Examining Activity, Courtship, and Social Behaviour in Zoo-Housed Wreathed Hornbills: Potential Insights into Pair Compatibility

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please see attached document.

Comments for author File: Comments.pdf

Author Response

Summary:

The authors studied the behavior of a pair of wreathed hornbills (Rhyticeros undulatus) at ARTIS

Amsterdam Royal Zoo. The hornbills were observed for a total of 15 hours over several 30-minute

observation sessions occurring at the same time of day within ~2.5 months. A combination of

behavior sampling methods was employed to determine a general activity budget, total frequency

of courtship behaviors, and enclosure use. The authors present the behavioral results on their own

as well as statistical analyses investigating factors contributing to greater or fewer frequencies of

courtship behavior. Enclosure use was examined as a function of type of perching available as well

as visitor presence. The authors call for a greater presence of case studies that can be analyzed

together to draw broader conclusions and/or multi-zoo studies to better understand how to

promote breeding success in hornbills in human care.

General concept comments:

Overall, this paper could be presented as a case study of one hornbill pair’s behavior, but it needs

significant re-working to fit this description. A general point to be made is that the results don’t fit

the title; the behavior presented in this paper help us understand some aspects of hornbill social

biology, but compatibility is neither assessed nor able to be determined based on this study. The

abstract and the conclusions summarize the study very well, but the rest of the paper needs to be

brought in line with these sections and the overall message of the paper needs to be re-calibrated.

 

Thank you for the comments. Our paper is a case study, and we have stressed that in our abstract. We have altered the title to make the paper’s aims clearer from the start, and we have edited the explanation of our rationale in the abstract. However, we respectfully point out that we have used the word “potential” (indicators of pair bonding) in our abstract and we conclude in our abstract that this is a case study, on one pair of birds, and others must validate our behavioural findings. Thus, this is a call-to-arms for others to extend such study across captive hornbill pairs. We believe that we have been clear, accurate and objective in how we have described our case study, and we hope that others can use our approaches to look at a wider sample of birds in more detail. We have clearly explained our behavioural sampling and recording and we have used appropriate methods (that we have used many times before) to collect these types of behavioural data. We hope our responses further elucidate our methods and approaches to this research. We thank the reviewer for their thorough and helpful review of our work.

 

Specific comments:

 

Abstract:

- A good summary of what I think this paper is presenting, but the inner sections don’t reflect

the simplicity or limitations of the study/results.

Thank you for the comment. We are pleased that you feel the abstract is a good summary. However, we are unsure as to what the reviewer means by the “inner sections”. If more specific feedback can be provided, we will happily action. As per our comment above, we have edited the abstract to clarify the paper’s rationale. We hope this better defines our reasons for undertaking this case study.

 

Introduction:

- Provides valuable background information. The only thing missing is a presentation of any

known information about natural seasonality of this species, which comes up in the Results.

Thank you for the comment. We are pleased that you find the introduction to be useful. We have included further information on what is known about the natural breeding seasonality of this species in the wild, although information on this scant. Please see paragraph starting line 69.

 

Materials and Methods:

- Line 120: Baers pochard, not Bears

Thank you for spotting this. The edit required is Baer’s pochard. We have edited this.

 

- Lines 142-143: it appears that you used a combination of scan sampling and all-occurrence

sampling. See further comments about whether this scan sampling should be described as

instantaneous, as you say here.

Thank you for the comment. We are unsure of the reviewer’s comment here. We have two birds; these are our focal follows. We recorded behavioural data at one-minute intervals (so instantaneously). Therefore, we have instantaneous focal sampling because we recorded what each individual was doing at that specific time point to obtain our behavioural data, as is standard practice in many behavioural pieces of work. Therefore, this gives us an estimation of the time (in minutes) spend on state behaviours.

 

- Lines 143-144: results should be considered with the limitation that observations occurred

within the same narrow time frame each day

Thank you for the comment. It is not convention to discuss results in the result section. The results section is for description of the analysis. We have clearly examined our limitations and required research extensions, to answer this and other limitations, in the discussion section.

 

- Lines 147-148: did the “state behaviors” actually last for full minute intervals? Did you only

score behaviors that were performed for at least 1 minute? Or were you sampling

instantaneously every minute (with behavior potentially changing in the time until the next

scan)?

We are unsure of the reviewer’s comment here. When you perform an interval sample you record the behaviour seen on the recording interval (i.e., in this case the one-minute period). Therefore, you have assumed that the behaviour occurred for that full minute. This is one of the commonest forms of behavioural sampling used for creating time-activity budgets however, one must acknowledge that the duration of the interval needs to be short enough to capture what the animal is actually doing but long enough to avoid pseudoreplication.

 

- Line 150: it seems that the “event behaviors” were sampled as all-occurrence events, rather

than continuously being sampled.

Thank you for the comment. Each occurrence of an event behaviours was recorded when it was observed and this data collection was continuously performed during the observation period (this is called event sampling). We specified the events that we wished to record and the timeframe they would be recorded in (i.e., continuous monitoring for each performance of an event behaviour during the 30 minute observation period). Again, this is a standard way of recording event behaviours, see Bateson & Martin Measuring Behaviour for more information. We have attempted to clarify our recording of event behaviours in the text to avoid any confusion.

 

Line 161: please note that the version number for R is missing. Also, not that this is

necessarily problematic, but I don’t understand why so many software packages needed to

be used for the analysis you present (especially both R and RStudio).

Thank for spotting the error. The version number has been included. As RStudio runs on the platform R it is important to state the version of both packages that are used to ensure repeatability of the methods.

 

- Lines 186-188: I like how the result of this is presented later in the paper, but why was

analysis limited to just the time birds spent at the highest point? Were other outcomes

analyzed for a visitor presence effect?

Thank you for the comment. We analysed this because these hornbills prefer to be at higher elevations. We have included such details in the introduction and therefore this use of height gives us an ecological relevance to our explanation of results and key discussion points.

 

Results:

- Lines 204-205: is there any idea of why “Other” behaviors would have been significantly

different? I know you point to this for follow-up research, but I wonder if there is any

anecdotal evidence that could point to a clue.

Thank you for the comment. We don’t have any anecdote or otherwise to extend here, hence why we pose this as a future research question for others to consider investigating further. However, we are pleased that the reviewer also believes that this worthy and useful research extension to consider.

 

- Line 207: this is a logical issue that comes up several times throughout the paper. How

could it be that the two birds would not spend the same amount of time within 1 bill-length?

I can see in Figure 2 that the median time spent within 1 bill-length is higher in the female

than the male, but I can’t logically see how this would work (if the female is within 1 billlength of the male, the male then must be within 1 bill-length of the female). The more

appropriate analysis of this behavior would be how many scans the two birds spent within 1

bill-length compared to other state behaviors (but, see comment in Table 1 regarding

whether this behavior should be included in state behaviors at all).

Thank you for the comment. We have clarified this in the ethogram. The behavioural state was when the birds where within one beak length of each other, however, if one bird was moving to be within a bill’s length and the other was then moving away, we show the bird who is choosing as being in with one bill length and the bird who moved away as not. We have clarified this behavioural state in our ethogram. We did this as we felt it would provide some useful information on motivational states. Is one bird more motivated to be close to the partner than the other and does this suggest that a compatibility difference may be present? We have added further clarification of this in our method section (see section 2.2). We apologise for not including the complete description of this behaviour in the original ethogram.

And thank you for the comment on how we should include bill length within our further analysis as this is what we have done in our paper. We have used this state as a term in our model to see if it impacts on the decision to perform other behaviours that are part of the hornbill’s overall time budget. We clearly state that in our model output as show in the data analysis section, highlighted for the reviewer in yellow in the main paper but also copied here. “The final model run was thus: displays and calls ~ sex + inactivity + within one beak length of partner + perching up high.”

 

- Lines 240-245: interesting result, but providing some background on the species’ natural

seasonality and relevant environmental cues would help understand any practical

significance and/or application.

Thank you for the comment. We have extended the information presented on species’ natural history in the introduction as per the comments from other reviewers. We have also included more information on natural history, to explain any sex differences in the “Behavioural Outcomes” section of the discussion.

 

- Lines 261-263: please include the p-value for the relationship between exploration and

calling, or a statistical indicator of data dispersion explaining why analysis is not possible.

Thank you for the suggestion. We have edited this to show that exploration was included in an original model but due to it being insignificant as a predictor it was removed from the final model run. Our raw data are also uploaded to figshare.com at the link provided in the data analysis section (section 2.3, last sentence) to show our actual data used for analyses.

 

- Lines 274-276: where is the evidence that visitor presence/habitat quality are the reasons

for section 5 being the least used area (or is this a hypothesis? It is presented as a

statement.)?

Thank you for spotting this. The sentence should say “may relate to” and is further expanded in the discussion. We appreciate the thoroughness of your review.

 

Discussion:

- Line 299: why is the level of inactivity in these birds considered “heightened”?

We have removed the word “heightened”, thank you for the suggestion.

 

- Line 301-304: this seems like a significant factor that would affect the female’s behavior.

Thank you for the comment. You are correct, hence why we have provided such prominence to this post-research finding to ensure there is objectivity in our consideration and evaluation of our results. We hope this emphasises our call for research extension into behaviour across more pairs / more zoos.

 

- Line 304: do you mean “can’t” be discounted?

Apologies, this should read as “cannot be discounted”. Thank you for spotting this and again, we appreciate the thoroughness of your review.

 

- Lines 310-315: see comment in results section. Asserting that the male and female spent the same time within 1 bill-length doesn’t make sense to me; the time spent within 1 bill length for the pair should be compared to time spent in different proximity categories.

Thank you for the comment. Respectfully, we have done this in the model that we have run and considered the impact of being on the same perch on the results that we present. You can see that we have included this social measure in our final model (see section 2.3, data analysis).

 

- Lines 342-345: the relationship between these two statements is unclear to me. If there is a

clearly preferred perching, why would the male have exhibited increased exploration to find

preferred perching?

Thank you for the comment. Animals will display preferences even when their ideal may not be there. So, whilst these hornbills are preferring to be high, perhaps would like to still go higher? Given the evidence we present in the introduction as to the overall height of their wild nest sites, we feel this may be true (see data on elevation of tree holes in the introduction). The male bird may be wishing to try and explore the enclosure to see if any, even more elevated perching options are available. Hence why we recommend research across zoos with different forms of housing, and also research on different styles of aviary too.

 

Conclusions:

- I feel that this was a well-written, appropriate summary of the paper. The inner sections

need to be significantly re-worked to fit this one.

Thank you for the comment on the relevance of our introduction. We are unsure as to what the reviewer would like us to change, given that the feedback is the conclusion is well written and appropriate. We express our limitations, explain the importance of considering bird health status (ref the female’s health condition) and we have emphasised the need for others to take our methods and apply them more broadly. We are more than happy to consider further evidence if direction was provided.

 

- Lines 467-473: These lines in particular represent what I believe this paper is: a case study

of two individuals/one pair that can contribute to a broader survey of hornbill behavior, but

not one from which specific conclusions can be drawn, besides general recommendations

about perching.

Thank you for the comment. We have used this feedback to alter the title, and hence focus, of the paper.

 

Figures and Tables:

Figure 1:

- Lines 131-132: in the caption, you state that visitors would “ideally” pass from section 5 to

section 3. Why is that ideal?

Thank you for the comment. We have edited this to remove the word “ideally” and replace it with the word “usually”.

 

Table 1:

- Were “Enclosure use details” and “On same perch as partner” also sampled

instantaneously every 1 minute?

Thank you for the comment. We have clarified how such data were collected in the methods section. Enclosure usage data (i.e. the zone each hornbill was in) was recorded as focal follows at one-minute intervals (marked on ZooMonitor) and the number of times a hornbill was seen on the same perch as another hornbill was simply counted.

 

- I think “With partner” should not be included in state behaviors, but rather moved to the

same category as “On same perch as partner.” If this is done, this new category could

function as a proximity-measuring category and provide detailed information about how

much time the birds spend in certain proximities, without “With partner” taking away from

measurements of other state behaviors.

Thank you for the comment. Birds are making an active choice to be within one bill length of each other. Therefore, this is a social state. We have not performed any social network analysis on these data as the sample size is only two and the dataset is not big enough overall. On same perch as partner is our basic metric of enclosure resource use, as we have described in the text. We would like to keep our behavioural classifications as they are as they were agreed between the authors of the paper and the Zoo when methods were reviewed and discussed.

 

- “Visitor number” is described as being measured continuously in this table, but the text in

the methods section states that it was measured instantaneously each minute.

Thank you for comment. We have clarified this in the methods. Hopefully this is now easier to follow. Visitor presence was continuously recorded during each 30-minute observation period.

 

- Formatting suggestion: make each behavior title distinct (e.g., in bold) from its description

for ease of reading/interpretation

Thank you for the suggestion. We have followed the journal’s formatting of tables here and we are not sure that bold in the description is allowed in the text within the table.

 

Figure 2:

- The y-axis and description of the data suggest that this is the total time in minutes these

birds spent performing the behaviors. Please see comment in methods section, Lines 147-

148: were these actually full minutes? Or were they instantaneous scans at the start of

every minute? In the latter case, the y-axis should more appropriately be “number of scans”.

Thank you for the comment. When you perform an interval scan, you add up the number of intervals that the behaviour was seen for. Therefore, you get an approximation of the amount of time (estimated from the interval length) for that behaviour. This is the reason for choosing an interval scan method – to get an estimation of time spent on behaviour when continuous recording of a state behaviour is not possible. We believe that we have presented our results in a fair and accurate way.

 

- Formatting suggestion: I suggest changing the outlier symbol from an asterisk to a dot or

something more neutral, as asterisks are frequently associated with indicating statistical

significance. Relatedly, I also suggest adding a visual indicator of statistical significance.

Thank you for the feedback. We have edited all graphs with outlier symbols accordingly and we have included indicators of significance on the appropriate figure.

 

Figure 3:

- I suggest displaying the data as counts rather than percentages in these charts, so that the

difference in total observations between the male and female can be seen.

Thank you for the comment. We included the key result (calling and display) count in the explanation of the figure originally, but we have extended this to include data on all these event behaviours.

 

- Lines 212-213: in caption, need to re-word to something like “Only the male bird performed

allopreening…”

Thank for the correction. We have edited this accordingly.

 

Figure 4:

- I don’t understand the significance of this figure. What does displaying the number of

observations of individual preening + perching on the same perch tell us? I can’t discern this

from the Results text either.

Thank you for the comment. This figure is explained in the paragraph immediately above it and we show that there is likely a general trend in preening difference that could be examined further. i.e., is this an indicator of comfort, or of a bird coming into breeding condition.

 

 

Figure 5:

- Displaying both the counts and means is confusing here; I don’t know which more clearly

illustrates what you would like to show, but I suggest choosing one.

Thank you for the feedback. This graph has since been redrawn based on other reviewer feedback, and the axis label and caption adjusted. We hope this clarifies the results shown here.

 

- Issues with x-axis categories: what is “Time together”? Is it within 1 bill-length of proximity?

Why are “On same perch” and “Sat on same perch” worded differently? In general, please

match these categories to the language you use in your ethogram.

Thank you for the comment. We hope our edits to this figure have clarified the result shown.

 

Table 2:

- Similarly to other areas in the paper, the units displayed for each predictor should be either

verified to actually be minutes, or modified to “scans”.

Thank you for the comment. Again, this is the estimation of time from the instantaneous focal sampling because each observation estimates what that bird was doing in that one-minute interval (hence why it is called a state behaviour).

 

Figure 6:

- Line 258: the caption describes “potential relationship” for all four plots, but three of them

were shown to be significant; suggest re-wording to be clear about this.

Thank you for the comment. These figures are correlations; therefore, we cannot infer causation. Hence our use of the word potential. We do not wish to suggestion causation without further study; hence we wish to keep our wording here which we believe to be an accurate and objective description of what we have found in this case study.

 

- Is there an explanation for why the relationship between perching up high and number of

calls/displays was highly positively correlated, yet the trendline shows no slope?

Thank you for the comment. The model that shows a significant effect of being high up on performance of courtship considers multiple factors (predictors) and can pick out the impact of these factors on the output variable (i.e., the hornbill’s behaviour). The correlation is a more blunt description between two factors (in this case, the perching and the courtship). These factors do not seem to correlate but show a relationship when other variables are included. This shows the importance of considering the multifactorial nature of animal behaviour – taking into account and trying measure all of the potential influences on the animal’s activity patterns.

 

 

Figure 7:

- Similar to Figure 6, this is described as a “potential relationship”, but in the text of the

results section it is shown to be a significantly positively correlated relationship.

Thank you for the comment. Do you mean Figure 8? If so, as per our answer to the question pertaining to Figure 6, this is a correlation – a descriptive result – we are suggesting a possible relationship, but this is not causation. As is statistical convention, a correlation is only suggestive. Further testing is required for causation to be determined, and we expand on this need for further research in our discussion.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This is a well-written manuscript with clear description of the methods and good presentation of the results with valid conclusions. Although the level of novelty and potential applications is restricted to this particular pair of hornbills in this particular setting, the methods are replicable in other zoos. Therefore, this manuscript provides a basis for design of future studies in order to validate these findings across a larger sample size.

 

Methods and Materials

Line 108 - 110: Is the  Research Department and Bird Department at ARTIS Amsterdam Royal Zoo an IACUC? Does the journal require IACUC approval for this project to be accepted? Please verify.

Author Response

This is a well-written manuscript with clear description of the methods and good presentation of the results with valid conclusions. Although the level of novelty and potential applications is restricted to this particular pair of hornbills in this particular setting, the methods are replicable in other zoos. Therefore, this manuscript provides a basis for design of future studies in order to validate these findings across a larger sample size.

Thank you for the comments and kind words on the paper. We are pleased that you feel the paper makes a useful template for others to follow.

Methods and Materials

Line 108 - 110: Is the  Research Department and Bird Department at ARTIS Amsterdam Royal Zoo an IACUC? Does the journal require IACUC approval for this project to be accepted? Please verify.

We followed the ethical review process at ARTIS Zoo for this research. The project was observational with no direct manipulation therefore no further permissions from the Zoo were required to conduct data collection.

Reviewer 3 Report

Comments and Suggestions for Authors

See the uploaded PDF for the reviewer comments.

The study is well designed and analyzed, but due to the small sample size and the lack of breeding progress during the study period the connection of the study inferences to the objectives is bit weak. Shortening the paper to be more focused will make it effective.

Comments for author File: Comments.pdf

Author Response

The study is well designed and analyzed, but due to the small sample size and the lack of breeding progress during the study period the connection of the study inferences to the objectives is bit weak. Shortening the paper to be more focused will make it effective.

Thank you for the comments and useful feedback. We have actioned all comments from the annotated pdf although we were unsure as to the highlights added to the number of pheasants and ducks in the aviary with the hornbill. Thank you for commenting that the study is well designed and our data are appropriately analysed. We have clarified our objectives to state that we focussed on behavioural data collection - we were not looking at actual reproductive output, and so we hope this strengthens the theme of our discussion. If the reviewer could give more details here on how the paper should be made more concise that would be helpful. We feel that we have provide sufficient evaluation of our study to help others structure appropriate research questions for further studies.

in ex-situ conditions in ARTIS Amsterdam Royal Zoo, Amsterdam

Thank you for the comment and suggestion; we have simply included "zoo-housed" into the title to be more concise.

A sentence about the breeding age and the season in the wild in the Introduction would help understand the study better.

Thank you for the feedback. We are not too sure where the reviewer would like this to go, so we have attempted to include what scant information there is on wreathed hornbill nesting biology in the wild in the section of the introduction that discusses this species specifically. We have included information on months that nesting and fledging is documented, as well as nest site characteristics.

mention that the bird is categorised as “Vulnerable” by IUCN

Thank you for the feedback. We have included this statement in the text.

a note on their lifespan and typical breeding age would be useful

Whilst we agree that this would indeed be useful information, the EAZA husbandry guidelines for hornbills (Galama et al) state that "Average life expectancy is not known for any hornbill species" so we are loath to include erroneous information from unverified sources. We have included a statement in the discussion that such data are required.

Explanation about the “Display (count sum)” in the graph will help to clarify if the trend is due to progression of months or due to limited observations in the month of February

Thank you for the feedback. We have removed the counts from this graph to just show mean values. We hope this shows the pattern more clearly. We have also edited the caption and explanation of the figure.

 

Please explain “historic” beams

We have removed the term historic throughout the manuscript. Thank you for the feedback. This was overly pedantic description of what the aviary looked like. This should now be clearer.

Southern ground hornbills are from different family and ecologically quite distant to be a good comparison.

Thank you for the feedback. We are not using this species as a comparison, we are using it to support our point that behavioural displays are likely to be very species specific. We state this at the start of the sentence - that we need to interpret hornbill behaviour by each individual species.

References or data for recommending number of choices for female hornbills is insufficient.
Add more evidence or remove this altogether

Thank you for the comment. These data do not exist. We are stating this as a future research question that might help us understand hornbill behaviour and breeding outcomes under human care more clearly. We politely request to keep this statement because it needs to be investigated in the future and we are trying to encourage further study.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Please see attached file.

Comments for author File: Comments.docx

Author Response

We have responded to all reviewer comments in the Word document. Please see attached.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

No further comments.

Author Response

Thank you for finalising your review. We appreciate the useful and helpful information in developing the paper further.