstrate that environmental enrichment effectively improved welfare in captive yellow-breasted capuchin monkeys by reducing aggression and promoting exploratory behaviours across individuals, regardless of their personality scores. These results reinforce the central role of enrichment in fostering naturalistic behaviours that align with the conservation and educational missions of zoological institutions. While personality traits did not directly predict behavioural responses, individual differences remain crucial for tailoring management strategies that optimise welfare outcomes. Future studies incorporating larger samples and long-term monitoring are needed to assess the stability of these effects and to explore how different enrichment types, social structures, and visitor presence may further enhance welfare and conservation value.” [lines 561-570]

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The discussion still needs to be further revised to more fully explain the influence of various factors, such as sex and age. And the limited value of the results should be thoroughly analyzed and clarified.

Author Response

Responses to Editor and reviewers

Editor comments

I believe that this paper still needs correction and revision. The authors have not fully answered all of the questions and comments from revisions during both rounds of review. I recommend that the authors revisit their paper, and revisit all of the reviewer comments and ensure that the paper is edited fully to so that all reviewer comments have been thoroughly addressed and considered, and the paper amended accordingly. I also believe that ethical review of the methods should have been included.

Ferreira et al.: Thank you for the opportunity to further refine our manuscript. We have carefully considered your comments and the reviewers' feedback from both rounds.

Addressing reviewers comments: We have performed a comprehensive audit of all previous comments. We ensured that every suggestion was not only considered but explicitly addressed in the updated text. (see below).
Ethical Review of Methods: We appreciate your pointing out this, however, we have included a specific section under 'Methods' ('Ethical note') clarifying the ethical oversight of this study as follows:

This work followed the “Principles of laboratory animal care” [NIH publication No. 86-23, revised 1985] and was approved by the Committee of Ethics for Animal Use [CEUA] at the Universidade Estadual de Santa Cruz [proc. #0072015].

We believe these revisions have significantly strengthened the manuscript and hope it now meets the journal’s high standards."

R1.

Ferreira et al.: We appreciate the reviewer’s insightful comment. This new version has been modified by removing the personality variable to accommodate reviewer 2. Therefore, there have been more substantial changes to the discussion. However, we have sought to address your recommendation. We have revised the discussion to more thoroughly address the factors of sex and age and to explicitly clarify the scope of our results.

On Sex and Age: We expanded the discussion to explain that the observed uniformity in time allocation likely reflects the shared environmental pressures and social stability of the group, which may overshadow demographic differences in a captive setting. [L265-267]

On the Limited Value of Results: We have added a section acknowledging the study's limitations. We clarify that our findings offer a robust 'group-level' perspective on environmental enrichment, which is valuable for institutional management. However, we now explicitly state that these results may have limited generalizability to different social contexts or solitary housing. These clarifications can be found in the updated 'Discussion' section (Lines 304-312])."

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript appears to have undergone some revision but issues persist with the manuscript. While the authors have clarified the measuring of personality dimensions a priori in order to correlate these with responses to enrichment, upon a second review, another problem with the inclusion of a personality aspect to the study has emerged. This problem relates to the personality dimensions considered in the study. As pointed out in the previous draft, it is not clear why other aspects of personality were omitted from the study and a specific focus on neuroticism and extraversion was used. Further to this, the definition of what constitutes neuroticism in this study seems at odds with the literature and the prevailing understanding of the term and what it reflects. Generally speaking, neuroticism is a measure of emotional reactivity and one which can manifest in a range of different behaviours, including tonic immobility and self-directed behaviour, and sometimes aggression. Aggression is definitely not a hallmark of neuroticism. Thus, there is no direct link between the expression of aggression and neuroticism; if anything, in the primate literature, aggression is more closely linked with extraversion and dominance than it is with neuroticism. Based on this alone, the study appears to be considering extraversion only in the personality dimensions component of the study, again posing the question of why other aspects of personality have not been considered. While I take the authors justification for not conducting explicit personality tests due to restrictions imposed by the institution housing the animals, this does mean that personality dimensions cannot be fully explored, which then limits what value exploring personality in this context has. As suggested in the previous draft, the personality aspect of this study should be removed and the study should be reframed as an exploration of environmental enrichment in a species of high conservation concern.

There are also issues with the predictions as outlined. Why would high-neuroticism individuals exhibit a greater reduction in aggression than low-neuroticism individuals? If anything, assuming aggression is an indicator of high-neuroticism as the authors do, as a personality dimension it should remain stable and thus high-neuroticism individuals should display a consistent pattern of higher aggression, regardless of context; the consistency of the expression of behaviour across contexts is literally the definition of personality. Why would abnormal behaviour be linked to extraversion? There is little evidence in the literature to suggest that this is the case and, unless there is a clear association, this should be removed from the predictions. While reduced abnormal behaviour is a common outcome of enrichment studies, linking this to personality dimensions seems a stretch. The thinking behind the predictions is not clear or lacks grounding in the existing literature of the field. The authors must justify these predictions in a manner that the reader can understand and is logically consistent.

In addition to the major issues outlined above, many of the previous problems with the manuscript persist. Notably, the specific feedback which was provided to the authors seems to have been largely ignored, as evidenced by the persistent problems with formatting, the unnecessary use of acronyms, a methodology which is still ambiguous at times and tables and figures which are ambiguous or confusing. I am not convinced by the authors argument that each behavioural category should be tested independently, especially when the results are summarised as proportions of behaviour displayed (implicitly comparing time/frequency of one behaviour against others). The discussion is also somewhat repetitive and should be rewritten to synthesise the findings and present them holistically.

Author Response

R 2:

Comments and Suggestions for Authors

Ferreira et al.: We deeply appreciate the reviewer’s critical assessment regarding the personality dimensions. We agree that the conceptual limitations and institutional restrictions on formal personality testing would indeed limit the value of such an analysis.

Following your recommendation, we have completely removed the personality component from the manuscript, including all related objectives, predictions, methods, and discussion points. The study has been reframed as a focused exploration of behavioral responses to environmental enrichment in a species of high conservation concern (Sapajus xanthosternos).

Furthermore, as suggested, we reanalyzed the behavioral data using a single multivariate mixed-effects model. This approach provides a more robust and biologically meaningful assessment of the effects of enrichment while ensuring conceptual coherence throughout the manuscript. [L 75- 88; L178-202]".

We sincerely appreciate the reviewer’s critical feedback regarding the clarity and presentation of our manuscript. We acknowledge that the previous version contained ambiguities that may have obscured the rigor of our study. To address these concerns, we have performed a comprehensive revision of the text, figures, and tables:

Methodological Clarity: We have significantly clarified our analytical approach. We now explicitly state that our choice of a joint mixed-effects framework was specifically designed to account for the inherent interdependence of behavioural time budgets. This approach allows us to address covariance among responses and, crucially, to avoid inflating Type I error rates—a common pitfall when using multiple univariate models.
Interpretation of Results: We have refined the Discussion to better reflect the nuances of our data. While descriptive trends suggested shifts in aggression and exploration during the enrichment phase, we have clarified that these did not reach statistical significance when analysed collectively. We now discuss this stability as a finding in itself, potentially reflecting the 'buffer' effect of captive management.
Sexual Dimorphism: We addressed the ambiguity regarding sex-based differences. Contrary to expectations based on wild Sapajus populations, our integrated analysis confirmed that males and females in this study allocated comparable time to major behavioural categories under the current housing conditions.
Formatting and Acronyms: We have conducted a thorough review of the manuscript to remove unnecessary acronyms (eg. CEUA, GLMM) and standardize formatting. Tables and figures have been redesigned to ensure they are self-explanatory and directly support the statistical findings reported in the text.

We believe these substantial revisions have resolved the previous ambiguities and significantly improved the transparency of our methodology and results.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed the comments and revised the manuscript accordingly

Author Response

R3:

Comments and Suggestions for Authors

The authors have addressed the comments and revised the manuscript accordingly.

Ferreira et al.: Thank you for your time.

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

With this revision, the authors have markedly changed the content of the manuscript. The authors have removed the personality component of the study, which makes for a much cleaner study with a simpler design and interpretation. There are a number of aspects which require some further work though. Particularly the authors should address the study predictions because the current justifications for these predictions are not clear. That is not to say that they are inherently wrong but rather that the thinking behind them is not evident in how they are presented presently. The presentation of the results also needs work in order to ensure that they are explicit and clear for the reader. The discussion section also feels somewhat disjointed and would benefit from being synthesised and streamlined. The authors are encouraged to compare their results to the literature more broadly and use examples wherever possible to illustrate their points. Given the nature of the findings, it would also be helpful to focus on parsimonious explanations first before looking at alternative explanations for the patterns which emerged from the study. With some polishing, I think that this manuscript has the potential to be a publishable study.

Comments for author File: Comments.pdf

Author Response

Responses to reviewers

15^th March 2026

Manuscript: 3774072

We thank the reviewer for their careful reading of the manuscript and for the constructive comments, which helped improve the clarity and quality of the paper.

Selene Nogueira

Abstract

Page 3

Lines 55-56 What is meant by "descriptive by-eye shifts"? This is not intuitive. Does this mean that shifts were observed but these were not statistically significant? If so, this should be rephrased to clarify the meaning.

Ferreira et al.: We thank the reviewer for this comment. We have revised the wording to clarify that these were visually apparent, non-significant trends and removed the phrase “descriptive ‘by-eye’ shifts” for clarity. (lines 36-40). We have also further clarified this in the Results (lines 245-250) and Discussion (lines 302-308).

In this new version the we reworded the abstract as follows: “By-eye inspection of data suggested that environmental enrichment was associated with an increase in affiliative / play behaviour and a decrease in time spent awake but inactive, a proposed marker of depression-like states. However, when behavioural categories were analysed simultaneously within the multivariate framework, overall behaviour time budgets and behavioural diversity were found not to change significantly across experimental phases. This indicates that for S. xanthosternos, the enrichment protocol used here did not provide sufficient novelty or complexity to alter established activity patterns. Integrated analytical approaches are needed to further evaluate the effectiveness of enrichment strategies to ensure they are tailored to specific cognitive and social needs of complex species;

Introduction

Page 6

Lines 128-131 The logic of this prediction is not clear. Why would sexual dimorphism translate into differences in behavioural expression per se? More explanation is needed here. Is there evidence of existing differences in non-captive populations? If so, how are these differences attributable to dimorphism specifically and not inherent sex differences? It seems that social effects or sex-linked behavioural priorities would be stronger effects than differences in size and appearance.

Ferreira et al.: We appreciate the reviewer’s insightful comment. While we agree that sex-specific reproductive strategies are fundamental, we argue that in Sapajus xanthosternos, physical dimorphism acts as a functional mediator of these priorities. The significant difference in body mass and craniodental morphology between sexes (approx. 35%) imposes distinct biomechanical constraints and opportunities. For instance, larger body size in males facilitates the exploitation of high-resistance food items (extractive foraging) and defensive behaviors, whereas the smaller frame of females allows for more efficient movement in the forest periphery. Thus, dimorphism is not merely a morphological trait but a physical substrate that enables niche partitioning and dictates the energetic costs of different behavioral responses (Fragaszy et al., 2004; Souto et al., 2011).

We added more explanation in the ms as follows: “Sexual dimorphism in Sapajus xanthosternos is hypothesized to drive divergent behavioral responses, as distinct morphological constraints between males and females facilitate niche partitioning and influence the energetic budget allocated to foraging and social interactions [32]” . (lines 78- 81).

Lines 131-132 This prediction seems backwards - surely one would expect a difference in the activity profiles of adults and juveniles, given the differences in their needs and social roles? Social grooming, infant care and competitive interactions are going to be much more prevalent in adults than in juveniles and similarly, juveniles are going to engage in play more often than adults. This prediction requires clarification/revision.

Ferreira et al.: We fully agree with the reviewer and apologize for the lack of clarity in our previous wording. Our intention was not to suggest that age-related behavioral differences are absent, as ontogenetic stages undeniably dictate distinct biological needs and social roles (e.g., increased play in juveniles and reproductive efforts in adults). The original sentence was a drafting error and has been corrected in the revised manuscript to reflect that we indeed expected systematic differences between adults and juveniles based on their developmental requirements and social dynamics.

In this new version we reworded as follows: “In contrast, we predicted systematic differences in behavioral time allocation between age classes, as juveniles are expected to engage more frequently in play and exploratory behaviors, for instance, whereas adults prioritize social grooming, infant care, and competitive interactions due to their specific reproductive and social roles. [32, 33].” (lines 82-85).

Materials and methods

Page 17

Lines 259-260 So were data recorded as the total amount of time expressing each behaviour per 100 minutes of observation? This should be clarified. Was this what was used in the analysis? Why were the data not rather recorded as raw numbers of minutes/seconds per session?

Ferreira et al.: We thank the reviewer for this comment. Behavioural data were recorded as the total duration (in seconds) that each individual spent in each behavioural state during the 100-minute focal session. These raw duration values (seconds per session) were used directly in the analyses. As all sessions had equal duration, no further standardisation was required. We have clarified this in the revised manuscript.

We reworded as follows: “For the behavioural coding assessment, video-recorded focal observations (100 min per individual) were analysed by a single observer who was blind to the objectives of the study. The observer recorded the total duration (in seconds) that each individual spent in each behavioural state during the 100-min session using CowLog 3.0.2 [34, 35].”( lines 177-180)

Lines 265-267 This statement feels out of place and is written in a different tense to the rest of the manuscript. Were statistically significant effects examined through post hoc analyses? This should be made explicit.

Ferreira et al.: We thank the reviewer for this helpful comment. We have revised this section to ensure consistency in verb tense and to improve clarity. We now explicitly state that post hoc analyses would have been conducted only if statistically significant multivariate effects were detected. As no significant multivariate effects were found, no post hoc tests were performed.

This has been clarified in the revised manuscript as follows: “…We choose this approach because behavioural time budgets are inherently interdependent, as time allocated to one category constrains time available for others. Analysing categories jointly allowed us to account for covariance among behavioural responses and to avoid inflating Type I error rates associated with multiple univariate models. Post hoc tests would have been conducted if statistically significant multivariate effects had been detected; however, as no significant multivariate effects were found, no post hoc analyses were performed.” (lines 189-195).

Page 18

Line 274 The specific interaction effects should be stated explicitly. Without this, the reader will assume that all possible interaction effects were included. This will also have implications for how the model will run.

Ferreira et al.: We thank the reviewer for this important comment. We have revised the manuscript to explicitly specify the interaction terms included in the model. The model included the interaction between behavioural category and experimental phase (Category × Phase), as well as the interaction between sex and age class (Sex × Age). No other interaction terms were fitted. This clarification has been added to avoid any ambiguity regarding the model structure and the text was reworded as follows: “We included behavioural category, experimental phase [A1, B, A2], sex [female vs. male], and age class [adult vs. juvenile] as fixed effects, as well as the interaction between behavioural category and experimental phase [Category × Phase], and between sex and age class [Sex × Age].” (lines 185-188).

Lines 292-293 The meaning of this statement is not clear.

Ferreira et al.: We thank the reviewer for highlighting the lack of clarity in this sentence. The text has been revised to clearly state that abnormal repetitive behaviours and inactivity were excluded from the activity budget used to calculate behavioural diversity. The revised sentence reads: “... Therefore, abnormal repetitive behaviours (e.g., pacing) and inactivity were excluded from the activity budget used to determine behavioural diversity.” (lines 209-210)

Results

Pages 18-19

Lines 296-300 The contents of the table being referred to here appear to be summary statistics for different behavioural expressions. Perhaps it would be more helpful to frame this as a summary of the time budget of the population considered rather than a 'descriptive analysis', as is stated here. Ultimately, it is the same thing but readers will intuitively understand what is meant by a time budget, while descriptive analysis is less clear. It might also make sense to link this directly to the finding of the GLMM in order to illustrate that heterogeneity in the time budget is statistically significant, even if the data are at different scales (i.e. broader behavioural categories in the GLMM).

Ferreira et al: We thank the reviewer for this helpful suggestion. We have revised the text to frame Table S1 explicitly as a summary of the population time budget rather than as a “descriptive analysis,” as this terminology more clearly reflects the behavioural allocation patterns being presented. In addition, we have now explicitly linked the time budget summary to the results of the multivariate GLMM to clarify that the observed heterogeneity in behavioural time allocation was formally tested in the inferential analysis. This revision improves the conceptual coherence between the descriptive summary and the statistical modelling.

In this new version we reworded as follows: ”A detailed summary of the population time budget across behavioural categories is provided in Supplementary material [Table S1]. Estimated marginal means indicated that general activity accounted for the largest proportion of behavioural time across all experimental phases, followed by exploratory and affiliative/play behaviours, whereas aggressive and abnormal represented a comparatively small proportion of the activity budget [Figure 2]. Behavioural time allocation remained broadly consistent across the control [A1 and A2] and enrichment [B] phases, although descriptive trends suggested subtle category-specific variation….” (lines 215-224).

Lines 301-303 This is the only statistically significant finding in the study and seems to warrant its own figure in order to meaningfully convey the finding. It does not seem appropriate to refer to the vertical axes on a different finding's figure to illustrate this.

Ferreira et al.: We thank the reviewer for this comment. We have removed the reference to the vertical axes.

Lines 308-313 This feels like it would be better positioned in the context of the behavioural time budget above, rather than after describing the non-significant model effects. It would present the results more coherently if the descriptions of differences were linked directly to their statistical effects.

Ferreira et al.: We thank the reviewer for this helpful suggestion. To improve the coherence of the Results section, the paragraph describing the differences in behavioural time allocation (estimated marginal means) has been moved to follow the descriptive time budget paragraph. We reworded as follows:

“The time budget of behavioural data indicated marked heterogeneity in time allocation among behavioural states across experimental phases. Locomotor and general activity behaviours accounted for the largest proportion of observed time, whereas aggressive and abnormal behaviours occurred infrequently. A detailed summary of the population time budget across behavioural categories is provided in Supplementary material [Table S1]. Estimated marginal means indicated that general activity accounted for the largest proportion of behavioural time across all experimental phases, followed by exploratory and affiliative/play behaviours, whereas aggressive and abnormal represented a comparatively small proportion of the activity budget [Figure 2]. Behavioural time allocation remained broadly consistent across the control [A1 and A2] and enrichment [B] phases, although descriptive trends suggested subtle category-specific variation. By-eye inspection of Figure 2 suggests that affiliative/play behaviour showed higher estimated values during the enrichment phase relative to A1 and A2. In contrast, inactivity tended to decrease during the enrichment phase and remained lower in A2 compared to A1. Whilst aggressive behaviour showed a downward trend during and following enrichment, this was from a baseline of a very low amount of time spent performing this behaviour. No other patterns are detectable.” (lines 213-224)

Page 20

Lines 325-328 This feels like it would be better positioned as a discussion point, rather than a result. It involves interpretation, rather than just presenting a finding.

Ferreira et al.: Thank you very much for this comment. We removed from this section to discussion section.

Discussion

Page 20

Line 340 But none of the shifts across the enrichment phases were statistically significant. This must be rewritten to accurately reflect the findings. Moreover, the responses of the animals appear to be consistent across conditions, so citing behavioural plasticity here does not seem warranted; if anything, their response remains remarkably unchanged.

Ferreira et al.: We thank the reviewer for this important observation. We agree that the previous wording could suggest stronger effects than those supported by the statistical results. The sentence has been revised to clarify that no statistically significant differences were detected across enrichment phases.

In addition, the reference to behavioural plasticity has been removed, as the reviewer correctly noted that the behavioural responses were broadly consistent across conditions.

We reworded as follows: “Contrary to our initial predictions, the multivariate framework detected no systematic differences in behavioural time allocation between age classes. While we expected distinct activity profiles due to the differing social roles and biological needs of juveniles and adults, our findings suggest that, in this specific context, functional behavioral categories may remain relatively stable across these ontogenetic stages.” (lines 251-254).

Page 21

Lines 349-352 The opening sentence of this paragraph relates to sex differences and then the second sentence relates to enrichment phases. What is the link between these? It is inconsistent and does not follow logically. It also does not seem realistic to contrast the experimental response with the behavioural plasticity of free-living populations, given the vast differences in context. The overall message of this paragraph is not clear as a result. This should be revised to make the message clearer for the reader.

Ferreira el al.: We agree with the reviewer that the transition between the discussion of sex differences and the effects of environmental enrichment was not sufficiently clear. We have revised the paragraph to create a more logical flow, explicitly linking these findings. We now argue that the lack of sex differences and the stability across enrichment phases both point to a common theme: the constraints imposed by the captive environment on the behavioral plasticity typically seen in the wild. We have also tempered the comparison between experimental and free-living populations, acknowledging the contextual differences as suggested.

In this version the text was reworded as follows: “Despite expectations associated with sexual dimorphism in Sapajus spp. [32], sex did not significantly influence behavioural time allocation when behavioural categories were analysed simultaneously. Our findings reveal a notable stability in the activity budget of S. xanthosternos across experimental phases, contrasting with the high behavioral plasticity typically reported for Sapajus spp. in wild settings. In natural habitats, these primates exhibit significant seasonal and sex-based variations in time allocation, primarily driven by fluctuations in food availability and reproductive demands [21]. This indicates that, under the housing and management conditions of the present study, males and females allocated comparable proportions of time to the major behavioural categories. Social dynamics, enclosure design and resource availability may have contributed to this similarity, although further studies explicitly addressing these mechanisms are warranted.” (lines 260-268)

Page 22

Lines 370-372 While individual variation may certainly have attenuated the impact that the introduction of the enrichment had on the groups, this is not evident from the data that are presented here and is certainly not the only possible explanation. It is more likely, based on what is presented here, that the enrichment was simply not effective in eliciting significant behavioural change. This might suggest that the animals already experience an adequate degree of stimulation or that the type of enrichment provided here was not appropriate for the needs of the animals in their current context. It may also be that the number of hours were insufficient to detect a meaningful change in behaviour. Alternatively, perhaps the summarisation of the data prior to analysis attenuated the possible effect?

Ferreira et al.: We appreciate the reviewer’s insightful comment regarding the interpretation of our non-significant results. We agree that attributing the lack of population-level effects primarily to individual variation was speculative, as our data did not explicitly support this. Following the reviewer’s suggestion, we have revised the Discussion to include alternative and more likely explanations, such as the potential lack of effectiveness of the specific enrichment used for this context, the possibility that the animals’ baseline stimulation was already adequate, and limitations regarding the duration of behavioral observations.

In this new version we included the highlighted suggestions as follows: Environmental enrichment was associated with descriptive shifts in specific behavioural categories, increased affiliative/play behaviour and reduced inactivity during the enrichment phase. These patterns are consistent with previous studies showing that enrichment may promote engagement in species-typical behaviours such as play, and decrease levels of inactivity whilst awake which has been suggested to be a marker of a depression like state in non-human animals [44-47]. However, when behavioural categories were analysed simultaneously within the multivariate framework, none of these phase-related differences reached statistical significance. Although individual-level variation could theoretically attenuate effects detected at the population level [7, 20], the present results do not provide clear support for this explanation, suggesting that other factors may have contributed to the stability observed across enrichment phases. One possibility is that the enrichment provided was not sufficiently powerful to induce measurable changes in the overall behavioural time budget. Alternatively, the animals may already have experienced a relatively adequate level of environmental stimulation within their housing conditions, thereby limiting the magnitude of behavioural shifts following the introduction of additional enrichment. It is also possible that the total number of observation hours constrained the detection of more subtle behavioural adjustments, while the aggregation of behavioural data prior to analysis may have obscured short-term or transient responses. Taken together, these considerations suggest that, although enrichment may influence the expression of particular behaviours, the broader organization of the behavioural time budget can remain relatively stable under certain management conditions. (lines 281-292).

Tables and Figures

Table 1

The table should be formatted to indicate clearly the delineations between the descriptions provided. As it is formatted now, it appears as a single long paragraph, rather than distinct descriptions which correspond to specific behavioural states. Just adding in horizontal lines between categories would do the trick. The behavioural definitions should be tightened up. The definition for chasing uses the term 'chase' and is thus circular. The definition for 'Threatening' includes agonsitic vocalisations, so how does this differ from 'Agonsitic call'? Object manipulation's definition also uses the term 'manipulates', making it circular. The definition of 'Digging' includes an interpretive aspect ("appears to be searching for objects or invertebrates") which should be removed because this is not descriptive but rather an inference. Similarly, the inclusion of "to remove ectoparasites" in the definition of social grooming should be removed. For solitary play, replace the word 'lonely' with 'solitary'; how would the locomotor patterns of solitary play in this definition be distinguished from stereotypy? How are climbing and hanging different, based on the definitions provided? The label of 'Stand still' does not seem appropriate, given that the definition includes sitting and lying.

Ferreira et al.: We thank the reviewer for this helpful suggestion. The table has been reformatted to improve clarity and readability. Horizontal lines have been added between behavioural categories so that each behavioural state and its description appear clearly separated rather than as a continuous paragraph.

In addition, all behavioural definitions were carefully revised and tightened to ensure that they are purely descriptive and avoid circular or interpretative wording.

Figure 2

This does not seem like the best way to communicate the data being displayed here. The graphs all show different scales on the vertical axes and this makes interpreting the scales of change difficult for the reader. Perhaps presenting pair-wise comparisons as bar graphs (i.e. A1 VS B, B VS A2, A1 VS A2) with all behaviours on each plot would be more effective? Given that the interaction effect was not significant, it might be better to use a different approach. Combining the behaviours all into a single plot using different colours might be better. Doing it this way would make the differences between behaviour types clear for readers.

Ferreira et al.: We thank the reviewer for this suggestion. We have revised the figure accordingly, replacing the previous faceted plots with a single bar graph displaying all behavioural categories on a common y-axis and differentiated by experimental phase. This improves clarity and facilitates direct comparison among behaviour types.

Author Response File: Author Response.docx

Article Menu

Behavioural Time Allocation and Responses to Environmental Enrichment in Zoo-Housed Yellow-Breasted Capuchin Monkeys (Sapajus xanthosternos)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI