Tải bản đầy đủ - 0 (trang)
2 Oxygen Isotope Studies: Factors Influencing delta18O of Tree Cellulose

2 Oxygen Isotope Studies: Factors Influencing delta18O of Tree Cellulose

Tải bản đầy đủ - 0trang

38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



813



Fig. 38.1 Plant physiological and environmental/climatic factors influencing oxygen isotope ratios of plant cellulose



38.2.1 Atmospheric/Environmental

Processes

This section describes atmospheric and environmental

factors influencing plant cellulose d18O. Among these,

d18O of precipitation and atmospheric water vapor,

and relative humidity are important atmospheric factors whereas soil hydrological processes are crucial

environmental factors. d18O of precipitation and its

subsequent evolution during various soil hydrological

processes determine d18O of water trees take up

through roots and use for photosynthesis. d18O of

water vapor and relative humidity modify the source

water d18O during transpiration. As this modified

water is actually used for photosynthesis, a good

knowledge of what decides d18O of local precipitation

and atmospheric water vapor, and changes in relative

humidity during the growing season are required to

interpret the tree cellulose d18O record.



38.2.1.1 d18O of Precipitation

A detailed account of word-wide variation in the d18O

and dD of rainfall since 1961 is maintained by the

Global Network “Isotopes in Precipitation” (GNIP)



(IAEA/WMO 2006: http://isohis.iaea.org). On the

global scale, d18O of precipitation is largely governed

by ambient temperature. Decreasing temperature

results in reducing saturation vapor pressure of air

and drives the rainout process. As a result, the progressive precipitation associated with still lower temperatures becomes increasingly depleted in 18O. Dansgaard

(1964) demonstrated a linear relationship between surface air temperature ðTannual Þ and d18O of mean annual

precipitation ðd18 Oa Þ on the global scale as

d18 Oa ¼ 0:695Tannual À 13:6:



(38.2)



In the tropics, the typical relationship between d18O

of rainfall and surface air temperature is overshadowed by the amount effect, an inverse relationship

between d18O of rainfall and amount of precipitation

on the monthly (Dansgaard 1964; Rozanski et al.

1993) or individual rain event (Miyake et al. 1968;

Yadava et al. 2007) scale. Dansgaard (1964) explained

the amount effect in terms of (1) fractional removal

of heavy isotopes in the rain; (2) equilibration of

light rain (smaller drops) with enriched vapor below

the cloud base; and (3) higher relative loss of light

isotopes when raindrops evaporate below the cloud

base in arid region. A recent model by Risi et al.

(2008) suggests reevaporation of the falling rain, and



814



S.R. Managave and R. Ramesh



diffusive exchange with the surrounding vapor, and

recycling of the subcloud layer vapor feeding the

convective system as the likely causes for the amount

effect. Yurtsever and Gat (1981) analyzed rainfall and

its isotopic composition at 14 island stations in the

equatorial belt, and found a linear relationship (linear

correlation coefficient r ¼ 0.87) between mean

monthly d18O of rainfall (d18Om) and mean monthly

rainfall amount (Pm)

d18 Om ¼ 0:015 ặ 0:002ịPm 0:47 ặ 0:42ị:

(38.3)

The average rate of depletion in d18O of rainfall with

increase in monthly rain amount was À1.5 Ỉ 0.2‰

for 100 mm. This amount effect prevails in the precipitation in south and south-east Asia (Aragua´s-Aragua´s

et al. 1998). Yadava and Ramesh (2005) measured

rainfall amount and its d18O for all monsoon rain

events in 1999 at Jharsuguda (22 N, 84 E), central

India, and found a depletion rate of À9.2 Ỉ 1.1‰ per

100 mm rain for individual rain events and À2.2 Ỉ

0.8‰ per 100 mm of rain for monthly total rainfall.

The correlation of d18O of precipitation with air temperature (amount of precipitation) is mostly prevalent

poleward (equatorward) of 30 N/S (Fig. 38.2, Bowen

2008). More than 66% of the total stations with

recorded isotope data of precipitation, and more than

80% of the earth’s land surface exhibit relationships

between climate variables (temperature and precipitation amount) and the isotopic composition of precipitation (Bowen 2008). Figure 38.2 exhibits regions

where temperature and amount of precipitation correlate with d18O of precipitation. Trees growing in such

areas are likely to preserve information on past precipitation/temperature, and are most suited for isotope

dendroclimatology.

Rainout is a process by which a moving air mass

loses its water vapor through precipitation and the

remaining vapor becomes progressively depleted in

18

O. If at each stage within the cloud there is isotopic

equilibrium between rain and vapor, d18O (and dD) of

a moving air parcel can be modeled by Rayleigh

fractionation (Clark and Fritz 1997)

d18 Ovf ị ẳ do18 Ov ỵ e18 OlÀv . lnf ;



(38.4)



where d18 OVðf Þ and do18 OV are d18O of the residual

fraction f of water vapor and the initial water vapor,



respectively; e18 OlÀv the oxygen isotopic fractionation factor between the rain and vapor expressed in

per mil (‰) units. At about 25 C, d18O of accompanying rainfall is about 9‰ higher relative to the

vapor. An example of rainout effect in the Asian

region is illustrated by Aragua´s-Aragua´s et al.

(1998), who showed a progressive depletion of d18O

of rainfall along the trajectory of Pacific monsoon

from Haikou and Hong Kong (south of China) to

Lhasa (Tibetan plateau) where d18O of rainwater

changed from À7.2 to À18.3‰, demonstrating the

“continental” effect.

In addition to the amount and rainout (continental)

effects, d18O of precipitation also changes with season

(Fig. 38.3), plausibly implying different sources of

vapor. In India, the southwest (SW) and northeast

(NE) monsoons have different isotopic signatures

(Geeta Rajagopalan 1996). Yadava et al. (2007) analyzed d18O of daily rains for 2 years at Mangalore

(12.53 N and 74.52 E), which receives both the monsoons and showed that the NE monsoon precipitation

is relatively more depleted in 18O. This contrasts with

observations elsewhere in Southeast Asia, where summer rains are depleted in 18O relative to winter rains

(Fig. 38.3, Aragua´s-Aragua´s et al. 1998).



38.2.1.2 d18O of Atmospheric Vapor

d18O of atmospheric water vapor is one of the factors

that decide the d18O of leaf water and its effectiveness

increases with relative humidity (Roden et al. 2000).

Higher relative humidity facilitates towards achieving

isotopic equilibration of atmospheric water vapor with

leaf water. Unfortunately, the isotopic composition

of the atmospheric water vapor has not been monitored

as extensively as that of precipitation. In general,

atmospheric vapor is assumed to be in isotopic equilibrium with local rainwater and its isotopic composition calculated using the corresponding equilibrium

fractionation factor. It is known that d18O of atmospheric water vapor depends upon the moisture source

and its interaction with the surface (Gat 1996; Tian

et al. 2007). Further, mixing of water vapor from

ocean and local evaporation is likely to affect d18O

of local atmospheric water vapor. Srivastava (2009)

found the atmospheric water vapor, in general, to be

in equilibrium with ambient rain during the Indian

monsoon.



38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



815



Fig. 38.2 Significance

level of correlation between

long-term monthly average

precipitation d18O and

climatological parameters.

(a, b) Stations and grid area

with significant (p < 0.05;

black dots and dark gray

shading) and no significant

(white dots, light gray

shading) correlation

between temperature

(a) or precipitation

amount (b) and d18O.

(c, d) Mode (white

indicates temperature;

black indicates

precipitation) of the

strongest climatological

correlation at all stations

(c) or grid points with one

or more significant

correlation (d) (from

Bowen 2008)



38.2.1.3 Relative Humidity

Atmospheric relative humidity too plays an important

role in governing d18O of cellulose. During transpiration, lower ambient relative humidity increases the

evaporative enrichment of 18O in leaf water, which,



in turn leads to the synthesis of cellulose enriched

in 18O. It has been shown that in drier locations soil

moisture deficit affects the d18O of cellulose (Saurer

et al. 1997; Roden et al. 2005) as well.

In the tropics, quite often there might be a correlation between rainfall and relative humidity. If so, past



816



S.R. Managave and R. Ramesh



Fig. 38.3 Spatial distribution of the

difference (d18Osummer-d18Owinter) at Asian

stations. The dashed line separates the area

of positive and negative values of the

difference, and coincides with the

maximum northward extent of the

Intertropical Convergence Zone (ITCZ) in

the region during summer (from Aragua´sAragua´s et al. 1998)



rainfall could be reconstructed even in the case when

the amount effect is absent or only weakly present;

higher rainfall, through its effect on relative humidity,

would result in lower enrichment of the leaf water

d18O and that of the subsequently formed cellulose.

Further, changes in the atmospheric circulation pattern, especially with the movement of the Inter Tropical Convergence Zone (ITCZ), can induce seasonal

variation in relative humidity in some parts of the

tropics (Managave et al. 2010d).



38.2.1.4 Soil Hydrological Processes

While rain water is the source water for plants, various

subsequent soil hydrological processes operating on

the percolated rain water decide the d18O of water

available to plant. Understanding and quantification

of soil hydrological processes (viz. water percolation

in soil and its evaporation), however, is yet to receive

sufficient attention in dendroclimatology.

The isotopic composition of soil water primarily

depends upon the isotopic composition of precipitation

and ground water. In addition, environmental factors

(viz. relative humidity, solar irradiance), through their

effect on evaporation of soil water, also affect the

isotopic composition of soil water. Spatial and temporal heterogeneity in the isotopic composition of soil

water has been reported. Barnes and Allison (1989)

showed that an isotopic gradient exists in soil water

due to evaporation in the upper part of the soil. The



shape of this profile depends upon soil moisture content, soil texture, and changing isotopic composition

and amount of precipitation; the maximum heterogeneity in d18O and dD of soil water is observed at the

soil surface and it gradually decreases with depth.

Tang and Feng (2001) conducted a detailed field

investigation of these effects in Hanover, NH, USA,

where precipitation is more evenly distributed throughout the year, with isotopically enriched summer and

depleted winter rainfall. By measuring the temporal

variation in the isotopic composition of rainwater, soil

water at different depths and twig water from a maple

tree, they demonstrated that (1) soil water isotopic

composition is much less variable than that of precipitation, implying mixing of water from various precipitation events; (2) evaporation isotopically enriches the

surface soil water; (3) water from summer rains gradually replaces water of winter precipitation, the extent

of which depends upon the intensity and frequency

of summer precipitation; its influence decreases with

depth; and (4) d18O of twig water reflects that of the

soil water that has experienced evaporation.

Tsuji et al. (2006), in their study of d18O of coexisting tree species in Hokkaido, Japan, have shown the

importance of the rooting system of plants in controlling the plant’s isotopic composition: water uptake

characteristics of plants in the rhizosphere (a part of

soil affected by root system), in addition to the soil

water processes, ultimately decide the tree cellulose

isotopic composition. They found that shallow-rooted

tree species (spruce) responded to d18O of summer



38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



817



precipitation unlike deep-rooted (oak) trees, which

may not sample summer precipitation.

Isotopic analysis of xylem water in conjunction

with those of rain and ground water helps study

water utilization by plants. White et al. (1985) have

shown that a single white pine (Pinus Strobus) tree

could rapidly (within 3 days) switch its uptake from

shallower surface rain water to deeper groundwater.

Dawson and Ehleringer (1991), on the contrary,

showed that mature riparian trees growing close

to perennial streams use little or no stream water,

but instead prefer ground water. They explained this

behavior in terms of mature trees preferring reliable

ground water instead of unreliable surface or rain

water. The source water for the tree may change with

its age. Dawson (1996) showed that older (larger) Acer

saccharum trees had access to both shallow soil and

ground water while younger (smaller) trees of the

species depended on shallow soil water. Such results

are important as different sources of water often have

different isotopic compositions and decide the isotopic

composition of tree cellulose.



enrichment of the leaf water. Variation in the isotopic

composition of the leaf water (and hence that of cellulose synthesized from it) depends upon the isotopic

composition of source water (therefore a function of

ambient temperature or amount of precipitation) and

the extent of isotopic enrichment of leaf water due to

evaporation (therefore a function of ambient relative

humidity).

Plant physiological models (e.g. Flanagan et al.

1991; Saurer et al. 1997; Farquhar et al. 1998; Barbour

and Farquhar 2000; Roden et al. 2000; Barbour et al.

2004; Oge´e et al. 2009) are useful for the quantitative

interpretation of the isotopic composition of plant

cellulose.

Various models have been proposed to calculate the

isotopic composition of the leaf water. Such models

are based on detailed theoretical as well as laboratory

experiments. Based on the model of Craig and Gordon

(1965) describing isotopic fractionation during evaporation from an open surface and then incorporating

leaf boundary layer effects and diffusion through stomata, Dongmann et al. (1974) and Flanagan et al.

(1991) expressed the isotopic composition of the leaf

water (Rwl) as



38.2.2 Modeling Plant Physiological

Processes Affecting Cellulose d18O











ei e s

e s ea

ỵ akb Rwx

ẳ a ak Rwx

ei

ei

 

ea

ỵ Ra

;

ei

(38.5)



Trees take up the soil water through roots with little

isotopic fractionation (White et al. 1985) and transport

it to the leaf through the xylem. Analysis of xylem

water/cellulose shows (Ehleringer and Dawson 1992;

Lin et al. 1996; Schwinning et al. 2002; Evans and

Schrag 2004) that plants do preserve the isotopic composition of precipitation. The extent of isotopic enrichment of the leaf water depends primarily upon vapor

pressure difference between the intercellular air space

of the leaf and ambient air, which in turn depends upon

the leaf and air temperature, and relative humidity.

Effect of relative humidity on the evaporative isotopic

enrichment of the leaf water is lower at higher ambient

humidity (e.g. monsoon), a condition often characteristic of the main growing season in the tropics. But at

higher ambient humidity, the isotopic composition of

atmospheric water vapor through its exchange with the

leaf water becomes important. During the drier late

growing season (e.g. post-monsoon), lower humidity

creates higher vapor pressure gradients across the

leaf, which results in a higher evaporative isotopic



Rwl



Ã







where Rwl, Rwx and Ra refer to isotopic ratios

(18O/16O) of leaf water, xylem water and bulk air,

respectively; water vapor pressure of intercellular

leaf space is ei, of leaf surface is es and of bulk air is

ea; a*, ak, and akb are respectively liquid-vapor equilibrium and kinetic fractionation factors associated

with the leaf, and leaf-air boundary layer.

Alternatively, d18O of the leaf water at the site of

evaporation ðD18 Oe Þ with respect to that of the source

water was written by Farquhar and Lloyd (1993) as

D18 Oe ẳ e ỵ ek ỵ D18 Ov ek Þea =ei ;



(38.6)



where eà and ek are equilibrium and kinetic fractionation factors, respectively. D18 Ov is d18O of

atmospheric water vapor relative to the source water.

ea and ei are the same as described earlier.



818



In (38.5) and (38.6), ei depends upon the leaf temperature, generally assumed to be the same or linearly

related to the ambient temperature. However, Helliker

and Richter (2008), based on tree-ring d18O data

across the latitudinal range of 50 , demonstrated

the possible role of plant physiology in moderating

leaf temperature with respect to ambient temperature,

wherein a constant value of the leaf temperature

(21.4 Ỉ 2.2 C) was observed across a range 50 of

latitude from boreal to subtropical biomes.

The isotopic composition of the leaf water calculated using the above (38.5) and (38.6) is more

enriched than that of the observed bulk leaf water

due to the Pe´clet effect – transpirational advection of

18

O depleted (xylem) water to the evaporating site

opposed by the backward diffusion of 18O enriched

water from the leaf (Farquhar and Lloyd 1993). The

model of Barbour et al. (2004) includes the Pe´clet

effect. Determining the Pe´clet number for different

tree species involves the measurement of effective

path length; a model parameter that accounts for the

discrepancy between the d18O values predicted by the

Craig-Gordon model and measured bulk leaf water.

Generally, a detailed knowledge of the Pe´clet number

for a given species and for any part of the growing

season is unavailable. This difficulty is usually circumvented by either assuming a reasonable value of

the effective path length (and Pe´clet number) or by

using models (e.g. Roden et al. 2000) that ignore

the Pe´clet effect. The results of the latter option

are interpreted keeping in mind that the leaf water in

such models could be somewhat enriched in 18O

relative to actual values. This might be preferable

until the Pe´clet number for more species of trees

during various times of the growing season are carefully measured.

Sucrose formed by photosynthesis in the leaf

carries the isotopic composition of the leaf water,

modified by the Pe´clet effect. Subsequent biochemical

pathways leading to formation of cellulose are described by Farquhar et al. (1998). Based on review of

available data in literature, Sternberg (1989) reported

a value of 27 Ỉ 3‰ as the enrichment for cellulose

d18O of exchangeable oxygen in carbohydrate relative

to the water at site of synthesis. Further, it has been

shown (Sternberg et al. 1986; Saurer et al. 1997) that

45% of oxygen in cellulose is expected to exchange

with (xylem) water during synthesis of cellulose from

sucrose. This implies that ~55% of oxygen atoms in



S.R. Managave and R. Ramesh



cellulose carry the signature of evaporation (and hence

relative humidity conditions) taking place in the leaf.

Roden et al. (2000) outlined a mechanistic model

for interpreting d18O and dD of tree cellulose. Their

model considers exchange of oxygen atoms of sucrose

with the medium (xylem) water during synthesis of

cellulose. In their model, the final isotope composition

of tree celluloseð d18 Ocx ịis given by

d18 Ocx ẳ fO (d18 Owx ỵ eO ị

ỵ 1fO ) (d18 Owl ỵ eO Þ;



(38.7)



Where fO is the fraction of oxygen that undergoes

exchange with medium water, d18 Owx is xylem

water, d18 Owl refers to d18O of leaf water at the site

of sucrose synthesis and eO , the biochemical fractionation factor. The authors estimated fO to be ~0.42

based on model and experimental results.

It can be seen from (38.5) to (38.7) that the ea =ei

ratio, hence atmospheric relative humidity, and d18O

of precipitation play a crucial role deciding d18O of

cellulose. The importance of relative humidity in

deciding plant d18O is highlighted by Sheshshayee

et al. (2005). In the tropics, there is enough variation

in relative humidity during growing season to leave

its imprint on the isotopic composition of cellulose

(Geeta Rajagopalan et al. 1999). Likewise, as a result

of the amount effect, precipitation amount is also

expected to affect tree cellulose d18O.

The interpretation of cellulose d18O data from tropics, however, may not be straight-forward; factors

such as the duration of the growing season in relation

to seasonal changes in relative humidity, and the

strength of the amount effect in rainfall are likely to

influence d18O values of trees. Managave et al. (2010d)

showed varying response of cellulose d18O of teak

trees in differing monsoon environments to monsoon

rainfall. Their work underscored the importance of

considering movement of the ITCZ and associated

changes in relative humidity and isotopic composition

of rainfall while interpreting isotopic record of tropical

trees. Teak from southern India shows a significant

negative correlation with the amount of rainfall, plausibly because of a more pronounced amount effect in

the region. Cellulose d18O of teak trees from western

and central India show a significant positive correlation with the amount of rainfall, contrary to expectation. The positive correlation between teak cellulose



38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



dD and local rainfall record has been also demonstrated for teak from western India by Ramesh et al.

(1989). This is possibly a result of rainfall-induced

changes in the length of growing season in relation to

the movement of the ITCZ. In India, the movement of

the ITCZ divides the growing season into two parts: the

humid monsoon period when the southwesterly oceanic air mass brings moisture to the continent and the

(relatively) dry post-monsoon period dominated by

northeasterly flowing continental air mass. Higher

rainfall leads to lengthening of the growing season

and teak grows until a period of the lower relative

humidity (post-monsoon season) leading to more

evaporative enrichment of the leaf water in 18O and

hence higher cellulose d18O values.



38.3 Carbon Isotopes in Trees

Isotopic composition of carbon fixed during photosynthesis of C3 plant material (d13CPlant) depends mainly

upon factors that control the CO2 concentration inside

the leaf and d13C of atmospheric air. Figure 38.4 illustrates various factors influencing d13C of plant cellulose. Farquhar et al. (1982) proposed an equation for

the d13C of plant material (d13CPlant) as

d13 CPlant ¼ d13 CAir ÀaÀðbÀaÞðci =ca Þ;



(38.8)



819



where a and b are constants and represent respectively

the carbon isotope fractionations during the diffusion

of CO2 through stomata (4.4‰), and during enzymatic

CO2 fixation (27‰); ci =ca is the ratio of the intercellular to atmospheric CO2 concentration; d13CAir is

d13C of atmospheric CO2; ci is further related to the

(photosynthetic) assimilation rate of CO2 (A) and

stomatal conductance ðgs Þ in the following way.

ci ¼ ca À A=gs :



(38.9)



Thus the main climatic parameters influencing d13C

of carbon fixed during photosynthesis are those

controlling gs and A. Dominant factors controlling gs

are relative humidity, temperature and soil moisture

content whereas A depends on the photon flux as well.

In the moist, high altitude environments d13CPlant is

governed by photosynthetic rate (Gagen et al. 2007)

whereas gs controls d13CPlant at drier environment.

Seasonal soil moisture status is shown to affect seasonal variations in d13C of tree rings (Leavitt 2002,

2007). Cloud amount can modulate photon flux

thereby affecting A and hence ci and d13Cplant. This

was illustrated by Ramesh et al. (1986b) who observed

a negative correlation between cellulose d13C and

cloud cover. The effect of cloud cover over the tropics

in limiting light availability for photosynthesis is illustrated by Graham et al. (2003), who showed that the

heavy cloud cover associated with La Nin˜a in the



Fig. 38.4 Plant physiological and environmental/climatic factors influencing carbon isotope ratios of tree



820



Republic of Panama resulted in the reduction of photosynthesis and vegetation growth. It was also reported

that at a given site both the net radiation and vapor

pressure deficit could affect d13C variations of tree

cellulose (Vaganov et al. 2009). Diefendorf (2010)

found that on the global scale carbon isotope fractionation during assimilation by plants is correlated

with mean annual precipitation indicating watersupply-mediated stomatal control on the leaf-gas

exchange processes as a dominant factor.



38.3.1 Non-Climatic Factors Influencing

d13C of Tree Rings

In addition to plant physiological responses, d13Cplant

values are affected by d13Catm and its concentration

ðca Þ, both changing over the industrial period. Addition of fossil fuel carbon to the atmosphere has

decreased the d13Catm throughout the industrial period

(Suess effect), with the decrease of 1.52‰ during AD

1850–1998 (McCarroll and Loader 2004). A calibration of the d13C record of tree rings with ambient

climatic parameters, without due attention to increasing ca and depleting d13Catm is likely to lead to an

erroneous reconstruction of past climate.

The post-AD 1850 decline observed in d13C record

of trees cannot be fully accounted by the decrease in

d13Catm alone (Waterhouse et al. 2004; Gagen et al.

2007; Loader et al. 2007). Changes in the plant physiological response of trees to the increased CO2 concentration could also be a likely reason. Before relating the

observed tree ring d13C variations to meteorological

data, it is necessary to remove the changes in tree ring

d13C values induced by both the declining d13Catm

values and increasing ca . A correction for the decrease

in d13Catm due to the Suess effect is made either by

fitting a smooth curve to tree ring d13C variation and

then subtracting the fitted values (Freyer 1986) or by

subtracting the d13Catm as obtained from ice cores and

instrumental data from tree ring d13C values (McCarroll

and Loader 2004). Any remaining decreasing trend in

corrected tree ring d13C (d13Ccorr) could partly be due to

increasing ca . This is accounted by (1) considering

potential linear discrimination from À0.05 to 0.1‰

per ppm change in CO2 concentration (Treydte et al.

2009) (2) by non-linear de-trending of low frequency

d13Ccorr (e.g. McCarroll et al. 2009).



S.R. Managave and R. Ramesh



A “juvenile effect” (depleted and gradually increasing d13C in the first few decades of trees) is another

non-climatic cause influencing d13C of tree rings

(Freyer and Belacy 1983; McCarroll and Pawellek

2001; Gagen et al. 2004; Anderson et al. 2005). The

explanations given for this effect are (1) trees grow in

shade during early years of growth (Francey and

Farquhar 1982) (2) trees ingest soil respired CO2,

depleted in d13C, during early age and the intake

reduces as tree gains height, and (3) reducing stomatal

conductance with age as trees gain height (McDowell

et al. 2002). It is not possible to detect the juvenile

effect of individual trees when the samples are pooled

before extraction of cellulose. If individual trees are to

be considered for d13C measurements, then the rings

corresponding to the juvenile effect are best avoided

(McCarroll and Loader 2004). This provides an advantage of using non-detrended d13C series and hence no

climatic signal is lost. Alternatively, the “regional

curve standardization” technique (e.g. Briffa et al.

2001) could be used to correct for the juvenile effect

(Gagen et al. 2008). Here tree ring series are aligned

according to their cambial age and a standardized

juvenile growth trend is obtained. This trend is then

used to remove juvenile effect.



38.4 Recent Developments in Isotope

Dendroclimatology

38.4.1 Post-Photosynthetic Processes

and Isotopic Composition

of Tree Rings

In addition to the processes occurring during photosynthesis, post-photosynthetic processes also modify

d18O and d13C of tree cellulose. A good mechanistic

understanding has been gained on isotopic fractionation during the formation of photosynthates (Farquhar et al. 1982; Roden et al. 2000; Barbour et al.

2004) (see (38.5), (38.6) and (38.8)). However,

understanding of the post-photosynthetic processes

is limited. Evolution of the isotopic composition of

photosynthates while being transported from the leaf

to the tree trunk needs to be quantitatively explored

for better interpretation of isotopic composition of

tree rings.



38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



It has been observed that d13C and d18O of newly

assimilated sugar is correlated with photosynthesisweighted ci =ca (Brandes et al. 2006) and photosynthesisweighted d18O of mean lamina mesophyll water

(Cernusak et al. 2005), respectively. The latter have

shown correspondence between phloem sap sugar

exported from leaves of E. globules and observed leaf

water enrichment when diurnal variation in photosynthesis was taken into account. During the night or early

morning, enrichment in d13C values of soluble organic

matter has been reported (Brandes et al. 2006; Gessler

et al. 2007a; Kodama et al. 2008). This necessitates

consideration of diel biochemical processes fractionating isotopes of exported sugar from the leaf. Further,

Gessler et al. (2008) have shown that d13C of leafexported carbon, integrated over the day (24 h), is

enriched with respect to the primary assimilate, possibly explaining the enrichment in d13C of tissues in C3

plants produced by non-photosynthetic processes compared to that produced by photosynthetic processes

(Cernusak et al. 2009 and references therein).

Diel variations in d13C have also been observed in

phloem sap sugar (Gessler et al. 2008). Comparative

studies of diel variations in d13C of water soluble

organic matter of leaf, twig phloem sap and trunk

phloem of Pinus sylvestris (Kodama et al. 2008)

showed a progressive dampening of variations during

transport from the leaf to the trunk. Similar observations have been reported by Betson et al. (2007) for

Picea abies and by Gessler et al. (2007a), for Eucalyptus delegatensis. Mixing of the different carbon pools

having different metabolic histories could be the likely

reason for this (Brandes et al. 2007). By monitoring

modifications of the isotopic signals of d13C and d18O

from the leaf to phloem and to tree ring in Pinus

sylvestris, Gessler et al. (2009) have reported the

uncoupling of d13C in the tree cellulose and the ci =ca

ratio at the leaf level, implying variable enrichment in

13

C of sugar due to phloem loading and transport. The

authors could trace time-lagged transfer of d18O signal

from the leaf water to the tree cellulose. Damesin and

Lelarge (2003) have demonstrated that the d13C of stem

of Fagus sylvatica is affected by isotopic fractionation

during transfer of sugar from leaves to stems and

during stem respiration. They further cautioned that

these processes may partially mask the climatic signal

inherited during photosynthesis. As freshly developing

xylem tissue carries the isotopic signature of phloem

sugars (Cernusak et al. 2005; Gessler et al. 2007b),



821



the dampened climate signal observed in phloem sap

is likely preserved in tree cellulose.

Further dampening of the climate signal could

occur during synthesis of cellulose in the stem where

45% of oxygen in cellulose is expected to exchange

with (xylem) water (Sternberg et al. 1986, and Saurer

et al. 1997). Meinzer et al. (2006) gave an account of

whole-tree water transport and storage properties in

coniferous trees Pseudotsuga menziesii (Mirb.) and

Tsuga heterophylla (Raf.) Sarg. They observed the

transit times for D2O labeled water from the tree base

to the upper crown ranged from 2.5 to 21 days, and the

residence times of labeled water in tree ranged from 36

to 79 days, with the values increasing with the diameter

of trees. This entails water from a given precipitation

event being taken up by roots continuously exchanging

oxygen isotopes with starch during cellulose synthesis,

thereby further dampening the temporal climate signal

preserved in the bark phloem starch.



38.4.1.1 Inter-Annual Transfer

of Photosynthates

Helle and Schleser (2004) invoked post-photosynthetic fractionation during carbohydrate metabolism

as a likely reason for the observed seasonally persistent intra-annual d13C pattern in broad-leaved deciduous species Fagus sylvatica, Populus nigra, Quercus

petraea and Morus alba. In their highly resolved intraannual isotopic measurements (40–100 measurements

within each ring), a tri-phase d13C pattern (Fig. 38.5)

was found with (1) values showing increasing trend

during the early growing season (2) a subsequent

declining trend up to the late wood (3) an increasing

trend at the end of each ring. Inconsistency between

the observed pattern and modeled carbon isotope fractionations during photosynthesis was found and it

was suggested that the isotopic fractionation during

the storage and remobilization of plant organic matter

could be the likely cause. The earlywood and the end

of the latewood appear to be affected by post-photosynthetic metabolic processes while the mid-section is

most probably affected by climatic influences on leaf

carbon isotope discrimination, making it more suited

for past climate studies.

The importance of storage and remobilization of

carbon pools in the tree has been experimentally

demonstrated by Kagawa et al. (2006a). The authors



822



S.R. Managave and R. Ramesh



Fig. 38.5 Seasonal carbon isotope behavior in

total organic matter of tree rings from Fagus

sylvatica (beech) (from Helle and Schleser 2004)



pulse labeled Larix gmelinii with 13CO2 for understanding the transfer of photoassimilates from 1 year

to the next. The earlywood of the current year was

seen to get contributions from previous summer,

autumn as well as the current spring, whereas the

latewood mainly contained photoassimilates from the

current summer/autumn. In another study, Kagawa

et al. (2006b) showed the non-needle starch pool gets

approximately 43% of carbon from previous year’s

storage. As the non-needle starch pool can be used

for xylem tissues, the interannual transfer of photoassimilates could be the likely reason for the autocorrelation observed in isotope dendroclimatological

studies (Kagawa et al. 2006b). This could seriously

limit the resolution of climate reconstruction at intraannual time scales.



38.4.2 Plant Physiological Models

and the Interpretation

of Cellulose d18O

The plant physiological models discussed earlier have

greatly helped quantifying the relationship between

climatic (and plant physiological) parameters and

cellulose d18O. Anderson et al. (2002) calibrated a

tree cellulose d18O series from central Switzerland

with d18O of precipitation (from the Swiss Network

for Isotopes in the Hydrological Cycle station), temperature, relative humidity and ring-width. With a

modified leaf-water model of Dongmann et al.



(1974) they found variations in the dampening factor

(f O ) from 0.27 to 0.49 and to be dependent on the

relative humidity. This factor accounts for the degree

of dampening of leaf water d18O enrichment as

reflected in the stem cellulose, which is likely a result

of the leaf water isotopic inhomogeneity and exchange

between sucrose and stem water during cellulose synthesis (Saurer et al. 1997). A comparison of observed

and modeled d18O of oak (Quercus robur L.) from

Norfolk, UK enabled Waterhouse et al. (2002) to

deduce that the trees used a constant mixture of

precipitation and groundwater. Berkelhammer and

Stott (2009) used a modeling approach to infer the

insignificant role of d18O of the source water pool and

the important role of the length of the growing season

in determining intra-annual d18O cycles of bristlecone

pine tree rings.

Importance of the forward modeling approach in

interpreting tree-ring isotope data has been demonstrated by Evans (2007) and Anchukaitis et al.

(2008). Given site-specific climate data, forward modeling can be used to quantify the dominant control of

and expected variability in the isotopic composition of

tree cellulose. The expected variability in the isotopic

record can be estimated using Monte Carlo simulations to randomly varying input parameters in the plant

physiological model (Evans 2007; Berkelhammer and

Stott 2009; Managave et al. 2010a, c). Managave et al.

(2010a) used this approach for detecting the record

of seasonally varying isotopic composition of rains

in teak trees from southern India.



38 Isotope Dendroclimatology: A Review with a Special Emphasis on Tropics



38.4.3 Response to Global Climate

Change

Duquesnay et al. (1998) have reported increased

water-use efficiency (a ratio of net photosynthesis A

to conductance for water vapor gH2 O ) from 23 to 44%

during the past century in beech (Fagus sylvatica L.)

as a consequence of increasing CO2 concentration.

Saurer et al. (2004) studied the effect of increasing

ca on d13C record of Larix, Pinus and Picea

trees growing in northern Eurasia, from Norway to

Eastern Siberia. They observed constant discrimination (d13Cplant – d13Catm) values of two periods,

1861–1890 and 1961–1990 CE, despite average

increase in ca by about 43 ppm. Constant discrimination values yielded constant ci =ca ratio for the given

duration indicating, the authors explained, increased

intrinsic water-use efficiency of trees by acclimatization to increased ca through decreasing stomatal conductance. Increase in the water-use efficiency by 34

and 52% for tropical species Cedrela ordata L. (tropical cedar) and Swietenia macrophylla King (big leaf

mahogany) respectively as a result of increase in ca

has also been reported (Hietz et al. 2005).

Treydte et al. (2006) using d18O variations of juniper tree-ring cellulose from northern Pakistan reconstructed precipitation history back to 828 CE. One of

their important results was the detection of unprecedented twentieth-century intensification of the hydrological cycle in western Central Asia, perhaps an

effect of global climate change due to anthropogenic

forcing of climate.



38.4.4 Inferences from Combined Carbon

and Oxygen Isotope Studies

d13Cplant is mainly controlled by intercellular concentration of CO2, ci , which in turn depends on stomatal

conductance, gs and photosynthetic capacity, A (38.9).

Both lower gs and higher A result in the reduction of ci .

Thus higher d13Cplant values observed in the plant

cellulose could be because of lower gs and/or higher

A. d18O measurements of plant cellulose can potentially help ascertain relative importance of gs and A in

influencing d13Cplant (Saurer et al. 1997; Scheidegger

et al. 2000). Equations (38.5), (38.6) and (38.7)



823



indicate that one of the factors influencing d18O of

plant cellulose is relative humidity. It affects the

degree of evaporative enrichment of the leaf water in

18

O and subsequently synthesized cellulose i.e. lower

relative humidity induces higher cellulose d18O. Thus

it follows that if the relative humidity is the dominant

control over stomatal behavior, there should be a positive correlation between d18Oplant and d13Cplant.

This approach assumes that the variance in d18Oplant

introduced by changes in d18O of source water and

atmospheric water vapor is minor compared to that

introduced by stomatal behavior controlled by relative

humidity. In other words, the correlated variations in

d18Oplant and d13Cplant can be considered as indicative

of dominance of relative humidity in controlling stomatal behavior.

Saurer et al. (1997) highlighted the importance of

drier conditions in deciding d18O and d13C of tree

cellulose from Switzerland using plant physiological

models. Further, combining plant physiological models for interpreting d18O and d13C of cellulose, the

authors showed relation between the slope of the cellulose d18O – d13C curve and the sensitivity of the

ci =ca (ratio of intercellular and atmospheric CO2 concentrations) of a plant to changing relative humidity.

A positive correlation between d18O and d13C of

plant cellulose in several studies support this. Saurer

et al. (1997) found that d18Ocellulose and d13Ccellulose of

beech, pine and spruce trees growing at a site with

differing soil moisture conditions are positively correlated for each species (r2 varying from 0.87 to 0.98).

Barbour et al. (2002) have also reported positive correlation between the d18Ocellulose and d13Ccellulose

records of Pinus radiata from tree sites with different

water balance in New Zealand. Within these sites, the

slope of the d18Ocellulose – d13Ccellulose curve, i.e.

change in d18Ocellulose per unit change in d13Ccellulose,

was found to increase with increasing vapor pressure

deficit of the individual sites. A positive correlation

between d13C and d18O of the organic matter of the

phloem sap of European beech (Fagus sylvatica L.)

was reported by Keitel et al. (2003). The authors

attributed the observed correlation to stomatal control

as a common driving force. Roden et al. (2005) found

similar response in deviations in d18Ocellulose from

stem water d18O (d18Ocellulose – d18Ostem water) and

d13Ccellulose along a precipitation gradient in western

Oregon, United States. If photosynthetic capacity controls d13Cplant, i.e. light-limiting conditions, there



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

2 Oxygen Isotope Studies: Factors Influencing delta18O of Tree Cellulose

Tải bản đầy đủ ngay(0 tr)

×