The authors have considered most suggestions and the protocol seems to be largely improved. I have one issue remaining.

Lines 218-220: You state that the time in proximity to the food reward is measured cumulatively for the duration of the experiment. I think this should be restricted to the first 10 minutes after the bird has entered the arena to give each bird 10 minutes. As it is currently written, some birds would have more than 10 minutes (the ones that enter earlier).

Thank you for the opportunity to review a revised version of this interesting and well-written stage 1 submission “Neophobia across social contexts in juvenile herring gulls”. The Authors have responded to all my queries and added the clarifications on relevant points in the manuscript. There are some minor remaining points that would help with clarity of the manuscript and allow for study replications; please see my specific comments below. I am looking forward to reading the results of this interesting and well-designed study. 
 
Figure 1 and elsewhere. I apologize for missing to mention this point in the previous review round, but I am wondering why all alternative hypotheses include reduced variance. While for social conformity the reason is clear, it is not fully clear why the Authors expect reduced variance also in risk-dilution and negotiation hypotheses.
 
L165-166 or elsewhere. It would be necessary to emphasize in the text that the housing groups of ten individuals will be divided into two subgroups, that will consist of known individuals (as it was emphasized in the answer to the Reviewer). 
 
L172-174. Thank you for clarifying the position of the control object. It would still be important to clarify whether the control object always (also outside of the control trial) stays behind the food plate, or is it moved to elsewhere in the compartment when novel object trials take place. Please also mention clearly in the text that the control object stays in the testing compartment during both the habituation and the test phase. L184. Please exchange to “six days” in this sentence.
 
L186-187. Please state what kind of marker will be used on animals.
 
L201-203. Perhaps I have misunderstood the Authors’ point, but it seems that the second 10 minutes count will only start once the last bird leaves the start box, and this is problematic as some birds may already touch the food by the point the last of them leaves the start box. It would be better if the second 10 minutes count is started for every bird separately (i.e., every bird will have slightly different ‘start’ and ‘end’ time, but all will have a test duration of ten minutes). Please try to clarify more in the text.

1A. The scientific validity of the research question(s). 

The research question has clear scientific validity as outlined in the introduction. The introduction places the study in the existing research and emphasizes the need to investigate novelty responses in group settings.

1B. The logic, rationale, and plausibility of the proposed hypotheses, as applicable. 

Hypotheses flow from the introduction, are clearly formulated and reflect earlier findings. 

1C. The soundness and feasibility of the methodology and analysis pipeline (including statistical power analysis or alternative sampling plans where applicable). 

Experimental procedures are reasonably well explained to assess statistical analyses. However, more details are needed about the procedure (see comments below) and suggestions have been made to improve the design. 

My main concern is about the timing of testing with respect to the age of the birds and their upbringing. Birds will be hand-reared, which is fine but requires more justification and consideration how it might affect outcomes. More importantly birds will be raised in groups of 10 birds and moved into aviaries while in their nestling stage. Both seem to be highly unnatural as clutch sizes of herring gulls on average do not exceed three eggs and birds would normally remain in their nest or nearby when on roof tops until 45-50 days old. Particularly the large group size in relation to natural conditions might affect behaviours and specifically neophobia (birds might be more neophobic in the single situation as they are not used to it, whereas in wild birds often only one nestling survives). Furthermore, testing is planned around day 30 which is during the nestling stage (herring gulls fledge around 45-50 days of age). I cannot see the biological relevance testing gulls at a stage when they would not encounter novelty. Actually, they would still be fed by their parents and might still be supplementary fed by the caretakers in the current experiment. More importantly, I would not be able to make any predictions how the birds would respond as it does not reflect any natural situation. To test the predictions made in the introduction, birds should be tested after becoming independent and when they are in the flight cages. This ensures that birds are feeding on their own (i.e., reducing variance within individuals due to maturation) and also allows testing predictions at a stage when they would encounter novelty. This might require additionally habituating birds to the testing arena but would be worth the effort. 

Furthermore, entire nests seem to be collected. However, the authors never mention any consideration of relatedness (i.e., keeping track of nest ID or only using one chick per nest). This needs consideration as relatedness affects responses (more related = more similar).

1D. Whether the clarity and degree of methodological detail is sufficient to closely replicate the proposed study procedures and analysis pipeline and to prevent undisclosed flexibility in the procedures and analyses. 

Methodologies are reasonably well explained but need more detail in places (see below)

1E. Whether the authors have considered sufficient outcome-neutral conditions (e.g. absence of floor or ceiling effects; positive controls; other quality checks) for ensuring that the obtained results are able to test the stated hypotheses or answer the stated research question(s). 

The authors have considered inter-observer scores and also have control conditions. However, the experimental design could be improved regarding the length of testing. Currently, experiments last 10 minutes with the variable of main interest measuring the time between leaving the start box and feeding. Birds will leave the start box at different times providing different amounts of time to feed (less time when they leave the start box later). As 10 minutes seem to be already quite short, it would be better to start the timer once the bird has left the start box to give all birds 10 minutes to feed. The time to leave the start box can be restricted to 10 minutes. Both would result in values of 600 seconds in case birds do not leave or do not feed within the 10 minutes, which would be much better to handle than artificially assigning 600 seconds in case the birds do not feed within their remaining time after exiting the start box. 

General comment:

This sounds like an interesting study with a solid theoretical base. My main concerns are regarding when experiments will be conducted in relation to the age of the birds. Furthermore, more details are needed when describing the procedures. Detailed comments are below.

Detailed comments:

Abstract:

Clear and informative.

Introduction:

Line 111pp: A) Why will juvenile gulls be used? Juveniles often respond differently to novelty than adults due to their lower experience. The use of juveniles should be better justified and background information provided about age differences in relation to neophobia and if possible in relation to social contexts. B) Furthermore, the authors mention control (food without novel objects) and experimental trials (food with novel objects). From the introduction it is unclear why this approach has been chosen. It would be nice to see a bit more about this approach. For example, many studies calculate the difference between feeding without and with the novel object or include both measures to control for differences in normal feeding latencies. It might also be mentioned that the difference between the two measures is a very good reflection of the actual fear invoked by the novel object (which is not always the case when the time to feed with object is taken without consideration of the time to feed without object). A paragraph critically evaluating both approaches (considering feeding times without objects or not) would help to understand and justify the own approach. C) Finally, the authors plan to use hand-raised birds. Again, more background information should be provided how origin (wild, hand-raised, captive bred) might affect responses. For example, being hand-raised or wild born has been shown to have substantial effects on neophobia responses (Neophobia Threshold hypothesis, Dangerous Niche hypothesis; Greenberg R (2003) The role of neophobia and neophilia in the development of innovative behaviour of birds. In: Reader SN, Laland KN, editors. Animal innovation. New York, Oxford:  Oxford University Press. pp. 175-196; Greenberg R (1990) Feeding neophobia and ecological plasticity: A test of the hypothesis with captive sparrows. Anim Behav 39: 375-379). 

Material & Methods:

Lines 139pp: It seems entire nests are collected. How is relatedness considered in testing and analyses? Individual identification is mentioned further down, but does this include nest ID? Closely related individuals are likely to respond more similarly, which can considerably affect results. Ideally, only one egg per nest should be used to have independence of data. In case, this is not possible, nest ID has to be considered in any analyses.

Lines 148-151: Will nestmates be assigned to different groups? This would be ideal. Why are 10 birds combined? This seems to be a quite unnatural group size for the nestling stage.

Lines 153-155: Will the same 10 birds from the rearing stay together or will they be assigned to new groups? Again, having nestmates in different groups would be important. What about sex? Sex can affect neophobia. Will you sex birds and assign them into groups considering sex? With five days old, gulls would still be in the nest. How are birds kept until they fledge? Are they basically on the ground? This setup seems to be a major deviation from what the birds would experience in the wild (regarding number of birds and rearing conditions). Given that you are interested in social effects, this does not seem to be a good idea. Ideally, you should replicate wild conditions regarding number of birds together and rearing condition (i.e., keeping them in nest-like conditions until they fledge). 

Lines 157-159: Why do you test the gulls before they fledge? The first time the birds would encounter novelty in the wild is after they have fledged (although gulls nesting on roofs might wander around on the roof before fledging and experience some novelty). More importantly, as they are fed by their parents in the wild, they would not encounter novelty next to their food as it is directly transferred from the parent to the young. While it might be easier to test the birds before fledging, I cannot see the biological relevance and actually would not know what to expect/predict. A better approach would be to move the fledged birds into their flight aviaries and get them used to feeding alone. Once this has happened tests can commence. This approach would have much more biological validity and would be comparable to other studies. 

Lines 162-164: Birds are allocated into two groups of five in their home cage. How is this done? Are they physically separated and if so, how is this done? With random allocation, how do you consider nestmates and sex?

Lines 164-165: You mention a control object. This approach has never been mentioned before. Why do you use a control object rather than the feeding dish alone as control? Please justify your choice. Also, where will the control object be placed? Next to the food or somewhere in the enclosure? In the latter case it would still elicit neophobic reactions as the move of the object constitutes a change in the environment. 

Tables and Figures: Have the table name and description above the table. I think you mention table 2 before table 1. Please revers the naming that table 1 is mentioned before table 2. For figures have the title and description below the figure.

Lines 178-179: Earlier it is mentioned that all birds are marked individually with colour ring combinations. Why is there additional marking necessary and how does this marking look like? 

Lines 184-186: Test conditions will last for 10 minutes only. The authors mention several studies having similar durations, but they are not on gulls. Have you done any experiments confirming that 10 minutes are enough for gulls? Birds often wait hours to approach food when a novel object is present. 

Lines 191-192: How many birds are moved in the group condition? How do you assess latency to feed in the group control and experimental condition? Do you have a focal bird and record its time or do you record latency to feed for all individuals? Do you record order of feeding to account for other birds feeding first and therefore affecting responses in birds feeding later?

Lines 202-204: Birds will have different times to leave the start box. This means that birds differ in the time available to feed from the food. As 10 minutes is already very short, it might be better to start the 10 minutes once the focal bird has left the start box to give all birds 10 minutes. In case, a bird does not leave the start box, one could terminate this phase at 10 minutes. These changes would make coding easier in case the birds do not leave the start box or do not feed. Alternatively, when you are interested in the time between making the food available and feeding then it would be better to measure the time between start of the experiment and the time to feed. 

Lines 211pp: Will individual IDs be considered in one or the other way to calculate repeatability of responses on the individual level? How will you consider responses of other individuals?

Thank you for the opportunity to review this interesting and well-written stage 1 submission. The study aims to assess neophobia in 80 juvenile herring gulls, in individual and social setting (i.e., in groups of 5 individuals). This is a very timely research question, and the Authors describe it well, placing it in a rich body of previous literature. The Authors pose clear research hypotheses, accompanied with very nice illustrations. Methods are well-designed and tackle well the research questions. The proposed data analysis plan uses current statistical techniques. However, there is some room for improvement regarding the description of the proposed methods and data analyses, to make the procedure easy to follow and fully replicable by others. Please see my specific comments as well as copy-edits below. I am looking forward to reading the results of this study.

Specific comments.

Page 4. L121, L133-136. The Authors will test 80 individuals of two species, as at the time of testing, the species cannot be ascertained yet. This may create problems in the group compositions, even if only 10% of the total number of individuals happen to be lesser black-backed gulls. Why not simply testing the animals later (i.e., before their release in weeks 8-10, see L159) if the species can be determined accurately then? 

Page 5. Task design. Is placing a control object for three days in the home enclosure enough for the animals to habituate to it, and thus consider it as a control object? If possible, I would suggest having a control object for longer time in the home enclosure, especially as in some groups the control trials will be done immediately after the three habituation days, but in other groups some days later (which may lead to animals getting dishabituated in the meantime). 

Page 5. Objects. As two proposed objects are of the same colour (blue), would it be possible to exchange them for, e.g., an orange object (i.e. as this colour is also present in the multi-coloured ball)? Objects 3 and 4 seem to be objects that are commonly used for cleaning – can the Authors be sure that animals will consider them equally novel as a blue folder or a multi-coloured ball?

Page 6, L188. Why placing a food bowl in front, and not on the side of the novel/control object? What kind of food and how much of it will be provided in the food bowl?

Table 1. Is the “additional buffer” included in the “zone of interest”? It is not clear to me how/if the “additional buffer” will be coded.

Table 1. Perhaps you can include in the “test arena entry” description that both feet need to be outside the start area AND that the bird needs to be within the test arena - I think that this would make the variable more comparable between the birds.

Page 7. Video coding. Does the trial end when the bird starts eating, or is every trial 10 minutes long? If it is the latter, how will the Authors treat variables if there are several events of eating? In this vein, will you measure all times that the animals spend within the zone of interest, or just how long the animal spent there the first time (i.e. before it first started eating)?  

Page 7, Statistical analyses. Previously in line 96-97 it is stated that repeatability will be measured, but there are no details on the specific analyses/tests that will be done. It is also stated that the variance between individual and group trials will be compared, but there are no details about the proposed analyses.

Page 14-15. “Rationale” should include more information about the power analysis, and “Sampling plan” part should include more information on the number and composition of the tested individuals and groups. For instance, as all groups will include 5 individuals, sexes will not be balanced, and perhaps some groups will include more than one lesser black-backed gulls. Do the Authors have a plan how to deal with these? Further, more details should be given regarding the test design (how the novel object trial versus control trial will be done, that the tests will be randomized, etc.)

Copy-edits.

Page 1. Abstract should include other expected outcomes of the tests, apart from the reduced variance (that are stated elsewhere in the text).

Page 3, line 83. “don’t” -> “do not”

Page 3, line 84. “neopobia” -> “neophobia”

Page 3, line 87. “there exists” -> “there is”

Page 3, line 89. “though” can be deleted.

Page 4, line 114. “After all, ” can be deleted.

Page 4, line 115. “inconsistent findings” -> “inconsistent previous findings”

Figure 1. I would suggest to add the name of the respective hypothesis (instead of A, B and C scenario).

Page 4, line 122. “a a-priori” -> “an a-priori”

Page 4, line 130. I think that “However” can be deleted as the current sentence follows the logic of the previous sentence.

Page 5, lines 153, L157. “ad libitum” should be in italic.
Page 5, lines 155-156. “depending on weather conditions” -> please note what kind of weather conditions are supplemented with heating plates?

Page 5, line 178. “each individual get” -> “each individual gets”

Figure 2. It would be good to label the large testing area as “test arena” (perhaps as “G”).

Page 6, line 192 “and a new test begins” -> do you mean that a new test for (a) new bird(s) begins?

References. Please put Latin names of species in italic, make sure that every reference has all details (including journal name, issue, volume and page numbers), write full journal names, make sure that publications have doi identifiers

L263, L270, L294, L308, L320, L329, L335, L339, L368, L372, L375, L394, L397, L400, L403. “en” is redundant

L305. “to cite this article” is redundant

Please check details of reference L342-344, L363-367, L381-382, L386-387

Page 14. “latency measures”, “latency types” – please keep a consistent labelling otherwise one gets an impression that in one case it is about different variables, and in another that it is about variable distribution. Briefly mention what approach is proposed by Snijders and Bosker 2012; “a priori” should be written in italic

Page 15. “depending on the social mechanism at play” is a bit vague, would suggest to elaborate.