|The manuscript by Felgistsch et al. presents evidence regarding the ice-nucleating macromolecules (INM) in samples taken from several portions of birch trees in Austria. The revised manuscript improves clarity in a number of areas. My present concern is that the link between the evidence about ice nucleating temperature and particle number is almost totally decoupled from the spectroscopic evidence used to investigate the chemical nature of the samples. I think the data presented about various samples from the trees are interesting, and the authors have improved the manuscript by including at least hypothetical scenarios by which macromolecules within the tree wood could possibly contribute to soil or atmospheric ice nucleation. |
Line 17 of page 1 (abstract) and line 32 of page 11 (conclusions) state similarly that “the majority of the samples showing freezing temperatures close to those of birch pollen extracts, indicating a relationship between the INM of wood, leaves and pollen.” I have two concerns with this statement. First, the meaning of “close to” needs to be clearly defined in both of these instances. At least four of the leaf samples and several primary wood samples are closer to ultrapure water than to the pollen wash water, so the text should clarify how the reader should interpret “close.”
Second, without any presented statistics, it is hard to know how well to trust the second part of the statement, which uses the freezing temperature (i.e. Fig. 2/top) as support that INM from wood, leaves, and pollen are similar. This may or may not be true, but the evidence from Figure 2 does not seem to strongly and systematically support that statement. The authors also discuss this same point at the bottom of page 8 (line 41) where after discussing Figure 2 they state that “Based on these results, we hypothesize that the INM in birch trees are quite similar in pollen, leaves, primary wood, and secondary wood.” In contrast to their conclusions from the figure, I see a trend where the leaves are generally lower in freezing temperature than the other tree samples, the primary wood is frequently next coldest, and the secondary wood is slightly warmer still. This interpretation of the figure could be used to support a contrasting conclusion that INM from different fragments are *different*. Based on this important possible discrepancy I think the evidence is not presented to support the “indicating a relationship …” portion of the statement. If the statistical meaning of the terms are presented and supported, it could then be more defensible.
The authors also show IR and fluorescence spectra from fragments of tree TBA, but no data is shown for any other trees. If the authors wish to use the spectroscopic data to support any general conclusions, I do not see how this can be possible without systematically showing the data for all trees and for all sample types. Then the authors can attempt to generalize the results in some way. For example on page 10 (line 15) the authors state “The measured FTIR spectra indicate that the birch extracts are chemically similar to each other, and to pure birch wood.” Again the definition of “similar” is important here, in part because the majority of the molecular composition may be similar, but it is hard to know how this may or may not relate to the INM. One would guess that the INM are likely a small minority of the overall molecular composition, and so I see very little link between the ice nucleating data and the spectroscopic data. Both are interesting independently, but trying to tie them together is more problematic in my opinion.
I think the manuscript is an interesting contribution to the overall scientific literature, but I suggest the scope of the discussion and conclusion statements be given more quantitative support before publication.