Review of Sensor Technologies in Animal Breeding: Phenotyping Behaviors of Laying Hens to Select Against Feather Pecking

Round 1

Reviewer 1 Report

The paper tries to connect different developments and communities to a big issue on feather pecking. It is important that this review is produced. In general I have one big remark and some minor. The big remark is that the connection between the omics and sensor chapters is made in few sentences. I expect it in the chapter 4, but there are more examples given that might fit in the previous chapters. Maybe I expect more that from the FP case and the omics chapter you really can formulate the needs for sensing systems and the requirements to be fulfilled. This can then be used to discuss the sensing part whether that is possible on short, medium or long term, or in the end is not possible at all. Then the conclusions can be more concrete. Now you have to find the advices in some individual lines.

L2-4  + L59-61:  why not mention feather pecking instead of less damaging behaviour. ALign the key words and the title. Now same words are used.  In title you talk about sensor, but it is also used for identification of hens. For me this might be confusing. You need sensors for identification and sensors for measuring behavioural characteristics. Make this clear. 

L97 + L108: maybe a reference can be added

L125: formulate this as the goal of the paper.

L183-187:  is this the formulation from the needs coming from the omics world and the multifactorial complexity of feather pecking. Formulate it as needs.

L204: is there a common understanding of the location of the antenna, or is it part of the environment. Is this adding to the needed phenotype.

L219:  This is a nice overview. Can you combine it with some requirements so that we can see whether this is enough to provide.

L227: explain the word practically. In some circumstances this might not be enough.

L239: add 'applications for laying hens'

L325-333: try to put this in perspective of short, medium and long term. Now it can be seen as some speculation.

L343: What about multispectral vision techniques.

L367: elaborate on  different research scenarios

L383-388: be aware that laying hens use 3D and broiler 2D. Later on in paper the 3D aspect of laying hens is discussed.

L408: 'up to ten' was indeed a challenge, but not good enough for practical groups with thousands of hens.

L413: explain 'new approaches in PLF'. There is no reference and it can be discussed whether this all fits under the PLF concept .

L468: see main comment on this chapter,. Now it is a description of applications. Bring it more in line with needs and requirement from omics (chapter 2) potential solutions from sensing technology (chapter 3) and then discuss what additional might be needed. Maybe the whole data analytics, infrastructure, and BigData soltions can be discussed in this chapter.

L491 + L522: by mentioning 'proxy measure' you refer to chapter 2?

L573: mentioning 'sentinel' might be a good option. The question then arises how big that senstinerl group should be. This might come back as a requirement.

L600: in paper developments on Precision Livestock Farming are not shown and discussed. My advice is to stick in the conclusion section on omics and sensor technology and the expectations how it can be combined to contribute to the breeding programmes on feather pecking for laying hens.

L617: Is this appendix adding to the paper? The figure A1 is not on an Infrastructure? and figure A3 is not in the paper while it is mentioned in the text. For figure A2 it can be discussed whether it should be addressed to the management system if we are talking about a breeding programme and collecting phenotypic data. This is different stuff.

L671: not sure whether the word et al might be used over here.

Author Response

.

Author Response File: Author Response.docx

Reviewer 2 Report

Line 46: who is 'us'?

Line 72: remove brackets

Line 73: remove brackets and add 'in' and 'and':  in Sweden, Switzerland and the Netherlands

Line 91: I would say thousands, not hundreds 

Line 92-93: I suggest writing '...at high densities and that are difficult to distinguish from one another by human observers or camera's'

Line 93: replace 'visualizing' with 'identifying'

Line 95: replace 'like' with 'such as'

Line 99: remove 'identify individual animals in groups and to'

Line 104: replace with 'the system to distinguish between..'

Line 105: remove 'artificially' (isn't marking always artificial?); remove 'however'

Line 108: replace with 'have been used, fitting birds with a ...'

Line 116: remove brackets and replace 'e.g.' with 'such as'

Line 118: remove 'however' and 'is available that'

Line 169: remove 'other'

Line 173: put 'victims and controls' somewhere else in the sentence, e.g. 'gene expression in victim and control birds from FP lines'

Line 180: replace 'had' with 'is'

Line 184: replace 'a very-hard-to-measure-trait' with 'difficult to measure'

Line 186: remove brackets

Line 203: replace 'of the time a tag was' with 'with a time stamp,'

Line 209-211: I suggest moving this sentence 'Semi-passive tags … communication' to line 208, directly after '[51]', so that the tags are described from passive to semi-passive to active. 

Figure 1: Put the semi-passive tags also in the figure, or leave them out of the text altogether

Line 225: remove brackets and replace 'for example' with 'such as

Line 227: make it clear that this goes for HF tags only, for example by writing 'With HF tags, multiple animals can be… simultaneously'

Line 228: remove 'Also'

Line 229: replace  'Therefore they' with 'These tags'

Line 297: replace 'is relatively energy un-intensive' with 'uses relatively little energy'

Line 313: replace 'though' with 'although' and add 'the' before 'ability'

Line 320-323: I do not see the relevance of this application. How is presence of an animal near a specific resource related to FP? 

Line 352: maybe mention that this specifically goes for brown birds, that are difficult to distinguish from the background; white birds usually are no problem

Line 359: remove brackets and replace 'for example' with 'such as'

Figure 2 and 3: I have difficulties reading and understanding these figures. It is a lot of text, long sentences and the presentation does not make it any clearer. I would put long explanations in the text of the manuscript, and only words in a figure. I would prefer a straightforward Table, with rows for the types of camera's and columns with specifications, challenges and so on.  

Line 367-369: rephrase to 'Environmental factors like dust and ammonia affect longevity of camera's.'

Line 372: replace 'analyze' with 'analyzing'

Line 373: replace 'store' with 'storing'

Line 432-434: remove brackets

Line 438: replace 'levels' with 'level'

Line 440: remove brackets

Line 444: remove brackets and replace 'i.e.' with 'and'

Line 473: remove brackets ad replace 'e.g.' with 'such as' and add 'and' before 'swine'

Line 484: remove brackets and add replace by 'with provision of perches being legally mandated'

Line 526: How well did the system perform? Does this refer to the 0.29 accuracy?

Lines 540, 543, 544, 546, 551, 558: remove brackets

Line 619-620: this suggests that consumers want farmers to produce efficiently, and that that is a challenge for farmers. Is that true? 

Line 633: remove 'daily' since it suggests that data are stored or presented on a daily basis; this can be any time unit

Line 633: why focus on breeders and not on farmers? Most PLF systems are designed for farmers, to translate data into information so that farmers can make better decisions in their daily management. 

Figure A2: see above, I am not a fan of these schemes. Too many words, not clear. 

 

Author Response

.

Author Response File: Author Response.docx