More on Almagre Tree 31

I’ve done an interesting exercise today in which I’ve tried to do an estimated cross-section of Tree 31 – the one with the highly discrepant core widths. Pete has some new pictures and site information that I haven’t seen, so it will be interesting to see how they reconcile.

As noted before, Almagre Tree 31 has a large difference in growth rates between two cores only 45 degrees apart -see here . Here’s the measurement plot, showing a dramatic difference between a fast-growing radius and a slow-growing radius.

I’ve done a rough cross-section of Tree 31 to translate the information from the core into a cross-section. Mining engineers and geologists always convert drill core information into cross-sections and level plans. I realize that professionals in the dendroclimatology don’t need to do this – of course, traditionally they didn’t make location maps either, waiting apparently for the invention of GPS rather thanlow-brow methods like plotting sample locations on Forest Service topo maps, as a geologist might do. So at the risk of offending a dendroclimatologist by doing something as amateur as estimating a cross-section from available core, I’ve done the exercise anyway, articulating the assumptions as I go.

Up to the first half of the 19th century, the two cores were roughly in tandem. So I’ve hypothesized that the tree was roughly circular in 1835 with a radius of about 13 cm. Between 1875 and 2007, the slow-growing radius (31B – SW) has grown about 2.9 cm (the strip bark portion obviously growing at a rate of 0), while the fast-growing radius (31A – W) has grown about 10.0 cm. We only have two cores and, for the sake of argument, I’ve interpreted the fast-growing core as being more or less in the center of the bark strip. As an experiment, here’s a contour map showing a contour map of ring widths – you can see the consistent low growth leadgin up to 1875, with the lensy growth in the strip bark in the late 19th and 20th century. (This diagram illustrates some features, but the actual tree cross section is more rectangular – Pete’s sent me a sketch. It will take a little time to re-do this diagram to incorporate a more accurate cross-section and I’ll update this when I get a chance.)

So the tree is becoming more elliptical. In old strip bark trees, there’s plenty of evidence of the trees projecting along the axis of the ellipse so that they become increasingly elliptical and ultimately almost rectangular. That part’s been noticed. But I haven’t seen any specific discussion in the literature of the relative thickening in what seem to be the “early” strip bark stages. In this case, it’s almost like the strip bark tree is “trying” to achieve a circular arc, intersecting at the strip bark boundaries. That is the principle of this diagram – I hope that we’ve got enough information to reconstruct a cross-section as it doesn’t look like another examination will be possible before spring.

In Craig Brunstein’s interesting technical report on bristlecone morphology (online – see his Figure 4 for a wide variety of growth forms), his Figure 6 shown below illustrates a somewhat similar process of ellipticization for strip-barked limbs. This isn’t a trunk, but it shows rather nicely an ellipticization process.


Again, the distinction that I’m making here is this: the strip bark growth is not radially uniform, but appears to be much thicker in the center of the strip bark – at least during the first couple of centuries. In the case of Tree 31, it appears that the strip bark formed around 1835 plus-minus 15 years. So how does one obtain a homogeneous measurement for this sort of tree? If you have massive radial asymmetry as here, a simple average may not be very accurate. In addition, in examining the information from strip bark sites, it is remarkable how few trees have more than one core. My impression is that the available cores seem to come from the center of the strips and thus there is potentially a powerful bias here, about which I’ll have more to say. If bark stripping occurred randomly over the history of the site, then the problem might not be too severe; however, as I’ll discuss in my next post on this topic, we are getting the impression that strip bark occurrence is not random and intimately related to 19th century events.


  1. Steve McIntyre
    Posted Oct 23, 2007 at 1:37 PM | Permalink

    #3. Gary, the issue isn’t that it’s unreasonable to drill in the center of the strip. The question is: if you’ve got big radial asymmetry, how do you reduce the asymmetry to an index for that tree – assuming that you agree that the cores are not teleconnecting independently to the universe Mannian-style: do you try to adjust the measurements to a whole-bark basis somehow? IF people are going to use strip bark trees – against the recommendations of the NAS Panel – and the Team is addicted to strip bark trees, is there any rational way of adjusting for bark stripping? And we all know how the merry adjusters on the Team like to adjust.

    Thinking about it a little more, here’s a thought: let’s suppose – and these are just directional assumptions for now – (1) that the tree makes the same area of radial wood before and after stripping; (2) that half of Tree 31 was stripped in the early 19th century, previously whole bark; (3) the post-strip wood grows in a lensy way with the fast-growing radius at the centre of the lens. Then you’d have to divide the fast-growing radius by 4. Let’s suppose that this simple adjustment was applied to Tree 31 for all values after 1876 – when the second growth spurt commences. Then we’d get a graphic that looked like the one below (pink is core 31B, the slow-growing core unadjusted here). The raw fast radius is shown in grey; then the adjusted fast radius is shown dashed.

    But how would one go about estimating these adjustments? And what if stripping occurred in the 17th century as well? Does this graphic evidence that as well? IF one is even going to begin adjusting for strip bark, then you’d need measurements of the percentage of the circumference which is barked. Maybe it would be a good idea for the Team to get unaddicted to strip bark trees.

  2. Steve Sadlov
    Posted Oct 23, 2007 at 1:00 PM | Permalink

    This is personal poetic justice. I got a C+ in my field mapping course as an undergrad. What a potent lesson it was for me.

  3. Steve Sadlov
    Posted Oct 23, 2007 at 1:03 PM | Permalink

    Instead of ring width, you’re probably looking at some sort of scaled % of total radius at a given core. This is a deadly flaw in dendroclimo. If you say “the average ring width for a set of trees is X” that is a meaningless figure. Etc. You are onto something here Steve.

  4. Gary
    Posted Oct 23, 2007 at 1:06 PM | Permalink

    It seems reasonable to presume that most trees would have been cored near the center of the strip with the direction of the corer pointed toward the most likely location of the pith. This would yield the longest core, hopefully with the most rings. If the strip is wide enough the central sweet spot should be broad enough to allow for some error in this guesswork and still get the desired core. Problems would arise with narrow strips that have a very narrow sweet spot. With those the corer is more likely to come in at less than 90 degrees to the ring – something probably to be avoided. Unless there was prior knowledge of a better strategy, I think most people would follow this thinking. To really answer the question of what is the right radius might require the study of multiple cross-sections.

  5. Larry
    Posted Oct 23, 2007 at 1:31 PM | Permalink

    Define “sweet spot”.

  6. Dave Eaton
    Posted Oct 23, 2007 at 1:35 PM | Permalink

    Is it possible that given the two cores, you could derive principle axes or ‘eigenvectors’ for the tree and normalize that way?

    I’m absolutely ignorant of tree biology, but I wonder: since you have bark and stripped sections, is there anything in the physiology of trees that would make a comparison of these cores yield information about the history of the trees? My vague thought is that perhaps when one was uncertain as to whether a rapid growth were due to temperature or rainfall, a differential in the response of barked and stripped sections could resolve this.

  7. Gary
    Posted Oct 23, 2007 at 2:13 PM | Permalink

    #4, Larry – sorry, the “sweet spot” would be a point on the strip with a straight line path to the pith and the rings would have minimal arc over a maximal width of the strip. In Steve’s diagram at top, the line with 10.6 and 13 length measurements hits the spot on the outer arc.

    #6, Steve – Just taking it one step at a time. My presumption is only that most people probably will go for the center of the strip unless there is some compelling reason not to. I was setting the stage for how to approach the asymmetry issue so as avoid side-tracking on the geometry of arcs. I’ll work through your thought experiment this evening.

  8. jae
    Posted Oct 23, 2007 at 2:17 PM | Permalink

    we are getting the impression that strip bark occurrence is not random and intimately related to 19th century events.

    ?? Sheep?

  9. James Smyth
    Posted Oct 23, 2007 at 2:26 PM | Permalink

    It seems to me that another important question is the correlation coeffecients. There seems to be a lot of uncorrelated activity in the 1700 time-frame.

  10. Posted Oct 23, 2007 at 3:41 PM | Permalink

    Steve, your location map

    mentions “Sheep Mts” just to the left of your yellow flag markers.

  11. Posted Oct 23, 2007 at 3:06 PM | Permalink

    we are getting the impression that strip bark occurrence is not random and intimately related to 19th century events.

    Do sheep eat bark? Is it a tree disease like the current chestnut bleeding?

  12. Steve McIntyre
    Posted Oct 23, 2007 at 3:19 PM | Permalink

    Pete H has a different mechanism in mind which I’ll post on. I hadn’t thought about sheep for a while, but did mention 19th century sheep over-grazing in MM(EE 2005) as follows. Cripple Creek is near Mt Almagre; indeed the most direct road from CO Springs is the old Cripple Creek road. If sheep came with the gold rush, then there’s a plausible basis for late 19th century sheep overgrazing in the Mt Almagre – I have no evidence that this took place. One would have to try to figure out some way of testing. Maybe sediment cores in the small plateau lakes/reservoirs would yield info. Here’s what we said in MM05 (EE):

    pulse in woody plant growth throughout the American Southwest, attributed to
    overgrazing by sheep in the late 19th Century (see Figure 8), which in turn followed
    the extension of the railroads [Allen, 1998; Allen et al., 1998]. Sheep differ from other
    species in that they will completely destroy grasslands by eating down to the roots,
    leaving barrens [Allen, 1998]. Although Allen [1998] only documented the expansion
    of pinyon pines and junipers into terrain formerly occupied by 19th century grasslands,
    Allen (2004, pers. comm.) did not exclude the possibility of a similar effect involved
    in anomalous 20th century growth for bristlecone pines, but was unaware of any
    studies on the topic. There is a published reference to the introduction of large
    commercial sheep flocks in the late 19th century in the White Mountains CA [St. Andre
    et al. 1967], where the key sites of Sheep Mountain and Campito Mountain are
    located. The founder of the Sierra Club, John Muir, complained of the depredations of
    sheep in the Sierra Nevadas (adjacent to the White Mountains) as “hoofed locusts”
    [Muir, 1911]. Carl Purpus, a late 19th century botanical collector in the Sierra Nevadas,
    stated in 1896 that commercial flocks had cleaned out all grass to the top of Old Mt
    Whitney [present-day Mount Langley, which reaches 4,280 m] [Ertter, 1988]. Allen
    (pers. comm., 2004) said that there was a large commercial sheep trail at Jicarita Peak
    NM, another bristlecone pine site studied by LaMarche and Stockton [1974]. In severe
    high-altitude terrain, even after the departure of commercial flocks, a small population
    of bighorn sheep could prevent the re-establishment of grass (Leslie Thomas,
    Colorado Springs, landscape architect, pers. comm.) Since grass (and other herbs)
    compete with pines for scarce moisture, one can hardly exclude, on a priori basis, the
    possibility of a connection between anomalous 20th century growth rates of bristlecone
    pines and a growth release following 19th century overgrazing, as experienced
    elsewhere in the American Southwest.

  13. jae
    Posted Oct 23, 2007 at 4:30 PM | Permalink

    11: I just wondered, without thinking much, if sheep might “graze” on bristlecone pine bark. I know elk can play havoc with certain trees. I guess that’s really a stretch, though, because that would probably just cause the loss of bark in patches, not strips. Lightening does seem a more likely explanation, since it is often very intense up there. I’ve been frightened by St. Elmo’s fire several times in the high Rockies.

  14. Follow the Money
    Posted Oct 23, 2007 at 5:34 PM | Permalink

    Here’s some text from a bcp site that may be of direct interest here

    Another strategy for surviving is the gradual dieback of bark and the tissue that conducts water (xylem) when the tree is damaged because of fire, lightning, drought or damaging storms. This reduction of tissue that the crown has to supply with nutrients, balances the effect of any damage sustained. The surviving parts remain quite healthy. As an example, “Pine Alpha” at over 4000 years, is nearly four feet in diameter, yet has only a ten inch strip of living bark to support it.

  15. novoburgo
    Posted Oct 23, 2007 at 8:07 PM | Permalink

    Has anyone considered the benefits of sheep grazing nearby. I’m talking sheepdip here! Something you can purchase at your local Agway and your plants will love you for it. One sheep visit per year could go a long way toward making for a very good growth year. Consideration should also be given to the close proximity releases of CO2…enough to make your bristlecone bristle.

  16. Steve McIntyre
    Posted Oct 23, 2007 at 10:03 PM | Permalink

    As an experiment, here’s a contour map showing a contour map of ring widths – you can see the consistent low growth leadgin up to 1875, with the lensy growth in the strip bark in the late 19th and 20th century. The fast-growing Core 31A (from the West) is shown here as horizontal while slow recent growing 31B is shown from the SW. My impression is that Core 31B is too oblique in this rendering – I’ll experiment some more.

    Contour map using akima for Tree 31 ring widths. This corresponds to the bottom half of the diagram – core 31A is horizontal and core 31B is at 45 degrees.

  17. MrPete
    Posted Oct 24, 2007 at 5:56 AM | Permalink

    Steve, is it possible to map the contour map to the shape of the cross section?

  18. Stu Miller
    Posted Oct 24, 2007 at 9:02 AM | Permalink


    I, too, would like to see the contours on the cross section. Failing that, I need a bit of help with your map. What are the horizontal and vertical axes? The horizontal does not seem to be labeled, and the vertical label is cut off and unreadable. If I interpret the horizontal axis as some sort of time scale and the top of the graph as the 31a core, that kind of makes sense. Then is the left side of the triangle the 31b core with a different time scale? What is the curved dashed line thru the middle of the plot?

    Sorry to be a bit thick, but this form of contour plot is new to me.

  19. MrPete
    Posted Oct 24, 2007 at 10:38 AM | Permalink

    Here’s a (very rough) combined graphic:

  20. Steve McIntyre
    Posted Oct 24, 2007 at 9:57 AM | Permalink

    The contour map is on the same coordinates as the diagram, but only the bottom half of the diagram is shown (As there are only two cores – the top half of the digram is just a symmetric reflection.

    The cross section is more or less in cm, but it’s really just a sketch – don’t place more weight on it on that. I’m not entirely comfortable with all aspects of the sketch.

    As to what’s being contoured here – the hot spots are areas of rapid growth; cold spots are narrow growth. This sort of contour map is common in mineral exploration for ore grades. Yeah, I’ve never seen something like this applied to tree cross sections either, I’m just experimenting.

    What does the contour map show to me? well it shows the lens of high 20th century strip bark growth rather nicely to me, tho I am used to reading these maps. The peak in the fast core was in the early part of the century; it also shows that the slow core did not experience a commensurate peak.

  21. Bill
    Posted Oct 24, 2007 at 11:09 AM | Permalink

    “…Do you try to adjust the measurements to a whole-bark basis somehow?”

    James Erlandson posted a good article on Oct 21 in blog entry “8 Measured Trees…”, Partial Cambial mortality in high-elevation Pinus aristata, by A. Schauer, primary author. If I may be allowed to repost …?

    This might at least provide an approach to gathering high altitude cores, if no answer to the question of correcion.

    tree size and aspect of cambial death data were gathered from three Pinus aristata forests in central Colorado, USA. Stripping frequency tended to be higher for larger diameter classes. Partial cambial mortality exhibits significant directionality within each stand. Furthermore, cambial death was measured to be most frequent on the wind-exposed side of stripped trees in two of the three study sites and appeared to be at the third. Data presented here support the hypothesis that wind plays a role in the occurrence of partial cambial mortality in Pinus aristata. The mechanisms by which wind causes cambial mortality remain unclear

    The study went on to say that the density of the grove, its elevation, and the tree’s size bore directly on the stripping of bark. 30% and 33% of the trees at the study’s two high-altitude sites were stripped.

    The occurrence of exposed wood increases with tree size. No individuals over 80 cm dbh were totally covered in bark.

    If, as this suggests, the bark stripping (hence the ellipticism) commences at a given diameter of the tree’s life, there may be an optimum (maximum) age for trees at that altitude, after which their rings tend toward ellipticism, and should be eschewed from chronologies. From what’s been shown at this site, it would seem very difficult to correct for the kinds of environmental factors which must come into play.

    BTW: I’ve always figured that flagging, or krummholz formations of evergreens at and above timberline were the result of wind, even indicative of wind-direction. It’s an easy conclusion to leap to, though the mechanism may not be proven: wind-born snow lodges on the windward side of the sapling, cuts its sunlight, encouraging green growth in the opposite direction. Survivors’ larger trunk sizes force an exaggeration of the same process the following season. It seems to me the same process is at work with bristlecones.

    The result are beautful, natural bonzais, uniquely individual trees, but trees that would defy dendrochronologists’ efforts to “normalize” (if that is the correct term).

    Sorry for the lengthy entry. I’ll watch it in future.

  22. Steve McIntyre
    Posted Oct 24, 2007 at 10:49 AM | Permalink

    #20. Hey, Pete. you’re good at this stuff. BTW in terms of the coarse geometry of the tree, it seems unlikely that 31B would be so oblique to the surface, so the outer curvature in this area may be wrong. Also I don’t think that I’ve reconciled the info on tree diameter yet.

  23. MrPete
    Posted Oct 24, 2007 at 10:55 AM | Permalink

    Well, I did help architect the first PC GIS system; correlating multilayer maps comes a bit naturally.

  24. Dave Dardinger
    Posted Oct 24, 2007 at 11:00 AM | Permalink

    re: #20 Thanks MrPete!

    That makes it all clear instantly. In essence the tree slowed down growth drastically after it was damaged, but eventually the remaining strip of bark began to grow and even took-off for a while before settling down in recent years. One assumes that it’s now limited by resources: temperature/precipition, et. al. not by the infrastructure of the tree itself.

    BTW, this shows that while one picture is worth a thousand words, not all pictures are equal, no offence to Steve. Coming from a Chemistry background and used to seeing phase diagrams on triangular graph paper I, like others, was scratching my head trying to figure out what all the sides of the triangle stood for.

  25. Steve McIntyre
    Posted Oct 24, 2007 at 11:12 AM | Permalink

    #23. It really is an interesting synthesis diagram. It needs a little more work on the geometry, but it gets in a lot of elements, as you observe. Good diagrams take work. And it would have taken me a long time to figure out how to coerce the info from the akima diagram to the shape as Pete has done here. Hats off to Pete. (He’s also handy with a 4-wheeler over boulders.)

  26. Steve McIntyre
    Posted Oct 24, 2007 at 11:15 AM | Permalink

    #24. If one is going to use bristlecones or foxtails, then I think that the chronologies will need to be restricted to trees within well defined age limits which can be shown not to have been stripped. The trouble is that the Team is addicted to chronologies with strip bark so even obvious precautions meet with the fierce resistance of any addict.

  27. MrPete
    Posted Oct 24, 2007 at 11:41 AM | Permalink

    (The ‘warp’ tool in PhotoShop is your friend 😉 ).

    Back to my reports. Got a few photos prepped for Steve. More to go.

    Question for all: given a set of photos of a tree, from every (horizontal) direction, can you think of a way to accurately convert that to a “floor plan” showing a virtual slice? I can see how the bark/no-bark elements would show up, but converting to accurate trunk shape seems difficult at best.

    I’m wanting to bring big sheets of paper up there, and scissors, to cut out the trunk shape. Or, some kind of Very Cool circular laser rangefinder.

    But maybe there’s a better way, using the photos I’ve got?

  28. Dave Dardinger
    Posted Oct 24, 2007 at 12:16 PM | Permalink

    re: 27

    I’d talk to a physician, in particular a radiologist. The modern x-ray tools scan beams through a person and then rotate it around and then the computers can create any cut the doctor wants. A fullblown version of the software might be too expensive, but they’d likely be able to connect you to the proper place to go to find a cheap or free version with low resolution.

    Hmmm. For that matter, would it be possible to use one of those x-ray machines to produce an actual scan of a living tree and give you a useable cross-section directly? Would probably be too expensive for everyday use, but for special studies it might be better than finding a recently dead old tree and slicing it up for analysis. Though maybe you could do both and take an entire smallish tree to a lab and get the entire scan done at once.

    Actually I expect something like this has been done and is in the literature, but I’ve posted the idea here first, so someone else can do the literature search and let us know what was found. Then Steve can contact the authors and demand the entire 200 gigabytes of data to look at. 8)

  29. Posted Oct 24, 2007 at 1:03 PM | Permalink

    #27 MrPete:
    I think this is what you’re looking for:

    Take photographs. Get 3D models. $995.

  30. Yancey Ward
    Posted Oct 24, 2007 at 1:44 PM | Permalink

    Steve #26 and MrPete #27,

    A combination of these ideas would be a great research topic for the intrepid climatology student.

    If one is going to use tree rings for paleoclimate studies, a minimum standard should be that you use trees of similar ages for each year- a lot of trees need to be sampled, this I don’t doubt, but these adjustment techniques being used in combination with the problems now identified with growing asymmetries in older trees lends itself to the propogation errors- both accidental and otherwise.

    Also, it would be really nice if a good, non-destructive technique could be found to assess the total area of a tree ring rather than relying on just the width in one or more cores. Of course, the more cores you have, the closer you can actually approximate the area of a given ring.

  31. Posted Oct 24, 2007 at 2:14 PM | Permalink

    ground based laser scanning for the outer shape

    for the inside you could use mri scanning unfortunately that would mean “carying heavy equipment” 😀

  32. MrPete
    Posted Oct 24, 2007 at 2:17 PM | Permalink

    Thanks, JohnV! Very interesting. I’ve sent an inquiry.

  33. MrPete
    Posted Oct 24, 2007 at 2:30 PM | Permalink

    This also looks interesting: do-it-yourself 3D scanning using very simple tools. I’m thinking an improved version might be developed making use of an ordinary low-cost laser level!

  34. Larry
    Posted Oct 24, 2007 at 3:07 PM | Permalink

    IIRC, you can do that with a CAD rendering package. If you have access to a CAD operator, that may be cheaper than getting the photo tools.

  35. Spence_UK
    Posted Oct 24, 2007 at 3:09 PM | Permalink

    The art of reconstructing 3D models from 2D images is called “photogrammetry”.

    There are freebie tools around, a colleague of mine has some, not sure where he got them from though. Wikipedia has info (predictably):

  36. Geoff Sherrington
    Posted Oct 24, 2007 at 6:41 PM | Permalink

    I think in an earlier post I did not make a point strongly enough when talking about cleft grafting. Here, the whole tree is cut off, usually near the base, so that there is an abrupt cessation of all cambium material. Then a small piece of another plant of interest (a scion) is placed on the exposed cambium ring. Thus, the % of cambium from which new growth starts is only a 1% or 2% or so of the whole circumference. So I’m not surprised that trees grow with 30% of the bark removed. They can grow with 99% of the bark “removed”. The botanical character of new growth above the cut is dictated by the nature of the scion, not by the nature of the stock plant below. Since new growth starts from the circumference and can end up looking circular after some years and co-radial with the stock underneath, I would presume that the ring pattern would have its “centre” close to the edge and maximum ring widths from that edge to the direction opposite. Over the years, the higher parts of the new growth would be expected to revert to vinyl disc appearance, but not near the cut. The rate of radial growth of the graft is rather faster in the first few years than subsequent, indicating that the rings would not correlate with rainfall or temperature, being masked by larger factors related to survival respose.

    The strip bark trees are an intermediate expression of this extreme demonstration, the other extreme being healthy trees with no history of gross bark/cambium disturbance.

  37. Stu Miller
    Posted Oct 24, 2007 at 9:14 PM | Permalink

    Worrying about the response of tree rings to tree damage is all well and good, but I am hung up in a more fundamental concern. I am not a climatologist, so if I were going to try to deduce past temperatures from tree ring measurements, I would want to have some sense of the relative importance of water availability vs temperature on tree growth. As I see it right now, ring width is a function of at least growing season temperature and water availability. I suspect that the Almagre trees are pretty much limited to the winter snow pack for water.

    This leads to a questions for MrPete or Steve. First,is there by any chance a relatively close location where snow pack is regularly measured that would provide an indication of water availability? Second, if there is, how much does snow pack vary and how does it correlate with ring width?

    Obviously, this sort of study should have been done, and may have actually been done. If anyone knows of this sort of study having been done, I would be interested in a link or reference.

  38. MrPete
    Posted Oct 25, 2007 at 12:26 AM | Permalink

    Others have already pointed to important papers that show, not only is water of utmost importance, trees do essentially all of their growing in the early spring, in response to available water.

    Back to tree damage for a moment. The following photo (new tree #062) might be a bit interesting. It is a very nice example of what happens when heavy snows weigh down a thick needle-filled branch. The branch breaks, pulling a strip of bark with it. Clearly, this branch broke in the heavy snows earlier this year — the branch is still alive!

    As always, click to get to the gallery, and click on the magnifying glass to see the whole photo. In this case, a high resolution composite.

  39. Jonde
    Posted Oct 25, 2007 at 12:46 AM | Permalink

    Depth of snow has indirect influence on soil water availability.

    Case 1. Soil freeze before snow cover is thick enough to insulate the soil.
    In this case the roots cannot utilize the available water at spring because the soil is frozen and root activity cannot start in frozen soil. Thus, most of the water is lost as run-off, especially in steep ground, because soil is frozen and cannot storage any significant amounts of snow melt water. Here the frozen soil water that melts in the spring plays a major part during the early stages of growth, not snow cover melt water.

    Case 2. Soil do not freeze before snow cover is deep enough to insulate the soil.
    In this case, the upper soil layer temp rises very quickly after the snow cover has melted and melt water tends to stay in the soil for tree roots to absorb at the early spring. Thus, snow melt water plays a major role in spring soil water availability.

    Usually shallow snow cover melts quickly in the spring and thus do not have other negative influences on tree growth. But, if the snow cover is very thick and soil is frozen, then needles transpire the water that is stored in the tree trunk at previous autumn before roots can absorb new water. Thus, drought damage is imminent and growth of that year suffers.

  40. Gary
    Posted Oct 25, 2007 at 7:53 AM | Permalink

    #38 – MrPete
    That’s quite a tree with what seems to be less than 10% still-living tissue. Judging from its height, the loss of most of the bark must have been fairly recent.

    Nobody has brought it up yet, but the literature suggests that BCPs have “sectored architecture” meaning that shoot (trunk cambium and leaves) depend on the roots directly connected to them and that there is little if any lateral transfer of water and photoassimilate. If a root dies then the cambium above it also will expire. There also is some evidence that wind damage can kill the bark and presumably the roots below then also die. From your observations can you add anything to these findings?

  41. MrPete
    Posted Oct 25, 2007 at 9:15 AM | Permalink

    Gary #40,

    I can comment on part of your question. Take a close look at the overall tree in #38. Perhaps even click through to the gallery.

    Here are some observations I find interesting:

    * Notice how the tree loses diameter in “steps” as you look from bottom to top? You won’t find that in full-bark BCP’s. Look closely and you can see another branch split about halfway up. I’ll see if I can come up with a photo from another angle that shows it more clearly. And again, there’s another closer to the top.

    * From this and many other trees now in our photo collection (I have a *lot* more to assemble once I’m through some pressing work issues), it is clear that loss of limbs is a major factor in the lives of older BCP’s.

    * I can’t say for sure whether storm damage (snow-weight, wind, and falling trees) or lightning damage is more significant. But I suspect they both are, when it comes to these major impacts on tree growth. One set of guesses: if it looks like a clean tear on one side of the tree, that’s most likely storm-related damage. If it looks like a split from the top, often in a spiral, with the wood still there and not torn down, then that’s more likely a lightning hit.

  42. Earle Williams
    Posted Oct 25, 2007 at 11:37 AM | Permalink

    A short story on lightning strikes:

    About 15 years ago I was working with a geologist named Nancy on the ridgetops west of Raton, NM, conducting some surface geophysical surveys. Some afternoon clouds rolled in and after a while we could hear thunder in the distance. The storms looked to be a good five miles off so we continued our survey.

    We were using an induced electromagnetic profiling technique, where each person holds a rigid coil loop seperated by a fixed distance. We were, as I recall, 30 meters apart. I was holding the transmitter loop and Nancy was holding the receiver loop. The two loops are connected by a steel wire rope as well as by the data and power cabling. We were intent on completing the one transect we were on, which took us across a wide meadow and up a slight incline into some trees. With Nancy in the lead we would take a reading, advance to the next station and take another reading.

    Nancy was at the first station in the stand of trees and I was noting the instrument reading when we heard an explosion. I looked up to see Nancy with her arms over her head to protect her from the bits of branch and bark raining down on her. We both immediately droppped the equipment and scrambled to the truck as a torrent of rain was unleashed upon us.

    After the storm had passed and it was clearly safe we got out to investigate the lightning strike. The lightning did not strike the tree just behind Nancy as I had assumed. It struck a larger tree about 50 feet behind that tree. There was a slab of wood and bark about 4 inches thick had been blow off the side of the tree. This slab was about 2 feet wide and at least 8 feet long. There was probably an equal mass of wood and bark along that side of the tree that had blown off in small pieces, and this is what had struck the tree under which Nancy was standing and sent pieces of that tree raining down.

    I recall thinking at the time that the size of the bark slab was similar to that of a coffin, and that if the strike had come 10 minutes later it very well could have been. My conclusion then was that the incredible current surging through the tree created so much resistive heat it flashed the water in the tree into steam, thus creating the explosive force that blew out the side of the tree.

    I don’t think this was bristlecone pine country, but I wasn’t too interested in the trees except to be glad that they weren’t too close together. I expect that the tree that was struck could survive losing that much cambium and protective bark, but then again it wouldn’t suprise me if the trauma of the strike was sufficient to kill the tree entirely.

  43. Gary
    Posted Oct 25, 2007 at 4:31 PM | Permalink

    #41 MrPete
    I see the “stepped” morphology that does imply sudden damage. I think that a mapping of damage relative to magnetic north and slope might hint at the cause. A random distribution might suggest lightning while a preponderant direction could suggest weather as the cause.

    I looked at some of the other photos in the album. The wildflowers look spectacular. In temperate Eastern forests, spring wildflowers sequester nutrients, particularly phosphorus, while the trees are waking up from winter and the snow melt is carrying off the nutrients downslope. I wonder if something similar is happening in these mountains. BTW, pines technically don’t have flowers. What you photographed were very young cones (this season’s).

  44. MrPete
    Posted Oct 25, 2007 at 5:27 PM | Permalink

    We haven’t been up there in the springtime. Another curiosity for us: snowpack is notoriously unreliable here. The mountains had a nice snowfall right after I went up there. Now it is 79F and everything’s melting. That can happen all winter. A well-known (local) aspect of Pike’s Peak: the AdAMan club has hiked to the top every year on 12/31 for close to a century, to light a fireworks display.

    I must apologize for mis-hearing Leslie’s explanation about the BCP cones and “flowers.” I’ll ask her for a corrected description. (And all those photos are hers…)

    As to slope/direction of preponderant damage, two notes:

    a) The new photo set from trip #4 on 10/20 likely could provide some good data for this, although not every tree was photographed from top to bottom (the priority goal was ground-level bark). It will take some time to process…

    b) I suspect an intense understanding of microclimate weather would be needed. The mountain tops are full of tricky ridges and valleys, open spaces and relatively dense vegetation. And, we get huge storms from a wide variety of directions, especially along the Front Range.

    A permanent NOAA weather station is proposed for the top of Pike’s Peak. That would provide some interesting data!

  45. Gary
    Posted Oct 25, 2007 at 9:20 PM | Permalink

    Just needs a willing PhD student in Botany, some data recorders, and time to figure it out — if it hasn’t been done already. I remember my botany prof in 1970 telling us about the great age of BCPs. Somebody probably has looked at it and the results are buried in the literature.

    Pines (gymnosperms) arose earlier than the flowering trees (angiosperms) so they only have cones with a few simple parts (scales and ovules) instead of more complex flowers.

  46. MrPete
    Posted Oct 26, 2007 at 3:16 AM | Permalink

    Yeah, I got a mini-lecture on gymnosperms and angiosperms when I got home. I really am NOT a biologist…

    For those who have any interest:

    Gymnosperms (from greek for “naked seed”) are plants that produce seeds not encased in a “fruit” covering. I.e., they make something like a pine cone or similar directly-pollinated , which is why most are called “coniferous.” Usually softwood trees, usually evergreen, usually needles. But then there are deciduous, even broadleaf gymnosperms, like ginkgo.

    Angiosperm seeds are encased in a fruit (most plants.) For trees, usually hardwood, usually deciduous (leaves that die each fall), usually broadleaf. To keep us guessing 🙂 some evergreen angiosperms include live oak and rhododendron.

    (BTW, that’s not PhD stuff. Not even undergrad. Any Master Gardener would know. Just not me 😉 There are plenty of reasons Leslie is my better half … her “library stack” recall of plant and animal info is one!)

  47. Gary
    Posted Oct 26, 2007 at 7:40 AM | Permalink

    Plant phenology combined with psuedo-random occurrences like lightning strikes might require a bit of knowledge and experience to map out and explain, but I won’t quibble. It does cry out for investigation if environmental factors are so influential as they appear to be on the growth of BCP tree-mometers. Your mention of the diversity of plant growth habits (evergreen angiosperms and leaf-dropping conifers) tells us that nature has a lot of secrets to discover.

  48. MrPete
    Posted Oct 26, 2007 at 9:14 AM | Permalink

    Re phenology… Yes, it’s a very multidisciplinary field. Our sense is that multiple disciplines are going to be increasingly required for best understanding of observed phenomena.

    My “not PhD” comment was related to what’s needed for the field work. Deeper interpretation may well require very advanced insights.

    Related to all this, one might want to take a look at a link from the phenology article… relating to contributions of Citizen Science.

    We’re pretty amazed by the apparent difference between levels of field observation skill considered acceptable in dendroclimatology vs other bio-related fields. It’s not clear that the conclusions being drawn would pass muster in any high altitude forest ecology,

  49. SteveSadlov
    Posted Oct 26, 2007 at 1:19 PM | Permalink

    RE: #39 – My own experience in the Western US has been that in most years, there are heavy snows after the soil has begun to thaw. April and May are commonly beset by these late storms, in a few cases even early June. It has also been my experience that having the soil freeze prior to a snow pack getting started is rare. Tellingly, the (July 1 – June 30) precip years which are prone to either soil freezing prior to snow pack, or, soil thawing after the pack is gone in spring, tend to be years of drought. Years with ample moisture tend to have snow pack well prior to the soil freezing, and, a late renewal of the pack after it has begun to thaw. (Probably more amazing than the massive amounts of snow that make fodder for photos in places like Truckee is the way such large volumes of snow come and go in short amounts of time … I know our breathren along the Front Range are also quite used to this …)

  50. MrPete
    Posted Oct 26, 2007 at 2:03 PM | Permalink

    Our snows don’t hold a candle to the Sierras… I’ll never forget learning about 50 foot deep snow, about nordic ski cabins that have second, even third story doors to the outside, etc etc…

    the way such large volumes of snow come and go in short amounts of time … I know our breathren along the Front Range are also quite used to this

    But yes, I agree with the bottom line: our snow pack comes and goes. Rather unpredictable. Curious how to correlate “strange” snowfall with mountaintop tree growth. It’s tough enough predicting runoff for our precious reservoirs!

    Adding to a previous list above:

    (Case 1. Soil freezes before snow cover is thick enough to insulate the soil. // spring runoff lost due to frozen soil // growth only from released soil moisture; tree in danger of drought damage )

    (Case 2. Soil does not freeze before snow cover is deep enough to insulate the soil. // spring runoff provides water for growth )

    Case 3. Freezing rain/sleet creates thick ice layer, followed by snow. Partial melt throughout winter.

    Case 4. Repeated freeze/melt cycles with light snows throughout winter. Heavy spring snow followed by rapid melt.

    Case 5. Dry cold winter leaves ground bare but frozen. Spring thaw interrupted by occasional 1-2 foot snows. Spring snows typically melt to bare ground in 24 hours..

  51. jae
    Posted Oct 26, 2007 at 4:23 PM | Permalink

    We’re pretty amazed by the apparent difference between levels of field observation skill considered acceptable in dendroclimatology vs other bio-related fields. It’s not clear that the conclusions being drawn would pass muster in any high altitude forest ecology,

    Do most “dendroclimatologists” have a real expertise in dendrology?

  52. MrPete
    Posted Oct 26, 2007 at 5:23 PM | Permalink

    (My early posting got chopped somehow. I was listing several bio/dendro related fields.)
    jae, I can’t answer your question. I don’t know many dendroclimatologists myself.

  53. Euell Gibbons
    Posted Oct 30, 2007 at 10:25 PM | Permalink

    Ever eat a pine tree? Many parts are edible.

  54. Steve McIntyre
    Posted May 20, 2008 at 12:40 PM | Permalink

    I was looking to contact Craig Brunstein about his sampling at Almagre. Subsequent to our sampling last summer, he died at the young age of 56. Graybill, Lamarche and now Brunstein – all very early deaths.

  55. jeez
    Posted May 20, 2008 at 2:08 PM | Permalink

    Ahah! More evidence of [snip]!

3 Trackbacks

  1. […] but have nothing to do with CO2 fertilization or previously identified issues. We found (See here here ) that strip bark forms can result in enormous (6-7 standard deviation) growth pulses in one […]

  2. […] This is exactly the situation that we hypothesized at Almagre strip bark (our Tree 31 discussed here.) Here’s a ring width plot from the prior post. The glacier-scarred tree would yield a […]

  3. By Tamino and the Magic Flute « Climate Audit on Mar 22, 2012 at 5:07 PM

    […] but have nothing to do with CO2 fertilization or previously identified issues. We found (See here here ) that strip bark forms can result in enormous (6-7 standard deviation) growth pulses in one […]

%d bloggers like this: