Merits of Cellulose vs Whole Wood Analysis

17 Nov. 2002
Greetings to all on this tree-ring_isotopes listserve! Ed Wright just discovered a paper that considers the isotopic difference between holocellulose and alpha-cellulose results for d13C (title, authors, abstract below). Unfortunately, only the abstract is available on-line (Swansea Geographer v. 37 (2002)- pages not indicated on journal web site). Ed and I were considering
whether anyone had done a similar (published) comparison of d18O results between wood/tree-ring holocellulose and alpha-cellulose, but we are only aware of the comparison he did in his
dissertation. Anybody aware of such references?
Preparation of holocellulose from standard increment cores for stable carbon isotope analysis
Loader, N.J. 1,3, Robertson 2 , Lücke 3, & Helle 3. 1 Department of Geography University of Wales Swansea, Singleton Park, Swansea, SA2 8PP, U.K. 2 Quaternary Dating Research Unit, CSIR Environmentek, 0001 Pretoria, Republic of South Africa. 3 ICG-4, Forschungszentrum Jülich GmbH, Leo Brandt Straße, 52425 Jülich, Germany.
Time series of absolutely-dated oak (Quercus robur. L.) tree-rings (AD1970-AD1994) were collected from Sandringham Park, U.K. Stable carbon isotope and ring width data were compared between samples prepared for isotopic analysis using a conventional cellulose extraction (Loader et al. 1997) and a novel core based method. Results from this initial investigation demonstrate a high degree of similarity between the interannual variability in the stable carbon isotope ratios of conventionally prepared a-cellulose and holocellulose core time series. An isotopic offset is observed between the two sample series which could reflect the different origins of the sample material or the incomplete removal of the extraction solution. The drying process was found to distort slightly the absolute dimensions of the cellulose core, although the relative variations in ring width and wood structure were preserved. The ability to prepare cellulose from 5mm diameter increment bores could improve considerably the efficiency of isotope dendroclimatology.

Steven W. Leavitt
Laboratory of Tree-Ring Research
105 W. Stadium, Bldg. #58
University of Arizona
Tucson, AZ 85721
phone: 520-621-6468
fax: 520-621-8229


18 Nov. 2002
Hi Steve and All,
I am not aware of an article on d18O but the following article adds more information on the d13C story. I'm sure that many people are aware of this article, but I thought it might be good to get the ball rolling on the discussion. I'm not clear on whether Peter was using a combination of leaf and wood tissue or just leaf tissue in these analysis. He compares whole tissue, holocellulose, treated holocellulose (Ca-oxalate removed), alpha cellulose, and nitrocellulose. In most cases, holocellulose was not statistically different from alpha cellulose.

Van de Water, P.K. 2002. The effects of chemical processing on the d13C value of plant tissue. Geochimi. Cosmochim. Acta. 66:1211-1291.

J. Renee Brooks
Western Ecology Division
200 SW 35th St.
Corvallis, OR 97333
(541) 754-4684 (Office)
(541) 754-4799 (FAX)


20 Nov 2002
Hi Steve (and greetings to everyone on the listserve)

I recently had a graduate student ask me about published data on d18O in hemicellulose and was surprised to find very little published on the subject. Margaret Barbour's recent paper in Aust. J. Plant Phys (2001, 28:335-348) and Graham Farquhar's 1998 review in Howard Griffith's book
'Stable Isotopes' are some of the most detailed work on the d18O of organic matter, but I do not think they measured holocellulose or hemicellulose specifically. If one looks at the chemical structure of hemicellulose (see Buchanan et al. 2000 Biochem. & Molec. Biol. of Plants) they clearly have carbonyl groups which would be exchangeable. Thus we have always taken samples to alpha-cellulose. The number of carbonyl groups on hemicellulose molecules may be small enough to get away with not extracting them, but the added step is relatively trivial compared to the rest of the procedure. Some of us have used the Brendell et al (2000, Phytochem. Anal. 11:7-10) method to extract cellulose and feel that it really only produces holocellulose. In comparisons with d18O of alpha-cellulose we have found conflicting results (sometimes they differ and other times they do not). Julie Gaudinski has been putting the results of a number of these studies together and may submit a paper in the near future.

Cheers, John Roden