Malcolm Hughes and the Witness Protection Program

Malcolm Hughes coauthor Matthew Salzer) made a presentation entitled “Twentieth Century Bristlecone Pine Tree Rings near Upper Tree Limit Wider than in Recent Millennia”. This included a report on Sheep Mountain. He showed a picture of Matthew Salzer on Sheep Mountain and praised his work. He said that there was no difference between strip bark and whole bark chronologies and showed a graphic up to 2005 with relatively wide recent ring widths. Linah Ababneh’s name did not pass his lips (the Ababneh thesis showing non-nomalous 20th results discussed here, here here), nor did he discuss her work. In Ryan Maue’s felicitous phrase, it was as though she had been put in witness protection. The words “CO2 fertilization” also did not pass his lips in a discussion of possible explanations for the recent behavior.

He showed a photo BCPs at 3400-3500 m: it’s definitely dried looking than Almagre. He described the substrate “if you can call it soil”. He showed graphs from Lamarche et al 1984 and Graybill and Idso 1993. Sheep Mt was now said to have 300 samples and to go to 2005 (also mentioned Pearl Peak, Mt Wahington). Showed chronologies which visually were not small in recent years.

He showed a picture of a core along the lines of pictures shown for Almagre at CA showing relatively wide recent widths as compared to the 19th century. He showed a plot with 50-year bins back to 3000 BC or so showing that the present widths were the widest since about 2000 BC (when there was also a period of wide ring widths).

He showed a plot of whole bark and strip bark (# 12) showing no noticeable difference in chronologies. For reference, here is Linah Ababneh’s plot showing a material difference:

Abaneh Figure 3.Sheep and Patriarch strip and whole bark tree-ring width chronologies compared with Briffa et al. 1992 summer reconstruction based on tree-ring density. All chronologies are filtered after converting to z-scores to match the scale. Tree-ring data and Briffa et al. (1992) data are filtered using an 8-year low-pass filter (Fritts 1976). Confidence Intervals are based on the Briffa et al. (1992) series

He said that the increase could be attributed to temperature increase citing temperature stations at Barcroft and Crooked Creek.
Again for reference, here is Abaneh’s plot of ring width chronology:

Ababneh Fig. 5. Cold and warm periods as inferred from tree ring widths chronology (Ababneh, 2006, This study) fluctuations above and below the mean after normalizing, whole-bark and strip-bark chronologies are grouped together from two sites Patriarch Grove and Sheep Mountain.

Linah Ababneh’s result are different than the results presented at AGU. Malcolm Hughes was on Linah Ababneh’s thesis committee but did not mention her work which arrived at different conclusions than the ones presented here. Linah Ababneh’s thesis said that the data would be archived at ITRDB, but Hughes and her other thesis supervisors did not archive the data. David Meko said that they had lost track of Ababneh although she was easily located. When asked about the data by a CA reader, she said that she had legal advice not to provide the data to me.

Why didn’t Hughes acknowledge the efforts of Linah Ababneh in collecting Sheep Mt data, while acknowledging the efforts of Matthew Salzer? Why didn’t he discuss her conflicting results?


  1. pk
    Posted Dec 13, 2007 at 9:57 AM | Permalink

    Interesting considering Ababneh even won an award in 2002 for her work on the bristlecones.

  2. Peter D. Tillman
    Posted Dec 13, 2007 at 10:00 AM | Permalink

    A very odd affair. I’m sorry I no longer live in Tucson — it would be fun to go hang out at his dept. & pick up the gossip… Hint to Tucson CA readers…

    Cheers — Pete Tillman

  3. Bill F
    Posted Dec 13, 2007 at 10:10 AM | Permalink

    Was there any Q&A opportunity at the end of his talk where he could be asked about the subject?

  4. Mark H.
    Posted Dec 13, 2007 at 10:24 AM | Permalink

    It seems to me the future is in younger researchers, like Linah. They ought to be encouraged to keep their promises to provide data AND eventually they will find the courage to face down the fossils that will (in 20 years) be retired or dead.

  5. JamesG
    Posted Dec 13, 2007 at 10:45 AM | Permalink

    Interesting that Salzer brings back the MWP and LIA in North America, From one of his papers:
    “Noteworthy in the reconstructions are the post-1976 warm/wet period, unprecedented in the 1,425-year record both in amplitude and duration, anomalous and prolonged late 20th century warmth, that while never exceeded, was nearly equaled in magnitude for brief intervals in the past, and substantial decadal-scale variability within the Medieval Warm Period and Little Ice Age intervals”

  6. Gunnar
    Posted Dec 13, 2007 at 10:57 AM | Permalink

    >> she said that she had legal advice not to provide the data to me.

    I found this fact disturbing and interesting. If I understand correctly, she was a grad student at this university. Given that she could notpossibly open herself up to legal liability by publishing her own data, I’m speculating that the legal advice came from her personal attorney, and that the only source of legal threat comes from who ever owns the data (universtity?).

    If she collected the data herself, then it’s hard to see what she is afraid of. Do they still have some leverage over her? When one submits what amounts to a homework assignment, IP rights are not transferred to the school.

    Can someone ask her “what is the basis for the legal advice”?

  7. Larry
    Posted Dec 13, 2007 at 11:00 AM | Permalink

    10, if someone can mine the CA archives over the past few months, there was a more complete explanation. I seem to recall someone saying something about it being Hughes’ attorney, but I don’t remember whether that was speculation or information.

  8. BrianMcL
    Posted Dec 13, 2007 at 11:08 AM | Permalink

    Re Ababneh’s IP,etc.

    Can we please not go through all of that again? Surely it was done to death already.

  9. Larry
    Posted Dec 13, 2007 at 11:14 AM | Permalink

    12, ok, then please summarize.

  10. Posted Dec 13, 2007 at 11:25 AM | Permalink


  11. Bill
    Posted Dec 13, 2007 at 11:32 AM | Permalink

    I think it’s likely that the chronology cited was the same Salzer and Hughes used for “Bristlecone Pine Tree Rings and Volcanic Eruptions Over the Last 5,000 Years”, (2005). Their primary finding in that paper was that, “Many years of low growth identified… over the last 5,000 years coincide with known large explosive volcanic eruptions… (and) ring-width minima can be matched with known volcanic eruptions or ice-core volcanic signals in 86% of cases.”

    Assuming there were no volcanic eruptions in this presentation (?) one wonders at the serviceability of their chronology, to do double duty, one paper applying them to the proof that all cool years were forced by volcanoes, another to verify unprecedented modern warming

    It’s curious, also, how volcanic activity subsides (Salzer’s last citations of 1836, ’38, ’40 and ’42 don’t seem to interfere) with the advent of the modern hockey game.

  12. BrianMcL
    Posted Dec 13, 2007 at 11:36 AM | Permalink


    See also from comment 18 onwards.

  13. MarkW
    Posted Dec 13, 2007 at 11:42 AM | Permalink

    The data must be good. It agrees with the models.

  14. Gary
    Posted Dec 13, 2007 at 11:52 AM | Permalink
  15. Gary
    Posted Dec 13, 2007 at 11:53 AM | Permalink

    Let’s try that again.

  16. Gary
    Posted Dec 13, 2007 at 11:56 AM | Permalink

    A preview button sure would help…

  17. Bernie
    Posted Dec 13, 2007 at 11:59 AM | Permalink

    If we take Hughes and Salzer data at face value and discount Ababneh’s contrary findings then what we have is BCPGW not AGW, since none of wht Hughes argues undermines in any way Steve and Ross’s original results, namely, take BCPs out of the equation and the hockey stick disappears. If so, then the issue reverts to transparency in scientific research and Hughes and Salzer do not have a fig leaf to hide behind.

  18. tom
    Posted Dec 13, 2007 at 12:15 PM | Permalink

    ‘Why didn’t Hughes acknowledge the efforts of Linah Ababneh in collecting Sheep Mt data, while acknowledging the efforts of Matthew Salzer? Why didn’t he discuss her conflicting results?’

    Because it doesn’t fit the script

  19. Bill
    Posted Dec 13, 2007 at 12:22 PM | Permalink

    Re: Bill F, “Was there any Q&A opportunity at the end of his talk?”

    I would be interesting to hear their reactions to questions about the years:

    1815 – Mt. Tambora
    1883 – Krakatoa
    1902 – Mt. Pelee
    1980 – Mt. St. Helens
    1991 – Pinatubo

    Namely: “If the hundreds of other volcanoes you cite” can correlate with reductions in ring width, why don’t these eruptions?

    (Salzer was the primary author of Bristlecone Pine Tree Rings and Volcanic Eruptions Over the Last 5,000 Years”, (2005), and Hughes secondary).

  20. MarkW
    Posted Dec 13, 2007 at 12:30 PM | Permalink

    Obviously those volcanoes failed to teleconnect to the BCP’s properly.

  21. SteveSadlov
    Posted Dec 13, 2007 at 12:33 PM | Permalink

    1990s and early 2000s at Sheep Mountain … generally speaking, good snow pack. Decent mid spring snow dumps. Here are some real standout years – 1992 – 93, 1994 – 95, 1995 – 96, 1997 – 98, 2003 – 2004, 2004 – 2005, 2005 – 2006. Some of these were El Ninos and some were “wet” La Nina / neutral (typically accompanied by multiple strong Tonapah Lows).

  22. steven mosher
    Posted Dec 13, 2007 at 12:53 PM | Permalink

    a very bad man.

    some times old farts should just retire, before they become bad men.

  23. Gunnar
    Posted Dec 13, 2007 at 12:56 PM | Permalink


    Ok, that settles that issue. Sorry for bringing up something that’s already been dealt with. Those of us still trying to be productive don’t have time to read each and every thread. 🙂

    My guess is that she is a true believer and inadvertantly followed the evidence. Prior to Steve M, they really didn’t need to be so careful. What are the odds that a skilled auditor would take this much personal time?

  24. John A
    Posted Dec 13, 2007 at 1:36 PM | Permalink

    Is it too late to bring out the “Free Linah Ababneh” and the “Linah Ababneh is Innocent” t-shirts in time for Christmas?

  25. Willis Eschenbach
    Posted Dec 13, 2007 at 2:07 PM | Permalink

    Me, I got no dog in the fight, I just love the name “The Abaneh Thesis” … sounds like a Robert Ludlum novel, soon to be a major motion picture …


  26. David L. Hagen
    Posted Dec 13, 2007 at 2:10 PM | Permalink

    Linah Ababneh, Ph.D.
    appears to be constructing a web site at the Laboratory of Tree Ring Research, listing her contact as:

    Apparently not yet listed under People at Tree Ring Lab

  27. Gunnar
    Posted Dec 13, 2007 at 2:13 PM | Permalink

    >> Free Linah Ababneh

    Except for the Stockholm syndrome…

    She’s protecting that data, as if her job depended on it…

  28. Larry
    Posted Dec 13, 2007 at 2:27 PM | Permalink

    26, holy moly, how many dendrochronologists does it take to count tree rings?

  29. Pat Keating
    Posted Dec 13, 2007 at 3:01 PM | Permalink

    To be fair to Hughes, there may be a problem with the way the Ababneh data was collected, which might also explain the lawyer thing.

  30. R John
    Posted Dec 13, 2007 at 3:02 PM | Permalink

    They have created quite a little fiefdom there at U of A in BCP tree ring study. Not surprising, though, when you consider how much they push the AGW stuff down there.

    And as Anthony Watts also discovered, Arizona State University has one of the most notorious USHCN weather station in the middle of a parking lot.

  31. R John
    Posted Dec 13, 2007 at 3:06 PM | Permalink

    Oops… Forgot my link.


  32. R John
    Posted Dec 13, 2007 at 3:11 PM | Permalink

    Sorry, I’ve done this before, but seem to be forgetting something.

  33. BrianMcL
    Posted Dec 13, 2007 at 3:20 PM | Permalink

    Re your question in post #28
    “how many dendrochronologists does it take to count tree rings?”

    The answer, of course, is none as it is too hard and costs too much.

    Over at RC just now they’re hosting a discussion on changing Ptarmigan grazing patterns (seriously!) Part of my post asking when we can expect the same level of diligence to be applied to BCPs was deleted as “Off topic”. Go figure?

  34. Dan White
    Posted Dec 13, 2007 at 4:28 PM | Permalink

    John A. # 24,

    How ’bout we give a “Don’t Taze Me, Malcom!” shirt to Linah?


  35. SteveSadlov
    Posted Dec 13, 2007 at 4:51 PM | Permalink

    Yah, don’t taze me bro-has!

  36. Doug
    Posted Dec 13, 2007 at 5:11 PM | Permalink

    Re 34-35
    The Arizona system has not been very honest in this field. I did some research into them blocking Steve from their data base, and basically they stonewalled. Ignore whatever data you like Dr. Hughes, we’re behind you all the way!

  37. Larry
    Posted Dec 13, 2007 at 5:12 PM | Permalink

    35, it gets easier the more times you do it.

  38. Posted Dec 13, 2007 at 8:01 PM | Permalink


    I hope I got that right.

  39. Steve H
    Posted Dec 13, 2007 at 8:27 PM | Permalink

    Steve McIntyre:

    This is why I have supported, and will continue to fund your efforts. What you have done with the Bristlecone Pine Tree Rings, is beyond outstanding.

    Linah Ababneh’s result are different, and that is very informative. She concluded that rainfall was the most important aspect of tree ring growth.

    How do we settle this debate? Measure the sediment layers in the lakes closest to the sampled trees. High rainfall will increase the amount of sediment that was deposited in the lake.

    Why has such a simple measurement been totally ignored?

  40. Pat Frank
    Posted Dec 13, 2007 at 8:28 PM | Permalink

    I find it peculiar that strip bark and whole bark BCP’s should have the same ring width excursions over the same years and growth conditions. This result implies that each BCP maintains a constant growth rate in its remaining cambium, following bark-stripping. To do so, each such tree must renormalize its metabolic activity to its own percent of bark coverage. That is, as the bark is stripped off across the centuries, by whatever happenstance, each such tree scales down its metabolism linearly with the percent of remaining cambium, resulting in a constant depositional ring-width. That’s the only way strip bark trees could maintain a linear living-cambium ring-width correlation with whole bark trees. Naively speaking, I find that very hard to believe. I’d further bet that there is zero evidence that BCP’s, or trees of any sort, inversely re-normalize their metabolism to bark coverage.

    Not only that, but we’d have to assume that strip-bark BCP’s also lay down constant ring widths all the way across the remaining live-cambium surface. There should be no non-linear, e.g., horizontally conical, growth pattern across the remaining bark-strip. If there were, then a non-linear ring-width time series core would necessarily result, unless one judiciously chose to selectively core the region of the remaining living circumference that manifested rings remaining in concert with the underlying tree rings that reflect the pre-stripped, whole bark tree. That region would be somewhere around 25% in from the edge of the bark-strip (this point would vary with the shape and asymmetry of the ring-growth). To do this one would sample one’s way horizontally across the living bark, and then choose the core that showed the best linear extension from the underlying pre-stripped ring widths across the historical time region in which the bark was stripped. Doing so would, of course, amount to cherry-picking one’s core and imposing one’s conclusion.

    WRT the trees themselves, it seems much more likely that a strip bark tree would compensate for loss of bark by laying down thicker rings in the remaining living cambium. It shouldn’t take many core samples, or perhaps crosscuts, to check this.

    Tree response to bark-stripping could be further checked, however, by correlating leaf coverage with bark coverage. Leaf surface area is a measure of photosynthetic activity, and so should be a measure of metabolic activity. One could correct this further by renormalizing leaf surface area to chlorophyl greenness. If ring widths are constant no matter the bark coverage, then the (leaf photosynthesis)/(bark coverage) ratio should be a constant across BCP’s, whole bark and strip bark.

    One should also control for root-growth, but on first pass could assume that in mature trees metabolic activity is equi-partitioned, and root turnover and ring deposition are approximately linear as well. A plant biologist could probably make a rational judgment here.

    In any case, I think Malcolm Hughes’ claim about constant BCP ring widths no matter what barks requires considerable examination. The fact that there is so much secretiveness, hostility, and obstruction attendant to this work only encourages suspicion that something is not right. Science should be entirely transparent. Dr. Ababneh’s reference to a laywer’s advice on release of data is entirely inappropriate to the ways of science. It implies that she’s (been) frightened.

    • Pat Frank
      Posted Nov 17, 2009 at 11:50 AM | Permalink

      Re: Pat Frank (#40), The Pat Frank on comment 40ff isn’t the one — me — who has been posting for years here on sundry threads about AGW and tree ring thermometry, works at Stanford, and wrote the Skeptic article.

      [Jean S: Are you sure about that? #40 has the same originating IP as this one, and the e-mail address seems to be the Standford address you used to give at that time.]

  41. Bernie
    Posted Dec 13, 2007 at 8:45 PM | Permalink

    From a sheer logic your point is well taken. At the very least assuming that the bark stripping is in response to some form of stress there should be some impact on the growth rate. Perhaps we should wait to hear Hughes’s explanation. It appears a number of his colleagues focus on the impact of stress on trees – so I assume that they have had some input into his assertions. It is all very puzzling.

  42. Bernie
    Posted Dec 13, 2007 at 8:46 PM | Permalink

    Just based on logic and some pretty basic empirical assumptions your point is well taken. At the very least assuming that the bark stripping is in response to some form of stress there should be some impact on the growth rate. Perhaps we should wait to hear Hughes’s explanation. It appears a number of his colleagues focus on the impact of stress on trees – so I assume that they have had some input into his assertions. It is all very puzzling.

  43. Bernie
    Posted Dec 13, 2007 at 9:30 PM | Permalink

    Apologies for repeated message –

  44. Gerald Machnee
    Posted Dec 13, 2007 at 9:42 PM | Permalink

    We should not jump to conclusions. Hughes may have made an error in grammar.

  45. MrPete
    Posted Dec 13, 2007 at 10:08 PM | Permalink

    I’d ignore the Ababneh page(s) at LTRR — more than a year old and she’s no longer there.

    I agree with the above posting that suggests the likelihood she’s been frightened by someone or something. To that end, I would encourage following Steve’s earlier advice: leave the poor lady alone. There are bigger fish to fry.

  46. Gary
    Posted Dec 13, 2007 at 10:17 PM | Permalink

    #40-Pat Frank.
    Here’s something else to think about. Thuja occidentalis found in harsh growing conditions exhibit the same strip-bark habit found in BCPs. It might be inferred that the two species are reacting in the same way (ref 1). Experiments show that there is a relationship between the roots and the shoots called “radial architectural segregation” or “sectored radial architecture” that isolates sections of the tree crown to particular sections of the root system (refs 2, 3). If this is the case, it may be that the moisture and nutrient regimes for particular roots are controlling the growth for different sectors of trunk. It’s as if the plant has divided itself into nearly separate individuals.

    1. “…cores, cross sections, and intact trees of Thuja occidentalis up to 1,032 yr old were examined at sites along the escarpment. Radial growth rates as low as 0.05 mm yr were found, but greater rates were found in the very oldest stems, which were also characterized by extensive cambial dieback and strip-bark growth. Less than 2% of the potential cambial surface was present in some trees. Few specimens over 100 yr old were free of cambial disruption, and most had lost half of their cambium between the ages of 130 and 280 yr. Loss of cambium and morphological distortion appear to be the result of the accumulation of point sources of damage over time; exfoliation of rock from beneath trees is suspected to be the principal stimulus. These growth, developmental, and ecological properties are similar to those described for Pinus longaeva. Both species show constrained growth, cambial mortality, and gross modifications to morphology. In addition, both species show great age only when cambial mortality is extensive.” (P. E. Kelly, E. R. Cook, D. W. Larson. Constrained Growth, Cambial Mortality, and Dendrochronology of Ancient Thuja occidentalis on Cliffs of the Niagara Escarpment: An Eastern Version of Bristlecone Pine? International Journal of Plant Sciences, Vol. 153, No. 1 (Mar., 1992), pp. 117-127.)

    2. “Cliffs of the Niagara Escarpment, southern Ontario, Canada support a presettlement forest of stunted, long-lived, and grossly deformed Thuja occidentalis showing pronounced strip-bark growth. Previous studies had suggested that longevity and axis asymmetry in these trees were both influenced by root-mediated shoot mortality, but no quantitative data to support this hypothesis were available. To understand whether such a process could explain cliff-face tree morphology, whole plants were collected, serial sectioned, and morphologically analyzed. Dates of initiation and death of each root or branch were recorded, and the spatial and chronological relations between root and shoot mortality events were examined. The results showed that particular roots that had died were spatially connected to parts of the shoot system that died on average within 2.2 yr. Shoot mortality never preceded root mortality for roots and shoots that were spatially related to each other. A quantitative relationship was found between the size of the roots and shoots that were connected. These observations could be explained either by a disruption of the upward movement of water in hydraulically separate sections of xylem or by an interruption of the transport of photoassimilate to the roots. In either case, radial architectural segregation may permit the plant to survive in hostile and resource-patchy environments.” (D. W. Larson, U. Matthes-Sears, P. E. Kelly. Cambial Dieback and Partial Shoot Mortality in Cliff-Face Thuja occidentalis: Evidence for Sectored Radial Architecture. International Journal of Plant Sciences, Vol. 154, No. 4 (Dec., 1993), pp. 496-505)

    3. “Thuja occidentalis shows slow growth, strip-bark morphology, and extreme longevity when growing on cliffs of the Niagara Escarpment, Ontario, Canada. To test whether these features could result from sectored radial architecture, dye-infusion experiments were performed on trees in four different age/size categories. Safranin and/or crystal violet were infused for up to 2 wk into young lateral roots. Plants were then harvested, serial-sectioned, and photographed. When two dyes were infused simultaneously, different lateral roots were used. Trees were also infused with dye, wounded, and then resupplied with dye to determine whether hydraulic pathways were altered as predicted by commonly accepted models. The results showed rapid radially sectored dye uptake in trees older than 1 yr. Individual roots tended to supply dye to only part of the crown. Wounding had no influence on the pattern of dye distribution. Stem segments with opposing, overlapping, and adjacent saw cuts had hydraulic conductance values near zero. The results indicate that T. occidentalis has sectored radial architecture that almost completely isolates the hydraulic pathways to particular parts of the stem and crown. The results indicate that the survival, morphology, and longevity of this species are related to the capacity of this species to sector productivity and mortality in resource-patchy environments.” (D. W. Larson, J. Doubt, U. Matthes-Sears. Radially Sectored Hydraulic Pathways in the Xylem of Thuja occidentalis as Revealed by the Use of Dyes. International Journal of Plant Sciences, Vol. 155, No. 5 (Sep., 1994), pp. 569-582.)

  47. bender
    Posted Dec 13, 2007 at 11:50 PM | Permalink

    Gary, read the blog. Steve M knows just about everything there is to know about cedar, with several threads devoted to the topic. In fact it was Jacoby’s Gaspe cedars that MBH98 had inserted four placeholders into in order to ensure that the AD1404 chronology was included in the AD1400 step. Why did he do this extraordinary data manipulation? Because he knew the Gaspe cedars had a HS shape to them and he wanted to ensure that shape came out in the AD1400 reconstruction.

  48. MrPete
    Posted Dec 14, 2007 at 12:01 AM | Permalink

    ‘twould be interesting to have access to those papers. From what I’ve heard via Leslie H, while the connection may be as simple as sectored radial architecture close in to the tree, it gets far more complex and less predictable as you look further out to where actual uptake of nutrients occurs. There’s also a far more complex hormonal interplay between roots and shoots than was previously anticipated.

  49. Geoff Sherrington
    Posted Dec 14, 2007 at 2:07 AM | Permalink

    Re # 46 Gary – strip barks

    You quoted from Larson et al –

    In addition, both species show great age only when cambial mortality is extensive.”

    Is this reverse causality at its best? Should it not be that the species show cambial mortality when at great age (as they approach death)? What proof is offered that reverse causality can be ruled out?

    It is tricky to assume too much about tree growth if you have not had hands-on. I can cut a tree of some genera completely off at the base, a foot above the ground. Many times, it will throw out suckers which can become quite ancient in their own right. Some can seem to fuse with the base to give a circular shape that looks as if it was never disturbed. Now, imagine the ring counting problem if you sampled the stump as opposed to the sucker. Or, for a nasty twist, if you went through the junction with your corer at an angle.

    I have often seen trees decapitated by tropical cyclones growing again next Spring and I have to assume that a fair portion of the trees I see in cyclone areas are not good dendro candidates because the count has started from zero again above a certain height. Cyclone Tracy at Darwin stripped some to much of almost every tree for 50 miles square. Lightning, termites and fires can cause similar effects.

    There is no guarantee that the tree you sample today is the tree that was there 200 years ago. I have seen many historic trees older than 200 years, but dendro does not work on them because of the sucker problem. They are recorded in history books that have survived, especially seen in China. One was there when Kubla Khan came through.

    Trees can adjust metabolism to cope with cambium disturbance. The example of cutting off at the base is a fairly drastic cambium interference. The tree will commonly not die immediately following beheading. Indeed, in ornamental horticulture at least, this is a standard way to prepare for a cleft graft. The cambium will spread over the gap between the bottom stock and the inserted scion, usually in weeks to months, and telecommunicate again. The tree will grow, but with an interrupted history of rings above the graft.

  50. Geoff Sherrington
    Posted Dec 14, 2007 at 2:26 AM | Permalink

    Re # 48 Mr Pete

    I agree. Some hormones and many nutrients can be applied as foliar sprays, with little to do with root involvement. I’d guess that a die sprayed on the outside of leaves might penetrate into twigs, but I have not tried this. Chlorophyll can take this path in the Fall.

    The more I read CA, the more I feel that what is NOT said in climate science papers can be as important as what is admitted.

  51. kim
    Posted Dec 14, 2007 at 5:18 AM | Permalink

    Come back tomorrow!

  52. George M
    Posted Dec 14, 2007 at 7:48 AM | Permalink

    In reviewing earlier parts of this discussion which are in previous threads, I encountered the information that some part of her thesis work was funded by an outside entity (Bristlecone section of a Calif. nature group – sorry I didn’t copy the exact name). Maybe that is part of the release problem, two groups have to sign off on it?

  53. Gary
    Posted Dec 14, 2007 at 8:04 AM | Permalink

    I’ve been reading the blog for over two years. My point isn’t the HS or the Gaspe cedars; it’s the plant biology. Radial sectoring means that shoot cambia of individual bark strips are dependent on the conditions of their supporting roots. This adds another level of uncertainty to the idea of trees as thermometers – one individual is likely recording different conditions in different sectors. Suppose a root is damaged – cambial growth declines until the injury is repaired. Suppose a root grows into a moisture pocket – cambial growth is enhanced. Another root/cambium sector is subject to neither extreme. All three conditions would produce different ring structures and different interpretations of climate.

    #49-Geoff Sherrington.
    Yeah, that sounds backwards. You get the point about plant responses to environmental conditions within the constraints of their genetically controlled limits.

  54. Steve McIntyre
    Posted Dec 14, 2007 at 9:01 AM | Permalink

    #46, 53. At a very early stage in our studies, Ross and I spent a day with Doug Larson and PEte Kelly (who are both at the University of Guelph, as is Ross) and got a good lesson in the radial sectoring issue. One of our Almagre samples was an exposed root, from which we took a core. The root had quite a distinct low-freq pattern from the stems, something that we’ve not attempted to interpret so far. Larson and Kelly very much thought in terms of cliff cedars almost “mining” for nutrients and moisture and occasionally getting into good ore zones. It also seems quite plausible to me that the competition between bristlecones in these dry areas is at the root level rather than the canopy level.

  55. bender
    Posted Dec 14, 2007 at 9:09 AM | Permalink


  56. Posted Dec 14, 2007 at 9:14 AM | Permalink

    Geoff, #49

    In forestry, this is known as coppicing

    It has a long history of usefulness. Many ancient English houses are built of oak from coppiced trees.

  57. bender
    Posted Dec 14, 2007 at 11:10 AM | Permalink

    Evidence of threshold temperatures for xylogenesis in conifers at high altitudes
    Rossi, Sergio; Deslauriers, Annie; Anfodillo, Tommaso; Carraro, Vinicio

    Oecologia, Volume 152, Number 1, May 2007 , pp. 1-12(12)

    Temperature is the most important factor affecting growth at high altitudes. As trees use much of the allocated carbon gained from photosynthesis to produce branches and stems, information on the timing and dynamics of secondary wood growth is crucial to assessing temperature thresholds for xylogenesis. We have carried out histological analyses to determine cambial activity and xylem cell differentiation in conifers growing at the treeline on the eastern Alps in two sites during 2002-2004 with the aim of linking the growth process with temperature and, consequently, of defining thresholds for xylogenesis. Cambial activity occurred from May to July-August and cell differentiation from May-June to September-October. The earliest start of radial enlargement was observed in stone pine in mid-May, while Norway spruce was the last species to begin tracheid differentiation. The duration of wood formation varied from 90 to 137 days, depending on year and site, with no difference between species. Longer durations were observed in trees on the south-facing site because of the earlier onset and later ending of cell production and differentiation. The threshold temperatures at which xylogenesis had a 0.5 probability of being active were calculated by logistic regressions. Xylogenesis was active when the mean daily air temperature was 5.6-8.5°C and mean stem temperature was 7.2-9°C. The similar thresholds among all trees suggested the existence of thermal limits in wood formation that correspond with temperatures of 6-8°C that are supposed to limit growth at the treeline. Different soil temperature thresholds between sites indicated that soil temperature may not be the main factor limiting xylogenesis. This study represents the first attempt to define a threshold through comparative assessment of xylem growth and tissue temperatures in stem meristems at high altitudes.

    • bender
      Posted Sep 25, 2009 at 9:45 PM | Permalink

      Re: bender (#57),
      I wonder if soil temperature was limiting root initiation at Yamal until the mid 1970s. I wonder if that is the process responsible for “divergence” in general. North slopes vs. south slopes. Root initiation is a temperature-thresholded process. A tiny bit of warming could have an amplified effect on root, and therefore radial, growth – by significantly boosting the length of the root growing season. Cold soil is bad.
      I believe I proposed that mechanism for bcps in California here at CA … but before I understood the stripbark process.
      I wonder how much divergence there will be within the Yamal samples. And if there is, if it can be correlated to site aspect. Whatever happened to Martin Wilmking? Didn’t he promise to figure all this out?

      • Steve McIntyre
        Posted Sep 26, 2009 at 8:59 AM | Permalink

        Re: bender (#72),

        bender – cold soil is discussed in some of the early literature on BCP growth (Fritts did a technical report in the 1960s).

  58. Paul Foote
    Posted Dec 14, 2007 at 11:48 AM | Permalink

    Temperature is the most important factor affecting growth at high altitudes.

    Of course, water, nutrients and wind have no effect at all.

    Xylogenesis was active when the mean daily air temperature was 5.6-8.5°C

    Translation: “Growth was was active when it was warm enough to rain.”

    Cambial activity occurred from May to July-August and cell differentiation from May-June to September-October. … The duration of wood formation varied from 90 to 137 days

    And this of course doesn’t correlate at all with the rainy seasons.
    What hogwash.

  59. SteveSadlov
    Posted Dec 14, 2007 at 12:09 PM | Permalink

    E: #38 – “I see nothing!”

  60. SteveSadlov
    Posted Dec 14, 2007 at 12:13 PM | Permalink

    RE: #39 – In the case of Sheep Mountain, I suspect that snow pack survival is the controlling factor. Years where the snow pack does not survive until a certain point in June are lower growth, years where it survives to that point are higher growth. The moisture content of the poorly developed soil in that locale is going to depend, ni most years, on the snow pack factor. Generally speaking, summer convection is unrealiable and generally cranks up too late in the season to greatly influence growth. Prime time is June 15 – Aug 1, so snowpack is going to determine it.

  61. SteveSadlov
    Posted Dec 14, 2007 at 12:18 PM | Permalink

    RE: #46 – Niagara Escarpment is in a Humid Continental Climate with a summer precip max and where, generally, the snow is gone before the soil thaws. Sheep Mountain is in a Mid Latitude Steppe / Desert Climate with some Mediterranean influence, with a winter precip max and in many years, there are late spring snow events after soil thaw has commenced. Apples and oranges.

  62. SteveSadlov
    Posted Dec 14, 2007 at 12:21 PM | Permalink

    RE: #57 – Totally different climate than Sheep Mountain. BCPs are more like cactus than other types of pines / other conifers.

  63. beng
    Posted Dec 14, 2007 at 12:53 PM | Permalink

    RE 40:

    Pat says:
    I find it peculiar that strip bark and whole bark BCP’s should have the same ring width excursions over the same years and growth conditions. This result implies that each BCP maintains a constant growth rate in its remaining cambium, following bark-stripping.

    Pat, from my observations over the decades, there’s little transverse movement of sap/nutrients across bark. IOW, if a circumferential section of tree-bark dies, the roots & branches in-line w/this section also eventially die (couple of yrs).

  64. Pat Frank
    Posted Dec 14, 2007 at 7:35 PM | Permalink

    #63 — beng, thanks. I gathered something of that from the interesting abstracts about radial sectioning that Gary posted (thanks, Gary) in #46. But the question is whether the remaining living circumference continues monotonic ring growth; that is, growth that remains a linear extapolation of the ring width trend prior to the bark stripping.

  65. Geoff Sherrington
    Posted Dec 15, 2007 at 12:13 AM | Permalink

    Re # 64 Pat Frank

    A reason for my mention in # 49 of decapitation then cleft grafting is that there really has to be some radial movement of sap to achieve a final circular pattern the same diameter as the stump, from a graft initially a few mm wide at one point on the cambium circumference. So radial transport cannot be ruled out entirely, though it might not be major. How major and how fast remain unknown to me.

    Re # 53 Steve McI

    Agreed in degree, but photosynthesis in canopies is a valid process and part of the reason for growth. The leaves do play some part in ring width; without them one is on an asymptote to zero. Whether leaves play a controlling part, or whether it is the roots, is one core aspect of the sub-problem. Would be interesting to run the radial sectoring investigation upside down; that is, remove and keep removed part of the canopy and see if, over the years, the rings below the removed part ceased to function. My guess is that they still function, but at reduced activity. A related problem is whether the cambium above the strip-bark wound still functions.

    Simply look at the complexity of the sulphur cycles and pathways and their interactions with other growth controllers. SO2 forms aerosols, might change cloud/light availability, forms acid in soil that can affect availablity of about a dozen nutrients, releases CO2 from carbonate nodules and rocks, can be absorbed by both roots and leaves, could change predator levels and growth hormone effects, soil biology levels, accelerates soil from rock conversion rate and clay content, with attendant ion exchange capacity, water retention properties etc etc.

    As a general observation, some root diseases like armillaria fungus and phytopthora start at a root point, but the tree seems to die overall from tender new leaves inwards. So these pathogens seem to have some ability to avoid radial sectoring.

    The whole subject of tree rings is complex. I claim to know only a little about it, but it seems that each increment of research reported by reconstructionists opens up multiple new lines for inquiry. Another problem is that short studies of plants, like dye penetration, might not relect growth processes over centuries, (as I’m sure you’ll agree is a big problem in reconstructing geology). Until these various complexities are understood, I tend to forget about temperature because that is approaching a structured investigation from an erroneous direction.

    The oft-repeated assertion that high altitude trees under stress to survive and very old, are most suitable for tree ring studies, smacks a bit of selective sampling. Where is the boundary between OK and not OK? Is it sharp or gradual?

  66. Geoff Sherrington
    Posted Dec 15, 2007 at 12:17 AM | Permalink

    Re # 53 Gary

    Getting the logic backwards. After a windy storm in Tasmania, a TV crew interviewed a local farmer, saying how strong the winds were. “Yep”, he commented. “Blew down a lot of trees that hadn’t been blown down before”.

  67. MrPete
    Posted Dec 15, 2007 at 4:36 AM | Permalink

    At a very basic level, the roots vs leaves question is obvious:
    * When our fruit trees’ roots are attacked by moles, the trees die – 100% of the time, and rather quickly.
    * When devastating hail strips the leaves from our trees***, they recover, 100% of the time. Can’t say the same for the tomato crop, but hey, this is Colorado 😀
    ***We don’t have earthquakes, hurricanes or tornados here (we just call it “high wind warning: 60mph winds with 90+mph gusts.”) But we get lightning and we get hail. Several years ago, sixteen inches of hail fell in one storm. Our greenhouse survived. A friend’s mini-observatory did not. This year we had a storm with a LOT of golf-ball hail, and quite a few billiard-ball hailstones. Hail like that strips trees of all leaves – “winter” trees in summer, with a strange green “snow” on the ground 😦 … oh, and our truck required replacement or repair of every body panel plus some glass. Fortunately, our new Class IV roof was fine.

  68. KevinUK
    Posted Dec 15, 2007 at 4:50 AM | Permalink

    #22 mosh

    Thanks for that you tube link. I’ve not been paying any attention to that man behind the curtain fo quit esome time now thanks to our resident ‘toto’.


  69. David Holland
    Posted Dec 15, 2007 at 7:52 AM | Permalink

    “Bristlecone pine paleoclimatic model for archeological patterns in the White Mountain of California” (2007)
    Linah Ababneh

    Available online 14 September 2007. Quaternary International

    Archeologists recognize that prehistoric human populations were a significant variable in the operation of past California ecosystem processes, and that a changing environment created constraints and opportunities for prehistoric cultural adaptation. The tie between paleoclimatology and archeology is an evolving interdisciplinary field of study that aims to facilitate a better understanding of prehistoric subsistence-settlement patterns relative to climate variability. A paleoclimatic model based on temperature inferences and precipitation reconstruction from tree-ring widths of bristlecone pine in the White Mountains of California is used to explain the distribution of prehistoric alpine villages in the White Mountains. The paleoclimatic model is visually and statistically compared with calendar ages converted from calibrated 14C dates from the village sites. Results elucidate the role of water availability in the frequency and intensity of settlement, as 88% of the calendar dates associated with 14C dates fall within wet periods (α=0.05, p=0.005). The coincidence of the calendar dates around above average precipitation periods confirms that long-term available resources are contingent on the availability of moisture, the key to life in the desert environment.

  70. Pat Keating
    Posted Dec 15, 2007 at 9:44 AM | Permalink

    67 Mr. Pete

    a LOT of golf-ball hail, and quite a few billiard-ball hailstones. Hail like that strips trees of all leaves – “winter” trees in summer, with a strange green “snow” on the ground

    Yes, we had that once in McLean, VA (close to DC) when I lived there. There were ruts in the ‘green snow’ on the roads from traffic, just like you get with the white stuff. I saved one of the hailstones (bigger than a golf ball) in my freezer but it sublimed away…..

  71. Bill
    Posted Dec 15, 2007 at 12:56 PM | Permalink


    ***We don’t have earthquakes, hurricanes or tornados here (we just call it “high wind warning: 60mph winds with 90+mph gusts.”)

    I know when you say “here” you’re not talking about the plains east of us front-rangers, towns such as Holly. The town was “totalled”.

    My mom grew up (’30’s) in Eastern Colorado towns where tornado shelters were fairly common. The following link, (p. 3) shows tornado counts since 1988, which I believe is the year I watched from my condo window in west Denver as a funnel cloud dropped out of the sky, worked its way down Broadway a short distance (whacked a car dealership I think) then disappeared. Holly event was the first tornadoe-related death in Colorado since 1960, according to article.

    No stories like Dorothy’s, but we get them. Very pedantic of me, but fun in a “twisted” sort of way, to look it up.

    Pete: You must be reading the entries here about root/shoot compartmentalization. The Brunstein article about bristlecone growth characteristics (Steve Mc. cited some time ago, along with the cross-sectional shapes, in regard to coring anomalies) has some interesting observation about “sediment damming” on upslope sides of bc pines, and its relationship to the root formation (the latter causing as well as responding to the former), soil moisture retention, branch and strip-bark growth formation, undercutting of downslope root anchorage, long-diamer bearing, etc. (too much to repeat here). One thing I took away was that about 26 of 33 old trees (aged 1,000 – 2,000 years), or nearly 80%, had the bark strip on the upslope side.

    It’s a big PDF file (23 MB), but his pictures/sketches (really artwork) are superb. Very slow-going, very un-Acrobat-ic, which may have discouraged some readers.

  72. bender
    Posted Sep 25, 2009 at 9:51 PM | Permalink

    20th century divergence is such a ubiquious pattern. Occam’s razor says it has to be a universal mechanism. Alipne soil is universally cold. The Ababneh update trends up. Just not as far up as Graybill. Polar Urals trends up. Just not as far up as Yamal. Other Yamal chronologies trend up too.

    • Steve McIntyre
      Posted Sep 26, 2009 at 7:06 AM | Permalink

      Re: bender (#73),

      bender, Esper discusses the Divergence Problem in a recent Climatic Change commentary – as an adverse comment on an article by Craig Loehle on reconstructing temperature from upside-down U-curves. An exchange that I’ll try to report on.

      • Dave Dardinger
        Posted Sep 26, 2009 at 8:09 AM | Permalink

        Re: Steve McIntyre (#74),

        BTW, how is Craig’s surname pronounced, Lo -el le, Lo -ly, Lo -el, Lole or some other way?

  73. bender
    Posted Sep 26, 2009 at 5:04 PM | Permalink

    Steve, are there enough samples in the raw Yamal data, and enough metadata – to check if the slope of the uptick varies according to aspect? I bet the metadata don’t even include aspect. But depending on the precision on that lat/lons, you could look it up from a digital elevation model.

    • Steve McIntyre
      Posted Sep 26, 2009 at 6:54 PM | Permalink

      Re: bender (#79),

      metadata – you kidding me??? For example, the Avam-Taymyr metadata doesn’t even show which site a sample is from – and the sites are 400 km apart. I’ve started a new Yamal thread.

  74. Steve McIntyre
    Posted Nov 17, 2009 at 10:57 AM | Permalink

    ping – the new Salzer, Hughes paper appears to be a writeup of Hughes’ 2007 AGU presentation.

    I’ve also received some correspondence in the past year from a scientist who knows Linah Ababneh and vouches for her in very positive terms.

    • Posted Nov 17, 2009 at 2:31 PM | Permalink

      Re: Steve McIntyre (#81),

      Any sign of any supporting data? Or is Hughes still working on them?

      Steve: Supporting data has been placed online in commendable fashion. I’m working on an analysis.

      • RomanM
        Posted Nov 17, 2009 at 6:41 PM | Permalink

        Re: John A (#83),

        Steve: Supporting data has been placed online in commendable fashion. I’m working on an analysis.

        I have been trying to read the tree ring files using both your progamme (make.rwl) and the dplR library with hiccups in both methods on some of the rwl files. I am writing up my own script.

        I also noticed a lack of uniformity in the files with regard to end-of-data values and number of digits in the values. Are there soem glitches in the files as well or is it just me?

        Steve: Roman, I’ll post up collations in 2 minutes to save you time on clerical work. I had to do a couple of tweaks to make.rwl. You need to use the widths option widths=c(7,5,rep(6,10)). Also I had to specify the first item as a factor using colClasses. Ill also upload modified make.rwl in two minutes. QCed collations in my usual style will be in CA/data/salzer.

        • RomanM
          Posted Nov 17, 2009 at 7:18 PM | Permalink

          Re: RomanM (#84),

          I got my script working. There is a small glitch with mwa_rwl, and only in one spot I haven’t identified yet.

          My method uses readLines and then I parse the lines to extract the data. It is quite fast in its execution.

        • RomanM
          Posted Nov 17, 2009 at 8:32 PM | Permalink

          Re: RomanM (#85),

          Found the glitch in mwa_rwl. It wasn’t in my R script.

          There are two consecutive cores labelled MWA069x in the file. They start in the same year, but consist of two different sequences of ring widths and end at different times.

      • bender
        Posted Nov 17, 2009 at 8:28 PM | Permalink

        Re: John A (#83),

        Supporting data has been placed online in commendable fashion. I’m working on an analysis

        See? Anyone can do it!

One Trackback

  1. By Kill It With Fire | Watts Up With That? on May 30, 2011 at 2:17 PM

    […] failed replication, as shown by the Ababneh Thesis. Not only that, but one of the authors of M2008 (Malcolm Hughes) had to have known that, because he was on her PhD committee … so the M2008 study used […]

%d bloggers like this: