Dear Editor,
Below we respond to all of the reviewers comments individually. The main issue
was a misunderstanding. We have expanded the section on hydration of the
specimens because reviewer 3 missed it and wrongly assumed that the specimens were dry. Even after the many additions for the reviewers, we have shortened the paper by about 10% and shortened the explanation section of the discussion by 25%.
Robert Schofield
Responses to Reviewers' comments are starred:
Reviewer #2: This paper covers all the bases, often at very great length, and is
interesting, but really does go into too much, not speculation exactly, but expatiation. I wonder whether the authors have ever sat back and said 'Can we put this
more succinctly? I think they should try to do so
I have some other comments, often, but not always, minor, that should at least
be considered by the authors. The numbers refer to page (my page numbering;
the authors have committed the heinous crime of not numbering their pages!)
paragraph and line respectively.
General. It gets very tedious, particularly on pages 6, 7 and 8, to have the flow of
the text interrupted by the addresses of the manufacturers. I estimate there are
10 lines-worth of names, websites etc, on pages 6-8. If we have to have, e.g. the
website of the producer of the permanent marker, could they not be signified by a
small subscripted number and spelt out in the back?
* We have removed the marker manufacturer and other sources that we thought
weren’t specifically important. We have left some seemingly minor items because
we thought the source important, such as the very sharp Persona razor blades.
In general, though, we would have preferred to leave all sources, as we don’t always know which sources are essential: for example, when we recently tried to
repeat one of our earlier studies, we were surprised to find that the specimen
mounting did not hold with a different brand of 5 minute epoxy gel, but did hold
when we changed to the original brand. We will leave the question of footnotes to
the editor.
1 'Corresponding author. Put dialling code for USA (001?) on the telephone
number?
* Done
2 1 6 'For the first time.' For Bromine, or all halogenated cuticles'
* To our knowledge, this is the first time that fracture resistance, abrasion resistance, or dynamic mechanial properties have been measured on any biological material fortified with halogens or Zn, Mn, Cu or Fe. We are also not aware of
any other direct mechanical comparison of these materials to mineralized tissue.
2 1 9 'biomineralized' Surely if it is a 'tissue' it is by definition 'bio', and 'mineralized should be sufficient.
*”Biomineralized” was changed to mineralized throughout the paper when it was
used in conjuction with “tissue”, “cuticle”, etc..
3 1 11 an otiose 'bio' again.
*removed
3 1 14 'than' should be 'from' ??
*reworded
4 1 4 No need for period in middle of 'Cancer magister'
* removed
4 1 7 'un-calcified' should be 'uncalcified'
* changed
5 2 last. Again an otiose 'bio'
* removed
5 3 5 'halogens, chlorine.' Is this i.e. or e.g.?
*reworded
6 1 1 'Lined' Is this a method of catching the crabs? Not known to me, anyhow!
*reworded to make the common name more evident
6 1 penult. 'high-humidity' If it is high humidity it is presumably pure water. Being
kept in sea water is quite different. Does this matter?
* We have not looked for differences in mechanical properties between samples
kept in sea water vs. high humidity chambers. However, all samples were treated
similarly: the autotomized claws were kept in sea water until mounting. The
mounted specimens were then kept in high (100% at equilibrium) humidity
chambers until testing. See the expanded section on reducing changes in hydration.
6 2 1 'intercomparisons' Surely comparisons are by their nature 'inter'
* I agree, but I had a reviewer of a previous paper change a “comparison” to “intercomparison” because it was more precise…..changed to comparison here.
6 2 all. Just where one might hope for some detail (as opposed to the manufacturer of permanent markers) we are left dangling. What are the salmon teeth
made of: dentin, eameloid or enamel or what? Why this odd collection of hard
tissues?
* We have pointed out in the discussion that the hardness and modulus of elasticity sampled the enameloid cap, while the fracture test sampled the whole tooth.
The specimens were selected because they were calcified and uncalcified “tool”
materials in vertebrates. This justification has been added to the experimental
section.
6 4 1 (and elsewhere) 'Newtons' should be 'newtons'
*fixed
7 2 all. I'm unclear about this machine compliance business. It says near the end
of the paragraph that the machine compliance was 'smaller' than the energy
stored in the stiffest specimens. But this seems to imply that say up to 30-40% of
the total deflection could be caused by machine compliance. Although the curve
goes down to zero slowly, and therefore the net effect of the machine compliance
may be negligible in regard to work, it will certainly affect the shape of the loaddeformation curve, and one would like the authors to be a bit more up-front about
its effect.
*We did produce force-deformation curves that were corrected for machine compliance using the data for the rigid glass control to obtain the deflection of the
machine as a function of force. As the reviewer points out, the area under the
corrected curves was the same as the area under the uncorrected curves: the
corrected curves rise more rapidly (reach the peak force with less displacement
since the machine deflection is subtracted) but the extra area early on is balanced by the extra area under the uncorrected curve that reaches the peak force
at a greater displacement. For a specimen that breaks with a single fracture, the
uncorrected curve has a vertical drop-off, but the corrected curve tails off (both of
the curves reach the same final displacement because the deflection of the machine goes to zero as the force goes to zero). We chose to use the uncorrected
curves in the figure, because the tail in the corrected curve might give the misleading impression that there were multiple fractures for the calcified cuticle
specimen. A sentence has been added to the figure caption noting that the plots
are un-corrected.
The machine compliance can lead to an overestimate of the energy of fracture: if
the fracture happens quickly and its duration is small compared to the computer
sample rate or the mechanical response time of the load cell, then the area under
the tail in the corrected plots will be too large. Based on the area under these
tails, we expect that this overestimate could have reached up to 25% for the stiffest calcified cuticle samples, if they broke quickly in a single fracture. This is the
reason for desiring quasi-static fractures. Any resulting systematic bias in our results, estimated to be less than a few percent, would be in the direction of decreasing the difference between calcified and brominated cuticle. This estimated
bias is so small compared to the experimental error, that it was not included. A
sentence has been added to the discussion of compliance with this estimate of
potential bias.
7-8 general. I am unclear what kind of test the authors think they were doing. It
looks as if it is a 'work of fracture' experiment, with a specimen pre-notched to
ensure that the applied loads could be low, and fracture quasi-static. It certainly
wasn't for instance a K test (I am not saying it should be) but it would be good to
be quite clear about all this.
* We have added a parenthetical to the description pointing out that we use the
quasi-static work-area method with notched samples in tension.
8 5 3 'The modulus' The authors talk as if it is obvious what modulus is measured
by the Oliver-Pharr technique, but in fact it is not clear, particularly in what may
be a strongly anisotropic material like this. The Oliver-Pharr modulus is a reasonable kind of modulus to talk about, but it would be good to have this point
clarified.
* We do describe it as the O-P modulus in the data table, and we have added OP to the original description in the text.
9 3 end. 'We did monitor..' What were the results of the monitoring. That there
was no need to worry, or that some points were excluded? What? We should be
told.
* We have added a sentence noting that the few cases yielding different values
were caused by samples that were loose in the mounting cement. Thus even for
samples that are homogeneous, this cross-check is valuable.
12 2 2-3 It is unclear what is the relationship between the 'laminae' and the
'tubes' It says that the laminae in TEM are about 7 microns thick. Why do they
not appear in SEM? Are the tubes contained within the laminae, or what? Is the 5
micron diameter the internal or external diameter, or what? This description is not
clear.
* We have added a clause pointing out that the 7 nm laminae are visible in the
SEM image as thin sheets. The higher resolution TEM image was required to
measure the thickness of the laminae.The 5 micron value is approximate and not
particular to the ID or OD.
12 1 second half. This discussion of fracture anisotropy would be better placed
later, in the mechanical results and discussion.
* We would like to keep this discussion here, where it is likely to be near the figure that makes the similarities with wood grain obvious.
12 3 last 3. Is it significant that the Br is incorporated more quickly that Zn into
ant and scorpion cuticle? This sentence should either be expanded to show its
significance or deleted.
* We have added a clause that points out that the difference suggests a different
biochemical processes – chlorine, the halogen associated with zinc, is deposited
after ecdysis.
12 4 4 '±' Is this standard deviation, SE, 95% confidence interval, range, or what?
There may be a statement about this earlier, but I could not find it. Sentence in
brackets. Just say 'N = 7' ?
* both items were changed
13 2 all. Again, this sentence about anisotropy in fracture could be kept till later.
* We would like to leave this near the figure.
15 1 2-end. Surely the wordiness can be reduced. 'We found the calcified material as about 30% more resistant than the bromine-rich cuticle (P = 0.038, Wilcoxon two-sample test).' Something like that!
* done
15 2 all. Again, surely this could be reduced in length.
* done
15 6 top. The wording for the orientations must be exactly the same in the text
here and in the legend in table 1, which they aren't.
* removed extra
16 1. Again, why are we dealing with leaf-cutter ant mandible? The rationale for
using insect vs. crustaceans must be spelt out a bit more.
* We have added the explanation: there is no cuticle in the examined crabs (and
most crustaceans) that is not calcified or brominated.
17 1 Perhaps a reference to the physics involved in impact would be good here.
*We have added a general reference to vibrations and waves. The issue we are
addressing in this paragraph is: what test frequency would probe the dynamic
properties associated with an impact? If the impact was cyclical, the answer
would be the fundamental and possibly higher harmonics of the Fourier decomposition of the surface displacement cycle. If the cyclic indentation were approximately sinusoidal (as opposed to saw-toothed), the period of the fundamental
would be about four times the period from zero to maximum indentation. For a
single cycle of this displacement, the Fourier decomposition has a range of frequenies instead of discrete frequencies, but the frequency with the largest amplitude is still the same as in the cyclical case. We have modified the text slightly to
improve the explanation.
19 2 What is a particle 'size' ?
* changed to diameter
Figure 3 legend. Line 2. 'transparent' should be 'translucent' (as elsewhere)?
* changed
Figure 5 legend . I could not see how this was meant to work. The notch seemed
to be on the compressive side of a very long piece of calcified cuticle. It is not
clear how a crack would spread from the tip of such a notch. If the cuticle has a
modulus of 35 GPa, how can it be bent into the arc of a circle like this? A baffling
figure.
* Explanation was added to the legend: the specimen is not stressed, the curve is
the natural shape. There should be little compression in this tension test.
Figure 7 legend. I so not see the 'chitin-free non-calcified layer' on the lower lefthand corner of the image,
* arrow added to figure
Figre 11 legend. Work per unit area does indeed have the units of N/m, but this is
terribly non-intuitive. Why not keep it at J/m2
* N/m were used to show that it is the area under the N vs. m curve – units have
been changed at reviewers request.
Table 1 There is lots of wasted space here, which could be eliminated to make
the numbers etc. visible. The grid lines are too thick; what is bold 'M' ?(No doubt
we are told somewhere, but it should be mentioned in the legend.)
* We think that this must be re-formatted by the journal itself.
Table 2 legend. How can probabilities be less than equal to something that has 4
significant figures? Why not the same number of significant figures for low and
high frequencies? Why not have the values at the different frequencies shown
the same way (1.3e-4; 1.8e-7 or whatever)?
*done
Reviewer #3: At first glance, the manuscript by the Schofield group is a wideranging analysis of the bromine enriched "spoons" of crab chelae. As such, it is
the first study to systematically tackle the intriguing question regarding the advantages of bromination in invertebrate scleroproteins. Although the text is articulate, it is extremely prolix. Considering the single-spacing of the present edition,
something pared down to half the length would be nice. The tone of the writing is
somewhat annoying; does it have to be so peevish? The reader gets the impression that this group thinks it walks on water, when in fact the manuscript shows
several foibles and misconceptions.
Specifics:
1. The authors are too vague about sample hydration during mechanical testing.
Surely they realize that crab carapace functions exclusively under a fully hydrating environment. Of the four mechanical tests described (abrasion resistance,
fracture energy, AFM indentation and dynamic testing), only fracture energy tests
were described as performed at "high humidity" (what % humidity?). I assumed
the others had been done on dry specimens, because nothing was mentioned
about wet testing. If this is true, what biological relevance do the dry measurements have? Although a few tissues (e.g. enamel) have mechanically similar behaviors in wet and dry state, it would be an outrageously risky generalization.
No mechanical tests were made on dry samples, as mentioned in the first paragraph of the experimental section. Nevertheless, we have expanded this section
with more details to avoid confusion. Autotomized claws were kept in sea water
(similar to the laboratory environment and to the natural environment for a minor
portion of the day) and prepared specimens were kept in a high-humidity chamber (100% at equilibrium). As one of the highest intertidal organisms in the environment in which we collected them, P. crassipes is out of sea water for the majority of the day, but hides under rocks in a humid environment thought to be
similar to the high-humidity chamber. Most specimens were tested within an hour
or two of autotomizing, and no specimens were tested more than 8 hours after
autotomizing. The specimens were removed from the high-humidity chamber for
the duration of the test (typically a few to 45 minutes) and tested in laboratory
humidy, ranging from 45% to 70% RH.
To reduce drying rates during testing, the cut surfaces of AFM specimens were
coated with epoxy and the abrasion resistance specimens were wetted during
testing.
To address the question of whether the mechanical properties of arthropod cuticuticular “tools” similar to the crab “tools” change rapidly after death or during
exposures to unsaturated air during the testing period, some time ago we made
AFM measurements on the whole sting of a live anesthetized scorpion (one of
the easiest cuticular “tools” to test live). After the first measurements, the scorpion was cut from the mounted sting (exposing the interior of the sting cuticle to
air) and we made a time series of measurements. Two types of cuticle were tested, including cuticle very similar to the ant cuticle and probably to the brominated
cuticle of this paper; the greatest changes are reported here. Three hours after
dissection, during which the sting remained in the test apparatus, the modulus of
elasticity differed from the live test by 20% (not quite a significant difference for
our test: p = 0.07), while the hardness differed by 7% (not significant: p=0.43).
One week after dissection, using more data points, the modulus of elasticity differed by 15% (p = 0.03), but the hardness did not differ significantly. There are
many similarities between the scorpion cuticle and the crab cuticle (e.g. mechanical properties similar to Br region, protein matrix, similar ultrastructure, similar
interior) that suggest that even if we had not tried to maintain the natural hydration state of the crab specimens, the results would not have differed dramatically.
2. The authors argue that Br is uniformly speciated throughout the spoon struc-
ture. But it is clearly not uniformly DISTRIBUTED (Fig. Inset). Highest levels are
in the core, and the significance of this, vis-à-vis mechanical properties, should
be addressed.
*We have no evidence that the Br concentration is not uniform. This is a whole
specimen with varying thickness, not a slice of uniform thickness. If such a specimen has a uniform Br concentration, we expect the projected Br density (areal
density) to be lower near the edges of the specimen where the volume sampled
by the beam is lesser, and we expect the areal density of Br to be greater near
the center where the sample is thicker, just as the figure shows. Although we
qualitatively see no deviations from what we would expect for a uniform concentration of Br, we do not know whether or not this is true quantitatively. For this
reason we can not address this question. We have added an explanation to the
text.
3. WGA is specific for N-acetylglucosamine not chitin. Ovalbumin, the usual positive control for WGA, has very little in common with chitin. If authors want to be
confident about chitin presence, they should confirm it with other tests (wide angle X-ray, glucosamine in amino acid analysis, etc).
*Confirmed with X-ray diffraction and added to the text