Review for "The seasonal reproduction number of dengue fever: impacts of climate on transmission"

Completed on 13 Jan 2015 by Cory Morin.

License: http://creativecommons.org/licenses/by/4.0/

Login to endorse this review.


Comments to author

Basic reporting

• The introduction is fairly well done but I would appreciate if the author addressed a couple of minor issues:
o The last sentence of the first paragraph states that females obtain blood to gain energy for egg laying. While this is true, it would probably be more appropriate to replace “energy” with “nutrients” or say “energy and nutrients”.
o Towards the end of the second paragraph the author discusses the effects of “global warming” on dengue vectors. I would suggest replacing “global warming” with “climate change” because it is a more general term and encompasses possible precipitation changes which will undoubtedly influence vector population dynamics.

• This article does need extensive grammatical editing. I understand that English is not the author’s first language but often the grammatical errors make it difficult to read and understand. I suggest the author work with someone in the journal’s editing office or their university who speaks English as their first language.

Experimental design

Overall I think the author used an appropriate and interesting methodology. The equations used in the study seem appropriate and are justified by the literature. However, I do have a few questions and suggestions listed below:

• At the end of page 3 it reads, “…hatched pupae..” which is incorrect since larvae emerge from the eggs and complete 4 larval instar stages before becoming pupae and then finally adult mosquitoes.

• It should be made clear that c is actually the inverse of the EIP. When I first looked at equation 3b I thought it was incorrect until I realized the mistake was in how c was described. For instance, at 30°C c = 0.1. What this really means is that the EIP is 10 days at 30°C and the rate of completion of the EIP at 30°C is 0.1.

• Rainfall has a large impact on vector population dynamics and it is addressed in the paper using carrying capacity K, however, I would appreciate a little more detail describing K and its calculation. For instance, how are KE, KM, and the minimum egg carrying capacity determined? Additionally, there is very little information about φ and how it is calculated and what types of values it takes.

• Table 1 is very useful for looking up parameter names and values while reading through the methods section. I think this table could be expanded for other variables and parameters as well. I often found myself searching through the paper to remind myself of variable names, values, and equations (a, bm, bh, s, c, µm, Km, KE, NH, SH, IH, RH, NM, LM, I, M, NE, EE, IE). A separate table or expansion of Table 1 with these variable/parameters, their meanings, and their values/equations would be very helpful.

• In most instances the model was only run over a single year using averages. It might be worth also running the model under warmer/cooler and wetter/drier conditions. This could be done by using the lower/upper quartile temperature and/or precipitation for each month. This may yield interesting results.

• The model produces interesting results, however, no validation analysis has been performed or referenced. Is it possible to do such an analysis? If so, this would greatly enhance confidence in the results. It seems monthly temperature values were used as input but the model seems to be run using a daily time step. Is it possible to run the model using historical temperature data? If so, could the results be compared with dengue case data or mosquito population data from Chiang Mai?

• How were the initial values for the model results in Figure 3D chosen?

• In the second paragraph under the subtitle “The Effects of Climate to Dengue Transmission” is EM supposed to be LM?

Validity of the findings

• Under the section “Sensitive Analysis of Parameters” is says that vector control represents the only method to control outbreaks. I would say “vector control and vector-human contact reduction” or something similar because these are two different things but they are both discussed in the paper. The former refers to minimizing the vector population while the latter refers to avoiding contact with the vector.

• The author concludes through the sensitivity analysis that the best way to limit dengue transmission is through avoidance of the mosquitoes. This is because the model shows that the largest drop in dengue infections occurs when b is at half its value compared to when K is half its value or µm is doubled. However, this does not consider the difficulty in achieving the values. For instance, it might take a lot more time and money to reduce b by half than to reduce K by half. Assessing the effort to achieve these values is far beyond the scope of this study but it is worth mentioning some of the possible benefits and challenges of each of the intervention methods being modeled.

• The decline in the mosquito population later in the year is due to exhaustion of food supplies for the immature mosquitoes in the model. It seems to me, however, that the food supply would be constantly replaced by organic materials from leaves and other insects that fall into the habitat. Is there much evidence of food supplies being depleted later in the year? If so, what mechanism is replacing the food for the next year that is not replacing food throughout the year?

• The limitations section is rather brief and I believe there are some other limitation that should be mentioned. I understand, of course, that most can’t be avoided but it is worth discussing the limitations to put the results in perspective and prepare for future research. Some examples are listed below:
o No validation analysis was performed so it is difficult to assess the accuracy of the model results
o Much variability in the parameter values/equations exist but many are not well known or studied. I believe the author did their best to choose appropriate values but it should be acknowledge that there might be great variability in the values over time and space.
o The model was run using average monthly climate conditions but weather and climate can vary from day to day and year to year. This will not be reflected in the model results in the present study.

Comments for the author

Overall I like the methodology of the paper and believe it has the potential to produce interesting and useful results. I do feel, however, that extensive grammatical editing needs to be done before I can fully judge the quality of the manuscript.