Dear Prof. Ma,

Your manuscript has been reviewed by two external reviewers with relevant expertise and your responses to the comments are adequate. New experiments and data mining were also conducted, and your new figure 1 showing the experimental flow diagram is particularly welcomed. However, there are still major concerns that preclude publication of your manuscript at its current form

(1) First of all, please provide all information about pyrosequencing. Please make a supplementary table showing the number of raw sequence, read length, high-quality sequence number and so on. In addition, I strongly suggest that each of sequence data (individual location) be assigned with an unique accession number. In other words, please submit your high-throughput sequencing data to the GenBank on the basis of individual sample tested in this study, but not simply as a whole with data pooled from all samples. This will be of great merits for others who might be interested in your data.
(2) Table 1 is a great supplement with very useful background information. However, this table needs to be re-organized and it is certainly not well adjusted for publication
(3) Please clearly define the two different strategies, particularly the drawback for the indigenous microbial flooding technique. The authors state that “Indigenous microbial flooding technique improves oil recovery by introducing oxygen and salts through water-based injection to stimulate reservoir microorganisms. Despite the validity in field trial, this technology also has some limitations, in particular, instability during microbial flooding process.” It is not clear here. The injection of exogenous microorganisms and of ex-situ products has drawback due to the sieve effect of strata on microbial cells. So the oxygen and salts are introduced to stimulate the indigenous microorganisms. However, to my present understanding the migration of reservoir microorganisms would be subjected to sieve effect of strata belowground. So what does the author really mean about “instability during microbial flooding process for the indigenous microbial flooding technique”
(4) The lack of control. Please clearly state that it is difficult to document the indigenous microbial community structure” somewhere in the text, and I personally feel that backflow water sample can not be used as a control (see comments below). In addition, water-injection with some nutrients is said to stimulate the growth of certain indigenous microbial functional guilds. But it is rarely discussed.
(5) At the place of first appearance, please define what is “injection and production water samples of water-flooding petroleum reservoirs”. Most people would not have time reading every words of your manuscript. Therefore, it would be very useful if you could clearly mention what the purpose is for injection well and production well in line 23-24(very briefly)
(6) Line 31. This statement seems to be contradictory to the previous one. If the community structure exhibited large differences between the injected and produced water samples (Line 31), a large number of microbial populations would not have been shared in the conglomerate reservoir (in contrast to what have been observed in Line 29-30)
(7) Line 37. The term “relative abundance” could be removed
(8) Line 38. Because the authors do not determine oxygen concentration, it cannot be placed.
(9) Line 137. Please clearly state what “Backflow” is. Is this water flowing upward through an independent well?
(10) Line 200. This section needs to be described and discussed in greater detail.
(11) Line 368. I think there is no solid evidence in support of the sequencing depth effect in this study. The authors need to be careful about this statement. Provided that the number of sequence reads is more than 3000, the core structure of microbial community might be captured regardless of sequencing methods.
(12) Line 467. In Conclusion section, please clearly state what is the main difference, i.e., the taxonomic identity of microorganisms that exhibited large differences. In addition, the term of “structure” might be used throughout the manuscript because the authors conducted additional experiments to quantify the population size.
(13) Line 475. The authors need to be careful about this conclusion because the water backflow would most likely be different when compared to injection water. The reason is that the soil and subsurface sediment might serve as filters, leading to the difference between injected water and backflow water? Please carefully explain this point.
(14) All figure legends need to be re-organized. It is expected that your figure legends will be quite detailed and very precise. In fact, from the figure title and the axis labels of a graph/table the reader should be able to determine the question being asked, get a good idea of how the study was done, and be able to interpret the figure without reference to the text.

Kind regards

Zhongjun Jia

Dear Dr. Ma,
Thank you for submitting your manuscript to BG.
The manuscript has been reviewed by two external experts and the Editor (myself). All reviewers find that your study is of interest for general readers, and the comments have been adequately addressed.
Before publication, there are few minor concerns for your attention as follows
(1) Line 28. I guess it is not “injected” and “produced”, but it is water in injection well and production. Please be careful about the usage of “produced” throughout the manuscript.
(2) Figure 6. please note that the circle should have been within box.
(3) Figure 6. please replace “I” with “bacteria”, and replace “II” with “archaea”

Kind Regards
Zhongjun Jia