Investigating Masking Effects of Age Trends on the Correlations among Tree Ring Proxies

Round 1

Reviewer 1 Report

This work involves a tremendous amount of data: multiple field sites sampled, multiple number of trees per site, two species of trees, tree ages upwards of 300 years, and measurements of four different variables (width and three stable isotopes).  Hence the large amount of graphics in supplementary materials.  Congratulations.

A central tenet in dendrochronology has been that multiple tree-ring predictors of some environmental feature not be overly correlated with each other, which would cast doubt on the significance levels in multi-variate regression analysis.  The central conclusion here is that the four tree-ring variables tested are reasonably independent of each other, especially after some detrending or another has been applied.  Good news, which merits publication.

Comments:

The expression "cambial age" seems odd.  It's not that a tree's cambium has age, as the cambium is just living cells that must die and replace themselves through time.  A recent article in Forests (Biondi, From Dendrochronology to Allometry, Forests 2020, 11(2), 146; https://doi.org/10.3390/f11020146) spoke to this: "cambial age, a value that refers to a tree ring, not to the vascular cambium, whose physiological ability to divide and form new tissue defies the commonly held notions of aging and senescence."  Alternatively, if cambial age is just a tree's age, or the age of the tree for any given tree ring, then why not call it that?  Young trees are young, old trees are older.  The outermost rings of an old tree were applied recently, but an old tree is considered old because it has 300 rings.

Line 76: first first-difference?

Figure 2h: The pattern of labeling these graphs was broken here.  Should be TRW, instead of spline.  More generally, I would at least experiment with giving each row its own label (e.g., ND for row 1), and each column its own label (e.g., dD for column 1).  The entire figure is TRW, so that could be a figure label.  All the x-axis labels are the same, so that could be a bottom label.  With that, the many titles of each subfigure could be eliminated, allowing the data parts to be expanded.  Just a thought.

Line 325: "reserves . . . may only be used for gapping stress."  Gapping stress?  I've not ever seen that term before.

Author Response

This work involves a tremendous amount of data: multiple field sites sampled, multiple number of trees per site, two species of trees, tree ages upwards of 300 years, and measurements of four different variables (width and three stable isotopes).  Hence the large amount of graphics in supplementary materials.  Congratulations.

A central tenet in dendrochronology has been that multiple tree-ring predictors of some environmental feature not be overly correlated with each other, which would cast doubt on the significance levels in multi-variate regression analysis.  The central conclusion here is that the four tree-ring variables tested are reasonably independent of each other, especially after some detrending or another has been applied.  Good news, which merits publication.

We are grateful to the reviewer for the nice words and the helpful comments

Comments:

The expression "cambial age" seems odd.  It's not that a tree's cambium has age, as the cambium is just living cells that must die and replace themselves through time.  A recent article in Forests (Biondi, From Dendrochronology to Allometry, Forests 2020, 11(2), 146; https://doi.org/10.3390/f11020146) spoke to this: "cambial age, a value that refers to a tree ring, not to the vascular cambium, whose physiological ability to divide and form new tissue defies the commonly held notions of aging and senescence."  Alternatively, if cambial age is just a tree's age, or the age of the tree for any given tree ring, then why not call it that?  Young trees are young, old trees are older.  The outermost rings of an old tree were applied recently, but an old tree is considered old because it has 300 rings.

Thank you for raising this issue. Indeed the cambium never ages! However, the definition given by Swiss Federal Institute for Forest, Snow and Landscape Research WSL is: “Age of a section of stem. The first year corresponds to the first tree ring outside the pith.” Thus, we use it to indicate the ring number from the pith, and to avoid confusion with the chronological age. Following your suggestion we avoided the word cambial when not necessary, and we add a definition of cambial age.

Line 76: first first-difference?

We have removed the repeated word, thank you.

Figure 2h: The pattern of labeling these graphs was broken here.  Should be TRW, instead of spline.  More generally, I would at least experiment with giving each row its own label (e.g., ND for row 1), and each column its own label (e.g., dD for column 1).  The entire figure is TRW, so that could be a figure label.  All the x-axis labels are the same, so that could be a bottom label.  With that, the many titles of each subfigure could be eliminated, allowing the data parts to be expanded.  Just a thought.

We modified fig 2 as suggested, indeed it looks easier to follow. Thank you

Line 325: "reserves . . . may only be used for gapping stress."  Gapping stress?  I've not ever seen that term before.

It was a mistyping, sorry. We changed the sentence in: “be used for biotic or abiotic stresses”

Reviewer 2 Report

This study focuses on the relationship of TRW and the main isotopes used in forestry (except N) for larch and pine trees and finds periods where tree age might influence the relationship between them. The authors test different detrending methods and find that  after spline and first difference time series detrending the correlation values were not age related. Additionally, the study also analyzes the relationship among isotopes. All this analysis extends for hundreds of years of almost two hundred trees. The manuscript is well structured and written, despite the multiple combination of TRW, each isotope and two tree species. At first glance, I found it strange that the authors could pinpoint the year 100 as the turning point of the relation being significant or not,. but after reading it, I realized that it is because of the way they divided the group ages. I would suggest that the authors us the term 'approximately' or 'around 100 better.

One more issue concerning the terminology is the use of the term 'juvenile' isn't younger than 100 years old a bit too long of a period?

I would also suggest that you show the graphs following the order of when they are mention in the text to facilitate and easy flow for the reader.

Minor comments:

Line 20: extra space not needed.

Line 22: change the sentence to 'The correlation between d13C and dD was the exception, as it became significant only after age >100yr.

Line 27: erase 'we also provided also'

Line 28: correct 'trends'

 Line 70: 100 years sharp?

Line 87: Erase 'previous' all studies are prior to the authors's.

Line 89: extra space.

Line 111: Replace 'have been' by were.

Line 130: erase 'the' before each.

Results:

Line 160: replace (iiii) by (iv)

Line 157-162: No need to mention the figures here you do it rightly in lines 157-162 for figure 2 and Lines 192-205 for figure 3.

Fig 1 should be brought just right after Line 109.

Fig.2 The captions should be self-explanatory. Please mention what a,b, ....k, and l are as you did in Figure 3.

Line 337: ... leading to...

Line 399-344: This interpretation is interesting but I am not still convinced to which period of growth the actually refer when they mention the 'juvenile phase.

Author Response

This study focuses on the relationship of TRW and the main isotopes used in forestry (except N) for larch and pine trees and finds periods where tree age might influence the relationship between them. The authors test different detrending methods and find that  after spline and first difference time series detrending the correlation values were not age related. Additionally, the study also analyzes the relationship among isotopes. All this analysis extends for hundreds of years of almost two hundred trees. The manuscript is well structured and written, despite the multiple combination of TRW, each isotope and two tree species. At first glance, I found it strange that the authors could pinpoint the year 100 as the turning point of the relation being significant or not,. but after reading it, I realized that it is because of the way they divided the group ages. I would suggest that the authors us the term 'approximately' or 'around 100 better.

Thank you for the positive comments and the suggestions to improve the work. In fact 100 yr is just a reference point, therefore we added 'approximately' or 'around' as suggested.

One more issue concerning the terminology is the use of the term 'juvenile' isn't younger than 100 years old a bit too long of a period?

We agree that juvenile for a 100 yr period may sound odd. However, for pine trees that grow at a low altitude the juvenile phase for carbon isotope last about 50 yr (Young et al. 2011), and in other studies of isotope trends, it was stated that the 'juvenile' phase of the pines at the tree-line lasted longer. Our database has several trees that reach an age older than 700 cambial age, 100 can be regarded as young. In the introduction, we add a sentence to define the juvenile phase as lasting about 100 cambial years

I would also suggest that you show the graphs following the order of when they are mention in the text to facilitate and easy flow for the reader.

Thank you for the suggestion, but the graphs of fig 1 and fig 2 follow the same logical order, with the different detrending methods on rows and the different variables on the columns. Therefore, a modification of their order may not help visualization of the graphs.

Minor comments:

Line 20: extra space not needed.

Corrected, thank you

Line 22: change the sentence to 'The correlation between d13C and dD was the exception, as it became significant only after age >100yr.

Changed, thank you

Line 27: erase 'we also provided also'

Corrected, thank you

Line 28: correct 'trends'

Corrected, thank you

Line 70: 100 years sharp?

We added the word around before 100 years

Line 87: Erase 'previous' all studies are prior to the authors's.

Done, thank you

Line 89: extra space.

Done, thank you

Line 111: Replace 'have been' by were.

Done, thank you

Line 130: erase 'the' before each.

Done, thank you

Results:

Line 160: replace (iiii) by (iv)

Done, thank you

Line 157-162: No need to mention the figures here you do it rightly in lines 157-162 for figure 2 and Lines 192-205 for figure 3.

We have removed the mention to the figures, thank you.

Fig 1 should be brought just right after Line 109.

We agree, but for the submission, we have respected the format of MDPI that requests all figures to be placed between the results and the conclusion sections. In the editing phase, the position of the figures will be changed.

Fig.2 The captions should be self-explanatory. Please mention what a,b, ....k, and l are as you did in Figure 3.

We modified captions of fig 2 as suggested, indeed it looks easier to follow. Thank you

Line 337: ... leading to...

Done, thank you

Line 399-344: This interpretation is interesting but I am not still convinced to which period of growth the actually refer when they mention the 'juvenile phase.

We changed 'juvenile phase in: approximately first 100 cambial years