A System Dynamics Model of Supply-Side Issues Influencing Beef Consumption in Nigeria

Round 1

Reviewer 1 Report

General Remarks

This paper uses simulation modelling of an all-Nigeria beef supply and beef consumption model to look at options to increase beef consumption per capita in Nigeria. The key argument made is that to increase consumption, home production of beef must increase.

A few major issues arise;

• The global narrative on beef cattle numbers and links to sustainability (the title of the journal) focus upon issues linked to rising numbers of cattle and the contribution to resource use issues and recently the contribution to global greenhouse gas levels. The case is made here that there is a need to increase protein consumption per capita in Nigeria and beef production is the focs of paper to do this. Given the broader global issues, this paper needs that context within introduction and some discussion points on this issue and effects of the Policy-driven simulations produced.

• Following on this, the paper has gaps in descriptions/details of Nigerian beef production; numbers, production methods (there are non-defined categories of semi-intensive, ranching etc), current carcass weights (the paper uses ‘yields’ and very significant changes in Policy 3 & 4), slaughter ages (paper suggests changes), and current grassland systems (changes in Fertiliser use are suggested). The relationships with dairying are not covered or mentioned (over 10% of the 20M cattle in Nigeria are dairy and high numbers of male calves and cull cows will feed into beef meat trade). The other gap is the failure to mention any numbers of proportions of imports and exports, compared to home production. The modelling responses could also usefully include changes in cattle numbers, land area used by cattle providing a useful indication of the associated resource use and greenhouse gas emission impacts

• Some of the modelling data is troubling. Whilst the authors seek to re-assure that the model works well with historic data, the Policy scenarios lead to surprisingly rapid changes in the beef system. Some of these seem implausible, increasing carcass yield (which must be weight?) by double over a few years is unrealistic biologically, but even more in terms of changes by the farmers and the inertia created by modifying herd genetics with only <1 calf per year. Figure 8 looks wrong, so maybe it is, but there is little discussion of its challenges. There is nothing wrong with modelling extreme responses, but the first question is to question the response data, but when included in results section there needs to be adequate discussion. Appendix A also has all the graphs having non-unitised Y axes, and some of the shapes of the response curves appear counter-intuitive and few explanations are given

• There is also little comment on demand and price issues, rather the presumptions that human intake can be largely driven by supply. This argument needs a stronger introduction, justification of the modelling and discussion of the results of the key driving force of the modelling.

Detailed Comments on sections

Introduction

Section would benefit from comment about global cattle numbers and current issues of sustainability of these levels..resource use and climate change impacts and then the issue/conflict with low levels of BCPC

A more neutral phraseology should be adopted about change, so that instead of ‘benefit’ better to note the changes as ‘increases’ etc

At no point do we have information on Nigerian herd sizes for beef and dairy (and inter-relationship with dairy is never mentioned – male calves and cull dairy calves – wordsearch for ‘dairy’ has no results in text. Yet according to FAOstat there are 2.2 million dairy cows in Nigeria out of a population of 20 miliion cattle – a substantial flow of prospective beef animal s – c 1 million male calves that are, or could become beef slaughter animals

The section also needs some basics of the Nigerian beef herd, with these figures being part of a starting point for any modelling. Ideally these would be historic figures too, as the modelling exercise starts in 1990. It’s unclear how this exercise could be undertaken without these herd numbers and slaughter numbers

Section also talks about free-grazing, semi-intensive and ranching systems – these need to be defined properly for a Nigerian context, they will means different things to different readers around the world

2.1 Protein consumption

Put all figures in kg, there is reference here to pounds

63.76 – this is meaningless precision, round up to a sensible number

2.2 Modeling..

Here the referencing is using poor style e.g . .’…the study by [30]…’ and later ‘In contrast, [35] looked at…’ This is not good referencing style. The sentence should refer to the work and if needing to refer directly to the reference say ‘Smith et al.. [nn]…’

2.3 Challenges

‘it has been reported…’ this sort of sentence needs a reference. This is one of few sections of paper that actually refers to cattle numbers and regional variation

Methods

There is little detail here about modelling methods, assumptions and inputs and outputs

4.2 Stock flow

Here the dairy herd should fit in. Pasture is referred to, but again there is little mention of this key resource, except for the saying the numbers in the modelling came from FAO. Pasture areas, inter-relationships with cropping land and scope for improved pasture management are all relevant to a change in beef system output, yet are rarely mentioned in any quantitative terms. The reader has no idea whether pasture land in Nigeria is 1%, 10% or 80% of land area.

4.2.1 Pasture and feed module

Without a starting point in Introduction about pasture hectares, it is difficult to appreciate the scale of change needed when the re I change in the feed issues, such as feed density. Biologically, and based on history, it is difficult to envisage an increase in carcass yield (whether this is kg or kg carcass per kglive weight see later) without an increase in grassland quality, which inter-relates to land area

4.2.2

Gestation period is not a variable that can change – any changes will have minimal effect relevant to a model with much larger uncertainties

‘empties’ , not good English..

Throughout

‘Cows’ in English terminology refers to adult female breeding stock (after first calf to be precise). Often it looks like cows refers to all cattle, including growing males. Throughout careful use of cattle, cows and male and female growing stock should be referred to..

Figure 6

This mentions ‘import’ but we have heard little about these. Some appreciation of the scale (relative to home consumption) and flexibility (could it rise/fall?) is important..

Beef market module

Carcass yield is mentioned here and elsewhere, especially Policy 3 where it is a key driver/outcome of this Policy model , is never defined. Ordinarily, it would be carcass yield = carcass weight as % of liveweight ..kg/kg. But here it seems to be used as carcass weight. If carcass weight is whty is being meant then use it. Throughout defining units needs the units e.g. kg/ha, kg/hd

4.3 Policy test

Policy 1 – ‘proper’ . this is not good terminology ..use ‘improved’ or ‘enhanced’

Policy 3 ‘a carcass yield slope of 0.2’ is not clear what the ‘yield’ value is, nor here what the slope refers to. If carcass yield is in fact carcass weight then the changes in a 12 year period, as per Figure 8 are extra-ordinary. From the text it appears that this supposes a rise in animl performance to that in the USA. Given the different genetic/breed base in Nigeria, such a change to the breeding population change (the breeding cows and bulls would need to change from mainly indigeneous cattle) AND the production methods so much so as to be biologically impossible, as well as practically inforeseeable. A doubling of national carcass weight at slaughter over 12 years, given the structure of West African agriculture is difficult to foresee, and whilst models can achieve much to identify what might need to change to achive certain outcomes, here as a realistic policy objective Policy 3, then this just appears silly. A much slower rate of change, involving even a rapid rate of change by breeders/farmers, but slowed by the slow rate of change in the breeding herd (unless all cow are slaughtered and replaced by imported stock) is unrealistic

Policy 1 – increased feed growth also looks unrealistic, except at individual farm level, again because of the slowness of individual farmers to adopt any measures and thus have a singular effort. The biology of this speed of change also looks unrealistic in Nigeria.

5 fold increase in beef consumption, based on home production changes, ith a starting population of 20 million cows suggests the modelling process is dubious.

Limitations and future research

The point about whether ‘such cows need more feed’ is odd as the modelling should have involved a feed element and thus would have modelled it. An increase in carcass yield (presumed to be weight) can only come about through either an increase in biological efficiency or through increased intake over the life of the animal, both of which could/should be modelled with the appropriate assumptions.

‘effects poultry parts’ an example of some poor English and meaningless terminology

Conclusions

Paragraph 3 , the recommendations should be a series of actions rathrr than the terminology/jargon used within the modelling shorthand and should be more understandable as stand-alone. Even after reading the paper in depth it is difficult to understand what is being recommended. The last sentence of this paragraph is very unclear indeed.

Appendix A

These graphics, which are integral to the modelling are largely meaningless. The x axis of nutrition index has not been adequately explained and all the y axes are the same as the title rather a proper legend with a unit based scale. So for example graph 2 has a y axis about fertility. This flattens out at a plateau of 0.95 but without units difficult to understand. It could be calvings per cow per year, but 0.95 is unrealistically high.

Slide 6 is shown as a bell shaped curve, with both feed density index and the y axis difficult to understand. Slide 5 shows a similar curve for a y axis that might be beef consumption per human but a) all standard assumptions are that consumption rises throughout with increase in income per capita, especially from such a low base and b) the units (0-2, but with no units) don’t match any consumption figures given throughout.

Its not clear whether these graphics are outputs from model or year inputs to models. Either way, they do not make too much sense with poor labelling

Appendix B each term needs a clear plain English definition with clear units – quite a lot of the factor/variables

The hyper links do not go to stated websites but to a single

Author Response

Dear Reviewer, it appears that I need to attach again our responses to your first round of reviews (seems there is an error in the Sustainability system).  Please disregard this message/document, as you already responded within a second round of comments.  However, I wanted to make sure you had these responses just in case.  

Thanks you again and take care.

Best,

Jeff

Author Response File: Author Response.pdf

Reviewer 2 Report

My comments are included in the attached file.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

We thank you for your thoughtful comments that we used to improve the quality of our manuscript.  Please find attached a response to your comments as well as the response to the other reviewers' comments for your information.  

Jeffrey P. Walters

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript is very difficult to follow and make constructive comments because the layout is not with line numbers and clear distinction between sections. The Introduction is mixed with parts of materials and methods and the section of Literature Review & Point of Departure is confusing.  The authors propose a System Dynamics Model of Beef Consumption in Nigeria claiming that it could be used in future research to analyze consumption from other animal sources. The later statement cannot be supported by the contents of the manuscript.  The two research questions put forward are arbitrary and the whole manuscript is based on assumptions that neglect the socioeconomic dimension of beef consumption. The manuscript is repetitive and unfocused with an unreasonable large literature review and very limited information in the Material and methods section. The results and Discussion section is overloaded with Figures that are not properly explained and justified. Overall, the manuscript should be extensively revised in all sections. The authors should reconsider the major objective of this manuscript and adjust their modeling approach accordingly.

Author Response

Dear Reviewer (2),

We thank you for your thoughtful comments that we used to improve the quality of our manuscript.  Please find attached a response to your comments as well as the response to the other reviewers' comments for your information.  

Jeffrey P. Walters

Author Response File: Author Response.pdf

Reviewer 4 Report

Thank you for giving me the opportunity to review this article. I will discuss the article based first on the method used and then comment on the main research theme and research questions.

The article follows a standard presentation similar to other publications on dynamic systems. Dynamic systems method is used to analyze systems with many components, which are interrelated allowing for better understanding on mutual relations. Graphics and flow charts also provide easy to understand visualizations which facilitate introduction of policy scenarios and their evaluation. For the purpose of this article, it is a good technique. The authors present a conceptual model, followed by a causal loop diagram, stock flow diagrams, and policy scenario visualizations.

The research goal of the article is to understand why beef consumption is low in Nigeria. The two subquestions that form this research are i) identify the key drivers and how the interact as a system, and ii) identify ways to improve consumption levels. Hence my understanding is that the authors will document consumption oriented drivers and highlight why the system as is, is unsustainable. The authors proceed by presenting information on protein consumption in Nigeria showcasing the low levels can have adverse effects in health outcomes, nutrition and development. This information is important to frame the problem. The next section presents information on modeling livestock sector dynamics, and the challenges facing the beef industry. From this point on the authors focus on the production side. The dynamic system emphasizes production providing little to no information on consumption. It better presents a supply chain rather a consumer-oriented focus research, and this is one of my main criticisms of the paper. Yet, this information is crucial in understanding the connections and potential fractures in the supply chain of beef in Nigeria. Perhaps a change in title can solve part of this problem. A similar article was published by Suryani et al. 2016 in Jurnal Teknologi which can help with framing the consumption. 

Following up on that comment, the authors would have to present more information on consumption, demand by region (and I mention region broadly—it can be at a city level, area level, etc.) and also income distribution. At one point they relate per capita income with beef consumption, expecting a rising middle-income class to affect beef consumption. The authors document consumption drivers in page 3 where they discuss challenges of the beef industry, but these are not properly highlighted and incorporated in their model. This issue needs to be addressed and made more clear for the paper to properly showcase its contribution. Please elaborate more on the focus on the supply side and the choice of the parameters used in policy vs the demand side.

The link between consumption per capita and demand needs to be better documented. It is also unclear to me why population has a negative contribution to beef consumption per capita in Figure 3. With an increase in population you may have more consumption. Also, I am getting mixed signals as to what is the connection between consumption per capita and preferences on protein consumption and seafood consumption. Please address this when needed to be consistent with your contributions. 

Regarding the validity of the method. Since it requires simulations, the weights and priors to the main variables needs to be documented fully. Many of the parameters are either calibrated so literature needs to be used to back up these estimates, or are based on the authors’ judgement. This does not allow for verification of the results, let alone replication of the results.

Minor comments listed by page not by importance:

1. pg 1.
   • The first paragraph provides information on consumption and production. I would separate these and offer more literature on consumption since this is the focus. 
   • Please change the subquestions and do not present them as a list.
2. pg 2. section 2.1:

• Parts of it are repetitive. Please revisit the information presented.
• please make sure that the acronyms are introduced after the full 

3. pg. 3

• section 2.2 Modeling Beef Sector please change it to Livestock sector or consider deleting the second paragraph
• section 2.3. Do you have newer information on the industry?
• section 2.3 Not clear the connection between production and consumption. In general increased production does not lead to increased consumption

4. pg. 4

•  section 2.4. Please review the list i), comma is not needed; livestock system models without the s; last sentence should be effectively increase not increasing
• section 2 should be section 3 and please specify the Data you are using
• section 2 on methods, please specify the SDGs goals you are referring 
• figure 2, please align the content

5. pg. 9 

•  for carcass yield, other parameters to be considered? Transportation and shrinkage is a big issue in livestock

6. pg. 11 I am not sure I understand the reason to compare with developed countries like Germany and the US, and not a country that shares more similarities with Nigeria.

7. pg 15. Please provide a reasoning at to data lack that is mentioned in the last bullet point

8. Reference style. Please review the following references for style:1, 3,9, 27, 34,48

Author Response

Dear Reviewer (3),

We thank you for your thoughtful comments that we used to improve the quality of our manuscript.  Please find attached a response to your comments as well as the response to the other reviewers' comments for your information.  

Jeffrey P. Walters

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I am pleased that the authors have taken account of most of my comments, and I can see how the changes have been made and I think provide a more rounded storyline, including more of the issues of Sustainability (the title of the journal). There is more information about beef systems in Nigeria, about imports and the descriptions and definitions and labelling are much clearer.

I can see significant improvement, but still have worries that the modelling outputs look 'wrong' and unachievable given the biology of cattle, the limits on production improvements with a huge population of small-holder farmers and a huge population of low performance cattle. The outputs of the models just do not look realistic. I appreciate the authors do not want to re-run the models, whatever the comments of reviewers.

I can see some improvements, but still see some sections that need better writing style. The main model outputs and conclusions drawn upon them seem unrealistic. There remains no thorough discussion of the issues raised by having an increase in the National herd of Nigeria/and an increase in size (the idea of increasing carcass yield (which is actually 'weight') - all adding to the environmental burden at local and global scales has not really been well considered (to match the added comments in the Introduction - which are good).

My overall reaction is to perhaps move from my earlier recommendation  towards a more favourable response - but only with some re-working of the assumptions, the models and their outputs.

Author Response

Dear Reviewer:

We thank you for providing a second round of insightful comments.  We have done our best to address your comments in the attached document (and as reflected in the revised manuscript), which also presents the responses to a second reviewer who also participated in the second review round.  Per your request, we have also heavily edited for English grammar and flow.  

Thank you again and I hope that you are well despite this trying moment in history.

Best,

Jeff

Author Response File: Author Response.pdf

Reviewer 4 Report

Thank you for working on the revised draft.

I saw changes in the content but that was minor so I cannot see how that improved the manuscript to an extent that it fully conveys your message. You make the adjustments referring to a supply-chain but how is that tied well to your previous draft, that still needs to be explained. 

Portions that were added require your attention for grammar and spelling.

Author Response

Dear Reviewer:

We thank you for providing a second round of insightful comments.  We have done our best to address your comments in the attached document (and as reflected in the revised manuscript), which also presents the responses to a second reviewer who also participated in the second review round.  Per your request, we have also heavily edited for English grammar and flow.  

Thank you again and I hope that you are well despite this trying moment in history.

Best,

Jeff

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Overall

This paper has been improved again though I have some reservations about the model and the modelling outputs. 

There are still quite a lot of English issues - I have attached an annotated pdf which include a large number of suggestions, but manuscript should be checked again before final submission. A few more technical issues arise on this further reading of the paper - apologies that I could have noted some of these in earlier reviews. These are listed below

Line 64-66 – Ranch systems – this is odd terminology …do other authors use this English term in descriptions to other Nigerian or Sub-Sahara Africa countries.? In UK/US ‘ranch’ would describe an extensive system. Here ‘ranch’ seems to be used for intensive ‘farm’ based systems. Review this term and use the correct English and definition but the way I read the paper the term ‘farm’ might be best used.

Line 139 This sentence referring to Gebre does not make logical sense. Re-phrasing to the sentence below (which is from the Gebre abstract) could be a more useful sentence to include.

“The simulation also demonstrates that breeding for heavier body weight was considerably more profitable than the baseline scenario”

This is because I could not find in their paper that they had made a comparison between breeding systems with heavier genotypes and fattening systems, rather it was between breeding systems with the heavier lamb outputs and the baseline breeding systems

Line 149 This sentence adds nothing - it tells us about the total livestock production, which is noted significant. At line 143 though it is noted that the beef sector is small. Without some indication of how beef relates to rest of livestock then it adds little.

The proportion of meat protein that comes from beef, as opposed to poultry and pork (much of Nigeria is Muslim, but far from all) is not discussed at all. Increased poultry production might be a way to increase meat protein consumption? Suggest either restrict directly to beef or include a dsicussion point about alternative sources of meat protein

C. lines 425. Having now included issues of environmental sustainability in the Introduction, then the Discussion section should at least note that all policies that increased output of beef meat, would increase greenhouse gas emissions from the Nigerian beef herd. Larger beef herd, more productive pastures, larger carcasses that can come mainly with higher feed intakes, will all have more feed being eaten by cattle and greenhouse gas emissions (from both methane and nitrous oxide). Some comment would be still be worthwhile that makes the point that policies that may increase beef intake per capita may conflict with global demands for reduced greenhouse gases emissions from food production

Line 428. ‘In addition, some assumptions made under the policy scenarios would be difficult to implement given inherent biological, sociological and ecological constraints aside from those noted earlier’. This seems to be the main sentence that discusses the achievability of the model policies and outcomes. I think this sections still needs more commentary and enlargement. Whilst the models are ‘what-if’, the outcomes of models (as opposed to the assumptions feeding into them) of doubling important output factors, are still I firmly believe unrealistic. Doubling carcass weight (with  a  national herd of  20 million cattle) through increased feeding and genetics within 10 years might appear possible in a modelling context, but given the diversity and small-scale nature of African beef production, then maybe the paper should say more explicitly that ‘we understand that the projected scales of outcomes of the models are unrealistic in practice within the timescales, but we believe they show the direction of travel that production could follow if the policies were adopted’. This would be a more honest approach, rather than models with outcomes that could not be  achieved, even if the political will and resources were put into overdrive, which of course is equally unlikely.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

Thank you for another round of insightful comments and edits.  We have attached a document summarizing the changes we have made based on your comments.  Your comments, and the comments of the other reviewers, have greatly improved the quality of this manuscript - thank you.

Best wishes,

Jeff Walters

Author Response File: Author Response.docx

Reviewer 4 Report

Thank you for working on the manuscript. It has greatly improved and I do see that you have incorporated and addressed properly reviewer comments.

Some minor things to correct:

• line 38 add a comma before which
• line 45 should be overarching questions
• line 47, typos 'determine what are...'
• line 48 proposing, or consider rewording lines 47-49 since it doesn't flow well
• line 68, you have two full stops
• line 137, list item 1 needs a space
• line 230, subsection heading should be in a different line

Author Response

Dear Reviewer,

Thank you for another round of thoughtful comments and edits. We are convinced that the improvements to this manuscript are due to the thoughtful and rigorous insights from you and the other reviewers. 

The minor edits that you propose were made the manuscript.

Thank you again for your time, and we hope you are doing well in light of the craziness with COVID-19.

Best,

Jeff Walters

-------

Summary of edits:

• line 38 add a comma before which - added
• line 45 should be overarching questions - changed
• line 47, typos 'determine what are...' - changed
• line 48 proposing, or consider rewording lines 47-49 since it doesn't flow well – changed to improve flow
• line 68, you have two full stops – revised to flow better
• line 137, list item 1 needs a space - added
• line 230, subsection heading should be in a different line - edited