Modelling the response of net primary productivity of the Zambezi teak forests to climate change along a rainfall gradient in Zambia

Ngoma, Justine; Braakhekke, Maarten C.; Kruijt, Bart; Moors, Eddy; Supit, Iwan; Speer, James H.; Vinya, Royd; Leemans, Rik

doi:https://doi.org/10.5194/bg-16-3853-2019

Articles | Volume 16, issue 19

https://doi.org/10.5194/bg-16-3853-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/bg-16-3853-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 16, issue 19

Research article

|

08 Oct 2019

Research article |

| 08 Oct 2019

Modelling the response of net primary productivity of the Zambezi teak forests to climate change along a rainfall gradient in Zambia

Justine Ngoma, Maarten C. Braakhekke, Bart Kruijt, Eddy Moors, Iwan Supit, James H. Speer, Royd Vinya, and Rik Leemans

Download

Final revised paper (published on 08 Oct 2019)
Supplement to the final revised paper
Preprint (discussion started on 24 Oct 2018)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'review', Anonymous Referee #1, 14 Dec 2018
- AC1: 'Response of Net Primary Productivity of Zambezi teak forests to climate change along a rainfall gradient in Zambia', Justine Ngoma, 23 Jan 2019
RC2: 'Comments for bg-2018-421 manuscript', Anonymous Referee #2, 03 Jan 2019
- AC2: 'Response of Net Primary Productivity of Zambezi teak forests to climate change along a rainfall gradient in Zambia', Justine Ngoma, 23 Jan 2019

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (26 Jan 2019) by Paul Stoy

AR by Justine Ngoma on behalf of the Authors (29 Jan 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (02 Feb 2019) by Paul Stoy

RR by Donal III O'Leary (01 Mar 2019)

Suggestions for revision or reasons for rejection

The authors present an interesting paper on an important topic - the capacity for forests to offset increases in atmospheric carbon concentration through the mechanism of NPP and carbon sequestration. I think that this will be a useful contribution to the literature, however I have some concerns in the validation and quantification of the results, and suggest non-trivial revisions to the presentation of the material.
One concern is that the discussion of default vs. local parameters for the model is unnecessary. In my opinion, the model should not be run with default vaules, and the presentation of those results does not add value to this manuscript. It appears that a previous reviewer has raised this concern before, and it was not heeded.
A small but substantial concern is in the validity of the NPP change results. I am not an expert on this model, and I do not understand specifically what the margin of error is for NPP estimates. It seems to me that the model estimates fairly small changes in NPP, which may be within the margin of error for the model, and therefore statistically insignificant. I would like to see this rigorously addressed in the methods, results, and discussion.
I am curious about the use of terms such as 'correlation' in the absence of statistical tests and test statistics (results). Perhaps this is nit-picking, but I suggest avoiding these terms as they can be misconstrued from a statistical perspective.
I have noted other small editorial concerns such as the appropriate use of particular terms, and made suggestions for wording that I would consider more appropriate.
I see great value in the science conducted here, and I think that this will form a substantial contribution to the literature with further revision. It is clear to me that the authors have invested countless hours of thought and effort into this manuscript, and I hope to see them successfully revise this into a publishable manuscript.

Referee Report: PDF

Hide

RR by Anonymous Referee #1 (12 Mar 2019)

ED: Reconsider after major revisions (12 Mar 2019) by Paul Stoy

ED: Referee Nomination & Report Request started (22 Mar 2019) by Paul Stoy

RR by Anonymous Referee #3 (11 Jul 2019)

Suggestions for revision or reasons for rejection

This is my second review of this manuscript, and I am happy to see that it has been substantially revised in a way that is quite responsive to my previous comments. After reading through the authors’ responses, I think that I understand this manuscript much better, and I think that the authors’ arguments are well founded and generally supported. I have some suggestions for further revision, though I would like to say that I think this manuscript is much improved.

My main concern at this point is concerning the use of the word correlation. After reading through this again carefully, and considering the authors’ responses to my previous review, I think that this term is perhaps being used incorrectly, though it should be fixable and will not likely substantially change the heart of the results. In the manuscript, as well as supplementary figures S1-S5, the authors use the term correlation where I believe they are performing a bivariate linear regression. While they are very similar, their results and interpretation vary. The results shown here have a scatterplot with a best-fit line, as well as associated r2 value and p-value. I presume that these values were generated with a linear model (such as lm() in R). The resulting r2 value represents the amount of variability within the dependent variable (y-axis) that is explained by the independent variable (x-axis). The linear regression is predicated on a causal relationship between the variables (which seems to be true for these comparisons, though it isn’t explicitly stated as such). A linear regression will often also report p-values for the slope and intercept, with the null hypothesis being that, separately, the slope and intercept are equal to zero. It appears that the p-value reported herein is for the slope of the line. Therefore, having a p-value which is small for the slope of the line in these bivariate regressions is evidence that the slope is non-zero, though this is not the same thing as having a statistically significant correlation (though one usually goes with the other).
On the other hand, the authors could use a one-tailed Pearson’s product-moment correlation test (presuming their data satisfy the test assumptions, otherwise a Spearman’s would probably work fine with similar results) to determine if there is a significant correlation with P<0.05 (or their chosen alpha level). It is likely that the results of the correlation test will agree with the slope of the regression line. I think that your story will stay the same either way, but I would prefer to see the term correlation used to describe a correlation test, or to rephrase the manuscript to describe the linear models appropriately (because they are not technically correlations).
Finally, figures S4 and S5 have a few panels showing P<0; which would be better as P<0.001

Line-item notes:

Page 2 Line 15: looks like there are two spaces between: “respectively especially” but maybe that’s the justified formatting.

Page 2 Line 20: I would remove the comma after ‘thus’

Page 3 line 10: I would remove comma after “Reduced rainfall”

Page 3 line 12: This paragraph begins with a statement about regional variability, however this is followed by statements that seem to describe Africa as a whole.

Page 5 line 8: suggest “80 tree species” or otherwise specify as appropriate.

Page 13 line 28 – minor suggestion; I think it would be more clear to say that NPP increases at Namwala and Kabompo sites, but decreases at Sesheke

Page 17 Table 3: perhaps I’m missing something, but these changes in NPP % appear to differ from the values listed in the Conclusions (Page 19 section 4.4)

Hide

AR by Lorena Grabowski on behalf of the Authors (01 Jul 2019) Author's response

ED: Publish subject to minor revisions (review by editor) (16 Jul 2019) by Paul Stoy

AR by Justine Ngoma on behalf of the Authors (30 Jul 2019) Author's response Manuscript

ED: Publish as is (12 Aug 2019) by Paul Stoy

Download

Article (4053 KB)
Full-text XML

Short summary

The Zambezi teak forests are a source of raw material for the timber industry. Through application of the LPJ-GUESS vegetation model, we determined the forests' response to climate change at the wetter Kabompo, drier Sesheke, and intermediate Namwala sites in Zambia. While increased CO₂ concentration enhances forests' productivity at Kabompo and Namwala, the decreased rainfall will reduce forests' productivity at Sesheke by the year 2099, resulting in reduced raw material for saw millers.