Stable Isotopes in Dendroclimatology: Moving Beyond ‘Potential’ - Dendroclimatology - page 157

to climatology (Hughes 2002 ) . Such was the time commitment involved in obtain-
ing a few isotopic measurements, they were exclusively based upon unreplicated
'single-tree' series (e.g., Craig 1954 ; Farmer and Baxter 1974 ; Epstein and Yapp
1976 ; Libby et al. 1976 ). However, early studies provided evidence of the link
between moisture stress and 13 C enrichment and detected a relationship between
narrow rings and enrichment (Craig 1954 ) . Craig ( 1954 ) carried out his study on
Sequoia , from a relatively arid environment, and thus was able to identify the simple
link between isotope ratios and moisture stress typically found in such an envi-
ronment. Libby was perhaps a little overoptimistic in entitling her Nature paper
'Isotopic Tree Thermometers' (Libby et al. 1976 ) .
Park and Epstein ( 1960 , 1961 ) demonstrated that plants discriminate against the
heavier carbon isotope during photosynthesis and identified the biochemical path-
ways where this selection occurs. Early observations also indicated that varying
amounts of resins and waxes would contribute to isotopic variability in different
wood components because they differ in volume (Wilson and Grinsted 1977 ) and
are mobile between rings (Long et al. 1979 ) .
Epstein et al. ( 1977 ) first noted that it might be possible to transfer the work on
isotope ratios in precipitation, carried out by Dansgaard ( 1964 ) and others, to look
at archived oxygen and hydrogen isotope ratios in tree rings. The lack of a theo-
retical framework to explain stable isotope variability in plant materials meant that
interpretation of these early series was often oversimplistic, with little understand-
ing of the biological mediation of the signal (Loader et al. 2007 ) . Thus the proxies
could not be fully exploited until the development of appropriate models provided a
framework for interpretation.
Leavitt and Long ( 1982 ) reviewed the current 'state of play' in stable isotope
dendroclimatology and expressed concern at the apparent wide range of fraction-
ation effects and environmental variables that could be found in measured ratios,
questioning the merit of stable isotope dendroclimatology in much the same way
as Hughes ( 2002 ) later did. However, publication of the 'carbon isotope model'
by Farquhar et al. ( 1982 ) and a subsequent paper describing the application of the
model for the interpretation of tree-ring data by Farquhar et al. ( 1982 ) effectively
explained why there is natural variability between trees and why different trees can
carry different climate signals. Larger sample sizes were also found to provide more
coherent results (Freyer and Belacy 1983 ; Leavitt and Long 1984 ) .
A mechanistic focus is also observed in the early literature on water isotopes,
with measurements often carried out in conjunction with leaf water analysis using
terrestrial plants (e.g., Epstein et al. 1977 ; Burk and Stuiver 1981 ) . The understand-
ing of the relationship between

18 O of cellulose and soil water was subsequently
based on the idea that exchange might occur between carbohydrate and stem
water, with additional fractionations during pre-cellulose remobilization (Cooper
and Deniro 1989 ) .
The need for increased replication without an increase in the cost per sample
became a strong driver of multi-tree pooling methods. Many of the first stable iso-
tope studies managed to analyze multiple trees at annual resolution, at a time when
stable isotope measurement was still expensive and time consuming, by pooling

δ