Completed on 14 Apr 2016 by Kevin Geyer. Sourced from http://biorxiv.org/content/early/2016/03/17/043372.
Login to endorse this review.
Author response is in blue.
This is an interesting study, and research addressing the controls on ‘relic DNA’ concentrations in soils is important! We provide the following comments in an effort to strengthen the manuscript for publication:
1. Clarification of the pH effect on relic DNA is important - conflicting information exists in the text (lower pH soils have more relic DNA; Lines 207-208, Table 1) and figure S7 (lower pH soils have more similar communities after PMA treatment). Given these relationships, it appears that your data suggest a negative relationship between the quantity of relic DNA and community dissimilarity. Is this the case? In our opinion the presentation of the data obscures this point. For example, samples are ordered differently in Fig. 1 and 2, and regression of community dissimilarity against pH is in the supplement. Why not directly regress community dissimilarity against relic DNA quantity?
2. Why was relic DNA quantity converted to a binary variable for regression with edaphic factors (Table 1, Fig. S6)? Why was 20% chosen as a cutoff for relic DNA presence/absence? If the assumption is that PMA treatment is not quantitative this should be discussed prominently in the manuscript, as this affects interpretation of results.
3. Sample storage should be discussed more explicitly as this could potentially affect the number of viable cells. How long were samples stored before PMA treatment? In addition, a test of the effect of soil moisture on relic DNA quantity would be useful for interpreting the data as this could be an important determinant of the number of viable cells.
4. A broader discussion of the significance of these findings for other biogeographical work on soil microbes would be valuable, as it calls into question the validity of earlier DNA-based approaches (e.g., how diversity relates to soil properties).
-- Kevin Geyer, Eric Morrison, Serita Frey (University of New Hampshire)
Excellent comments. We will address them all in future drafts.
1) If I recall, there was no relationship between % relic and dissimilarity. Because of this we chose to change the order of the soils to clarify the relationship between 16S dissimilarity and ITS dissimilarity. We will report this more explicitly in future drafts.
2) Binary was chosen because, due to measurement error, some of the mean % relic DNA values are negative. Since this is not technically possible, we used binary to accommodate those values. We chose less than 20% based on the distribution of the data which showed two distributions in the % relic DNA. The natural cutoff was 20%. We will clarify this distinction in future drafts.
3) The collection dates and processing dates are listed in Table S1. All samples were shipped overnight on ice (never frozen) and kept at 4'C until processing, if necessary. The longest gap between collection and processing was 7 days. Those samples had minimal relic DNA (29,30,31).
4) Good point. We will clarify.
Thanks for taking the time to comment.