Almagre: the Graybill Photos

Christine Hallman of the University of Arizona LTRR has 4 photographs from Almagre from the Graybill collection. Although we didn’t have these matches in mind, there are some beautiful matches in Pete Holzmann’s photos, which Pete has collated below.

Here is the 1983 and 2007 view of the general Graybill Almagre site, looking to the east:

Looking to the south now. (The 1983 view is one ‘ridge’ closer to the mountain, and thus does not have so many trees in the way):

Now here’s an amazing coincidence. At left is Graybill’s photo of a BCP being cored; at right, Pete H. . The shots are from slightly different angles, so the background of sky and trees is not quite the same. But the tree branches are unmistakable and this is clearly the same tree. Even more remarkably, this happens to be our tree #30 (Graybill 84-55), which is the one that has the long chronology that we posted about in our first post (and one of the tagged trees for which we’ve been unable to reconcile tag numbers to archive ids so far.)




  1. Jim
    Posted Oct 16, 2007 at 5:46 PM | Permalink

    Just a location question

    Is that boulder in the plains at the base of the mountain?

  2. SteveSadlov
    Posted Oct 16, 2007 at 6:46 PM | Permalink

    Any idea why Graybill cored so far up the tree?

  3. Jonde
    Posted Oct 16, 2007 at 7:29 PM | Permalink

    Until first branch knot, the tree trunk is slightly inclined. This can cause the formation of reaction wood, etc. altered wood cells because of gravitational pull and trees attempt to upright the stem. It is not possible to use the tree rings where reaction wood is formed in dendrochronological analysis. After the branches the tree trunk seems to be straight.

  4. Steve McIntyre
    Posted Oct 16, 2007 at 7:45 PM | Permalink

    #3. Many bristlecone sites are on slopes – wouldn’t that mean that “reaction wood” is a real risk to the data sets? (I’ve never seen the issue mentioned in bristlecone literature.)

  5. steven mosher
    Posted Oct 16, 2007 at 7:59 PM | Permalink

    RE 3 fascinating

    explain more. is reaction wood compositionaly different? or visibly different?

  6. Geoff Sherrington
    Posted Oct 16, 2007 at 8:03 PM | Permalink


    There has to ne a natural mechanism that makes trees grow more or less vertical and for roots to go down. It’s ‘gravitrophism’ for searches. If a tree root, for a simple example, starts to grow the wrong way, then then there are small particles in the growth cells that act as level sensors and trigger the release of growth hormones that thicken the root on the side that forces the root in the natural direction.

    It follows that if you sample the trunk of a tree now vertical, for dendro reasons, then you have to assume that the tree was always vertical, which is a bit of an assumption for trees several hundred years old. It adds to the noise.

    I have seen trees in China that are 4-600 years old from history records. Many of them are now a couple of feet above the surrounding agricultural fields, with small pedestals built around them. I presume without proof that this is because of deflation of the surrounding soil after centuries of cropping. If that is the case, one could expect significant changes in the ability of the tree’s roots to deliver nutrients and so perhaps long-term differences in tree ring patterns. More noise.

  7. Jonde
    Posted Oct 16, 2007 at 8:50 PM | Permalink

    Even in the sloped ground trees grow vertically towards the sky and tree rings tend to be “normal”. In some cases, like thawing, erosion, avalanche, etc., the trees might tilt. If the tree tilts, it starts to correct the position back towards the vertical position and this causes the bend or curve in the stem. Until the trees has corrected its position, wood cells formed are effected by the gravitational pull. Wood in the affected part the stem is called reaction wood.

    Two links. First one see 6d for illustration. Second one has a good photo to show the reaction wood in cross section of the white fir (ltrr page).

    It might be that the tree rings of Graybill 84-55 are “infected” to the extent that it is impossible to cross-date the tree rings.

  8. Pat Frank
    Posted Oct 16, 2007 at 8:52 PM | Permalink

    #3,6 — If the tree knows geometry, it might be possible to renormalize the ring widths using the cosine of the angle the tree is out of vertical. Has anyone tried something like this?

  9. Henning
    Posted Oct 17, 2007 at 8:18 AM | Permalink

    Its a combination of factors that cause trees to grow excentric or oval shaped trunks. Gravity (on changing slope angles) is just one of them. Prevailing wind condition is another, probably more dominant.

  10. Hans Erren
    Posted Oct 17, 2007 at 8:29 AM | Permalink

    #3,6 — If the tree knows geometry, it might be possible to renormalize the ring widths using the cosine of the angle the tree is out of vertical. Has anyone tried something like this?

    You don’t know if the position angle hasn’t changed more often during the last millennium, so it’s an arbitrary correction (very useful though for adjusting towards the “right” answer. 😉 )

  11. Henning
    Posted Oct 17, 2007 at 8:34 AM | Permalink

    P.S. Frankly I can’t see how you could account for any of these without taking two cores in the same height at different ankles and calculate the average width of each ring from these two samples. Whether or not that makes treerings a better temperature proxy – I wouldn’t know. It seems to me like treerings in general could probably serve as a valid correction basis for borehole termeometry but beyond that I’d doubt it.

  12. Posted Oct 17, 2007 at 9:00 AM | Permalink

    #1 – That’s Colorado Springs. If you look closely, you can see me waving. 😉


  13. Mark T.
    Posted Oct 17, 2007 at 9:03 AM | Permalink

    #3,6 — If the tree knows geometry

    I’ve heard it is a required class in Baumgarten.


  14. Mark T.
    Posted Oct 17, 2007 at 9:08 AM | Permalink

    Cool, another Springs resident. We’re growing in number. Our plan to take over the world is gaining momentum! 🙂


  15. jae
    Posted Oct 17, 2007 at 9:20 AM | Permalink

    Reaction wood in softwood tree species takes the form of increased amounts of latewood on the underside of the “lean” of the tree and is called “compression wood.” This adds mass to the underside of the bole and gives it more compressive strentgh. The presence of reaction wood probably renders the core useless for these types of studies, since the amount formed year-to-year could vary due to the amount of stess on the tree (the size of the bole and amount of lean will dictate how much wood is formed). In the case of hardwoods, the reaction wood is formed on the UPPER side of the bole and is called “tension wood.” The presence of reaction wood is one of the major reasons for severe warpage of lumber.

  16. Posted Oct 17, 2007 at 9:25 AM | Permalink

    It is common practice to renormalize using trigonometric methods in dendroclimatology. Using this method one can compute ring widths to .0031459mm. Also, by differentiating the cosine function, one can determine instantaneous growth rates for any hour of any day.

  17. MrPete
    Posted Oct 17, 2007 at 11:44 AM | Permalink

    Re: reaction wood, assymetric trunks, etc…

    Some observations that may be helpful:

    1) The collected data may be sufficient to further inform some of the stated hypotheses. Most of the trees were sampled from more than one side. The provenance data includes tree lean, surface slope direction, and side of tree cored for every sample. Plus, there are photos of every tree (not from all sides. With more people, maybe that would be good, just like SurfaceStations 🙂 )

    2) Some trees can only be sampled from certain directions. The strip bark trees in particular are only growing where bark exists.

    3) At the extreme would be tree #31 (Graybill tag #84-56, see below). It has TWO elongated oval trunks (split at the base). The easterly trunk, being measured in the photo, ended up useless (immediate rot). We cored the western, vertical trunk. Even that one was more than two feet in one direction, less than one foot in the other. Partly this is due to the stripped bark, so it only grows on the east and west sides. Partly due to prevailing wind, I’m sure (strong westerlies). And partly due to the generally south-facing slope of the site. (Click to get to the gallery, click again on the magnifying glass for a closeup view. Grainy but visible.)

  18. Steve McIntyre
    Posted Oct 17, 2007 at 12:44 PM | Permalink

    I’ve got some measurement data last evening for two cores from this tree, which I’ll post up separately. Pete, which way is this photo facing?

  19. MrPete
    Posted Oct 17, 2007 at 12:48 PM | Permalink

    This photo is facing north-northeast.

    With all this interesting discussion, I’m wishing we had a compass with us to accurately determine the angles (slope, tree lean, sample entry angle, photo direction). As it is, we only have approximations.

  20. MrPete
    Posted Oct 17, 2007 at 12:53 PM | Permalink

    Thinking out loud: we may go back and do just that. It would improve the data set, and would correct for any observer errors as well. Without an actual compass, it would not be that hard for an observer to get their directions mixed up. Yes, even on the Front Range of Colorado. Around here, everyone knows which way is west. There’s this big mountain range that’s kinda hard to miss :-)…

  21. Steve Sadlov
    Posted Oct 17, 2007 at 12:54 PM | Permalink

    Some stuff on Cali BCPs:

  22. Steve Sadlov
    Posted Oct 17, 2007 at 12:55 PM | Permalink

    Ignore that first link, browser is misbehaving ….

    This is what I meant to post.

  23. Steve Sadlov
    Posted Oct 17, 2007 at 1:16 PM | Permalink


    “Annual precipitation is less than 12 inches (30cm), most of which arrives as snow in winter. On a summer’s day the amount of precipital moisture in the air is about half a millimeter, the lowest ever recorded anywhere on earth.”

    “These ranges have more ties with plants and animals to the east in the Great Basin, than their close neighbor the Sierra Nevada, only 10 miles (16km) westward. ”

    “The average max.-min. temperatures range from about 70°F (21°C) to 37°F (3°C) at the base, and from 36°F (2°C) to -26° (-32°C) in the alpine zone. Precipitation averages 4 in. (10 cm) at its base to 20 in. (50 cm) mostly as snow, along the crest. Due to the varied topography, precipitation can and does differ greatly in localized areas. Winds blowing along the crest can blow most of the snow from some areas, leaving little for trees like the bristlecone at the treeline – 11,200′ (3414m). Thunder and lightning storms occur frequently in the high country. Winds are strong at the crest, both summer and winter.

    “The dry climate and high altitude make this region a rare environment. The rapid changes in elevation create abrupt habitat and species changes. The soil quality is poor, and at its poorest in the alpine zone. This factor combined with a short growing season, results in sparse and delicate flora. Recovery from disturbance is slow, said to take more than 100 years!”

    “The geologic makeup of the White Mountains is of quartzitic sandstone and granite bedrock. A large part of the soils on these slopes have been swept away by the extreme conditions. Also present are extensive outcrops of dolomite (limestone) a very ancient rock first laid down under water 500 million years ago, then slowly uplifted through time. Numerous fossils of this period can be found here. It has been speculated that one could have walked chest deep across the early Paleozoicsea located in the region at that time. The dolomite is low in nutrients but of a higher moisture content than the surrounding sandstone. Because these soil types inhibit the growth of other plants, they provide a competition-free arena for the slow-growing bristlecone pines.”

    “On dry windswept mountaintops of the Great Basin in the western United States grow earth’s oldest living inhabitants, the bristlecones (Pinus longaeva, Pinus aristata).”

    [SteveS’ note – now I can see why the California ones look sparser than the Co ones – they are different subspecies. P.Longaeva are found in the Great Basin whereas P. Aristata are found in the Rockies]

    “Needles can live twenty to thirty years thus adding new foliage to that already on the tree, picture of needles takes little energy. The long-lived needles provide a stable photosynthetic capacity to sustain the tree over years of severe stress. ”

    “The oldest bristlecones live in the most exposed sites, with a considerable amount of picture of old tree space between each tree. The longevity of the bristlecone needles and the inability of other plants to grow in the dolomite soil make for little leaf litter or ground cover. This distance in between, combined with the lack of ground cover, is how a tree can sustain a lightning strike, catch fire, and not have the fire spread to surrounding trees. ”

    “Simply put, dendrochronology is the dating of past events (climatic changes) through study of tree ring growth. Botanists, foresters and archaeologists began using this technique during the early part of the 20th century. Discovered by A.E. Douglass from the University of Arizona, who noted that the wide rings of certain species of trees were produced during wet years and, inversely, narrow rings during dry seasons.”

    “New wood grows from the cambium layer between the old wood and the bark. In the spring, when moisture is plentiful, the tree devotes its energy to producing new growth cells. These first new cells are large, but as the summer progresses their size decreases until, in the fall, growth stops and cells die, with no new growth appearing until the next spring.”

    [SteveS’ note – in other words, by about August, the remnent moisture from snow pack is gone, leaving only whatever has come from monsoonal / local convection, which quickly is overconsumed, the growth stops for the year.]

  24. Steve Sadlov
    Posted Oct 17, 2007 at 1:23 PM | Permalink

    Final note, as mentioned before, the scenarios that I am familiar with which would give White Mountain BCPs more moisture during the key time frame in a given year are:
    1) A very agressive zonal pattern of great depth, which can overcome the Sierra rainshadow. We get this frequently during an El Nino and at times (maybe 4 – 5 times during the wet season) during some La Ninas (of the type where the rain line is ~ 36.5 N).
    2) A Siberia Express pattern (e.g. polar jet moving from Yukon into California, straight down from the north)
    3) Persistent cut off lows over Nevada during the summer months
    4) An early (May – June) monsoon that is far to the west (as opposed to one going up through the middle of the Great Basin)

  25. steven mosher
    Posted Oct 17, 2007 at 3:46 PM | Permalink

    S Sadlov

    Briffa did some mountain hemlock work in Lassen. With the eruption there in 1914 I bet those
    rings would be interesting.

  26. steven mosher
    Posted Oct 17, 2007 at 3:49 PM | Permalink

    OT. for sadlov.

    Click to access hemlock.pdf

  27. Murray Duffin
    Posted Oct 17, 2007 at 3:50 PM | Permalink

    In the Swiaa Alps pine and larch can be found on steep slopes, with the first inches of the trunk growing almost horizontally, and then bending to vertical in the next couple of feet, thus providing the distinctive shape of the Swiss Alphorn. Modern alphorns usually get their shape from lamination. No help for dendrochronology, but illustrates the tree’s drive to gow where it can and get vertical as quickly as possible. Murray

  28. steven mosher
    Posted Oct 17, 2007 at 3:56 PM | Permalink

    RE 26. actually other might like this as well. From a method standpoint.

  29. steven mosher
    Posted Oct 17, 2007 at 4:04 PM | Permalink

    OK last post on 26:

    Abstract: In this paper we review the ecology and physiology of mountain hemlock (Tsuga mertensiana (Bong.)
    Carrière) in the context of a dendroclimatological analysis. To better understand the relationship between mountain
    hemlock growth and climate variability throughout its range we have analyzed chronologies from 10 coastal sites, located
    along a transect extending from northern California to southern Alaska. The chronologies exhibit significant
    large-scale cross-correlations, with two distinct growth regions implied: chronologies from the northern Cascades in
    California, to the Queen Charlotte Islands, British Columbia, are correlated with each other but are distinct from Alaskan
    chronologies. While intervals of coherent reduced growth along the entire transect occur episodically throughout
    the record, intervals of coherent enhanced growth are less common. Response function analyses indicate that summer
    temperature is the most influential factor limiting growth throughout the study region, while winter precipitation is an
    additional limiting factor south of Alaska. Warm summer temperatures are associated with enhanced growth in the current
    year but with reduced growth in the following year. This response is believed to be a reflection of the energy required
    to mature cones initiated in the preceding year. The association with winter precipitation may reflect the role of
    deep, persistent snowpacks in regulating the duration of the growing season.”

    A Lassen versus white mountains study would be kinda neat S Sadlov. Your thoughts on climatic differences?

    Reconstructing California climate.


  30. MrPete
    Posted Oct 17, 2007 at 7:31 PM | Permalink

    By the way, re reaction wood on tree #30 / 84-55… the “matched” photo above is not particularly helpful for understanding the angle of the tree. That photo was rotated a bit to match the Graybill version, and seriously cropped as well.

    In general, I don’t know that these photos can be relied on to understand absolute lean. The terrain was often quite rough, and the camera did not have a built-in level ;)…

    You can see whether a trunk contains bends or twists of course. Below you’ll find the whole photo. (As usual, click to see in the gallery…)

    As I recall, tree #30 has a very slight lean overall. Not enough that we recorded it. We’d have to go back up and stare at these trees a bit to record a better set of lean/reaction wood data.

  31. Geoff Sherrington
    Posted Oct 17, 2007 at 6:56 PM | Permalink

    Seems that most discussion is about nutritional changes, not temperature. Though one is said to interact with the other in the literature, it is indirect and noisy. It is integrated heat, more than maximum or average temp, that more closely affects the growth. My sceptism about highly accurate temperature reconstructions remains. I’m more interested in the within-tree and between-tree variability, plus the between-observer variability, of the isotopes. Fascinating exercise, had to be done.

  32. Jonde
    Posted Oct 17, 2007 at 7:10 PM | Permalink

    Significant part of the radial growth (early wood cell formation) commence at the early part of the growing season, etc. spring and early summer. Late wood cells do not anymore have a significant influence on ring widths as is also easy to see from Pete’s photos. Because of this, late summer and autumn weather conditions do not anymore have influence on ring widths of the current year.

    This paper is quite good in explaining the phenomenon (see table 4). 90% of radial increment of pines took place before end of June in intra-alpine valley in Switzerland (dry conditions).

  33. Jonde
    Posted Oct 17, 2007 at 7:14 PM | Permalink

    Sorry bout the link. This should work better.

  34. Geoff Sherrington
    Posted Oct 17, 2007 at 7:42 PM | Permalink

    Just read the Abebnah Ph.D. posts. Simple fact explained again by her. If you have drought, trees don’t grow fast. If you have lots of water, trees grow faster. That’s why we water gardens. As to how this correlates with temperature, or even dominates it…..

  35. SteveSadlov
    Posted Oct 17, 2007 at 8:33 PM | Permalink

    RE: #26 – Lassen certainly gets more direct weather from the Pacific than the Whites. Also, during times like right now, with the rain line moving between 36 and 38 N, Lassen is getting much more rain, in general, than we are down here in the upper 30s N. Lassen also experiences a type of cold damp, which you see when the snow level is down around 2K at that latitude, which you would never get in the Whites. Finally, big difference in exposure to monsoonal moisture – the Whites are more prone to it (but given that its really May and June early monsoons that would matter, Lassen is far enough north as to expect standard Pacific storm rain that late in the season … 6 of one, half dozen of the other). Soils and bedrock obviously completely different.

  36. Jonde
    Posted Oct 17, 2007 at 8:43 PM | Permalink

    In fact there are “threshold temperatures” for different plants. I think this was mentioned previously. If the temperatures are between the lower and upper limits, the growth is not affected by the temperature, at least not in any significant way. Because of this, 1-2C differences in annual temperature are not easily visible in tree rings unless the changing temperatures are strongly influencing the periods of growing season when the temperatures are in the upper or lower thresholds. Even then I would say that only change in temperatures that can be seen in bristlecone ring widths is the change in spring temperatures (earlier snow melt date – earlier melting of soil and higher soil temperatures – earlier start of root activity, thus roots can absorb more soil moisture during the early wood cell development at early summer if there is water what to absorb).

  37. Steve Sadlov
    Posted Oct 18, 2007 at 11:08 AM | Permalink

    I disagree with #36. I doubt that at an elevation above 10K feet in such an inland part of California (e.g. Continental Climate), that even a 2 deg C higher than normal “average” temperature for a given spring is going to make any difference with when the snow pack is gone or the soil thaws. The controlling factor will be both the amount of winter (frozen) precip and how late it fell in the season. Ironically, these same factors will also control available moisture in July and August. A big, late snow pack means good moisture during July and August. (The “normal” begining of climatic Spring in that location is essentially late June.) The winter of 2006 – 2007 is a classic case of the worst case scenario in terms of moisture. Total precip was low, and, the normal big dumps of Feb and March did not happen. There was a slight recovery in April, but not enough to result in a decent snow pack. The meagre snowpack was gone by early June, and with the soil mositure already low due to the overall lack of precip (plus lots of wind) it was bone dry by the end of June.

  38. Steve Sadlov
    Posted Oct 18, 2007 at 11:13 AM | Permalink

    I would add that for White Mountain BCPs, the only liquid precip even experienced at the sites is due to very infrequent and unreliable local convective or imported monsoon showers. Most of these occur, when they occur, at the very end of the growing season, the peak being Aug and Sept. Very rarely, these will start earlier, say, in July or June. Even more rarely is a freak occurence of a late May monsoon, but that is so rare it is not worth discussing. Bottom line is, for all intents and purposes, White Mtn BCPs depend on snow pack for moisture.

  39. Jonde
    Posted Oct 18, 2007 at 7:21 PM | Permalink

    Steve you are right and, in fact, this was my point. Most of the radial growth will take place during the first month of the summer and is dependant on winter precipitation (snow). Positive response of ring width to winter precipitation is clearly seen in Siberian larch forests. I dare to say that the same rule apply to BCP.

    I was just hypothesising that if someone use BRCs to reconstruct past air temperatures, those reconstructed temperatures are most probably defined by spring air temperatures.

  40. Geoff Sherrington
    Posted Oct 19, 2007 at 7:10 PM | Permalink

    Re 39 Jonde

    You might be right. But remember, a tree cannot grow in the absence of water. Water is needed to take nutrients from earth to roots. So there is some water and there is some water variability (plausibly) over time. Which is more important, spring air temperatures or water supply variablilty, I have no idea, but I would not place all bets on just one horse. And of course, there could be a correlation between spring air temperatures and water supply, just to confuse the experiment – an experiment dealing with CO2 effects on hypothesised global warming.

  41. jae
    Posted Oct 19, 2007 at 8:40 PM | Permalink

    Mr. Pete:

    As I recall, tree #30 has a very slight lean overall. Not enough that we recorded it. We’d have to go back up and stare at these trees a bit to record a better set of lean/reaction wood data

    Slight leans make no significant difference. Reaction wood is only significant at fairly extreme leans, like maybe 20 degrees. I would not worry much about leans less than that.

  42. Jonde
    Posted Oct 20, 2007 at 6:20 AM | Permalink

    40 Geoff:

    I say you are right and if I should make a bet I would bet precipitation. It is just funny that Mann reconstruct air temperatures that do not have any influence on radial growth (annual means). If Mann insist to use BCPs to reconstruct any temperatures it should be early summer, not any 12 month means.

%d bloggers like this: