Natural Recovery and Liming Effects in Acidified Forest Soils in SW-Germany
Round 1
Reviewer 1 Report
Dear Authors,
the manuscript is interesting and the subject matter is current and well introduced.
Some keywords are not relevant while others are duplicated. Some sentences in the “Introduction” are long and unclear, causing trouble to the reader.
-Lines79-83; lines 88-91; lines 93-96.
The description of the measures and results of the 1983 and 1989/90 sampling campaigns could be included in a table and deleted from the text.
Line 167: gon?
The discussion should also be revised, the authors should avoid repeating the results already described.
There is a lack of general conclusions, the authors simply respond to the aims.
The manuscript needs minor revisions.
Author Response
Dear reviewer,
thank you for your helpful input, we greatly appreciate you taking your time to read our manuscript and advise us on those parts that needed improvement!
In response to your points:
Line 37: we removed the keywords “Picea abies” as irrelevant and “natural recovery” as a duplicate
Lines 79-97: long sentences were split in shorter ones and some words rearranged to improve readability in the introduction
Lines 152-176: we moved the description of our sampling methods from text into a table to provide a clearer overview of the different sampling campaigns. The results of 1980s sampling were originally described in Wilpert et al. (1993) using barplots and summaries in text form, without tables of actual measured values; thus we feel it would be difficult to derive the absolute values we would require to construct a satisfying table for the discussed soil chemical parameters.
Line 153 in Table 2: we used “gradian” instead, a synonym to “gon” in English (unit measurement of angle, 1 gradian = 0.9°)
In chapter 4.3 we attempted to focus and shorten the discussion, beyond this we did not see much redundancy in the explanations of our results. For each essential study parameter we intended to give first an overall summary and synoptic overview of our own results before discussing and comparing that with results from the literature, finally in an attempt to find explanation for the differences between our study site groups. We felt that this structure was essential in order to help our readers navigate what we see as a complex topic with very much interconnected soil chemical parameters. (Lines 476-477, 536, 558, 575-576, 588)
Lines 614-625: we rewrote our conclusions in order to provide a more general conclusion and outlook to our work
We hope this has improved the quality and readability of our work.
Reviewer 2 Report
Comments on the paper « Natural recovery and liming effects in acidified forest soils in SW-Germany” by Lelde Jansone, Klaus von Wilpert, and Peter Hartmann.
General comment
This study reports on the long-term development of acidified soils with their potential for natural recovery on established control plots, and the long-term effects of repeated lime application – at a collective of 10 study sites in growth regions and soil properties. The study is unique and increase the knowledge on soil resilience and its indicators. I really like this report (although I identified mainly 2 real weakness that shall be corrected) and encourage the editor to accept this publication with major changes, and at the condition the authors improve the manuscript substantially by taking into account the following specific comments.
Specific comments
L 126: “Directly adjacent to an untreated “control” plot a “limed” plot was established…” Was the studied “limed” plot the same size as the studied “control” plot, or the whole plots (size in Table 1) were under study? Please be more specific.
L 127: “up to 8% MgO, 4% P2O5 and 10% K2O”… what about the percentage of Ca and lime grade (particle size)?
L 134: “(CaMg(CO3)2 55/35)” : be clearer with the units of the two numbers (percentages?) and grade.
L 139: “the scientific investigations have accompanied the regular forest management practices here”. Can you briefly outline what was the regular forest management practices please? Breifly.
L 150: “Six to eight samples of the mineral soil were mixed into a single sample for the laboratory analysis…” Can you give more details about soil sampling design since the experimental areas had several hectares. Was it done along a transect? What was the spacing between sampling points, etc.
L 175: “…Eijkelkamp root auger (cylindrical, diameter 8 cm, length 15 cm) up to 60 cm mineral soil…” was this root auger instrument used for sampling at the 5 randomly distributed points as well as in 5 m distance from a single soil profile? So it was always volumetric soil sampling (except in the soil profile)? This section is not clear about how the root auger was used.
L 199: “…to explore the similarity of sites by their soil chemical parameters at 0-30 cm control plot mineral soil” and L 217: “The statistical analysis was applied to both site groups separately.” I really wonder why the authors decided to form groups. It can be interesting to explore the data to show similarities and discrepancies among and within sites, but I do not understand why the statistical analysis was done separately for the two groups. If soil texture is deemed an important feature, then why not use this criterion for grouping the sites? And this grouping was done on the 2015 soil data from the controls, which were more or less altered by the recovery from acid deposition reduction. This is a first real weakness of this paper. Why not use the data from the start of the experiment? All are questions raised by this intention of grouping…
L 242: Section 3.1. Liming effects in 2003. I found this section very short, while there is a lot presented already in the graphs that could be presented in this section. This is 20 years after the first liming treatment. It is of considerable importance to report the change (or no change) of the treatment after so much time. For instance: no effect of liming on O horizon pH (according to Figure 2!), strong effect on the mineral soil BS (according to Figure 4 [why the data on the O horizon is absent in this figure and the next ones???]), but no effect on mineral soil CEC (0 to 60 cm depth according to Figure 6!). This section should be expanded to report in more details on the soil response to moderate liming rate after 20 years.
L 471: “The combined effects of the first standard dosed and the second doubled dosage campaign have in fact been observed in the 2003-2010 observation period” I am not so sure I would buy that given the previous comment…
L 616: “We were able to evaluate the development in soil chemical parameters, specifically the acidity status and soil C and N that have been documented from 1980s until 2015 at our long-term study sites” : I would have liked to see these data from the 1980s in a table in the paper, but they are not!!! This is very unfortunate. This is the second real weakness of this paper, not presenting the 1980’s data and change over the first 20 years.
Author Response
Dear reviewer,
thank you for your time, expertise and helpful input, we greatly appreciate it!
In response to your points:
“Plot size” column in Table 1 was edited to make the distinction between “control” and “limed” plot sizes more obvious (yes, an entire plot was used in sample collection across the different sampling campaigns, and no, “control” plot was generally smaller than “limed”, likely because the lime application was done not only for scientific purposes but also as a forestry practice in 1980s).
Line 128: we added more detailed information concerning the exact % of the calcium carbonate mixture’s different components; the % of Ca-component varied in accordance (around 40% CaO)
Line 129: added lime grade (particle size)
Lines 136-137: clarified that 55/35 stands for 55% CaCO3 and 35% MgCO3 and added lime grade (particle size)
Lines 142-143: listed the standard forestry practices that we have seen at our study sites since establishment in 1980s
Lines 152-176: moved the description of sampling methods from text into a table in order to create more structured and detailed overview
Our intention for grouping the study sites was on one hand to reduce the high variance in our data due to inherently high heterogeneity of site specific properties (substrate, growth region, altitude, tree stand composition, water regime etc.). Also, we wanted to explore the potential source of similarities between the soil chemical parameters at our study sites early on, which we have come to believe is the soil texture, yes, but not strictly that – the acidity status i.e. the chemical properties themselves appear to pre-determine response rates of change. We found this very interesting. We chose the 2015 data because we required fine earth stocks in order to be able to directly compare one site’s element stocks to another site (where fine earth stocks are very different the element concentrations are not considered a suitable means of direct comparison), and this information was not available in the previous sampling campaigns. Also, the number of replicates was lower in the previous sampling campaigns (which is also the reason we could not apply parametric tests to the 2003 and 2010 data with n<30).
Lines 235-238: we attempted to formulate better why there is no statistical comparison between 1980s data and 2003, i.e. the initial 20 years after 1st treatment
We do agree that a proper statistical comparison between the initial 1980s soil data and 2003 follow-up sampling campaign i.e. the initial 20 years since treatment would have been greatly beneficial to the overall study. However, there are several reasons why we could not pursue this. Firstly, due to being a part of a much larger collective of a total of 174 study sites in 1980s, the ten sites in our present study were a part of different site collectives that had not been sampled in the same year but rather over two larger campaigns (1985/86 and 1989). The digital records available to us at present day appear to be incomplete in some cases and caused us some confusion. Secondly, the 1980s sampling depth was limited only to the upper 10 cm mineral soil (the reason being the initial liming effect was expected to reach no further than topsoil), making it impossible to compare any potential changes between then and 2003 in the soil profile below 10 cm depth, as we would have wished for, not only with respect to lime treatment effects but also to any possible natural recovery in the soil profiles. All in all, we relied on the expertise and integrity of the published results of our former colleagues to describe the initial site conditions from the establishment of this study in 1983.
Lines 246-254: we expanded the paragraph in order to discuss also the lack of significant differences in soil acidity parameters pH, BS as well as CEC by 2003 sampling campaign
We hope to have addressed properly your points of concern for the manuscript, and thus improved the quality and readability of our work.
Reviewer 3 Report
I carefully read the submitted manuscript. This study investigated natural recovery and liming effects to the acidified forests with four different soil types in South West Germany. Over all, I think the manuscript is well constructed, data are solid, and statistical tests were applied adequate. In addition, the large data sets of this study provide important and useful information for readers of the journal.
Here I suggest some minor comments
Material and Method
1: If possible, the location of study sites are pointed by map in appendix.
2:For the K-mean clustering, totally, how many samples were applied, 50 samples (10plots x 5 replication in each plots)?
Discussion
3:In discussion section and line 632-633; in liming effects, I think a table of comparison of previous studies and this study are helpful for reader.
That's all.
Author Response
Dear reviewer,
thank you for your time and helpful input, we greatly appreciate your assistance!
In response to your points:
Lines 649-650: we added a map of study site location in Appendix A Figure A1 (and Line 144: cited map in Appendix)
We also noticed our latitude/longitude format needed a correction in line 145 Table 1
Line 202: specified sample n=49, i.e. 10 sites x 5 replicates except for just 4 replicates at study site “Freudenstadt"
We decided against a table of comparisons of our study with previous studies. We feel that an adequately complete comparison covering all relevant parameters and relevant literature – enough for a table of comparisons – would blow up this study and should be done instead in a separate meta-study. The number of studies represented in our discussion's text is restricted to studies from Germany and its neighbouring countries - with the intention to have had the climate and other site conditions as comparable as possible to our study sites.
We hope this improves the readability and quality of our work.
