Stable Isotopes in Dendroclimatology: Moving Beyond ‘Potential’ - Dendroclimatology - page 153

to be extracted first, because they are mobile between rings; the methods used to
prepare laths for X-ray densitometry are adequate for this (Schweingruber et al.
1978 ) . In the most commonly used method for extracting whole wood to -cellulose,

lignin is oxidized by the in situ generation of chlorine dioxide (ClO 2 ), by using acid-
ified sodium chlorite in aqueous solution to yield holocellulose (Green 1963 ; Wilson
and Grinsted 1977 ; Leavitt and Danzer 1993 ) or treated further to hydrolyze hemi-
celluloses by using sodium hydroxide to yield

α

-cellulose (Green 1963 ; Loader et al.
1997 ) . Most laboratories use a rapid 'batch processing' version (Loader et al. 1997 ;
Rinne et al. 2005 ) . An alternative method for cellulose preparation, proposed by
Brendel et al. ( 2000 ) , and subsequently modified by Gaudinski et al. ( 2005 ) , which
is even faster and uses relatively benign chemicals, has been used successfully in a
number of studies (e.g., Poussart et al. 2004 ; Evans and Schrag 2004 ; Poussart and
Schrag 2005 ; Anchukaitis et al. 2008a ) .
Whole-wood and cellulose stable carbon isotope ratios, although offset, seem to
vary in parallel. Borella et al. ( 1999 ) concluded that a portion of signal is lost in
whole wood, but Loader et al. ( 2003 ) found that cellulose and lignin

α

13 C, from
oak, gave marginally weaker correlations with instrumental climate data than whole
wood from the same trees. The use of whole wood is also supported for

δ

18 O anal-

δ

18 O analyses of different wood components
in a global dataset of oak and pine to conclude that no climate information was lost
in analyzing whole wood over

ysis. Barbour et al. ( 2001 ) compared δ

α -cellulose and, depending upon local conditions,

concluded that extraction of -cellulose from wood samples might be unnecessary

for isotope studies. Some caution is required in working with subfossil wood, due
to the differential decay rates of lignin and cellulose (Schleser et al. 1999 ) . For
hydrogen isotope analysis, cellulose nitrate or isotopically equilibrated cellulose
must be analyzed, owing to the propensity for isotopic exchange in approximately
30% of the hydrogen atoms. Equilibration replaces all of the exchangeable hydro-
gen atoms with hydrogen of known isotopic composition and, in nitration, the
hydroxyl hydrogen is replaced by nitro groups (Ramesh et al. 1986 ; Filot et al.
2006 ) .
Methodological choices for sampling and measurement should be made on the
basis of the aims and objectives of the study and might usefully be viewed as a series
of questions (Fig. 6.1 ) in order to choose the most efficient methodology for sample
preparation.

α

6.3.1 A Note on New Measurement Techniques

18 O are in fact an average of five oxygen molecules, all
of which have different potentials for reexchange with stem water, by virtue of their
position in the cellulose matrix. Therefore, certain of the oxygen molecules more
reliably record the source water signal than others, which has been confirmed from
experimental measurements on crop seedlings (Sternberg et al. 2006 ) . For example,
one molecule exchanges fully with source and trunk water (named as oxygen-2)

Measurements of cellulose δ