Effects of the Application of Biochar to Plant Growth and Net Primary Production in an Oak Forest

Round 1

Reviewer 1 Report

This article describes a field study in which biochar is applied at 5 or 10 Mg/ha to coppice regenerated mixed oak forest.  Wood production is determined through use of found allometric equations describing stem and coarse root mass as a function of stem diameter, i.e. allometry from previous study was applied.  Litterfall was separated into leaf and reproduction fractions so that net production also included these tissue fractions.  Relative growth rates of individual trees (>20 cm DBH) responded positively to 10 Mg/ha, but not 5 Mg/ha.  Neither total production, nor any of the tree fractions responded to char except for the reproduction fraction in stands receiving 10 Mg/ha char.  Authors conclude that char has a positive response to tree growth and allocation toward reproduction.  The study is well conducted, the results add important data to a growing literature on application of biochar to forestry plots, and the manuscript is well written. 

Some literature is missing from that reviewed which misrepresents the response to char and preceding studies in forests.  Authors present biochar as a strictly positive crop enhancer and do not consider the full range of published biochar studies.  Cited meta analyses, can provide overall positive responses, yet when qualified there are numerous neutral to negative responses.  Furthermore, some meta analyses showing neutral to negative responses should be included and properly represented (e.g. Jeffery et al. 2017. Environ Res Lett. 12:053001. and Spokas et al 2011 J Environ Qual 41:973-989.). Much of the biochar literature is tainted by overzealous proponents.  In fact, char frequently has little to now effect.  It is necessary to objectively represent the full range of observed plant and soil responses to biochar applications.

Authors claim that their study is the first to consider large scale forestry experiments (L 75, L 370).  Yet recent published results were not cited (Sarauer_et_al-2019-GCB_Bioenergy 11_660, Sherman_et_al-2018-GCB_Bioenergy 10_246, Sherman & Coleman 2020 GCBB 12_223).  There may be room to claim a first, but be sure to include a qualifying phrase (i.e. “that we are aware of”) when making such claims.  It is important to cite and properly represent these articles in this paper so that readers are shown the majority of forestry biochar studies.

Presentation of statistical analysis results needs to be revised and expanded.  I am concerned that the denominator degrees of freedom in the one-way ANOVA of individual tree growth may be inflated by including individual tree measurements in the model.  In this study, the plot and not the tree is the experimental unit, which means that the ddf should be 8-11 rather than something over 100 if individual trees are included as the experimental unit.  A table of ANOVA results should be included showing the degrees of freedom, P-values and possibly F-stats for the char effect for all individual tree and plot level variables. 

Authors need to share more of the individual tree analysis and consider replacing relative growth rate in the individual tree model.  Individual tree RGR results in Table 2 shows a regression output with an intercept term, which is not informative.  Rather, we need to see the estimated least squares means for all three treatments and mean separation letters like that presented in Fig. 2. Authors describe that only “canopy trees” (>20 cm DBH) and not all trees are used in the individual tree RGR model. One reason provided for not including all trees is that RGR depends on tree size. Yet a range of size is still included from 20 to over 40 cm DBH, which contradicts the reasoning.  Furthermore, RGR calculation accounts for initial size differences by including initial diameter.  So, it appears that data have been selected to show where there is an effect rather than objectively considering all data.  It is fine to consider the 20-40 cm dbh class, which is like including “crop trees” that might ultimately remain at the end of rotation.  However, readers also need to see the results for the all-trees analysis at least in the ANOVA table.  An alternative to using RGR would be a model including char treatment as class variable and initial diameter as continuous covariate.  i.e. DBH2018 = Char + DBH2016 + e, where only Char is specified as a categorical variable.

Detailed comments follow listed by line number, table or figure number.

L 41. Cite Jeffery et al. 2017. Environ Res Lett. 12:053001. and Spokas et al 2011 J Environ Qual 41:973-989

L42-3. The production increase varies depending on studies including neutral to negative.  This section should be modified to properly represent the potential rather than repeating claims that char is all good.

L 47-8. Odd phrasing shold be revised.  In what context is woody biomass used as energy.  Antecedent phrases is about forest ecosystems.  Are they using woody biomass as energy?  Here I think maybe you’re talking about bioenergy production systems described in previous sentence.

L 51. The “although” phrase is a trainlling though on an alredy completed sentence.  If it is even necessary it should be made into a separate sentence.

L 68. Sarauer et al 2019 Reforesta 7_1 and Sarauer & Coleman 2018 CJFR 48_581 did find negative growth in tree seedlings.  Sun et al 2020 Forests 11_0711 show an interaction with char and rhizobia inoculation in black locust seedlings

L 75-6. Include Sarauer_et_al-2019-GCB_Bioenergy 11_660, Sherman_et_al-2018-GCB_Bioenergy 10_246, Sherman & Coleman 2020 GCBB 12_223

L 80. Rephrase.  I don’t understand why overuse would cause cause them to be transformed.

L 90. Not so. See Sherman and Sarauer papers.

L 132-4. Recommend deleting as unnecessary.  Simply describe measurments starting a year after application.

L 168. Flush rather than flash?

L 187.  Indicates individual trees included in model.  Therefore, must adjust denominator degrees of freedom to 11 or less to represent that plot is experimental unit. 

L 204-5.  Including means and errors in text is redundant with Table 1 and should be deleted as unnecessary.

L 212.  Here it appears criteria for ingrowth was height reaching 1.3 m (i.e. DBH), yet in L 135 and L 156 the ingrowth criteria seem to be reaching 5 cm dia at DBH.  These should be consistent with what was used and not contradict.

L 217-8.  Expand.  I don’t understand what weakened stems means in this context.

L 221-4.  What was the P-value for char treatment from one-way ANOVA? We need to know this for all response variables.  Suggest including anova table that includes all variables.  What was mean RGR value for controls?  Are means for C5 different from C10.  Here suggest including a figure that looks like Fig. 2.  We also need to see mean DBH2016 and DBH2018 by treatment to compliment RGR data.

L 235, 2389, 241. Again, delete data from text that is shown in Figure.

L 253-4.  I don’t understand how including trees that range from 20 to 40 cm DBH eliminates the effect of size.  We need to results of RGR for all trees to compliment the only canopy tree analysis.  Try using initial diameter as covariate in DBH2018 model.

L 256.  This is not an objective perspective and should be more balanced.  Biochar can have negative to neutral effects as well as positive effects.

L263-4. Need to represent Sarauer_et_al-2019 and Sherman_et_al-2018 here.

L 294-300. Fig 2 only shows differences in reproduction and there are no differences among char treatments in NPP of other fractions.  Discussion should not ignore statistical results and accurately represent means as being statistically equivalent.

L 336-7.  Delete.  Speculative without evidence. 

L 363. No information was presented on nutrient availability, so it is not possible to draw conclusions about such a mechanism. Okay to move to discussion, say L 324, but does not belong in conclusions for this study.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The study deals with the important and timely issue of the usage of biochar in woodlands, which could beneficially affect the stand growth as well as the carbon storage in forest ecosystems. Thus, the presented results are interesting both from scientific and practical point of view. The experiment was properly designed and the paper was well written. Thus, I do not have reservations of major importance, and I recommend publishing the article in Forests after minor revision.

My minor points are as follows:

• I suggest deleting “natural” in the paper’s title, because the word could mislead a reader. The studied stands are secondary coppice forests (e.g. l. 100); thus, I guess that the Authors by “natural” understand rather “naturally regenerated”. Secondly, by deleting the word the title will become shorter, which is beneficial. The word should be deleted / changed also in some other sentences in the paper body (e.g. l. 91 and in the caption of Table 1).

• 19/20: The content of the parenthesis should be corrected to be clearer to a reader.
• 84: Following the name “Murray” given after the Latin name the name of a name giver should be consequently given also when other Latin names are written (e.g. in l. 127).
• 100, 148: “deciduous broad-leaved forest” – please stay here “deciduous” or ”broad-leaved”, but not both.
• 110: The reference (Soil Taxonomy?) to the used soil classification system should be inserted.
• 112: “plots (plots” –> better: “plots (numbered from”
• Some additional information on the studied oak stands should be given in M&M. For example, was the stands even or uneven aged? What was the average age and approximate height of the studied trees?
• 128/129: “The aboveground biomass of the dwarf bamboo was measured in July 2017.” What do you specifically mean? What was done? In fact, the bamboo biomass was described in Results (l. 209-210); however, the method the Authors obtained the results was not described in M&M.
• 174, 228: “Quercus serrata” -> “Q. serrata”
• From Table 1 a reader knows that the values which follow the averages in the text are SE values. However, it should be clearly stated in M&M (e.g. after l. 194).
• 208/209: “there was a tendency for C5 to have a lower biomass (F2,9 = 0.46, p = 0.65)”. The statement could be deleted as it refers to the clear insignificant difference.
• Tables are not aesthetic. They should be reedited.
• Table 1: I) I suggest deleting “Plot” in rows of the second column, but, instead, to insert “Plot no.” in the column head; II) “speceis” -> “species” (the third column head).
• 248: I suggest replacing “Small letters” with “Lower cases”
• 249: Which post-hoc test (Dunn?) was used to examine the differences between mean values? It should be clearly stated in M&M.
• Some statements placed in Discussion (l. 253-256, 293-297) are the repetition of the results’ presentation placed in Results. The phrases should rewritten / deleted.
• 269: Do you really mean “lower site quality” here? It is surprising because fine textured soils form usually forest site of higher but not lower quality than sandy soils. Please, write more specific what you mean here.
• l .287: “ca. 10 yr” -> “ca. 10 yr old” (?)
• 334: “thre” -> “the”
• 341: “increases diminish” sounds awkward. Please, rewrite the sentence for the more reader friendly version.
• l. 370/371: Only here a reader is informed about the approximate age of the investigated forest stands. However, as mentioned before, the specific information on the age should be given earlier in M&M.

Author Response

Please see the attachment.

Author Response File: Author Response.docx