Graybill and Funkhouser [1993] on Bristlecones

I have recently located a copy of Graybill and Funkhouser [1993], Dendroclimatic Reconstructions during the past millennium in the southern Sierra Nevada and Owens Valley, California, which has been very hard to find. This appears to be Graybill’s last publication before he died. A detailed excerpt follows. Some key quotes:

Unfortunately the chronologies from the White Mountains do not correlate well with the southern California divisional climate data. Additional trials using averages and principal component scores of various subsets of the tree ring data with the divisional data were also unsuccessful…

Growth responses of residual chronologies from Sheep and Campito Mountains to spring and summer temperatures of the current prior year are either negative or not significant…

These climatic responses at upper treeline may be of limited value or biological significance in making inferences about climate-tree growth relationships if the hypothesized fertilization effects of anthropogenically induced CO2 processes resulting from the Industrial Revolution [Graybill and Idso 1993] are real. It is possible that the climatic processes that dendrochronologists most frequently investigate e.g. precipitation, temperature and drought) are being masked or overridden by the extraordinary growth increases over the last century or more….

Based on the analyses undertaken here, it can be inferred that available instrumental records do not provide an adequate sample of the trends or variability in climate required for accurate planning purposes or for evaluating the nature of climatic processes.

The article is cited in Hughes and Diaz [1994], where it is attributed to a volume edited by R. Lavendra, Proceedings of the Southern California Climate Symposium: Trends and Extremes of the Past 2000 Years. I’ve tried on several occasions without success to locate this in a university library. Through the miracle of google, I located this article (in a 1999 publication edited by Rose, M. and P. Wigand) as Technical Report 13, Natural History Museum of Los Angeles County (which was at the Royal Ontario Museum). Hughes is thanked for reviewing various drafts.

Graybill and Funkhouser present chronologies from the key bristlecone and foxtail sites: Cirque Peak, Sheep Mountain, Campito Mountain, Flower Lake, Timber Gap Upper, Timber Gap Lower, Onion Valley, Methuselah Walk and three sequoia sites mentioned in Hughes and Brown (Mountain Home, Giant Forest and Camp Six). All sites are within 30 miles or so of 37N; 118030’W.

The bristlecone site chronologies were developed by first fitting either a straight-line or negative exponential function to each ring width series to supposedly remove the non-stationarity due to normal growth processes. (There are some odd numerical processing issues in how this is done in ARSTAN, which I will visit some time.) The sequoia chronologies were done with a 40-year spline because of fire impacts. (This will remove any low-frequency information.)

Graybill and Funkhouser report:

“The five uppermost chronologies [CPF, SHP, CAM, FLF, TGU — the core of the MBH98 PC1] also share a dramatic and historically unique growth increase over the last 120 to 150 years, Similar trends have been observed in other arid site upper treeline bristlecone and limber pine chronologies from the Great Basin and Colorado [Graybill, 1987; Graybill and Idso, 1993]. Lamarche et al [1984] hypothesized that this phenomenon was the result of anthropogenic increases in atmospheric carbon dioxide coincidental with the onset of the industrial revolution. Increased availability of this resource (normally limited at theses high-altitude sites) for photosynthesis and the attendant increases in water use efficiency would both be beneficial for tree growth.

This trend does not appear in three upper treeline foxtail pine chronologies and one lodgepole pine chronology from the Sierra Nevada developed by Graumlich [1991]. Since our standardization methods were essentially identical, this cannot be considered a source of the differences. The growth forms of the trees sampled in the two studied were different and this likely is important.

Statistically significant autocorrelation was found in all the bristlecone and foxtail pine chronologies…An issue now considered involves possible sources of the autocorrelation in tree ring index chronologies. In many genera, current growth depends on food produced in one or more preceding years [Fritts, 1876]…

Unfortunately the chronologies from the White Mountains do not correlate well with the southern California divisional climate data. Additional trials using averages and principal component scores of various subsets of the tree ring data with the divisional data were also unsuccessful. ..
The strongest “local” climate signals in the White Mountains chronologies are related to moisture or moisture availability as affected by temperature (e.g. the strong negative response to spring temperatures.) Similar conclusions were reached in an intensive three-year study of bristlecone growth at various elevations in the White Mountains [Fritts, 1969]. This is not unexpected since the stands are on xeric sites bereft of substantial soils. Annual precipitation in the White Mountains averages only 16.6 inches…

Growth responses of residual chronologies from Sheep and Campito Mountains to spring and summer temperatures of the current prior year are either negative or not significant. The winter and spring temperature responses found at Campito Mountain are limited (~13% common variance), while the spring precipitation and drought responses are only marginally stronger (~18% common variance). These results are equivocal in terms of Lamarche’s [1974b] hypothesis that low frequency trends in the Campito Mountain chronology might provide a record of warm-season variation.

These climatic responses at upper treeline may be of limited value or biological significance in making inferences about climate-tree growth relationships if the hypothesized fertilization effects of anthropogenically induced CO2 processes resulting from the Industrial Revolution [Graybill and Idso 1993] are real. It is possible that the climatic processes that dendrochronologists most frequently investigate e.g. precipitation, temperature and drought) are being masked or overridden by the extraordinary growth increases over the last century or more. This is not a problem with chronologies developed from stands at lower elevations since they do not evidence recent growth trends that would suggest or defensibly allow for hypotheses regarding CO2 fertilization of growth. ..

Review of the climatic sensitivity (1) of each chronology; (2) of various chronology averages and (3) various PCA scores led to the conclusion that four potentially useful climate reconstructions were possible… The two strongest dendroclimatic relationships are between an average of the TGL foxtail pine and sequoia residual series and total November to April precipitation at (1) Grant Grove (r=0.72) and (2) California division 5 (r=0.71)…Dynamic difference in precipitation totals and variability across the short lateral distance from the Sierras to the arid Owens Valley. Mean annual precipitation at Grant Grove is 42.3 inches, while at Independence, the mean value is only 5.2 inches. A twenty fold difference in variance exists between the two records…An average of two foxtail pine residual chronologies (FLF and NVA) from the eastern slope of the Sierras is used to reconstruct annual (July to June) precipitation at Independence. The dendroclimatic relationship is not as strong as those described above (r=0.63).

Cirque Peak (CPF) is the only residual series with a temperature series adequate for reconstruction. There is a moderately positive response with average fall-winter temperature at Grant Grove (r=0.58). Climatic sensitivity to temperature during this part of the year probably reflects a critical need for lengthy periods of photosynthesis. Scuderi [1987] reached a similar conclusion using a chronology that he developed from the same stand. A more recent study [Scuderi 1993] reconstructs temperature for a different period (June to January)). Because our chronology (1) does not evidence strong relationships with any of the divisional climate data and (2) the reconstruction based on it is statistically weaker than the others we develop here, these data may be of limited usefulness in defining climate-tree growth models or formulating generalized statements about climate variability on a regional or greater scale.

Based on the analyses undertaken here, it can be inferred that available instrumental records do not provide an adequate sample of the trends or variability in climate required for accurate planning purposes or for evaluating the nature of climatic processes.

The caution placed on Lamarche [1974b] here is important as this very old publication is still cited as authority for use of bristlecone site chronologies as a "temperature proxy". It would have been nice to have had this reference in hand for our earlier articles, but it certainly re-inforces Graybill and Idso [1993] and the points made in MM05a and MM05b.

137 Comments

  1. TCO
    Posted Oct 3, 2005 at 2:44 PM | Permalink

    Interesting. Like to see how you move this forward into a publication.

    What about Funkhill or Idso? Are they alive? Have you corresponded?

  2. JerryB
    Posted Oct 3, 2005 at 2:57 PM | Permalink

    Idso hosts a website at http://www.co2science.org/scripts/Template/MainPage.jsp?MerchantCode=CO2ScienceB2C&Page=Index

  3. TCO
    Posted Oct 3, 2005 at 3:36 PM | Permalink

    Not to get all Dano or anything, but those guys seem a bit strangely rinky-dink. If we are going to suck secret money off of the oil companies, you’d think we could come up with something better looking than four Idsos with inflated self-conferred titles…

  4. Steve McIntyre
    Posted Oct 3, 2005 at 3:55 PM | Permalink

    I’ve corresponded with Idso. He was flabbergasted to learn that his series were the heart of the MBH98 reconstruction.

    I’ve requested some information from Funkhouser and never got a reply. He’s an associate of Hughes and I’m blacklisted.

  5. Willis Eschenbach
    Posted Oct 3, 2005 at 4:58 PM | Permalink

    TCO, at some point you are going to have to stop the ad hominem attacks (#3, “rinky-dink”, “inflated, self-conferred titles”) and discuss actual science … but I’m not holding my breath.

    If you truly don’t want to “get all Dan0 and everything” as you claim, then please …

    don’t.

    We don’t need it, and neither does your reputation. I welcome your comments (if you have some) on the Idso’s science. Your ad-hominem comments, on the other hand, only reduce the chances that someone will actually listen to you when you get around to discussing the science, and not the people doing the science.

    Science, you see, doesn’t care what titles the Idsos want to give themselves, and neither should we. Science is concerned with replicability and verifiability, not reputability. If a man says the earth is flat, it doesn’t matter whether his title is self-conferred or he is a PhD from MIT — he’s still wrong.

    And if a man says the earth is an oblate spheroid, it doesn’t matter if he styles himself “Emperor Norton, Prince of Galactic Climate Change” — he’s still right.

  6. TCO
    Posted Oct 3, 2005 at 6:27 PM | Permalink

    We don’t need no stinkin’ “stopping with the ad hominems”! And you got to admit that site is kinda weak. I mean if Steve is going to harsh on the geographers for not being swinging dick math jocks, then he needs to go after Idso. That dude is a geographer and “we”* have beaten on the weak-ass geographers and other tree-hugger lite types.

    And* form a devilish Stalin-Hitler pact with Mann. After all the dude is a physics jock (those guys always try to spread out and take over other people’s feilds…that could be part of the problem actually…seque to the Crooked Timber issue with physicists in social research…)

    *taking a little license here…hehe…
    **I’m deliberately stinkin’ starting a paragraph with “and”.

  7. JerryB
    Posted Oct 3, 2005 at 7:31 PM | Permalink

    TCO,

    Just a brief hint: your way beyond ditzy proclivities are showing.

    You might try reading a few hundred of the Idsos’ thousands of summaries, and get some idea of just how idiotic your statement “And you got to admit that site is kinda weak.” seems before you choose further to exhibit your apparent cluelessness of their work.

  8. TCO
    Posted Oct 3, 2005 at 8:48 PM | Permalink

    I did. It just seemed sorta lite. No bandpass filters. No grass plots waving their tumescent alpha trees.

  9. bender
    Posted Sep 22, 2006 at 10:20 AM | Permalink

    BCP correlations with spring temperature are negative??!! What the … ??? Ring width is larger when temperatures are colder?

    Indeed, Graybill & Idso (1993) Table 2 report the same thing, for both full-bark (r=-0.52) and strip-bark (r=-0.41) samples, compared against spring temperature. (The positive response is to winter tempearture, is far weaker, and is only in the full-bark series!)

    Moreover, Graybill & Idso (1993) Fig. 6, p. 89 say that the full- & strip-bark “chronologies are indistinguishable until about 1870″ after which, they speculate, the “full-bark trees have a larger, more active root system than do the strip-barks, which would provide a somewhat greater sink for new carbon”.

    Putting it together. Perhaps there was an 1870 spring thaw-freeze that killed off large parts of the strip bark root systems, thus making them the stem-carbon sinks that they appear to be? [Note, now, that all previous posts on root dieback (maples, birches) due to frost all of a sudden come into play.]

    This is so implausible and yet so parsimonious that it’s simply got to be checked out.

  10. bender
    Posted Sep 22, 2006 at 11:36 AM | Permalink

    Whitebark pine is another high-elevation pine species in the Rocky Mountains. Has anyone looked at its responses to temperature?

  11. jae
    Posted Sep 22, 2006 at 11:49 AM | Permalink

    bender: here’s one.

  12. jae
    Posted Sep 22, 2006 at 12:04 PM | Permalink

    Spam filter got me. I’ll try again. Here’s a whitebark pine study. It doesn’t fit well with the Team’s proxies, though.

  13. bender
    Posted Sep 22, 2006 at 12:18 PM | Permalink

    Excellent – check out that Fig. 5. Looks like this Biondi character seems to know what he’s doing …

  14. jae
    Posted Sep 22, 2006 at 12:40 PM | Permalink

    Yes. Maybe he actually found some trees that preserved a temperature record.

  15. bender
    Posted Sep 22, 2006 at 12:48 PM | Permalink

    Biondi et al. (1999) point out (first sentence of section 3) that the high-elevation whitebark pines are positive responders, while the lower-slope Douglas-fir are negative reponders. This illustrates TCO’s “ideal”. Most importantly, the postive responder – whitebark pine – does not yield a HS temperature reconstruction. Not even close.

    Has Steve M read this? I guess so – this is probably the paper he referred to yesterday.

    This is brutal.

  16. bender
    Posted Sep 22, 2006 at 1:08 PM | Permalink

    Wondering if there are any whitebark pine studies from areas colder than Idaho. Here’s one on whitebark pines in California. Note the HS in Figure 2. But if these pines are responding to rising temperature, then why are not the pines in colder Idaho responding even more strongly?

    We need some whitebark pine data from Canada. Because at the moment it appears this genus is responding to some regionalized southwestern US phenomonon, not GW.

  17. Dano
    Posted Sep 22, 2006 at 1:48 PM | Permalink

    15:

    ‘bandwidth’ bender wrote:

    Because at the moment it appears this genus is responding to some regionalized southwestern US phenomonon, not GW.

    ‘bandwidth’ should take some forestry classes. Why? He’d then be familiar with the references cited. Why is this important? He’d know something about the discipline. Why is this important? He could tell something about the study. Why is this important? He’d know then that the italicized is incorrect. Why is the italicized incorrect? The authors studied a small area in a particular climate classification, which is not in the ‘southwestern US’. Why is this important? Looking at the discussion on pg 198, they discuss factors in other areas. Why is this important? This is a study that found similar results as those conducted in other areas, and discuss these findings despite at least two possible confounding conditions that affects krummholz growth and vertical stem development in that particular area. Conditions that likely may be different in Idaho (depending upon site selection and soil moisture, as Idaho is generally a continental climate and not maritime) and less affected by one of the factors discussed.

    Those wishing to…er…audit the classifications need ArcMap and the data are here in .e00 format.

    HTH,

    D

    Steve: Dano, yourad hom schtick is both tiresome and incorrect. bender has taken more than a few forestry classes and I dare say that his formal qualifications in this area exceed your own.

  18. bender
    Posted Sep 22, 2006 at 2:02 PM | Permalink

    Hmmm, I would have thought California and Idaho alpine sites would have an alpine microclimate, which might be as important as the maritime or continental macroclimate.

    There are lots of gaps in my reading, no question. And yes, auditing requires bandwidth. Thanks for the ad hom attack and the wasted bandwidth as a useful contrast to my approach.

  19. bender
    Posted Sep 22, 2006 at 2:05 PM | Permalink

    Dano, educate poor bender with a bandwidth-wasting “cite dump” of everything you’ve got on BCP and WBP ecophysiology & dendrochronology? Thx.

  20. bender
    Posted Sep 22, 2006 at 2:24 PM | Permalink

    Thanks for pointing me to p. 198 of that paper, Dano. You’re a real time-saver. Although I wish you’d copied and pasted the first sentence of the final paragraph from the conclusion:

    The existence of significant interactions between ecological traits and climate variables that result in abrupt changes and reversible responses cautions against general projections of global change that assume simple, linear, ecological responses.

    This will be news to some readers at CA. But of course not to any person who’s ever taken a forestry course. All foresters are instructed as freshmen to read that line on that page in that paper in that literature.

  21. Steve McIntyre
    Posted Sep 22, 2006 at 2:34 PM | Permalink

    #15. bender – since you mention Millar et al 2004, don’t forget the estimable study Millar et al 2006, discussed here http://www.climateaudit.org/?p=585 and an earlier preview noted up here http://www.climateaudit.org/?p=104. Millar concluded that the MWP in California was substantially warmer than at present – a conclusion at odds with the BCP ring width chronologies.

  22. bender
    Posted Sep 22, 2006 at 2:51 PM | Permalink

    As Dano correctly points out, Steve, these papers are all news to me. I ain’t no BCP ecologist. I’ll check them out before saying anything more on the subject.

  23. jae
    Posted Sep 22, 2006 at 2:54 PM | Permalink

    Here’s a dissertation abstract regarding temperature/growth relationships in whitebark pine (and larch) in Idaho. Looks like some divergence in the last years, though….

  24. Dano
    Posted Sep 22, 2006 at 2:55 PM | Permalink

    17:

    Continental snow often contains less moisture than maritime snow; that’s one of the minor reasons why there is a ‘continental’ and ‘maritime’ classification [airmass modification by land is the big reason for the difference]. Your big snow depths with high water content are most often in the Cascades and Sierra, not Selkirks, Sawtooths, Rockies (except, of course, in drought years an’ stuff). This difference has lots of contrasting polysyllabic effects like your big words in 18, and results in scrutiny at generalized, sweeping conclusions like what I italicized.

    ———-

    19:

    Lets just note that you made a conclusion without reading the paper.

    Anyway, the bolded line isn’t news to anyone who manages resources, or studies anything in the natural sciences or ecology, or even to freshmen biology students.

    But it’s good you read the lines. Now, read the lines about the PDO and interannual variability during particular periods next, and reflect on what that might do to individual tree response above and below snowline when the variability period changes (and the snowline period changes, and what that means to meadow invasion in this context). That’s why you want to have some botany and ecology classes, as these classes help you contextualize growth with population dynamics and (wrt dendro) why sampling choice matters. And the spatial response of PDO will be different. and. and.

    ———-

    Generally, for all:

    Here’s the definition of ad hom. Folks maluse ad hom a lot as a tactic.

    Pointing out a conclusion was made because one lacks knowledge isn’t ad hom.

    See, ad hom would be something like: “your argument is wrong because you expend much bandwidth”. Not ad hom is something like: “your argument is wrong because of x, y, z, and by the way, you expend much bandwidth”.

    See, I pointed out you didn’t understand what the paper was doing.

    That’s not ad hom, despite your wish for it to be so.

    HTH,

    D

  25. bender
    Posted Sep 22, 2006 at 3:03 PM | Permalink

    Dano,
    1. Thx for the defn.
    2. Do you have that BCP/WBP cite dump ready yet?
    3. Was that Constance Millar 2005 manuscript referred to by Steve M ever published?

  26. bender
    Posted Sep 22, 2006 at 3:11 PM | Permalink

    Re #24 Beat Dano to it:

    Millar, C.I., J.C. King, R.D. Westfall, H.A. Alden, and D.L. Delany. 2006. Late Holocene forest dynamics, volcanism, and climate change at Whitewing Mountain and San Joaquin Ridge, Mono County, Sierra Nevada, USA. Quaternary Research. In Press.

    I must be “trying too hard” again.

  27. Dano
    Posted Sep 22, 2006 at 3:13 PM | Permalink

    20:

    Aside from the spp. being different, the aspects and microclimates as described by Millar are different as well. Nonetheless, why would it be odd that certain sites might be warmer? Sites that are, as described by Millar, different than the White Mt BCPs?

    You, Steve, could ascertain your implicit assertion by going out and coring something after you formulate a hypothesis. Let us know when you post up your study plan.

    Best,

    D

  28. Steve McIntyre
    Posted Sep 22, 2006 at 3:19 PM | Permalink

    I have a copy of thed Biondi et al 1999 pdf online here . This is co-authored by Hughes. We cited it in our NAS presentation in part because Biondi was one of the NAS panelists.

    Dano – there are lots of tree rings already cored. They don’t need my personal attention pleasant as the enterprise would undoubtedly be. Maybe you could inspire Hughes to archived his 2002 results from Sheep Mountain – on your hypothesis, this ring widths should be wide beyond imagination.

  29. bender
    Posted Sep 22, 2006 at 3:22 PM | Permalink

    Millar et al. 2006 now In Press

    Late Holocene forest dynamics, volcanism, and climate change at Whitewing Mountain and San Joaquin Ridge, Mono County, Sierra Nevada, CA, USA

    Constance I. Millar, John C. King, Robert D. Westfall, Harry A. Alden, Diane L. Delany

    Abstract
    Deadwood tree stems scattered above treeline on tephra-covered slopes of Whitewing Mtn (3051 m) and San Joaquin Ridge (3122 m) show evidence of being killed in an eruption from adjacent Glass Creek Vent, Inyo Craters. Using tree-ring methods, we dated deadwood to AD 815–1350 and infer from death dates that the eruption occurred in late summer AD 1350. Based on wood anatomy, we identified deadwood species as Pinus albicaulis, P. monticola, P. lambertiana, P. contorta, P. jeffreyi, and Tsuga mertensiana. Only P. albicaulis grows at these elevations currently; P. lambertiana is not locally native. Using contemporary distributions of the species, we modeled paleoclimate during the time of sympatry to be significantly warmer (+3.2°C annual minimum temperature) and slightly drier (àƒ⣃ ‹’€ ’24 mm annual precipitation) than present, resembling values projected for California in the next 70–100 yr.

    Apologies if everyone knew this already.

  30. bender
    Posted Sep 22, 2006 at 3:30 PM | Permalink

    Re #27
    I could not open that Biondi pdf, Steve M. You sure it’s ok? (Maybe it’s me.) Helpful Dano to the rescue?

  31. jae
    Posted Sep 22, 2006 at 3:53 PM | Permalink

    Another abstract from some publication (unlisted) pertaining to Idaho.

    A composite whitebark pine tree-ring chronology from 24 trees from four sites was used to develop a 1028-year long reconstruction of spring temperature for the Sawtooth-Salmon River region of central Idaho. The chronology was calibrated against Ketchum and New Meadows, Idaho US Historical Stations, April-May average monthly temperature using half-sample calibration-verification tests for the period that contained historic climate data, 1909-1992. The chronology accounted for 41% of the variability in the climatic data and successfully simulated medium to high frequency trends. A 19th century cold period coincides with the “Little Ice Age.” Neither the instrumental nor the proxy temperature records show evidence of warming in the 20th century.

    Note the half-sample calibration-verification method. This looks more reasonable to me than the normal calibration/verification split.

    It is looking like AGW is bypassing Idaho. Some people think Idaho is out of this world; maybe so.

  32. bender
    Posted Sep 22, 2006 at 3:59 PM | Permalink

    Dano, since you know that Millar et al. (2004) paper so intimately, could you tell me why, in Fig. 2, only site b is a post-1940 “positive responder”? Because that site dominates the “All Sites” in panel g.

    Or do you think that looking this closely at whitebark pine is a waste of time?

  33. jae
    Posted Sep 22, 2006 at 4:03 PM | Permalink

    Here’s a “poster-session” type presentation by Millar, but couldn’t find a date. Very strong evidence for a very warm MWP and cold LIA in the Sierras.

  34. bender
    Posted Sep 22, 2006 at 4:21 PM | Permalink

    Thanks for all the reading materials you can procure, jae.

    It’s funny how Dano instantly got so visibly upset about this whitebark pine (=WBP) line of inquiry. Funny because he’s half-right: my original question about whitebark pine was very innocent. I had no idea people had worked on this species. No idea the results contradict the BCP strip-bark work. Where there’s smoke …

    I thought the cool Dano character existed to gather and collate information on topics. But when asked for that kind of information: nothing. Nothing but flame. Very uncool. Must be touching a nerve.

  35. bender
    Posted Sep 22, 2006 at 4:34 PM | Permalink

    Re #30
    That’s a master’s thesis from Utah State.

  36. Dano
    Posted Sep 22, 2006 at 4:50 PM | Permalink

    33:

    ‘bandwidth':

    the line of inquiry was because your assumptions were wrong. You could have used a paper about any late seral tree in that ecotone and I’d have said something, as you think that throwing numbers around is sufficient to gain understanding of ecosystems. If that were the case, there’d be no need for the discipline of forestry or ecology.

    And if you’re going to use different spp. and different sites with different microclimates to make an assertion about a different species on a different site in a different microclimate (viz. 20 and 33), then the discussion should be moved elsewhere, as it is not, as some are wont to dudgeon here, ‘scientific’.

    Lastly, credit where due: jae has provided two helpful links here today. Somebody should ask the major professor in 30 whether he’s a Bayesian.

    HTH,

    D

  37. bender
    Posted Sep 22, 2006 at 5:06 PM | Permalink

    Apologies, Dano. I don’t have the time or talent to decode your mysterious ramblings. Maybe you’re trying too hard?

    1-Do you agree with the the bolded statement in #28, or not?
    2-Could you give me your cite dump on BCP & WBP ecology?

    Best,

  38. jae
    Posted Sep 22, 2006 at 5:10 PM | Permalink

    Here’s a literature review concerning growth rates of subalpine conifers. CO2 fertilization, nitrogen deposition, etc. are discussed. One very interesting possible explanation for the recent rapid growth of bcps is the amount of snowpack.

  39. jae
    Posted Sep 22, 2006 at 5:11 PM | Permalink

    Must be posting too much. Karma got me again. Once more:
    Here’s a literature review concerning growth rates of subalpine conifers. CO2 fertilization, nitrogen deposition, etc. are discussed. One very interesting possible explanation for the recent rapid growth of bcps is the amount of snowpack.

  40. bender
    Posted Sep 22, 2006 at 5:20 PM | Permalink

    Thanks again, jae. (Just as I thanked you in #33 – which, Dano will note, comes before #35.)

  41. TCO
    Posted Sep 22, 2006 at 5:24 PM | Permalink

    Is there a good review paper on all aspects of bcp responses? Review of evidence pro and con rather then Stevian-biased snippet selections? Is it imagined to be different then other tree species?

  42. bender
    Posted Sep 22, 2006 at 5:25 PM | Permalink

    That’s what I’ve been asking for already, TCO. Dano’s the source. Ask him. He seems unresponsive to me.

  43. jae
    Posted Sep 22, 2006 at 5:56 PM | Permalink

    39: No, I don’t think that ALL bristlecone’s are considered “different” from other species. The problem seems to be limited to just the Sierra bcps, which I see as different for the following reasons:

    1. They grow in very arid environments (with cactus), and therefore would be expected to be moisture limited more than temperature limited.
    2. The experts in the field consider their recent growth rate increase to be odd (“different”) and not temperature related (see #37).
    3. Their presence or absence in reconstructions makes way too much difference. A reconstruction involving numerous chronologies should not hang on just one of them. The outliers should not exert the dominant effect.
    4. It appears to me that many, if not the majority, of tree species (including bcps at other locations) at high altitude/latitude are showing the opposite effect (the divergence problem)

  44. bender
    Posted Sep 22, 2006 at 6:07 PM | Permalink

    jae, it’s the authoritative review I believe TCO is after, not opinion.

  45. Steve McIntyre
    Posted Sep 22, 2006 at 6:08 PM | Permalink

    #37. Teh reference to PEterson is interesting – if you look at the MBH98 Corrigendum and the list of series that were deleted from the network, you’ll notice the deletion of many Peterson series (the wa* series). Mann’s purported explanation in the Corrigdendum for the exclusion of the various tree ring series is false. NAture was informed of this prior to the Corrigendum but printed the false explanation anyway, (the Corrigendum was not peer reviewed.)

  46. Steve McIntyre
    Posted Sep 22, 2006 at 6:13 PM | Permalink

    I am unaware of any comprehensive review of bristlecone; despite TCO’s snarky comment, in our EE 2005 article, I diligently collected articles on bcps, including some now relatively obscure references in American Midland NAturalist from the 1960s. I would have cited any adverse references and discussed them in order to be comprehensive. I would certainly have discussed any comprehensive review of ecophysiology had I located it.

  47. Steve Sadlov
    Posted Sep 22, 2006 at 6:18 PM | Permalink

    RE: #9 – My experience with the Eastern Sierra and Western Great Basin has been this. During Winter-Spring seaons where there are mostly Cold Fronts coming down from Alberta and Yukon, there is ample snowfall east of the Sierra crest. Where there are mostly Occluded Fronts coming in straight from the West, off the Pacific, the Western Slope gets the snow and east of the crest is left with a deficit. Also, since air masses coming from the West have had their moisture squeezed out by orographics, by the time they get to the Whites there is not much left. Furthermore, besides the latent heat pickup from the orographics, the descent over the cliff of the Eastern Sierra downslope constitutes a mini Chinook. So, perhaps Springs that are on the average colder are the ones where the Cold Fronts are coming down from the North. More snow, better preservation of the snow pack due to colder temps. Better moisture availability from June (from the melting snow, sometimes until August or September) onward after the Pacific High moves north and the annual and expected Summer dry season sets in. Besides snowpack the only moisture during summer here in the SW US is from Monsoon moisture – no way mid latitude systems can break though the Pacific High.

  48. bender
    Posted Sep 22, 2006 at 6:21 PM | Permalink

    Re #44
    That’s what I figured, Steve M. Your literature reviews look pretty thorough and objective to me. Had to ask anyways.

    Say, what do you make of this Graybill & Idso speculation on shoot:root ratios, whereby strip-barks are pouring all their C allocation into shoots, due to lack of roots? That’s what I really want to know. That’s what I wish Dano would comment on. I know he knows something about that topic.

  49. Steve Sadlov
    Posted Sep 22, 2006 at 6:23 PM | Permalink

    RE: #16 – Funny thing, here in California there are not very many native pines in areas influenced strongly by marine air. Coulters are probably some of the ones closest to the ocean. By the time you get to the real pine areas, most of the true marine influence is gone although the moisture content might be higher than certain continental air masses. Of course, some of the cT air that makes it into our interior during the summer is pretty danged moist in its own right. But no snow from that, only rain.

  50. Steve McIntyre
    Posted Sep 22, 2006 at 6:31 PM | Permalink

    Hard to say. Now that I think of it, the NAS panel did not canvas strip-bark forms as a form outside bcps. Kelly and Larson have investigated strip bark in cliff-dwelling cedars and attribute the strip bark form to the killing of parts of the root system by mechanical erosion of the cliffs. In our lengthy discussion with them, (I think that they said) that cliff-dewlling cedars grew very slowly both when they were old and young (and this is one of the reasons for longevity) and that growth pulses in cliff-dwelling cedars tended to occur when they were “mining” little pockets of nutrients that came their way from time to time. The bcp substrate must be very poor in nutrients. In our EE article, we did not hitch our wagons to CO2 fertilization although we noted it as there seemed to be other possibilities of important nonclimatic factors affecting bcp 20th century growth. I am intrigued by the broad pulse of 20th century growth in southwestern trees attributed to the introduction of 19th century sheep by Craig Allan although bcp’s are probably too high for this to be involved.

  51. bender
    Posted Sep 22, 2006 at 6:42 PM | Permalink

    Re #48
    Hmmmm, the Kelly & Larson observation on roots & strip-bark cedars could well be relevant. Did they actually state this in their papers?

  52. TCO
    Posted Sep 22, 2006 at 6:50 PM | Permalink

    I would be blown away if I examined the bcp work and it was really thorough and fair. I’m more familiar with you selecting things that support your viewpoint than with you thoughtfully analyzing a situation objectively.

  53. TCO
    Posted Sep 22, 2006 at 6:51 PM | Permalink

    But in any case, if it’s a good review, why haven’t you submitted it to J. Tree Science or equivalent?

  54. Barney Frank
    Posted Sep 22, 2006 at 6:56 PM | Permalink

    #45,

    My experience living on the west slope of the central Sierra is that high summer temps lead to numerous thunderstorms over the Sierra crest and eastward. Moderate temps do not. Whether this increase in thunderstorms with temp is widespread enough to impact ring width I don’t know.

    #46,

    Just a little observational data. A pine tree (or other) that has had a significant portion of the circumference of its bole killed by fire or scarring will very often show quite large ring width growth for many years after, due presumably to the growth being concentrated in a smaller area or perhaps as an attempt to cover the injury. This occurs with or without root injury. How long it lasts I don’t know, but it seems applicable to strip bark specimens for some period of time.

  55. Dano
    Posted Sep 22, 2006 at 7:04 PM | Permalink

    39:

    No to the review paper, yes to the responses.

    They are very hard to study, as their response time is so slow (growth rate at hi elevations slow – you can go downhill in the Whites and see differences, BTW). You have to tease ring characteristics out by dendrochronology to known events (hence the reason for the tephra paper). One jae link in this thread gives you clues to the work (App D), where they are teasing out the relationships after stand-replacing fires. Not the same as at treeline (as there rarely are stand-replacing fires there), but the difficulty of figgering out water relations and other aspects of competition and adaptation to stress are there. So the thread somewhere that asked about sheep grazing and all that is getting to difficulty in why the Sierra BCPs react differently than, say, other spp just a few miles away or BCPs in CO. That is: how do you show sheep grazing impacts in 1920 vs 1880? (the droughts in the late 1800s drove sheepherders up over the Sierra crest into the east side and brought invasive plants with them, and you can look for pollen clues, I doubt there’s been much work done in that, but there are plenty of grad students at UCB and UCD to go up there and poke around)

    BCPs, simply, have adapted to this odd (for the area) environment thru sheer…’will’, if you will. If you go up there, you’ll see few other plants. They literally grow in rock and talus-type soil (no good temperature regime up there for chemical weathering to soil), and the few plants around are mostly under their shade. BTW, even the G&I hint at the difficulty of teasing out whether the ~~divergence is due to water relations, or as Peterson (below) hints at, how hard it is to get data for a good relationship. Anyway, comparisons of BCPs to subalpine or subalpine-alpine ecotones is very problematic in my mind. Go up there and look hard, then go down to subalpine fir-whitebark pine stands and see the difference.

    41:

    That Peterson is a good linky, jae.

    There should be a 1a. – the climate regime of Sierra BCPs is different than NV, UT, and CO BCPs (little or no summer thunderstorms for moisture).

    Note the Peterson (brief) lit review similarities to the cited refs of the Millar et al. and the Whitewing. That is: these papers all are coming at the issue from different directions, yet use many of the same refs (the Millar is better for ecology IMHO); use the cited refs of a paper to learn more – that’s the best way to expand.

    Anyway, moisture limitations are often related to temps (Sierra being a good exception), which is what Peterson is hinting at with the increased growth, as well as reinforcing what I was saying about climate differences making the Sierra spp different comparatively.

    Best,

    D

  56. TCO
    Posted Sep 22, 2006 at 7:10 PM | Permalink

    Dano, you sound like you’re on Steve’s side…

  57. Barney Frank
    Posted Sep 22, 2006 at 7:12 PM | Permalink

    #53,

    Dano, how different are CA bcp from NV bcp? The White mountains are darn near on the border. In fact they cross it up north don’t they? Does NV really get significantly more summer precip than eastern CA?

  58. jae
    Posted Sep 24, 2006 at 8:59 AM | Permalink

    This National Biological Service article discusses factors affecting growth of the Sierra bcps and concludes:

    “Changes in snowpack duration, which affects length of growing season, are a more likely cause of growth increases. Unfortunately, the long-term relationship of snowpack to tree growth has not been adequately investigated because snowpack data are often difficult to obtain.”

  59. Dano
    Posted Sep 24, 2006 at 12:22 PM | Permalink

    57:

    The wrn NV BCPs are very similar, as their precip/climate regimes are similar. Once you start getting close to UT, you are more likely to pick up summer tstms from monsoonal moisture, and you have a continental climate regime.

    CO BCPs often get summer tstms, but they got snow this week and looking at two Skew-Ts [CO, CA (EDW a little warmer still, yesterday afternoon Salt Lake looks cold, so there’s illustration of your maritime – continental airmass difference] shows me the temps and this means the CO BCPs have likely shut down, but CA are still going (using the 700 mb temps, as currently 700 mb ~3150m].

    When you get differences like this, you usually develop different races within species – e.g. Pondo pines, IIRC, up here don’t interbreed well with Jeff pines, but where the two co-occur (CA), they interbreed better, so the WA Pondo may be a different race with different response to environmental stressors (this is also important when considering where to buy natives for replanting and why CA has so many different forestry nurseries).

    56:

    ‘Side’?!?

    Best,

    D

  60. Steve McIntyre
    Posted Sep 24, 2006 at 1:13 PM | Permalink

    #55. BTW the sheep issue seems quite relevant to me; the interaction of sheep to wood growth has been documented by Craig Allen in many SW US locations and one would need to exclude it for bristlecones/foxtails. There’s evidence of sheep at high altitudes in late 19th/early 20th century – see refs in our E&E articles.

    Here’s one study of pollen in the Sierra Nevadas.

    Journal of Biogeography
    Volume 26 Page 899 – July 1999
    doi:10.1046/j.1365-2699.1999.00330.xVolume 26 Issue 4

    Palynological evidence for 19th century grazing-induced vegetation change in the southern Sierra Nevada, California, U.S.A.Robert A. Dull1
    Summary
    Aim Stratigraphic pollen records are used to assess historic vegetation changes that have transpired in a North American mountain meadow since the introduction of Old World livestock species in the middle 1800s.
    Location Monache Meadows is located on the Kern Plateau in the Sierra Nevada mountain range, California, U.S.A. It is situated along the upper reaches of the South Fork Kern River in the southernmost drainage basin of the Sierra Nevada.
    Methods Short core samples of meadow sediments were extracted from five locations throughout Monache Meadows. These five samples are classified according to topographic position within the meadow — two upper meadow sites and three lower meadow sites. Stratigraphic analyses of fossil pollen from each core were used to assess vegetation composition before the introduction of European livestock (pre1850) and throughout the historic period (1850–present). The historic period geochronology is based on 210Pb dating of selected strata from each core.
    ResultsRiccia was a dominant taxon in the upper meadow before the introduction of grazing; Salix seems to have been more abundant in the lower meadow. Both Riccia and Salix decreased dramatically by c.1900, coeval with marked increases in Artemisia (upper meadow) and Cyperaceae (upper and lower meadows).
    Main conclusions Changes in meadow vegetation occurring during the latter part of the 19th century at Monache Meadows are attributed primarily to the introduction of European livestock (sheep and cattle). Other factors that may have contributed to the observed shifts in composition and dominance include changes in native herbivore populations and decreased fire frequency as a result of 20th century fire suppression policies.

  61. TCO
    Posted Sep 24, 2006 at 6:03 PM | Permalink

    I would feel better about the seriousness of the confounding factors, Steve, if you showed them in comparison to other tree series. I just get a bad feeling that you are picking on the high datapoints to try to bias the overall. Not saying you are for sure. Just worried, you might.

  62. Dano
    Posted Sep 25, 2006 at 10:18 AM | Permalink

    60:

    Note how the…er…Dull paper chose particular sites for its analysis, and two of the five sites were unsuitable for analysis. I hope no one calls it cherry-picking or wants to audit it.

    Anyway, this paper is interesting, but keep in mind this paper is in a different biome than where BCPs are found, and it looks at mountain meadows. So this is a poor paper to roll up to a dry alpine ecosystem. And the author concludes that on that site the actual scenario may be a far more complicated interplay of causative factors [pg 910] than just grazing.

    Best,

    D

  63. bender
    Posted Sep 25, 2006 at 10:46 AM | Permalink

    Dano, any insights on this Graybill & Idso business of root mass and carbon allocation in strip-bark vs normal BCPs?

  64. Dano
    Posted Sep 25, 2006 at 11:35 AM | Permalink

    63:

    That paper’s not in JSTOR nor is that journal in my sub. What’s their passage state?

    Best,

    D

  65. bender
    Posted Sep 25, 2006 at 11:59 AM | Permalink

    Graybill & Idso

  66. Dano
    Posted Sep 25, 2006 at 12:03 PM | Permalink

    65:

    That link bombs Acrobat. I don’t see anywhere on the other thread where someone has fixed it.

    Best,

    D

  67. bender
    Posted Sep 25, 2006 at 12:06 PM | Permalink

    Works for me. Maybe you have the same issue as TCO does with the detection attribution review paper? An older version of reader or something?

  68. Bob K
    Posted Sep 25, 2006 at 12:17 PM | Permalink

    The pdf loads fine for me in acrobat 7.

  69. Dano
    Posted Sep 25, 2006 at 1:25 PM | Permalink

    Nope. Still bombing in IE and Moe-zilla. Updating now & we’ll see how that works.

  70. Steve McIntyre
    Posted Sep 25, 2006 at 1:32 PM | Permalink

    Jeez – Dano – after all your pontificating, you haven’t even read Graybill and Idso 1993. That’s ridiculous.

  71. bender
    Posted Sep 25, 2006 at 2:28 PM | Permalink

    :)

  72. Dano
    Posted Sep 25, 2006 at 2:47 PM | Permalink

    70:

    You’re trying too hard. You have no basis to make that assertion. Someone is asking me a detailed question on the paper and you’d think I’d want it in front of me to answer, wouldn’t you?

    63:

    1. I don’t mind their assertion, but they’d be better supported in making this assertion if they’d have included comparative foliage length differences as well as some root mass index (comparing an early-seral spp to an alpine spp needs justification, in my view). They also don’t really discuss N deposition nor snowpack changes** nor temp changes (if any) – that is: just CO2.

    2. Contra to above assertion in #48, the strip-bark are allocating metabolites to cambial growth vs foliar growth in full-bark. Also, G&I hypothesize that strip-bark form has less root mass, not more (lacking bark makes a tree continually allocate metablolites to trying to make wood). Anyway,

    G&I hypothesize that the CO2 fert is allowing the strip-bark to make wood (as that’s what they are missing) and the full-bark to make foliage, as they don’t need to allocate metabolites to wood (wood both protects and supports structure, bark is part of wood and, up there, protects against wind-driven ice more than anything, I imagine).

    This makes sense, but there are other factors that could cause growth other than just CO2, as I’ve discussed before – N fert., slower snowpack melting, etc (we’ve discussed these before, somewhere). These factors haven’t been ruled out. Further, these strip-bark forms are a subpopulation of the whole, and their responses may or may not be how that species responds.

    There are, of course, a number of studies that have shown CO2 fert. may be a factor in increased cambial growth (and a number that show that as long as N isn’t limited, increased cambial growth will occur).

    Lastly, the last line of their paper, seems to me, to be in there to support their hypothesis, but as we know there are plenty of findings now (see, the G&I was an early paper) to show there is evidence in other spp. that CO2 may be affecting radial growth.

    ==========

    Generally: is there a way to make the software accept longer HTML addies? I commonly can’t link because the address is too long. Or I suspect so, anyway, as the preview function bombs.

    ==========

    Best,

    D

    ** http://www.bioone.org/perlserv/?request=get-abstract&issn=1523-0430&volume=035&issue=03&page=0323

  73. Steve McIntyre
    Posted Sep 25, 2006 at 3:06 PM | Permalink

    On reading G and I, OK.

    For purposes of bristlecones as a valid temperature proxy, it doesn’t matter a whit whether the anomalous growth is due to CO2 fertilization, N fertilization or some other factor – only that it is not been established that it is due to temperature.

    Nobody on the Team has even ventured to defend bristlecones. Their only defence is on the basis of goofy invented rules in which they claim that it is “mathematically correct” to include PC series up to and including bristlecones (proving one more time that Mann is “not a statistician” – nor Ammann nor Reverend Wahl.

    Dano, I don’t see you stepping up to the plate with any evidence that bristlecones should be used either.

  74. bender
    Posted Sep 25, 2006 at 3:25 PM | Permalink

    Re #72:

    G&I hypothesize that strip-bark form has less root mass, not more

    That was my read on it, and Steve M’s as well. Others may have interpreted contra.

    “Since carbon availability as well as other environmental factors affecting tree growth are similar, if not the same, for all trees here, we suggest that the strip-bark forms are incorporating newly fixed carbon primarily into cambial growth. The full-bark forms may also be doing so to some degree, but they are likely using a substantial fraction of their fixed carbon for foliage and reproductive growth. Increased carbon allocation to roots is likely involved in both cases, as below-ground growth is generally stimulated to at least the same degree as above ground growth (Idso et al. 1988). However we suspect that the full-bark trees have a larger, more active root system than do strip-barks, which would provide for a somewhat greater sink for carbon in the former trees as opposed to the latter (Norby et al. 1992).”

    The question I want to ask a qualified BCP ecologist is why the divergence in the two chronologies can be traced back to 1870:

    “The two chronologies are almost distinguishable until about 1870, when the one based on strip-bark trees begins a sustained, low-frequency trend of radial growth iincrease.”

    and whether this divergence could be linked to the supposed root mass difference.

  75. bender
    Posted Sep 25, 2006 at 3:29 PM | Permalink

    Steve M, #30 is still on the books. Am I the only one can’t get the Biondi pdf?

  76. Greg F
    Posted Sep 25, 2006 at 3:30 PM | Permalink

    Generally: is there a way to make the software accept longer HTML addies? I commonly can’t link because the address is too long. Or I suspect so, anyway, as the preview function bombs.

    I assume your using the link button. It puts quotation marks around the web address. If you edit out the quotes it should work.

  77. Steve McIntyre
    Posted Sep 25, 2006 at 3:42 PM | Permalink

    Fixed. The address was Ok; the file was damaged. Re-loaded.

  78. bender
    Posted Sep 25, 2006 at 3:48 PM | Permalink

    Dano, you aren’t side-stepping the question in #10 about whitebark pines are you? I mean, I understand they are a different species than bristlecone pine (thanks for the lesson in taxonomy) and I understand that Idaho is not California (thanks for the lesson in geography) and I understand that they might respond to different things in different ways than BCP (thanks for the lesson in physiology). The question now is what do you make of the bolded statements in the published Millar et al. (2006) paper – the abstract being outlined in #29?

    Do you accept, or reject, the bolded statements in #29:

    Using contemporary distributions of the [various pine] species, we modeled paleoclimate during the time of sympatry [AD815-D1350] to be significantly warmer (+3.2°C annual minimum temperature) and slightly drier (‘ˆ’24 mm annual precipitation) than present

    You clearly know more about SW US pines than I do. Valuing your judgement, I would appreciate hearing your assessment.

  79. TCO
    Posted Sep 25, 2006 at 4:08 PM | Permalink

    73. Steve, are the bcps any worse then tree ring series in general in terms of possibility of confounding, non-temp factors?

  80. Dano
    Posted Sep 25, 2006 at 4:09 PM | Permalink

    73:

    My reading of the literature is that nobody knows, but temp and moisture availability are the strong determinants in radial growth. The G&I is interesting in that it went at it another way (hmmm…sounds familiar…) and it would be nice to see it – not replicated – but reproduced in 2-3 other spp to see if it’s valid.

    That is: no one can say either way right now. Make your idea happen, Steve – show how coring a tree to show the initial calibration concerns can’t be fixed by statistics (even though all kinds of stuff can be shown by running the numbers), and this negates all the other proxies too. With your mining background, you can probably make a borehole happen.

    74:

    The question I want to ask a qualified BCP ecologist is why the divergence in the two chronologies can be traced back to 1870 and whether this divergence could be linked to the supposed root mass difference.

    [I’m not a “qualifed BCP ecologist”, if there is one] I’d say root mass being lower is an indicator of lower “health” of the tree (they are old as h*ll, so who am I to say about health?) and is related to above-ground biomass – but maybe they are strip-bark because of their lower root biomass…hmmm. Conversely, the tree being strip-bark likely has a lower conductivity and thus conducts fewer metabolites, hence the lower root mass. I’d say any advantage that a stressed organism can use to exploit will be used – in this case, let’s say that something changed that changed the snowpack melt, where after 1870 there was a higher water content and it took ~17 more days to melt. That would be an advantage that the tree could use to exploit and make cambium – the full-bark has no need to exploit for cambial growth and so allocates metabolites to reproduction (what I would expect). The next step here is to compare seed numbers between the two to see if indeed the reaction is from a stressed organism – but a confounding factor is that trees under stress often have mast years, but here the BCPs are making cambium (rather than seeds?), so I don’t know what to think, unless this spp reacts differently than other pines due to its adaptative mechanisms (e.g., the environment is so harsh reproduction isn’t a good survival strategy – survival is).

    76:

    I type my HTML. I tried without quotes and that did it. Thank you!

    Best,

    D

  81. bandwidth
    Posted Sep 25, 2006 at 4:16 PM | Permalink

    Thanks. And the Millar et al. 2006 paper – specifically the statement in the abstract?

  82. bandwidth
    Posted Sep 25, 2006 at 4:23 PM | Permalink

    Re #79
    You need to distinguish between strip-bark and full-bark. Which do you mean?

  83. Dano
    Posted Sep 25, 2006 at 4:30 PM | Permalink

    78:

    This site is modifed Mediterranean (maritime) climate and IIRC Med climate only covers less than 8% of terrestrial ecosystems. Of course these trees on this site can have a microclimate warmer at times than others; there have already been some sites that have been identified as such.

    Best,

    D

  84. TCO
    Posted Sep 25, 2006 at 4:48 PM | Permalink

    Give me the answer for each, then.

  85. bandwidth
    Posted Sep 25, 2006 at 4:52 PM | Permalink

    Dano, I’m not trying to put words in your mouth, but are you saying that it is incorrect to assume that what which may be true for one tiny area of California (Millar’s MWP 3.2°C warmer than today) is not necessarily true for the globe?

    And regarding Biondi’s whitebark pine study, you are arguing that Idaho’s relatively dry continental climate makes whitebark pine there a poor temperature proxy?

    If I’ve got that right, then what would you expect in a colder, snowier location, say, in the northern Canadian Rockies? Shouldn’t whitebark pine there be more temperature-limited than in the southern Rockies?

  86. Ken Fritsch
    Posted Sep 25, 2006 at 4:58 PM | Permalink

    re: #73

    Nobody on the Team has even ventured to defend bristlecones. Their only defence is on the basis of goofy invented rules in which they claim that it is “mathematically correct” to include PC series up to and including bristlecones (proving one more time that Mann is “not a statistician” – nor Ammann nor Reverend Wahl.

    Was this where they justified the use of bcp because it improved the RE statistic? If so would not that confirm both their lack of statistical and dendrochronology credentials?

  87. Steve McIntyre
    Posted Sep 25, 2006 at 4:59 PM | Permalink

    Dano, if dendrochronologists don’t have a proper theory of bristlecones, they shouldn’t be used in reconstructions – end of story. They should never have been introduced into reconstructions; you should simply acknowledge this and “move on”.

  88. TCO
    Posted Sep 25, 2006 at 5:22 PM | Permalink

    87. Can you show that the bcps are any worse then tree series in general? If you are going to eject them from the island, then you need to set a criteria and eject all that meet that criteria. Not just get rid of the data that it serves your purposes to get rid of. Any kind of meta-analysis uses inputs with various flaws in it and attempts to get a better answer by combining it. If you removes certain peices and retain others, you need a comparative rationale. If your whole argument collapses to all the tree rings are bad, argue that (not some wolf in sheep’s clothings).

  89. bandwidth
    Posted Sep 25, 2006 at 5:25 PM | Permalink

    What’s curious is that both strip-bark and full-bark exhibit the same high-frequency pattern of fluctuations in ring width. The only difference is the steepness of the slope that is the blade of the hockey stick. But if strip-barks are no good, why are full-barks any better? [Which is what the NAS implies.]

  90. bandwidth
    Posted Sep 25, 2006 at 5:31 PM | Permalink

    TCO, the between-species comparison you keep asking for is irrelevant to some degree. When you have a definable sub-population of trees of a single species that is strange in form and strange in growth response, it is fair to disallow them. That is what the NAS said. If you had strip-bark oaks showing strange growth from full-bark oak, they should be disallowed too. And if a guy were half-man/half-robot I think it would be ok to keep him out of heaven for awhile until his true nature has been properly assessed.

  91. TCO
    Posted Sep 25, 2006 at 5:51 PM | Permalink

    How can you say bcps are strange? Strange implies comparison to others that are not strange.

  92. Steve McIntyre
    Posted Sep 25, 2006 at 5:53 PM | Permalink

    #9. bender, Kelly and LArson in their discussion of cedars say that the killing of parts of the root system is what causes strip bark to form. For their cliff-dwelling cedars, they attribute the killing of parts of the root system to purely mechanical causes – erosion of the cliffs exposing roots and killing them. Bristlecones in many of these areas seem to be on steepish slopes.

  93. bender
    Posted Sep 25, 2006 at 5:58 PM | Permalink

    Steve M, you comment on #9 the mechanism by which strip-bark samples are formed, but how about that wacky dual opposing spring/winter temperature response (in both types!)? I haven’t seen that discussed before. Did I miss it in previous threads?

  94. bender
    Posted Sep 25, 2006 at 6:01 PM | Permalink

    Re #91
    I’m merely parrotting what I’ve been told: the stip-barks are strange. Read G&I93.

  95. TCO
    Posted Sep 25, 2006 at 6:02 PM | Permalink

    92: Is that explanation a hypothesis thrown out or a pretty firm belief?

  96. bender
    Posted Sep 25, 2006 at 6:14 PM | Permalink

    Read G&I93 – the chronologies are widely divergent based on growth form. See Fig 5. (I bet you haven’t even looked at it.)

    Read NAS – it is recommended that no one use strip-bark trees as a temperature proxy. What more can you possibly want?

    Have you ever looked at a cross section of a stripped-bark tree? I bet you haven’t. For whatever physiological reason the growth rings are accelerated as the tree tries to close the strip wound.

  97. TCO
    Posted Sep 25, 2006 at 6:23 PM | Permalink

    1. I responded to figure 5 specifically, very recently.

    2. They also say a lot of stuff that y’all disagree with and have not done thorough review of the issue. What more I want is thorough explication in detail. Not convenient, selective appeal to an authority.

    3. No, that’s what I’m trying to learn about.

  98. bender
    Posted Sep 25, 2006 at 6:43 PM | Permalink

    If you’ve seen the Figure 5 and understood it then why on Earth do you keep asking why bcps are strange? A subgroup of them is strange because they can take on this strange growth form that shows strange growth patterns. You keep asking for the same evidence over and over, apparently because you can’t keep track of the facts the way most of the rest of us can. It was the same way with:
    -PCA
    -calibration & verification
    -what the Mannomatic mines for
    -B&C05
    -definition of sampling error in time-series context

    On any one of these issues you’ll fall back to the position that you were right all along when you can’t even recount what the discussions were about. You’re WRONG every time in any way that matters. Only in the most trivial sense of semantics are you ever right.

    If you are, as you say, trying to learn something, then go ask someone you trust. The people with the answers you seek are clearly not here. Because it’s the same damn question over and over again. Post the question once. Someone will answer when the answer is known.

  99. Barney Frank
    Posted Sep 25, 2006 at 6:50 PM | Permalink

    Apparently the strip of bark is usually on the leeward or uphill, protected side of the tree which considering the habitat makes sense. I doubt very much that mechanical root damage has much to do with BCPs strip barking.

    As an aside, many here probably already know this, but the oldest BCP on earth, at Wheeler peak, was cut down in 1964 after a dendro guy who was coring it broke his corer off in the trunk at 4000 years. USFS gave him permission to whack it off. Kind of ironically amusing in a sick sort of way to me, probably not so much to scientists and tree huggers.

  100. TCO
    Posted Sep 25, 2006 at 6:58 PM | Permalink

    Maybe we shouold do more sectioning. Trees are cheap. Would we run into a scarcity problem (of old trees for followup studies) if more sectioning was done vice coring?

  101. bender
    Posted Sep 25, 2006 at 7:08 PM | Permalink

    bcps are not cheap. Bloom Clubbers will probably tell you there are any number of metrics showing how these are among the most rare habitats on earth.

    The strip-bark willows of the central plains are as Barney describes: stripped by wind/dust abrasion. Radial growth goes up during a pulse of stripping, presumably due to reduced cambial carbon sink, with no change in roots.

  102. TCO
    Posted Sep 25, 2006 at 7:17 PM | Permalink

    98.

    Bender, I asked more then one question. The physical rationale (or known reason) for “strippedness”, is a different issue then the response of bcps to temp/CO2.

    Bender, posts 6,8, 10 in the Graybill and Idso thread (Feb2005), show that I saw figure 5 and gave some thought to it. You were in that thread. Didn’t you notice me? No kisses for you, Mom…

  103. bender
    Posted Sep 25, 2006 at 7:24 PM | Permalink

    … and then you forgot it, prompting you to ask, unnecessarily, “how can you say bcps are strange?”. And now I’ve reminded you, and now you see I’m right. And yet I’m sure you’ll forget again in a year’s time or less. And so it goes …

  104. TCO
    Posted Sep 25, 2006 at 7:32 PM | Permalink

    Mom, you haven’t adressed your forgetting that I was in that thread. Bad, bad, Mom.

    I’m making the point of bcp “strangeness” versus other trees. Even if you accept that the BCPs are diverging from each other, do other series show divergent behavior at different times?

    Figure 5 shows BCPs versus each other (subspecies). Actually the sample size is pretty low, so it may not even show that, could be region versus region.

    My point remains, if you are going to kick the BCPs off the island, you have to show that they are worse sinners then the general population. If you don’t, your argument just collapses to not liking tree rings in general and the BCP-specific examination is unwarrented.

  105. TCO
    Posted Sep 25, 2006 at 7:33 PM | Permalink

    In all seriousness, I REALLY did not forget Figure 5. We just have a remaining argument on implication.

  106. Barney Frank
    Posted Sep 25, 2006 at 7:34 PM | Permalink

    Would we run into a scarcity problem (of old trees for followup studies) if more sectioning was done vice coring?

    Definitely not. In the year 7006 we would have plenty again.

  107. TCO
    Posted Sep 25, 2006 at 7:36 PM | Permalink

    In the year 2525…

  108. bender
    Posted Sep 25, 2006 at 8:06 PM | Permalink

    Re #104 TCO, I know what point you’re making. I’m telling you it’s irrelevant.

    Reading over your older posts I see more clearly what you want. I sympathize, but I’m telling you: you aren’t going to get it. It would take a billion dollars and thirty researchers 20 years to disentangle all the individual effects that you want to see quantified, one by one. I agree this would be terrific. But this is not Steve M’s job. Stop asking him to do it.

    We don’t know what bcp’s are responding to. Sorry. It’s that bozo simple. If Steve wants to highlight all the major uncertainties surrounding bcp interpretation, that is a perfectly valid argument on its own. Could be CO2, could be a heightened response to temperature, could be climatic factors other than temperature. Could be more than one climatic factor. Could be more than one response to temperature! Now whose burden is it to prove that reconstructions work? Steve M’s? Or the reconstructionists? This is not a biology argument, it’s a systems argument (propagation of error through uncertain propositions that are inferentially linked to form a whole argument). You want biological science when the real problem is systems science.

    That dispensed with, once and for all, now maybe you can tell me what you make of the heterogegenous response to spring vs winter temperatures in Table 2. (Do me a favour and circle the relevant coefficients and scan and paste the image here so that I can see you’re looking at the right numbers?) AFAIK no one has discussed this. Tell me TCO, if bcps are a temperature proxy, what are they a proxy for: increasing winter temperature, or decreasing spring temperature? DOH! You see the problem now? Probably not …

  109. bender
    Posted Sep 25, 2006 at 8:18 PM | Permalink

    Re #107
    You think you’re going to get permissions to cut down 4000yr old bcps with explanations like that? Good luck. Tell you what: you get the permissions forms aok’d and I’ll core the trees myself, you lazy dog.

  110. TCO
    Posted Sep 25, 2006 at 8:44 PM | Permalink

    I guess I’m just looking for something like a percentage argument. Would sectioning lead to a 10% reduction of the mature population, 1%, .1%, etc?

  111. TCO
    Posted Sep 25, 2006 at 8:53 PM | Permalink

    I’m not asking Steve to solve the problem of the bcps*, just asking if the questions there are more or less complex then those of tree rings in general. I also think your post (where you leave in the possibility of temp effects) is more thoughtful and more one I agree with then Steve’s “bad apple” and “non-climactic” remarks.

    *I did some speculation on what would be interesting to know/see in a real review, but was not requiring Steve to perform that. However, I think that independant of extra field studies, that I have not even seen a good analysis of all the collected information and studies yet. And that would take a person a year or so. Not 600 man-years. And given that that has not been done, I’m not so jazzed about appeals to NAS authority…

  112. TCO
    Posted Sep 25, 2006 at 8:59 PM | Permalink

    1. I don’t know how to cut and paste in a pdf.

    2. I already responded to the Table 2 pattern. Can you please refer to that post specifically and what was inadequate rather then making me repeat?

  113. bender
    Posted Sep 25, 2006 at 10:23 PM | Permalink

    Lee, is that you?

  114. bender
    Posted Sep 25, 2006 at 10:35 PM | Permalink

    Re #112 A search of “Table 2″ indicates only my postings. PLEASE do NOT repeat yourself. A single line with a post # would be fine. That’s three characters.

  115. TCO
    Posted Sep 26, 2006 at 7:31 AM | Permalink

    post 70 in the Bunn et al thread (7JUN2005).

  116. bender
    Posted Sep 26, 2006 at 7:35 AM | Permalink

    Oh how, oh how could I possibly have missed that?! Can’t wait to find it.

  117. bender
    Posted Sep 26, 2006 at 7:58 AM | Permalink

    My stats gaps will show here, but I’m not clear what the higher Pearson correlations means intuitively for the issue at hand. Is the variability of the forcer included? (Is it normalized by std deviation?) For instance, if r for PDSI is heavier then r for temp, but temp variaion is much more then PDSI, then the tree would still be an effective tree in the sense of limiting stand, no?

    You’re not clear on how to interpret a correlation coefficient? You’re not even clear on how it’s calculated? Yet you’re perfectly clear on everything else in climate science to the point where you think your criticisms of McIntyre carry any weight? Good grief. These are no longer “tough questions” you’re asking. They’ve been asked, and answered, and you reject the answers because you don’t understand them. Asked a millllyun times, they become baseless insinuations.

    Go away. Read a stats book. Take a night class. Lurk. Learn. Don’t post. Good day.

  118. Otto
    Posted Sep 26, 2006 at 9:04 AM | Permalink

    Graybill and Idso Figure 6 is strong evidence that upper forest border BCP (Pinus longaeva) ring width is tracking decadal to multidecadal scale variability in warm season temperature during the preindustrial era.

    Otto

  119. bender
    Posted Sep 26, 2006 at 9:17 AM | Permalink

    Re #118
    No it’s not. The dotted line is not instrumental data. According to the Figure caption it’s the “reconstruction of average April-September temperature (dotted)”. So of course, its going to be correlatd with the data on which the reconstruction is based. Moreover, the two “series were smoothed with an 8-year low-pass filter” to enhance the impression of a strong correlation. Moreover the positive correlation shifts toward negative starting in the 1870s.

  120. Otto
    Posted Sep 26, 2006 at 10:01 AM | Permalink

    Re 119

    I disagree.

    The temperature reconstruction (dotted line) is a completely independent data set – a latewood density based reconstruction not BCP.

    It is smoothed because that is how you compare decadal-scale rather than interannual variability and the temperature signal in upper forest border BCP is in these lower frequencies.

    The lack of correlation in the post industrial period is, according to Graybill and Idso anyway, due to atmospheric CO2 fertilization effect on the BCP.

  121. Ken Fritsch
    Posted Sep 26, 2006 at 10:10 AM | Permalink

    It is smoothed because that is how you compare decadal-scale rather than interannual variability and the temperature signal in upper forest border BCP is in these lower frequencies.

    What is the rationale for looking for a signal in those lower frequencies?

  122. bender
    Posted Sep 26, 2006 at 10:57 AM | Permalink

    Re #120 I stand corrected, Otto. Although the Figure caption is ambiguous, the text on p. 91 confirms what you say: the temperature reconstruction is not from the source bcp data. Thanks for the correction. But how do you know the reconstruction was not based on other bcp data (the text does not say), and how do you know the validation reconstruction wasn’t cherry-picked, and, more importantly, how do you explain a dual temperature response (Table 2: spring (-), winter (+)) yielding an unambiguous temperature reconstruction? Did the (-) response just vanish somehow? It’s possible I’ve misunderstood something, so I would welcome any further insights you can offer.

  123. Otto
    Posted Sep 26, 2006 at 11:01 AM | Permalink

    re 121 What is the rationale for looking for a signal in those lower frequencies?

    One that I’ve heard that seems to make sense:

    Upper forest border BCP may retain their needles for many years. Needle length is partially determined by temperature and growing season length. The needles from many years combine to form the photosynthetic apparatus for a tree. This multi-year needle retention leads to high autocorrelation in the ring widths of upper forest border BCP. The temperature signal contained in the ring widths is, in effect, naturally smoothed by needle retention acting like a biologically imposed low-pass filter.

  124. TCO
    Posted Sep 26, 2006 at 11:16 AM | Permalink

    Is there a paper comparing coniferous and deciduous species that helps to demonstrate this rationale?

  125. Steve Sadlov
    Posted Sep 26, 2006 at 11:17 AM | Permalink

    RE: #83 – Dano I’m not so sure about your climate designations. Mediterranean, unmodified or modified, is only found in on the coast, in the coast ranges and the Great Valley, outer transverse ranges and debatably, the Sierra foothills up to the the winter tule fog line. All other climates in California are continental or subtropical. Sierra Nevada above the tule fog line are borderline between subtropical alpine / highland and continental alpine / highland, essentially determined by latitude. The other ranges and plains are clearly continental (further north) or subtropical (further south). Specific examples – Northern Sierra Nevada – continental upland / alpine. Mojave desert, midlatitude arid. Colorado Desert – subtropical arid. Etc.

  126. bender
    Posted Sep 26, 2006 at 11:37 AM | Permalink

    Re #123
    The mechanism makes sense and the search for a mid-range response is logical. What is not acceptable is smoothing the two series and then arguing that the correlation represents the strength of the annual-scale response. And so there goes one of the supposed advantages of tree-ring data as temperature proxy. Yes, they are “annually resolved” in terms of dating, but not in terms of reconstruction. What this means is that you can’t make inferences about one year being more warm than others. Inferences are restricted to the decadal scale.

  127. Phil B.
    Posted Sep 26, 2006 at 12:04 PM | Permalink

    Re # 123, 126, Shouldn’t they be prewhitening instead of Lowpass filtering the tree ring series. I have seen the idea of prewhitening discussed in the literature, but not any results.

  128. Otto
    Posted Sep 26, 2006 at 12:14 PM | Permalink

    Re 122

    But how do you know the reconstruction was not based on other bcp data (the text does not say),

    Pg. 87 describing the reconstruction they compared their regional BCP chronology to in Fig. 6, “The reconstruction of average April-September temperature was based upon variation in maximum latewood density measurements of samples from 53 stands of various coniferous species. All species were different than those used in the current study. Calibrations and reconstructions were accomplished for 23 points that are part of a 5 deg. lat. by 5 deg. long. grid of northern hemispheric temperatures.”

    and how do you know the validation reconstruction wasn’t cherry-picked,

    They compared their regional BCP chronology to the two closest reconstructed gridpoints.

    how do you explain a dual temperature response (Table 2: spring (-), winter (+)) yielding an unambiguous temperature reconstruction? Did the (-) response just vanish somehow?

    Graybill and Idso do not reconstruct anything in this paper. The set out to show a post 1850 CO2 fertilization effect in high alpine trees and almost inadvertantly present evidence that upper forest border BCP (Pinus longaeva) ring width is tracking decadal to multidecadal scale variability in warm season temperature during the preindustrial era.

    As for the r values in table 2, I don’t give them a lot of weight. Its only 31 years of data. The negative correlation with spring temperature probably reflects some years when the trees benefitted from delayed snow melt. The positive correlation with winter is more difficult to explain and may be spurious. I’m more inclined to accept the common low frequency patterns over 250 years between the density-based (non-BCP, non-ring width) temperature reconstruction and the high elevation BCP ring width.(Fig 6)

    As for a completely unambiguous temperature reconstruction from BCP – maybe not – trees are not thermometers (or rain gauges for that matter), but we don’t have thermomters or rain guages for most of the past.

  129. bender
    Posted Sep 26, 2006 at 12:29 PM | Permalink

    Thanks, Otto! Good answers. Especially the point about n=31 years. I missed that. (I just realized too that the n=31 time period, 1949-1980, occurs after the divergence from strong positive response had started. You would expect increasingly ambiguous responses as you move out of the 1800s.)

    Say … can you tell me why strip-bark and full-bark samples respond differently to whatever it is they’re responding to? Is it a matter of the full-bark samples having larger carbon “sinks” to fill? Is it due to differences in root mass, cambium surface area, foliage volume, or all three? Any literature on the subject?

  130. Steve McIntyre
    Posted Sep 26, 2006 at 12:52 PM | Permalink

    bender, an interesting comparison is between Graybill and Woodhouse chronologies at Niwor Tidge for limber pine discussed http://www.climateaudit.org/?p=623 . THe Graybill chronology is here. Although UCAR coffee would still be warm when they got to the Niwot Ridge site, no one has ever reconciled these discrepant chronologies in nearly 20 years.

  131. bender
    Posted Sep 26, 2006 at 1:11 PM | Permalink

    Re #130
    I’d seen that post before but had not noticed the red Graybill chronology there was for limber pine, not bristlecone pine. So limber pines are exhibiting the same strangely dichotomous response as bcps. Then I come back to my original question #10 about whitebark pine. Do they exhibit such dichotomous responses, and if not, is it because, as Dano suggests, they are in subalpine meadows, not truly alpine habitat?

  132. bender
    Posted Sep 26, 2006 at 1:15 PM | Permalink

    Re #130
    Did you ever ask Woodhouse about this remarkable discrepancy on limber pine?

  133. Dano
    Posted Sep 26, 2006 at 3:46 PM | Permalink

    85:

    Um, OK. I’m not saying that. If you want to use one data point to characterize the system, go ahead.

    87:

    I’ll wait for your work showing this is so, Steve, including your core sample database link in your paper and a dendro person as co-auth.

    125:

    Lots of deadlines rest of this week so minimal commenting, but this is a modified Med climate, as the precip regime dominates but altitudinal component makes it cold (vs typical subtropical). Millar et al state this in their study plan for the 2006 paper (can’t look for it right now), and you can see some folks ideas on this here** and here.

    Best,

    D

    ** Note how worst CA invasive plants come from the Turkey-Iran region…

  134. bender
    Posted Sep 26, 2006 at 5:32 PM | Permalink

    Re #133
    I don’t want “to use one data point to characterize the system”. I want to know what you make of the available data. I want to know what you would expect from a different species in a different system in a different area. And I want your cite dump.

    You’re too busy to answer all these wonderful questions about your favorite topic? So disappointing. You seemed so knowledgeable. Maybe you’ll have more time next week?

  135. Steve Sadlov
    Posted Sep 26, 2006 at 6:30 PM | Permalink

    RE: #133 – that is a highly unorthodox climate interpretation. It is really stretching things to label anything in the Sierra above the tule fog line “mediterranean.” Landis would have utterly disagreed with that if still alive.

  136. Dano
    Posted Sep 26, 2006 at 6:43 PM | Permalink

    134:

    I’m busy because I’ll be backpacking in BC and I have lots to do before then. At the risk of more bandwidth, your questions are overly broad, plus any ISI search will give you refs. I don’t know what point you’re trying to make here – spp react according to their location and climate regime – how is this news?

    135:

    Quickly, the precip regime is the key and the linkies I gave outline the idea. Not my idea, just passing it on. The lack of summer precip from monsoonal moisture gives weight, but I agree that the Koeppen doesn’t cover it, hence the DFG paper. As I said, Millar et al. were comfortable enuf with it – it’s in their study plan for the meadow invasion papers.

    Gotta go,

    Best,

    D

  137. bender
    Posted Sep 26, 2006 at 6:48 PM | Permalink

    No more bandwidth. If you don’t understand the questions and can’t answer them because they’re “too broad” and don’t want to share your references, that’s fine. No more questions. Have fun in BC.

Follow

Get every new post delivered to your Inbox.

Join 3,382 other followers

%d bloggers like this: