Review
Blackwell
Oxford,
New
NPH
©
1469-8137
0028-646X
October
10.1111/j.1469-8137.2008.02661.x
2661
5
0
Research
562???
Research
53???
ThePhytologist
Authors
review
UK
2008
Review
Publishing
(2008).Ltd
Journal compilation © New Phytologist (2008)
Research reviews
Research review
Oxygen stable isotope ratios of tree-ring
cellulose: the next phase of
understanding
Author for correspondence:
Leonel da Silveira Lobo O’Reilly Sternberg
Tel: +1 305 284 6436
Fax: +1 305 284 3039
Email: [email protected]
Leonel da Silveira Lobo O’Reilly Sternberg
Department of Biology, University of Miami, Coral Gables, FL 33124, USA
Received: 9 August 2008
Accepted: 22 September 2008
Summary
New Phytologist (2009) 181: 553–562
doi: 10.1111/j.1469-8137.2008.02661.x
Key words: cellulose, oxygen-18,
paleoclimate, stable isotopes, tree-ring.
Analysis of the oxygen isotope ratio of tree-ring cellulose is a valuable tool that can
be used as a paleoclimate proxy. Our ability to use this tool has gone through
different phases. The first began in the 1970s with the demonstration of empirical
relationships between the oxygen isotope ratio of tree-ring cellulose and climate.
These empirical relationships, however, did not provide us with the confidence that
they are robust through time, across taxa and across geographical locations.
The second phase began with a rudimentary understanding of the physiological
and biochemical mechanisms responsible for the oxygen isotope ratios of cellulose,
which is necessary to increase the power of this tool. This phase culminated in a
mechanistic tree-ring model integrating concepts of physiology and biochemistry
in a whole-plant system. This model made several assumptions about leaf water
isotopic enrichment and biochemistry which, in the nascent third phase, are now
being challenged, with surprising results. These third-phase results suggest that,
contrary to the model assumption, leaf temperature across a large latitudinal gradient
is remarkably constant and does not follow ambient temperature. Recent findings
also indicate that the biochemistry responsible for the incorporation of the cellulose
oxygen isotopic signature is not as simple as has been assumed. Interestingly, the
results of these challenges have strengthened the tree-ring model. There are several
other assumptions that can be investigated which will improve the utility of the
tree-ring model.
Introduction
Oxygen stable isotope ratios of tree-ring biomass can be used
as a proxy for climate during tree growth. The basic principle
of this tool is the observation that the oxygen stable isotope
ratio of precipitation (expressed here as the δ18O value) is
© The Author (2008)
Journal compilation © New Phytologist (2008)
related to temperature and at least part of this isotopic climate
signal is passed on to the isotope ratios of plant biomass. The
ongoing development of this tool can be roughly divided into
three phases. The first phase involved showing empirical
relationships between δ18O values of plant biomass and
climate. Libby et al. (1976) produced one of the earliest works
New Phytologist (2009) 181: 553–562 553
www.newphytologist.org 553
554 Review
Research reviews
that correlated the oxygen isotope ratios of tree-rings in two
German oak trees (Quercus petraea) with temperature. This
work was followed by several other studies that established
empirical relationships between oxygen isotope ratios of treering cellulose or wood and measured climatic parameters such
as temperature, relative humidity (RH) and precipitation
(Burk & Stuiver, 1981; Ramesh et al., 1986; Saurer et al.,
1997; Robertson et al., 2001; Rebetez et al., 2003) . Empirical
relationships between the oxygen isotope ratios of plant
biomass and climate have been used by some to project back in
time to before climate records were kept (Masson-Delmotte
et al., 2005; Ballantyne et al., 2006; Treydte et al., 2006). The
above studies, which relied on empirical relationships without an
understanding of the underlying physiological and biochemical
mechanisms, suffered from the weakness of not knowing
whether these relationships are robust through time, across
taxa and across geographical locations. This brings us to the
second phase: the construction of a mechanistic model to
explain the incorporation of 18O into plant biomass. This
second phase culminated in the Roden et al. (2000) tree-ring
model which integrated our understanding of oxygen isotope
fractionations associated with water passing through the plant and
the biochemical process of cellulose synthesis. This framework
model of tree-ring cellulose allows us to better understand
how the ecophysiological and biochemical behavior of trees
affects the oxygen isotope ratios of tree-ring cellulose. Several
assumptions regarding leaf temperature, the isotopic composition
of atmospheric vapor and biochemical fractionations, among
others, were made in establishing this model. The nascent
third phase of our understanding involves testing and
determining the impact of the above assumptions on the δ18O
values of tree-ring cellulose. In this review, I will discuss the
development of the basic tree-ring model (second phase)
and some recent developments which introduce us to the
third phase of understanding of tree-ring oxygen isotope
ratios. Finally, I will discuss future directions to improve our
interpretation of stable isotopes in tree-ring cellulose.
Development of a mechanistic tree-ring
isotope model
The development of a mechanistic model predicting the
oxygen isotope ratios of tree-ring cellulose based on climate
was the product of two parallel lines of study: elucidation of
(1) the changes in the oxygen isotope ratio of water as it moves
through the plant and (2) the changes in the isotope ratio of
oxygen as it becomes incorporated in the cellulose molecule
through biochemical effects. Studies on the changes of the
isotope ratio of water as it moves through the plant focused on
the only plant organ that contains water with a different
oxygen isotopic composition from that taken up by the plant
roots: the leaf. Interest in leaf water was stimulated not only
by Epstein et al.’s seminal work showing that the oxygen
isotope ratio of cellulose is related to that of water (Epstein
New Phytologist (2009) 181: 553–562
www.newphytologist.org
et al., 1977), but also by the observations that molecular
oxygen with the isotopic composition of leaf water is released
during photosynthesis (Guy et al., 1993) and that ambient
CO2 absorbed by the leaf isotopically equilibrates with leaf
water and is partly released back to the atmosphere (Farquhar
& Lloyd, 1993). The isotopic labeling of these two gases provides
an important geochemical tool with which to elucidate the
global oxygen and carbon cycle (Dole et al., 1954; Francey &
Tans, 1987; Farquhar et al., 1993; Bender et al., 1994).
Oxygen isotope ratios of leaf water
The isotopic composition of the evaporative pool of leaf water
undergoing transpiration (δ18OE) is given by the Craig–
Gordon equation (Craig & Gordon, 1965):
δ18OE = δ18OS + ε* + εk + ea/ei (δ18OA – εk – δ18OS) Eqn 1
The isotopic composition of the evaporating pool of water in
the leaf is affected by the ratio of ambient vapor pressure to
that inside the leaf (ea/ei), the δ18O value of atmospheric vapor
(δ18OA) and the equilibrium and kinetic isotopic effects (ε*
and εk, respectively) in addition to the isotopic composition
of stem water (δ18OS).
Several investigators measured the isotopic composition of
leaf water (δ18OL) under different RHs and, although the
isotopic enrichment trends were consistent with Eqn 1, leaf
water was often not as enriched as predicted by the Craig–
Gordon model (see Cuntz et al., 2007 for a review). This discrepancy was explained by the presence of a nonenriched leaf
water compartment (having a similar δ18O value to that of
stem water) as well as the presence of an enriched leaf water
compartment with a δ18O value similar to that predicted by
Eqn 1. Bulk leaf water isotopic composition was, therefore,
proposed to be the weighted average of the isotopic compositions
of these two water compartments (Flanagan & Ehleringer,
1991).
One of the weaknesses of the above compartmentation
hypothesis is the assumption that water compartments are
discrete and of a constant size. This assumption cannot fully
explain the pattern of leaf water enrichment. For example,
given the same vapor pressure deficit, water in a leaf that has
low transpiration rates becomes isotopically enriched compared
with that of a leaf that has high transpiration (Cuntz et al.,
2007). Interestingly, a common conversational error is to state
the opposite: ‘leaf water is isotopically enriched because of
high transpiration’. This pattern is best predicted by the model
proposed by Farquhar & Lloyd (1993) where the isotopic
composition of bulk leaf water is determined by the balance
between two water fluxes in opposite directions. In one direction,
from the apoplast towards the leaf vein, there is a diffusional
flux of the isotopically enriched water having the isotope ratio
predicted by Eqn 1 and in the opposite direction there is the
transpiration-driven convective flux of isotopically nonenriched
© The Author (2008)
Journal compilation © New Phytologist (2008)
Research reviews
stem water towards the apopolast. The net effect of convective
and diffusive flow is known as the Péclet effect (ρ) which is
dimensionless and given by the following equation:
ρ = (TL)/(CD)
Eqn 2
(C, the molar concentration of water (5.56 × 104 mol m–3);
D, the diffusivity of H218O in water (2.66 × 10–9 m2 s−1);
T (mol m−2 s−1), the rate of transpiration; L (m), the effective
mixing length in the leaf.) The effective mixing length is
dependent on the tortuosity of the diffusive pathway as well
as the physical distance from vein to apoplast (Barbour &
Farquhar, 2004). The bulk leaf isotopic composition is
predicted by the following equation:
δL = [(1 – α)δS] + [αδE]
Eqn 3
where α is given by (1 – e –ρ)/ρ. Leaf water can be thought of as
being determined by the fraction of isotopically nonenriched
water (1 – α) and the fraction (α) of enriched water. These
fractions, however, reflect a continuous gradient and not a
discrete compartment; this gradient is also dynamic and
subject to change depending on the transpiration rate of the
leaf. Note that the greater the transpiration (T ), the greater
the Péclet effect (Eqn 2), which reduces the contribution of
enriched water (α). The effect of transpiration in lowering the
δ18O values of bulk leaf water discussed above can, therefore,
be explained by this equation.
Biochemical effects during cellulose synthesis
The δ18O value of cellulose (δ18OC) is 27‰ (±4‰) higher
than that of the water (δ18OW) present during its synthesis
(Epstein et al., 1977; Deniro & Epstein, 1981). This constant
relationship was first hypothesized to be caused by the
contribution of one oxygen atom from water and two from
CO2 equilibrated with ambient water towards carbohydrate
synthesis. As the δ18O value of CO2 equilibrated with water
is ~40.0‰ higher than that of the water it equilibrated with,
the above process can be summarized by:
δ18OC = (δ18OW/3) + [2/3(δ18OW + 40‰)]
Eqn 4
which simplifies to:
δ18OC = δ18OW + 26.7‰
Eqn 5
However, if only pre-photosynthetic equilibration of CO2 with
water determines the δ18O value of cellulose, then one would
expect a temperature effect on the oxygen isotope ratios of
cellulose. This is because there is a significant and easily detected
temperature effect on the fractionation factor for the CO2/
water equilibration of 0.2‰ °C–1. DeNiro & Epstein (1981)
grew aquatic plants at different temperatures ranging from 25
© The Author (2008)
Journal compilation © New Phytologist (2008)
Review
to 35°C and found no significant differences in the fractionation
between cellulose and water as a function of temperature.
These results indicate that post-photosynthetic exchange
reactions occurring between carbohydrate intermediates and
water during cellulose synthesis play an important part in
determining the δ18O values of cellulose.
Sternberg & DeNiro (1983) hypothesized that postphotosynthetic exchange reactions during cellulose synthesis
occur during carbonyl hydration:
Eqn 6
They tested the magnitude of the oxygen isotope fractionation
between carbonyl oxygen and water for the carbonyl hydration
reaction in acetone, as a model molecule. They found a
fractionation factor of ~27‰, which is similar to the
fractionation factor between cellulose and the available water
during its synthesis.
Firm evidence for the importance of post-photosynthetic
reactions in the determination of the δ18O values of cellulose
was provided by experiments on heterotrophic cellulose synthesis
from tissue culture, Lemna gibba cultures and seed germination
(Sternberg et al., 1986; Yakir & DeNiro, 1990; Luo & Sternberg,
1992; Sternberg et al., 2006). In these experiments it was
observed that for a carbohydrate substrate, such as sucrose or
starch, approx. 40% of the oxygen destined for cellulose
became labeled by the water of the culture media or the water
available for seed germination. The exchange reaction is thought
to occur through the carbonyl hydration reaction during the
futile cycle in which fructose 1,6-Phosphate recycles through
triose phosphates (Hill et al., 1995). The above postphotosynthetic exchange of carbohydrate oxygen with water
during heterotrophic cellulose synthesis can be summarized as:
δ18OC = [φ(δ18OW + Δ)] + [(1 – φ )(δ18OSUB)]
Eqn 7
(φ, the proportion of oxygen that exchanged with water
during cellulose synthesis; Δ, the average fractionation for all
oxygen that exchanged with water during cellulose synthesis;
δ18OSUB, the average δ18O value of the oxygen in the original
substrate (sucrose or starch) that did not exchange with water
(Sternberg et al., 1986).)
Tree-ring isotope model
The tree-ring oxygen isotope model represents a synthesis of
knowledge concerning the physiological processes controlling
the oxygen isotope composition of leaf water with that
concerning biochemical isotopic effects during cellulose
synthesis (Roden et al., 2000; Barbour et al., 2004; Fig. 1). It
is well known that leaf water becomes isotopically enriched
during transpiration, with the Péclet effect being important in
New Phytologist (2009) 181: 553–562
www.newphytologist.org
555
556 Review
Research reviews
Fig. 1 Integration of physiological and
biochemical effects in a tree-ring model.
Physiological effects include leaf water
enrichment by transpiration. Stem water
which has the same isotopic composition as
soil water (δ18Oso) moves to the leaf and
becomes isotopically enriched according to
Eqn 3. Biochemical effects include the isotopic
identity of sucrose synthesized in the leaf and
translocation to the stem where exchange
occurs during cellulose synthesis. See text for
explanation of the equations. δ18OSUB, the
average oxygen stable isotope ratio (δ18O) of
the oxygen in the substrate (sucrose or starch)
synthesized by the leaf; δ18OL, δ18OS, δ18OE,
and δ18OC, the oxygen stable isotope ratios of
leaf water, stem water, the evaporative pool of
leaf water undergoing transpiration and
cellulose, respectively; α, the fraction of
enriched water in the leaf; P, phosphate;
Φ, the proportion of oxygen that exchanges
with water during heterotrophic cellulose
synthesis.
determining the isotopic composition of bulk leaf water. One
question that is critical to this model is whether the δ18O
value of sucrose synthesized in the leaf reflects the isotopic
composition of a particular leaf water fraction or that of bulk
leaf water. That it reflects bulk leaf water (as predicted by Eqn
3) was elegantly shown by the isotopic analysis of sucrose
exudates and leaf water from castor bean (Ricinus communis)
leaves (Barbour et al., 2000). In another study it was also shown
that there are no isotopic effects during sucrose translocation
(Gessler et al., 2007).
The tree-ring isotope model calculates the leaf water isotopic
composition based on the isotopic composition of stem water
and atmospheric vapor, vapor pressure ratios and the Péclet
number of the leaf (Eqns 1, 2 and 3). The δ18O value of
sucrose synthesized in the leaf is then calculated by:
New Phytologist (2009) 181: 553–562
www.newphytologist.org
δ18OSUB = δ18OL + 27‰
Eqn 8
Sucrose is translocated and, as it is allocated to cellulose
synthesis in the trunk (heterotrophic cellulose synthesis), an
average of 0.42 of the oxygen destined for the cellulose
molecule exchanges with stem water. The δ18O value of
cellulose can be calculated using Eqn 7, with φ as the above
proportion of 0.42 (Roden et al., 2000). Throughout this
model the oxygen isotope fractionation for the exchangeable
oxygen in carbohydrate substrates destined for cellulose synthesis
is assumed to average 27‰ relative to the δ18O value of the
water present during cellulose synthesis. It is also usually
assumed that the atmospheric vapor is at equilibrium with the
plant available water (stem water) and that leaf temperature
follows the ambient temperature. The latter assumption allows
© The Author (2008)
Journal compilation © New Phytologist (2008)
Research reviews
us to substitute RH for the vapor pressure ratio (ea/ei) in Eqn
1. With these assumptions, this model has only two major
unknowns: the δ18O value of the water available to the plant
(stem water) and the RH. Although this model represents a
leap in our understanding of tree-ring cellulose, it cannot be
used in paleoclimate reconstruction unless either the δ18O value
of plant available water or the RH is known. One method for
solving for the two unknowns was proposed by Jahren &
Sternberg (2003) using two simultaneous equations each
expressing the δ18O and the δD values of the nonexchangeable
hydrogen of cellulose.
The tree-ring isotope model gives expected results in some
cases (Roden & Ehleringer, 2000; Roden et al., 2005) and,
when inconsistencies are found, they are explained on the basis
of different water sources or the variation in the biochemical
processes (Waterhouse et al., 2002; Roden et al., 2005). In
one of the studies cited above (Waterhouse et al., 2002),
which used a best fit approach, the empirical values derived
for the biochemical fractionation factor between cellulose and
water (30‰) and Φ (0.46) were different from the previously
observed averages of 27‰ and 0.42, respectively. Many of the
above inconsistencies can be explained by further investigating
the underlying assumptions of the tree-ring model; the third
phase of understanding the tree-ring model.
Recent developments
Testing assumptions regarding leaf water
isotopic enrichment
Two of the principal assumptions regarding the isotope ratios
of leaf water are: (1) that the integrated δ18O value of
atmospheric vapor is the same as that of vapor in isotopic
equilibrium with plant water or annual average precipitation
and (2) that the average leaf temperature during carbohydrate
synthesis is the same as ambient temperature. The assumption
about the δ18O values of atmospheric vapor is adopted, in
part, out of necessity, because there are few measurements of
the isotopic composition of atmospheric vapor, particularly
on an integrative basis. To test this assumption one needs to
collect atmospheric vapor for a prolonged period at the same
time as monitoring the long-term isotopic composition of
precipitation. Collecting atmospheric vapor requires the use
of cold traps, flow meters and pumps over a sustained period of
time. In addition, collection of vapor is subject to fractionation
in the case of inefficient cold traps. This is a critical assumption
in that small errors in the estimate of the δ18O value of
atmospheric vapor can lead to large errors in tree-ring data. A
2‰ error in the estimate of the δ18O value of atmospheric
vapor as it propagates to the δ18O values of tree-ring cellulose
can be equivalent to up to 66% of the observed range in
tree-ring signals (Helliker & Griffiths, 2007). Epiphytes, which
have their water at equilibrium with atmospheric vapor, may
provide long-term proxies of the δ18O value of atmospheric
© The Author (2008)
Journal compilation © New Phytologist (2008)
Review
vapor (Helliker & Griffiths, 2007). Analysis of the organic matter
in Tillandsia usneoides (Spanish moss) in Miami spanning
127 yr and modeling approaches indicate the isotopic equilibrium
of atmospheric vapor with year-round precipitation (Helliker
& Griffiths, 2007). Tillandsia usneoides, however, only grows
in selected locations with relatively high humidity and it is
possible that in dry locations the isotopic composition of
atmospheric vapor does not reflect isotopic equilibrium with
year-round precipitation. However, note that the impact of
the error in the estimate of the δ18O value of atmospheric
vapor diminishes as vapor pressure deficit increases (Eqn 1).
The assumption that leaf temperature is the same as ambient
temperature allows one to use ambient RH for the vapor
pressure ratio term in Eqn 1. However, a recent survey
of tree-ring data encompassing a latitudinal range of 50°
indicated a surprising deviation between leaf temperature and
growing season temperature, particularly at higher latitudes
(Helliker & Richter, 2008). Helliker & Richter (2008), in a
survey of tree-ring cellulose, observed that the δ18O values of
tree-ring cellulose deviated considerably from that predicted by
the tree-ring model when it is assumed that leaf and growing
season ambient temperature are the same. This deviation was
particularly large at higher latitudes with lower mean annual
temperatures (MATs). However, the observed values can be seen
to be in closer agreement with modeled values if one considers
that leaf temperature is higher than the ambient growing season
temperature and MAT. Helliker & Richter (2008) observed a
remarkably constant leaf temperature of 21.4 ± 2.2°C for all
latitudes when they reverse-solved the tree-ring model for
temperature (Fig. 2), indicating that at higher latitudes the
leaf temperature of actively photosynthesizing trees could be
considerably higher than the growing season temperature.
Therefore, the assumption that leaf temperature is equal to
ambient temperature during photosynthesis may not be
applicable. There are several ways in which plants can regulate
leaf temperature, such as through cooling produced by
transpiration or heating produced as a result of the canopy
structure (Helliker & Richter, 2008; Woodward, 2008).
However, further corroborative evidence is needed regarding
the constancy of leaf temperature because there are alternate
explanations for the isotopic pattern observed by Helliker &
Richter (2008). Richter et al. (2008), for example, explained
the observed isotopic pattern on the basis of different effects
of RH or the isotope ratios of vapor at different latitudes. In
addition, the assumption that the biochemical oxygen isotope
fractionation during cellulose synthesis is constant may not
be valid, as a pattern of diminishing fractionation with decrease
in latitude has also been reported for aquatic plants (Sternberg
et al., 2007). The near-constant leaf temperature, if shown to
be real, allows us to discard the assumption that leaf temperature is the same as ambient temperature and replace it with
the constant leaf temperature of 21.4°C as observed by
Helliker & Richter (2008); at least for biomes having a
MAT < 15°C (Woodward, 2008). This would greatly simplify
New Phytologist (2009) 181: 553–562
www.newphytologist.org
557
558 Review
Research reviews
Fig. 2 (a) The difference between leaf
temperature derived from tree-ring oxygen
isotope ratios (open circles, gymnosperms;
closed squares, angiosperms) and ambient
growing season temperature as a function of
mean annual temperature. (b) Leaf
temperature derived from oxygen isotope
ratios of tree-ring cellulose using the tree-ring
model (open circles), and ambient growing
season temperature at each site as a function
of mean annual temperature (closed circles).
The gray bar represents the derived
temperature from all samples (±σ). Modified
from Helliker & Richter (2008).
the application of the tree-ring model. This finding should also
have implications for the use of the δ18O values of atmospheric
O2 and CO2 to budget atmospheric oxygen and carbon dioxide.
These budgets must take into account the fact that, if there is a
constant leaf temperature, the leaf water value will be significantly
different from that predicted if the leaf temperature is the
same as the growing season temperature. In addition, they
must take into account the fact that the isotopic equilibration
of CO2 and leaf water is determined by a fractionation factor
at leaf temperature rather than ambient temperature.
Testing assumptions about biochemical fractionations
during cellulose synthesis
The major assumption regarding the oxygen isotope ratios of
cellulose is that the constant 27‰ fractionation between
cellulose and water is caused by the exchange between carbonyl
oxygens of carbohydrates and water during cellulose synthesis.
However, Schmidt et al. (2001) suggested that the match
between the average 27‰ cellulose fractionation and that of
the carbonyl hydration reaction of acetone, as measured by
New Phytologist (2009) 181: 553–562
www.newphytologist.org
Sternberg & DeNiro (1983), is coincidental. They proposed
that the oxygen isotopic fractionation of cellulose relative to water
is the average of several different fractionation effects occurring
at different stages and in different oxygens of the glucose
moiety during the synthesis of cellulose. Schmidt et al.’s
conjecture was shown to be correct when Sternberg et al.
(2003, 2006) developed a technique for deriving phenylglucosazone from stem cellulose and analyzing the isotope ratio
of the oxygen attached to the second carbon of the cellulose
glucose moiety (δ18OC-2) separately from that of the other
oxygens (δ18OP-G , those attached to carbons 3, 4, 5 and 6).
The isotopic fractionation relative to water for the oxygen
attached to the second carbon (51.0 ± 5.0‰) differs from the
average for the oxygens attached to carbons 3, 4, 5 and 6
(22.5 ± 2.0‰) for cellulose synthesized from oxygen-poor
lipids (Sternberg et al., 2006). Nevertheless, the weighted
average fractionation for the whole cellulose molecule was
28.2‰ = (0.20 × 51.0) + (0.80 × 22.5), well within the range
of reported fractionations for cellulose synthesis in the literature.
There is a greater exchange between the oxygen attached to
the second carbon of the cellulose glucose moieties (70%)
© The Author (2008)
Journal compilation © New Phytologist (2008)
Research reviews
compared with the other oxygens in the glucose moieties
(33.6%) during heterotrophic cellulose synthesis (Sternberg
et al., 2006). However, there are biochemical reactions
particular to this oxygen which superimpose isotopic noise in
the climate signal of the cellulose molecule (Sternberg et al.,
2007). In a wide geographical survey of the δ18O values of
stem cellulose, its derivative and stem water, Sternberg et al.
(2007) observed that δ18OP-G was highly correlated with δ18OS,
but the correlation between δ18OC and δ18OS was lower
because of the isotopic noise in the oxygen attached to the
second carbon of the glucose moieties (Fig. 3). Therefore, it is
expected that the use of δ18OP-G (which does not carry this
Fig. 3 Oxygen isotope ratios of (a) stem cellulose (δ18OC), (b)
phenylglucosazone (δ18OP-G) and (c) oxygen attached to the second
carbon of the glucose moieities (δ18OC-2) as a function of the oxygen
isotope ratios of stem water (δ18OS). Open symbols represent isotope
ratios for samples collected in Arizona where the relative humidity
was very low. Regression lines are: δ18OC = (0.52 × δ18OS) + 31.6
with r = 0.76 and P < 0.05, δ18OP-G = (0.73 × δ18OS) + 30.1 with
r = 0.90 and P < 0.05, and (not shown) δ18OC-2 =
(−0.30 × δ18OS) + 37.6 with r = 0.14 and P > 0.05.
© The Author (2008)
Journal compilation © New Phytologist (2008)
Review
isotopic noise) will be superior to use of δ18OC in tree-ring
studies. Two observations confirm this expectation. First, it can
be shown that there is a much better fit (r = 0.89, P < 0.05)
between δ18OP-G as the dependent variable and δ18OS and
RH as the independent variables compared with using δ18OC as
the dependent variable (r = 0.78, P < 0.05; Fig. 4). Secondly,
by constructing a linear equation using generic values of
parameters for the tree-ring model (see Sternberg et al., 2007
for these values), it was shown that the linear factors for the
mechanistic equation predicting δ18OP-G as a function of
δ18OS and RH are undistinguishable from those of the best fit
linear regression equation (Sternberg et al., 2007). By contrast,
the factors of the mechanistic linear equation predicting δ18OC
as a function of δ18OS and RH are statistically different from
those of the best fit equation. Therefore, we cannot reconcile
the mechanistic model predicting δ18O values of cellulose
Fig. 4 Observed oxygen isotope ratios of cellulose (δ18OC) and
phenylglucosazone (δ18OP-G) versus that predicted by a best fit
multiple linear regression model using δ18OS and relative humidity
(RH) as independent variables for each of the above chemical
components. The multilinear regression prediction equations are:
δ18OC = (41.73) − (10.63 × RH) + (0.948 × δ18OS) with r = 0.78 and
P < 0.05, δ18OP-G = (43.87) − (17.47 × RH) + (1.076 × δ18OS) with
r = 0.89 and P < 0.05.
New Phytologist (2009) 181: 553–562
www.newphytologist.org
559
560 Review
Research reviews
based on δ18OS and RH with a best fit approach. The use of
δ18OP-G values of cellulose, however, successfully reconciles
the best fit approach of the first phase with the mechanistic
tree-ring model of the second phase.
Future directions
The tree-ring model has been improved by exploring the validity
of the common assumptions adopted during the development
of the model. Leaf temperature may vary much less than was
thought and derivation of cellulose to phenylglucosazone
removes biochemically based isotopic noise from the cellulose
molecule. There are several other assumptions of the tree-ring
model that need to be further explored and may lead to even
more improvements. Below, I comment on some assumptions
that are particularly critical in further development of the
tree-ring model.
Leaf water is under steady state
This assumption has been particularly relevant in partitioning
canopy evapotranspiration into transpiration and evaporation
using isotope ratios of water vapor over canopies (Lai et al.,
2006). It is possible that transpiration is well under way and
the leaf has entered steady state by the time carbohydrates are
being translocated to tree-ring cellulose synthesis. In a study
of oxygen and carbon isotope ratios of the Tasmanian blue
gum (Eucalyptus globulus), Cernusak et al. (2005) compared
the diurnal variation of the observed δ18OL with steady- and
nonsteady-state models. The observed δ18OL values were in
close agreement with both models during the day. However,
large differences between observed δ18OL values and the
steady-state model were seen at night. Therefore, even though
there is a discrepancy between steady- and nonsteady-state
models in E. globulus, this discrepancy is of no consequence in
terms of photosynthate production and the above assumption
holds. The validity of the above assumption must also be
verified with other species.
The δ18O value of cellulose from a tree ring is solely
determined by the average δ18O value of that year’s
photosynthate
This assumption may not be valid, at least for some species.
Empirical correlations between δ18O and δ13C values of treering cellulose and environmental factors have shown that the
correlation often improves with the inclusion of parameters
from the year previous to that represented by the tree-ring
cellulose (Reynolds-Henne et al., 2007). 13CO2 pulse labeling
experiments indicate that the label can persist in cambial
starch for up to 3 yr in slow-growing trees such as Larix
gmelinii (Kagawa et al., 2006). Kagawa et al. (2006) observed
that the carbon isotope ratios in earlywood had the isotopic
signal from the summer and autumn photosynthates of the
New Phytologist (2009) 181: 553–562
www.newphytologist.org
previous year, while latewood seemed to have the climatic signal
of the current year’s photosynthate. These elegant experiments
explain previous observations of the highly significant correlation
between the carbon isotope ratio of earlywood and that of the
latewood from the previous year (Robertson et al., 1997).
However, because there is considerable exchange of oxygen
during heterotrophic cellulose synthesis, it is likely that the
18
O signal of the previous year is dampened off as it is
translocated towards current-year earlywood. Pulse labeling
experiments similar to those of Kagawa et al. (2006), but
using 18O-enriched water, may be fruitful in determining the
influence of the 18O signal of previous years in earlywood.
The δ18O value of leaf water is sensitive to RH
Given the current theory on δ18O values of leaf water, δ18OL
may be more sensitive to stomatal response to environmental
factors other than RH. Stomatal closure, caused for example by
drought, could lead to a greater back flow of the evaporative
pool of water in the leaf and cause an overall enrichment of
leaf water (Cuntz et al., 2007). In addition, stomatal closure
could lead to higher leaf temperatures and increase the vapor
pressure gradient across the leaf. One beautiful example where
18
O in tree-ring cellulose reacts to stomatal closure rather than
RH was recently reported by Brooks & Coulombe (in press).
Nitrogen fertilization in a stand of Pseudotsuga menziesii
caused a flush of leaf growth which increased the foliar to root
area ratio and stomatal sensitivity. This higher stomatal
sensitivity was reflected in the δ18OC of tree rings subsequent to
fertilization. Further studies should concentrate on disentangling
the actual RH signal from the stomatal response signal.
The δ18O value of cellulose is solely determined by the
water available to the plant
There are no experiments that have been able to demonstrate
this. The slope between the δ18O values of cellulose and those of
the water available for photosynthesis in aquarium experiments
is always < 1 (Cooper & DeNiro, 1989; Yakir & DeNiro,
1990; Sauer et al., 2001), whereas, if the above assumption
is valid, one would expect a slope of 1. Several hypotheses
have been proposed regarding possible alternate sources of
oxygen for cellulose synthesis, such as O2 from photorespiration,
which was rejected by Cooper & DeNiro (1989), lack of
equilibrium between CO2 and water before entering
photosynthesis (Cooper & DeNiro, 1989; Yakir, 1992) and
the isotopic effect of metabolic water (Sternberg et al., 2007).
Acknowledgements
This research was supported by NSF grant OPP 0322364 and
DBI 0420533 to LS. I thank Tania Wyss for comments
and Brent Helliker for helpful discussions and for providing
Fig. 2.
© The Author (2008)
Journal compilation © New Phytologist (2008)
Research reviews
References
Ballantyne AP, Rybczynski N, Baker PA, Harington CR, White D. 2006.
Pliocene Arctic temperature constraints from the growth rings and isotopic
composition of fossil larch. Palaeogeography Palaeoclimatology Palaeoecology
242: 188–200.
Barbour MM, Farquhar GD. 2004. Do pathways of water movement
and leaf anatomical dimensions allow development of gradients in H218O
between veins and the sites of evaporation within leaves? Plant, Cell &
Environment 27: 107–121.
Barbour MM, Roden JS, Farquhar GD, Ehleringer JR. 2004. Expressing
leaf water and cellulose oxygen isotope ratios as enrichment above source
water reveals evidence of a Péclet effect. Oecologia 138: 426–435.
Barbour MM, Schurr U, Henry BK, Wong SC, Farquhar GD. 2000.
Variation in the oxygen isotope ratio of phloem sap sucrose from castor
bean. Evidence in support of the Péclet effect. Plant Physiology
123: 671–679.
Bender M, Sowers T, Labeyrie L. 1994. The Dole effect and its variations
during the last 130 000 yr as measured in the Vostok ice core. Global
Biogeochemical Cycles 8: 363– 376.
Brooks JR, Coulombe R. (in press). Physiological responses to fertilization
recorded in tree rings: Isotopic lessons from a long-term fertilization trial.
Ecological Applications.
Burk RL, Stuiver M. 1981. Oxygen isotope ratios in trees reflect mean
annual temperature and humidity. Science 211: 1417–1419.
Cernusak LA, Farquhar GD, Pate JS. 2005. Environmental and
physiological controls over oxygen and carbon isotope composition of
Tasmanian blue gum, Eucalyptus globulus. Tree Physiology 25: 129–146.
Cooper LW, DeNiro MJ. 1989. Oxygen-18 content of atmospheric oxygen
does not affect the oxygen isotope relationship between envrionmental
water and cellulose in a submerged aquatic plant, Egeria densa Planch.
Plant Physiology 91: 536 –541.
Craig H, Gordon LI. 1965. Deuterium and oxygen-18 variations in the
ocean and marine atmosphere. In: Tongiorgi E, ed. Proceedings of the
conference on stable isotopes in oceanographic studies and paleotemperatures.
Piza, Italy: Laboratory of Geology and Nuclear Science, 9–130.
Cuntz M, Ogee J, Farquhar GD, Peylin P, Cernusak LA. 2007. Modelling
advection and diffusion of water isotopologues in leaves. Plant, Cell &
Environment 30: 892–909.
Deniro MJ, Epstein S. 1981. Isotopic composition of cellulose from aquatic
organisms Geochimica et Cosmochimica Acta 45: 1885–1894.
Dole M, Lane GA, Rudd DP, Zaukelies DA. 1954. Isotopic composition
of atmospheric oxygen and nitrogen. Geochimica et Cosmochimica Acta
6: 65–78.
Epstein S, Thompson P, Yapp CJ. 1977. Oxygen and hydrogen isotopic
ratios in plants cellulose. Science 198: 1209–1215.
Farquhar GD, Lloyd J. 1993. Carbon and oxygen isotope effects in the
exchange of carbon dioxide between terrestrial plants and the atmosphere.
In: Ehleringer JR, Hall AE, Farquhar GD, eds. Stable isotopes and plant
carbo -water relations. New York, NY, USA: Academic Press, 47–70.
Farquhar GD, Lloyd J, Taylor JA, Flanagan LB, Syvertsen JP, Hubick KT,
Wong SC, Ehleringer JR. 1993. Vegetation effects on the isotope
composition of oxygen in atmospheric CO2. Nature 363: 439–443.
Flanagan LB, Ehleringer JR. 1991. Stable isotope composition of stem and
leaf water-applications to the study of plant water-use. Functional Ecology
5: 270–277.
Francey RJ, Tans PP. 1987. Latitudinal variation in qxygen-18 of
atmospheric CO2. Nature 327: 495–497.
Gessler A, Peuke AD, Keitel C, Farquhar GD. 2007. Oxygen isotope
enrichment of organic matter in Ricinus communis during the diel course
and as affected by assimilate transport. New Phytologist 174: 600 –613.
Guy RD, Fogel ML, Berry JA. 1993. Photosynthetic fractionation of the
stable isotopes of oxygen and carbon. Plant Physiology 101: 37–47.
Helliker BR, Griffiths H. 2007. Toward a plant-based proxy for the isotope
ratio of atmospheric water vapor. Global Change Biology 13: 723–733.
© The Author (2008)
Journal compilation © New Phytologist (2008)
Review
Helliker BR, Richter SL. 2008. Subtropical to boreal convergence of tree-leaf
temperatures. Nature 454: 511–514.
Hill SA, Waterhouse JS, Field EM, Switsur VR, Rees T. 1995. Rapid
recycling of triose phosphates in oak stem tissue. Plant, Cell & Environment
18: 931–936.
Jahren AH, Sternberg LdSL. 2003. Humidity estimate for the middle eocene
Arctic rain forest. Geology 31: 463 –466.
Kagawa A, Sugimoto A, Maximov TC. 2006. 13CO2 pulse-labelling of
photoassimilates reveals carbon allocation within and between tree rings.
Plant, Cell & Environment 29: 1571–1584.
Lai CT, Ehleringer JR, Bond BJ, Paw UKT. 2006. Contributions of
evaporation, isotopic nonsteady state transpiration and atmospheric
mixing on the δ18O of water vapour in Pacific Northwest coniferous
forests. Plant, Cell & Environment 29: 77–94.
Libby LM, Pandolfi LJ, Payton PH, Marshall J, Becker B, Giertzsienbenlist
V. 1976. Isotopic tree thermometers. Nature 261: 284–288.
Luo YH, Sternberg L. 1992. Hydrogen and oxygen isotope fractionation
during heterotrophic cellulose synthesis. Journal of Experimental Botany
43: 47–50.
Masson-Delmotte V, Raffalli-Delerce G, Danis PA, Yiou P, Stievenard M,
Guibal F, Mestre O, Bernard V, Goosse H, Hoffmann G et al. 2005.
Changes in European precipitation seasonality and in drought frequencies
revealed by a four-century-long tree-ring isotopic record from Brittany,
western France. Climate Dynamics 24: 57–69.
Ramesh R, Bhattacharya SK, Gopalan K. 1986. Climatic correlations in the
stable isotope records of silver fir (Abies pindrow) trees from Kashmir,
India. Earth and Planetary Science Letters 79: 66–74.
Rebetez M, Saurer M, Cherubini P. 2003. To what extent can oxygen
isotopes in tree rings and precipitation be used to reconstruct
past atmospheric temperature? A case study. Climatic Change
61: 237–248.
Reynolds-Henne CE, Siegwolf RTW, Treydte KS, Esper J, Henne S,
Saurer M. 2007. Temporal stability of climate-isotope relationships in
tree rings of oak and pine (Ticino, Switzerland). Global Biogeochemical
Cycles 21: 12.
Richter SL, Johnson AH, Dranoff MM, Taylor KD. 2008.
Continental-scale patterns in modern wood cellulose δ18O: Implications
for interpreting paleo-wood cellulose δ18O. Geochimica et Cosmochimica
Acta 72: 2735–2743.
Robertson I, Switsur VR, Carter AHC, Barker AC, Waterhouse JS, Briffa
KR, Jones PD. 1997. Signal strength and climate relationships in 13C/12C
ratios of tree ring cellulose from oak in east England. Journal of Geophysical
Research 102: 19507–19516.
Robertson I, Waterhouse JS, Barker AC, Carter AHC, Switsur VR. 2001.
Oxygen isotope ratios of oak in east England: Implications for
reconstructing the isotopic composition of precipitation. Earth and
Planetary Science Letters 191: 21–31.
Roden JS, Bowling DR, McDowell NG, Bond BJ, Ehleringer JR. 2005.
Carbon and oxygen isotope ratios of tree ring cellulose along a
precipitation transect in Oregon, United States. Journal of Geophysical
Research – Biogeosciences 110: G02003.
Roden JS, Ehleringer JR. 2000. Hydrogen and oxygen isotope ratios
of tree ring cellulose for field-grown riparian trees. Oecologia 123:
481–489.
Roden JS, Lin G, Ehleringer JR. 2000. A mechanistic model for
interpretation of hydrogen and oxygen isotope ratios in tree-ring cellulose.
Geochimica et Cosmochimica Acta 64: 21–35.
Sauer PE, Miller GH, Overpeck JT. 2001. Oxygen isotope ratios of organic
matter in arctic lakes as a paleoclimate proxy: Field and laboratory
investigations. Journal of Paleolimnology 25: 43– 64.
Saurer M, Borella S, Leuenberger M. 1997. δ18O of tree rings of beech
(Fagus silvatica) as a record of δ18O of the growing season precipitation.
Tellus Series B – Chemical and Physical Meteorology 49: 80 –92.
Schmidt H-L, Werner RA, Robmann A. 2001. 18O Pattern and biosynthesis
of natural plant products. Phytochemistry 58: 9–32.
New Phytologist (2009) 181: 553–562
www.newphytologist.org
561
562 Review
Research reviews
Sternberg L, Deniro MJD. 1983. Biogeochemical implications of the
isotopic equilibrium fractionation factor between the oxygen atoms of
acetone and water. Geochimica et Cosmochimica Acta 47: 2271–2274.
Sternberg L, Pinzon MC, Anderson WT, Jahren AH. 2006. Variation in
oxygen isotope fractionation during cellulose synthesis: intramolecular
and biosynthetic effects. Plant, Cell & Environment 29: 1881–1889.
Sternberg LDL, Anderson WT, Morrison K. 2003. Separating soil and leaf
water 18O isotopic signals in plant stem cellulose. Geochimica et
Cosmochimica Acta 67: 2561–2566.
Sternberg LDL, Deniro MJ, Savidge RA. 1986. Oxygen isotope exchange
between metabolites and water during biochemical reactions leading to
cellulose synthesis. Plant Physiology 82: 423– 427.
Sternberg LDL, Pinzon MC, Vendramini PF, Anderson WT, Jahren AH,
Beuning K. 2007. Oxygen isotope ratios of cellulose-derived
phenylglucosazone: An improved paleoclimate indicator of environmental
water and relative humidity. Geochimica et Cosmochimica Acta
71: 2463–2473.
Treydte KS, Schleser GH, Helle G, Frank DC, Winiger M, Haug GH,
Esper J. 2006. The twentieth century was the wettest period in northern
Pakistan over the past millennium. Nature 440: 1179–1182.
Waterhouse JS, Switsur VR, Barker AC, Carter AHC, Robertson I. 2002.
Oxygen and hydrogen isotope ratios in tree rings: How well do models
predict observed values? Earth and Planetary Science Letters 201: 421–430.
Woodward FI. 2008. Forest air conditioning. Nature 454: 422–423.
Yakir D. 1992. Variations in the natural abundance of oxygen-18
and deuterium in plant carbohydrates. Plant, Cell & Environment
15: 1005–1020.
Yakir D, DeNiro MJ. 1990. Oxygen and hydrogen isotope fractionation
during cellulose metabolism in Lemna gibba L. Plant Physiology
93: 325– 332.
About New Phytologist
• New Phytologist is owned by a non-profit-making charitable trust dedicated to the promotion of plant science, facilitating projects
from symposia to open access for our Tansley reviews. Complete information is available at www.newphytologist.org.
• Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged.
We are committed to rapid processing, from online submission through to publication ‘as-ready’ via Early View – our average
submission to decision time is just 29 days. Online-only colour is free, and essential print colour costs will be met if necessary.
We also provide 25 offprints as well as a PDF for each article.
• For online summaries and ToC alerts, go to the website and click on ‘Journal online’. You can take out a personal subscription to
the journal for a fraction of the institutional price. Rates start at £139 in Europe/$259 in the USA & Canada for the online edition
(click on ‘Subscribe’ at the website).
• If you have any questions, do get in touch with Central Office ([email protected]; tel +44 1524 594691) or, for a local
contact in North America, the US Office ([email protected]; tel +1 865 576 5261).
New Phytologist (2009) 181: 553–562
www.newphytologist.org
© The Author (2008)
Journal compilation © New Phytologist (2008)