Download The Application of Carbon Isotope Discrimination in Cereal Improve

226
Chapter 6
The Application of Carbon Isotope
Discrimination in Cereal Improve~ne~lt
for Water-Limited Environments
Anthony 6. Condor1
Graharn D. Farquhar
Greg J. Rebetzke
Ricliard A. Richards
Recently, two new cotnil~crcialwheat varieties were released for broadacre dryland production in Australia, which were very different from any
pl-evious varieties. The varieties Drysdale (released in 2002) and Rees (releaced in 2003) carry a broad cpectrum of disease-resistance genes and produce high-quality tlour, as might be expected for new wheat varieties ~ 0 1 ~ 1 peting in a demanding l-r~arket.But Drysdale and Rees are ui~iquebecause
they are the first varieties, of ally crop species, specifically bred for performance in dry environments using carbon isotope cliscrii-r~inationas an indirect selection criterion.
Drysdale and liees were bred by backcrossing the high transpiration efficiency trait, i.e., the highly efficient exchange of CO, for water through the
stotnata, into the widely grown, milling-quality variety Hartog. The backcrossing and selection process was based on the use of carbon isotope discrimination as a surrogate measul-e of transpiration efficiency, and resulted
in a population of elite lines, including the two new varieties, that outyield
the recurrent parent Hartog by 5 to 10 percent on average. The yield advantage tends to be greater in drier environillents and less in wetter environments. These new wheats are the first of what we anticipate will be several
varieties bred for high transpiration efficiency and yield using carbon isotope discrimination as a seco~ldarytrait.
These landmark releases come 20 years after it was first proposed that
the isotopic coinposition of plant carbon in C3 species should reflect differDr.ought Ad~lptationi r l C ~ r c n l s
0 2006 by The Haworth Prcss. Inc. All rights rcscrvcd.
doi: 10.130015781-06
1 72
DIIO I!C;H 1'AL)AI'T.l l'l Oil L1. CEIIE14LS
cnces in the transpiration efficiency of leaf gas exchange (Farquhar et a].,
1982) arld that carbon isotope analysis may therefore prove a useful tool in
breeding for improved water-use efficiency and yield in dry environments
(Farquhar and Richards, 1984). In this chapter, we discuss the philosophy
and practice behind the developtnent and use of carbon isotope discrimination as a se~ondilrytrait for cereal improvement and draw out some of the
major conclusions and challenges that have emerged as the concept has progressed froill biophysical theory to the reality of new wheat varieties in
frzr~rlers'fields.
CARBON ISOTOPE L)ISCRlrl.lIIYAII'ZO,V:A PHYSIOLOGICAI,
IMAKKER FOR HIGH TKANSPIKATIQIV EFFICIENCY
In this section we present an overview of what is meant by the term w r borl isotope (li.scrirrzirllrtio?~,
the basis for the association between carbon
isotope discrinlination and transpiration efficiency, and why carbon isotope
discrimination rnight be considered a useful secondary trait for yield irnprovenlent of dryland crops. The treatment ic not exhaostive. For a rnore
conlplete coverage of the concepts covered here. the reader is referred to the
early publications by Farquhar et al. (1982) and Farquhar and Richards
(1984), and to more recent reviews such as those by Farqohar et al. (1989).
Hall et al. (1994), Condon and Ha11 (1997), Brugnoli and Farquhar (2000).
Condon et al. (2002), and Condon et al. (2004).
Cnrborz Isotopes irz tlze Biosphere
Carbon accounts for approximately 40 percent of plant dry weight, and is
assimilated into plants by the process of photosynthesis. During photosynthesis, C0, from the atmosphere diffuses into the leaf interior through the
thousands of tiny stomata1 pores in the leaf epidermis. The CO, is then assimilated to generate the sirnple sugars (carbon skeletons) that are a substrate for downsti-eanl syilthesis of the nlultitude of orgarlic compourlds that
are important for plant growth. The carbon in atnlospheric C 0 2 and
throughout the biosphere occurs as two stable (i.e., nonradioactive) isotopic
for~ns.The nlost coillmo~~
form is 12C, which accounts for about 98.9 percent of the C in at~llosphericCO,. The other stable isotope, l3C, makes up
about 1.1 percent of atmospheric CO,. The proportion of 13C in the biosphere is sufficiently large that very slnall variations in the 13C112C ratio can
be measured accurately.
Early measurements of the l'CI12C ratio revealed that the C-isotope
composition of plant dry matter was different from the composition of the
atiuosphere on which plants Seed. Plants were found to contain fractionally
less 13C and thus relatively more 12Cthan the atmosphere. It has also been
~ e considerable variation e x i ~ t sin the l C / ' 2 C ratio
known fi)r sorne t i ~ i that
of plant dry matter (Brugnoli and Farcluhar, 2000). This variation has several implications for crop productivity, hut before expanding on these. we
firat need to explore more closely the processes that cause variation in the
C-isotope con~positionof plant dry matter. Before doing so, the principles
involved in carbon isotope analysis and how the data are expressed are
briefly summarized.
Isotope Antrlvsis nrzd Ter-~ninology
The isotopic composition of plant carbon is most often measured using
isotope-ratio Inass spectrometry (Preston. 1992). The technique first requires the production of a pure sample of CO? gas from the plant material.
This is usually achieved by combusting a small sample of dried. finely
ground plant inaterial at high temperature. The C 0 2 produced is then purified. for example by gas chromatography, and introduced into the mass
spectrometer wl~erethe C 0 2 nioleculcs are ioni/ed at high voltage and focused into a fast-moving beam. The isotopes are separated on the basis of
mass-charge ratio by passing the beam of ions through a strong magnetic
field.
It is technically difficult to accurately measure the absolute isotopic
con~positionof plant rnaterial because 13C is present in such low amounts
(approximately 1.1 percent of total C at natur:ll abundance). Nevertheless,
differences in conlposition between samples can be ~neasuredwith useful
precision. The isotopic composition ic therefore expressed by comparing
the molar abundance ratio, 13C/12C, of the plant sample (R,,} to the value of
the molar abundance ratio in a standard (R,). Carbon isotope compocition
(6°C) is labor cnlcolated as (R[, - R,)IR, (= R,,IR, - 1). For historical reasons. the standard for carbon has been Pee Dee belemnite (PDB), a carbonaceous rock. although a synthetic replacement is now provided by the International Atomic Energy Agency (IAEA). Values of 61TC,,, are often
expressed as the value times 10' or per mil so that they appear as whole
numbera. The 6''Ck,DBof plant material has :I v:ilue that is small and negative, in the range of approxiruately 1 0 per mil to 3 0 per mil. This is because
thc 17C/12Cratio ol' plant material is less than that of PDR. The atmosphere
has a value of 6"Cp,, that is also slightly negative, about -8 per mil relative
to PDB.
Carbon isotope composition (6l3Cp,,) has proven to be a ~isefulempirical measure, but as ~1nderst:indingdeveloped of the processes causing varin-
Ecl~lations6.1 and 6.2 represent an essentially instantaneous process of
discrimination. It is known that the value of c,/c,, is not constallt and neither,
therefore, is the amount of discrimination. Rather, discrimination responds
to nunlerous short-tern~and long-tenn environmental infloences, fluctuating depending on the extent that the stornatal step dominates or the carboxylation step dominates. These processes are also subject to some level
of genetic control. Measuring A'" in dry matter of C-, species provides an
assirnilation-weighted average value of c, / c , over the life of the plant material being analyzed.
Carboil isotope discrillrination by plants of C, species is small co~upared
with that by plants of C , species (Figure 6.1) and reflects to a much larger
extent the discrimillation that takes place at the stomata. A small amount of
variation exists in C-isotope composition of C4 plants resulting fro111 discrilllinntion duriilg C-fixation. This variation may be of siynificallce for
plant productivity and will be considel-ecl in more detail toward the end of
the chapter. Because of the greater potential for variation in C3 plants, discrilnillation hy C3 species has attracted far more attention niid forins the
basis for ilrost of what follows
Variation in Cctrhon Isotope Cornpusifion
Awzorrg Metabolites
The values of A13C measured in plant dry matter principally reflect discrimination during photosynthesis. yet in all plants sonre potential for additional small changes in C-isotope composition exists as C flows through
biochemical pathways downstraaln from photosynthesis to form val-ious
metabolites. For example, lipids tend to he depleted in I3C compared with
bulk plant dry matter, whereas the reverse is true for carbohydrates (Figure
6.2). The reasons for these differences in C-isotope cornposition among metabolites are primarily due to fragmentation fractionations (Tcherkez et al..
2004) associated with transfer of C atorns during the biochemical processes
that generate these various compounds. These processes will not he discussed in detail here, however, Brugnoli and Farclohar (2000) provide a useful summary. It is possible that further, additional fractionation of C isotopes
may take place during processes associated with transport of metabolites
and respiration since these processes also often entail some I-erlrrangement
of organic molecules.
Leaf sugars
Starch
Cellulose
Amino acids
Proteins
Lignin
Lipids
5
FIGURE 6.1. Relationships for C3 and C4 plant species between carbon isotope
discrimination during photosynthesis and the ratio of intercellular and atmospheric concentrations of Con (cJc,). The boxes represent the range of typical
data points obtained in leaf gas exchange studies in which carbon isotope discrimination was determined "online" by measuring the carbon isotope composition of the gas stream before and after it passed over the leaf. Source: Adapted
from Brugnoli and Farquhar, 2000.
10
15
20
25
Carbon isotope discrimination (103 x A%)
30
FIGURE 6.2. Carbon isotope discrimination in various metabolites in C3 species. The boxes represent the range of variation found in the literature, with the
vertical bars being the mean values for each compound. The dashed line represents bl3C for bulk dry matter. Source: Adapted from Brugnoli and Farquhar,
2000.
.4pi)liiofb,rr (?/G~rbiitlLotope Discriminntion in (>,recrl lr?~pioi~~~nrcttt181
level transpiration efficiency, fl,is directly related to only one co111ponent. W , the transpiration efficiency of biomass production, bat as will be
discussed in following sections, A f l also has the potential to influence each
of the other three components in tile yield framework.
VARIATION IN A13C OF C, CEREALS
Genotypic Variatiolt irz A13C
Substantial genotypic v:lriation in A13C has been obqerved in many C3
cereal species. Numerous studies have shown variation in A13C of at least 2
per mil, occasionally closer to 3 per mil, in bread wheat (Condon et al.,
1987: Condon et al.. 1990; Ehdiiie et al., 1991: Sayre et al., 1995), durulll
wheat (Araus et a]., 1998; Merah et al., 1999; Royo et td.. 2002 ), and barley
(Hubick and Farquhar, 1989; Craufurd ri 81.. 1991; Acrvedo, 1993; Voltas
et al., 1999), and in various species of rangeland (Johnson et a].. 1990;
Johnson and Bassett, 1991) and t ~ ~grasses
rf
(Ebdon et a]., 1998). Considerable genotypic variation is known to exist in A13Cwithin rice (Dingkuhn
et al., 1991; Condon et al., 1999), and no obvious reasons explain why similar variation in Al3C should not be present in other cereals such as oats, rye,
and triticale.
For healthy, unstressed plallta of C3 species, average values of c,/c, are
usually close to 0.7 (Farquhar, Ehleringer, and Hubick. 1989). This is equivalent to an average value of Al3C of ca. 21 per mil (froin Equation 6.2.
Given this average value oEA13C,what does a range of k1per n i l genotypic
variation in AnC mean in terms of pote~rtialvariation in A/T! This call be
calculated by malung further use of Equatio~l6.2 (which relates variation in
A13C to variation in c,/c,) and Equation 6.6 (which relates variation in
to variation in c,/r,,,). From these it can bc dctcl-~uincdthat. in a relative
sense. A/T of a genotype with a A'" value of 20 per mil could be approximately 1.3 times greater than A f l of a genotype with a Al3C of 22 per mil.
Thos large potential gains in Aflare associated with relatively low valoes of
Al3C. However, various reasons explain why it is onlikely that all of there
gains will be realized in crops grown in the field. These searons, discucsed
in detail in following sections, relate to a series of con~plicationsthat arise
as processes of exchange of CO? for water are scaled up froin the level of instantaneous fluxes at the stolnata to crop growth and water use in field canopies. Many of these complications relate to the physiological basis of variation in A I T and A/T.
Crrus~sof Genotypic ifariatiolz in A13C in C, Cereals
As indicated earlier, variation in (;/c, and thus Al3C can result fro~nvariation in stomatal conducmnce. in pliotosynthetic capacity, or a cornbillation
of both. In bread wheat. stomatal cond~ictanccand photosynthetic capacity
contribute approxirnntely equally to genotypic variation in A13C(Condon et
al., 1 990; hlorgnn and LeCain. 199 1 ). The data prcbentcd in Table 6.1 summarize obbervations rnnde on 6 of 14 bread wheat genotypes studied by
Condon et al. ( 1990). Among these genotypes. both stomatal conductance
and photoryntlietic capacity varied by 1.3-fold, while A13C varied over a
range of 1.8 per mil. For pairs of genotypes with the qame conductance
(e.g.. cvv. Veery 3 and Sunstar, Hartog and Quarrion), AIJC was lower in the
genotype with higher photosynthetic capacity. For pairs of genotypes with
the same photosynthetic capacity (e.g., Veery 3 and Quarrion), Al3C was
lower in the genotype with lower stornat;il conductance. It is safe to assume
that genotypic variation in Al3C will be the same for other cereals such as
durum wheat, barley, and rice since all of these species show a siniilar range
of variation in A13C 3s found in bread wheat.
It is also likely that considerable genotypic variirtion may exist in both
condilctance and photosynthetic capacity that is not revealed by me:isorin,g
A T . This is because of the overall tendency in plants of C j species for conTABLE 6.1. Variation in stornatal conductance, photosynthetic capacity, and carbon isotope discrimination among representative Australian semidwarf wheat
genotypes.
Genotype
Veery 3
Sunstar
..
Hartog
K1056
Quarrion
M3844
Stomata1 conductances Photosynthetic capacity
(rno1.m-2. s-1)
(rnrnol-m-2. s-?-Pa-')
0.55
1.66
(%,)
21.O
0.49
0.46
1.56
1.60
20.2
19.8
0.48
0.43
1.68
1.79
19.5
19.3
Source: Adapted from Condon et al., 1990
astornatal conductance and photosynthetic capacity (initial slope of the relationship between A and ci) were measured on flag leaves of well-watered and fertilized plants.
3 ~ 1 was
3 ~measured on sink tissue for C assimilated by the flag leaf, i.e., growing ears enclosed in the flag leaf sheath.
o
ductance and capacity to covary, such that c,/c,, is maintained close to a
value of 0.7, at least in leaves of healthy. unsti-essed plants (Farquhar,
Ehleringer, and Huhick, 1989). This tendency for conductance and capacity
to covary is illustrated in Table 6.2, which suminal.izes data on AI3C and related photosynthetic parameters obtained on a diverse collection of 41 rice
genotypes grown under periuanently flooded conditions in southeast Australia. Among all 41 genotypes a range in A l C of recently grown leaf material froin 19.2 pel- mil to 21.1 per mil existed. Although substantial variation
in leaf conductance (measured in this study as leaf porosity using a visconsflow porometer) was seen, very little relationship was apparent between
Al3C and conduct;~nce.In fact, for the bulk of the 41 genotypes. considerable genotypic variation in conductance existed that was reflected in nl~~lost
no change in A')C. This was because conductal~ceand capacity (inensured
as leaf N content) tended to covary. When genotypes were arbitrarily split
into three groups on the basis of N content, which is a robust measure of
photosynthetic capacity in rice (Peng et al., 1995). they showed no difference in average A13C of leaves. Rather, diflerences in average N content
aillong the groups were rellected in parallel differences in leaf conductance,
canopy temperature depression (CTD), and ' 8 0 conlposition (8lXO)(Table
6.2).
Similar observations of covariation of stomatal conductance and photosynthetic capacity were made by Fischcr ct al. ( I 998) on a "historic series"
of semidwarf bread wheats released by CIMMYT. In this study. the genotypic variation in stomata1 conductance exceeded the variation in photosynthetic capacity. with the result that more recent, high capacity, high cunductance genotypes had highel- values of Al3C than earlier, low conductance,
TABLE 6.2. Summary of variation in photosynthetic traits among 41 diverse rice
genotypes grown under continuous flooding in southeast Australia, 1999.
Conductance
-.
(per second)
1.8 i 0.2
1.4 i 0.2
1.1i0.2
-
N-content groupa
HighN (n= 14)
Intermediate N (n = 13)
LowN ( n = 14)
N content (%)
2 3.40
3.03-3.40
53.03
D13Cb (%)
20.3 +- 0.1
20.2 ? 0.1
20.3i0.1
CTD ("C)
5.6k 0.2
5.2 + 0.2
4.9iO.l
6'80 (%.)
25.02 0.2
25.8 i 0.1
26.0+0.1
Source: Condon et al 1999.
;
a Mean values of photosynthetic traits
(L standard error) are shown for three equally sized groups. separated
on the basis of leaf N content.
b ~ 1 3 C6180
,
(oxygen isotope composition with respect to standard mean ocean water [SMOW]) and stornatal
conductance were measured on upper canopy leaves Stornatal conductance was measured using a viscous-flow porometer (Rebetzke et al,, 2001), CTD (canopy temperature depression) using an infrared thermometer.
r
f
r
o
t
s
or
o
r
I
r
/I
tr
t
181
low capacity genotypes. The range in AfiC ainong the w11e:its studies by
Fischer et 211. (1998) \\Jiis 1 per mil.
The me;lrnremrnt of AJ3Cprovides an integrated estinlirte of genotypic
variation in c,/c,, and. potentially. A n : Unfortunately, A13Cprovides no information on whether I-, /c,~is v:l~yingdue to v:lriation in stomatal conductance or due to variation in photosynthetic capacity, and this infornl:~tion
may be inlpoi-tant for several reasons. One reason is that it is unlikely that
the uimpliqtic I-elationuhipbetween A1'C and Afldescribed by Equation 6.6
will f~illyapply in sit~~ations
in which variation in AliC reilects variation in
stomata1 conductance. Thic is b e c a ~ ~ any
s c genotypic difrerence in stomatal
corlductance will usually also be reflected in a dilference in leaf teniperature. Genotypes \\~ithlorver condrlctnnce will have warmer leaves (e.g., ?&ble 6.2) :ind therefore larger vapor pressure gr:ldients driving Lvater from the
leaves (Farquhar, Hubick, et nl.. 1989). A a result, transpiration per unit
conductance will also be greater (Equation 6.4). Ultimately this means that
the term ( f i t i - M.~,) in Equation 6.6 is not an independent rariable. and any
gain in ,4A" will be less than expected from the lower value of A13C that rcwits from the lower conductance. For individual leaves, this situation will
become wol-ue the less the air is ?tin-ed around the leaves. This is because
the air in the unstirred boundary layers around the leaves of a low-conduct ence genotypc will be warmer and drier than thc air in the boundary layers
:[round thc leaves of a higher-conductance genotype. Thc situation is predicted to becoine worse still when leaves form extensive, dense canopies,
\och ar in a well-fertilized cereal crop. This is largely because even less stirring of air around l e a e s will occur in such a canopy, although ndditio~lal
reasons for this exist that are explored more fully elsewhere (Cowan, 1988;
Jones. 1993). Despite the complications of elevated leaf temperature, in
most situations genotypes with low A13C resulting frorn lower stomatal
conductance I\ i l l still tend to have higher Am, but the gain in Am will be diluted sotnewhat.
01. collrsc another conwquence of low AI3C as\ociated with low ston~:ltalconductiince exists. Lower A1'C and higher A/l.resulting from low
5to1n:lt:ll conduct;ince are likely to be associ:~tetlwith lower photobynthetic
rate per unit leaf ilre:l illld possibly a slower r:itr of crop growtll. Low AI'C
i ~ \ \ ~ ~ i : with
l t ~ dIO\\J ~t0111:ltiil conductance inay therefore result in lower
productivity if no \tleong limitation to growth exists froill lack of water.
The colnplications associated with low stomatal conductance are not anticipated if low Al3C is associiltad with high photosynthetic capacity
(greater draw-down of ci per unit conductance). In this case no change in
leaf teinpeleature should occur, so the I-csponse of N T t o a change in AuC
sho~ildfillly correspolld to thnt given in Equation 6.6. Also. the expectation
would be for greater photosyntlletic rate per unit leaf area and, therefore,
prob;lbly a faster rate of crop growth, although as discussed in a following
section. the latter may not necessarily follow.
The values of A'3C nieasurecl in dry matter sanipled from plants of C3cereals under stress are alillost invariably lower than At3Cvalues ineasured on
unstressed plants. Abiotic stresses such as soil-water deficit (e.g., Farc~obnr
and Richards, 1984; Ehdaie et al., 1991; Condon et al., 1992: Merah et al.,
1999), soil colnpaction (Masle and Farquhar, 1988), soil aalilrity (Isla et al.,
1998: Rivelli et al., 2002), and low humidity (Condon et al., 1992) all result
in lower values of A13Cbecauhe they result in some degrcc of stoiuatal closure, causing c, /c,, to be lower. In some circurnst:lnces, abiotic stresses such
as salinity may be severe enough to cause damage to the photosynthetic apparatus, especially if' they are accompanied by conditiolls of' high light (e.g.,
James et nl., 2002). In such cases, a tendency toward a slightly higher value
of (.,/(.,,may exist due to lower photosynthetic capacity, but a co~~currel~t
tendency toward lower c,/c, will probably exist due to stornatal closure. In
practice it is difficult to resolve these opposing influences on c,/c, and
A ' T , even using sophisticated gas-exchange measurements.
It might be expected that N starvation would be reflected in h i ~ h e values
r
of h13C since N starvation should i-ecult in lower photosyothetic capacity,
but experiments conducted with wheat and barley Sound virlually no difference in A13C between N-sufficient and N-starved plants (Condon et al.,
1992; Kang ct al., 1 996; Robinson et al., 2000). This is probably because N
starvation was applied early in development. with two outcomes. First, the
plants adjusted leaf area, through snialler leaves and fewer tillers. so that
leaf N content was maintained at relatively high lcvcls (certainly less dif'ferent than the difference in applied N). Second, it is likely thnt stomata1 conductance fell to balance the photosynthetic capacity of thc leaves and coilstrain c,/c, near 0.7. Relatively low stomatal conductance is a comnlon
observation for N-starved cereals (Wong et al., 1979).
Values of A1jC meauured on unstressed plants of C3 cereals are typically
in the vicinity of 19 to 22 per mil. For p1;ints subjected to soil-water deficit.
A13C values measured on organs that have grown when plants were under
severe l r r c have been fi)und to be as low as 12 to 14 per mil. Many of the
lowest valves of A13C have been measured on grain of' teinperate cereals
grouln in field e~pcrirncntuin very low rainfall environments or in container
expcrjmcntu in ~vhiclisoil waler deficit became gradually more scverc (c.g.,
Condon ct al.. 1992; Voltas et al., 1999: Rotwl-ightet al., 2001). It is not curpri~ingthat a gmd~taldecline in the values ofA1C measured usually occur5
on rccentl y formed dry matter of temperate cereals grown in rain fed enviroi~mentc.since it is t19traIfor soil water availability to decline and the vapor
presture deficit of the air to increate as the growi~lpseason PI-ogrescec.Both
of these stresces will resolt in stomatal c1osilt-e and lower values of Al3C
(ConcIon et al., 1 992; Figure 6.3).
The very low values of A13C nleasured in grain have prompted hypotheses on 1-ecyclingof respired C 0 2 within the grain-enclosing stnictures of the
ear (Arauu et al., 1993; Gebbing and Sclinyder. 200 1 ) and suggestions that
an enllnnced contribution to C02-fixation from PEP c:lrboxylase present i n
br:lcts and ylurnes rnay exist (Bort et al.. 1995). Whereas any enhanced contribution fro111PEP carboxylase rem:iins moot. it is highly likely thnt some
recycling of respired CO, does take place within the ear. The effect of this
t
o
0
17
16
A
A ~ ~ C
0
Soil water
Conductance
v
vA
B
v
Days after sowing
FIGURE 6.3. Season-long changes in water in the 1.8 m soil profile, stomatal
conductance, and values of A13C measured in most-recently expanded organs
of wheat grown under rainfed conditions at Moombooldool, SE Australia in 1985.
Data points shown for soil water depletion, stomatal conductance, and Al3C are
mean values for eight genotypes. Note that changes in stomatal conductance
measured on source organs are detected as changes in 6136 in sink organs
sampled several days later. Source: Adapted from Condon et al., 1992.
recycling on the Al3C value of grain is difficult to resolve since the carbon
in the grain may be derived from numerous sources, including current
photosynthate from leaves and ear structures and assimilates stored earlier
in the sterns and leaf sheaths. Eaclr of these C sources may have quite different Al3C sig~lilturesthat largely derive fro111the environmental conditions
prevailing at the time the C was acquired.
Grain AI3C values are not always very different iiorn leaf A13C ~ ~ I L I C S .
Mean grain A'" values of approuim:ltely 19 per mil have been measured in
field experiments on durum wheat given supplemental irrigation (Merah
et al.. 2001; Royo et al., 2002). In a study of 11 advanced breedillg lines of
rice grown under flooded padi conditions in southeast Australia (Condon
et al., 1999). the rrlcan A13Cvalue of grain harvested at maturity (mid-April)
was 19.5 per mil (genotype range 18.6 to 20.3 per mil). This was very close
to the mean Al3C value of 19.8 per mil measured on recently formed leaves
sampled in early January (genotype range 19.4 to 20.2 per mil). The slightly
lower values for grain Al3C than leaf Al3C could easily be attributed to the
high starch content of the grain as compared with the relatively l~iglllipid
content of the leaves. Starch is known to have a lower A]" value than lipids
(Figure 6.2). Indeed, given the difference in chenlical conlposition betweell
leaves and grain. it is sulprising that the difference in AI3C was not greater.
Gerlotype x Environrnest Zrrlei-actioizfor A1"C
Genotype x environment (G x E) interi~ctionsfor AI3C are potentially
large because erivironment influences can have a very large effect on A13C
values measured in plant dry matter, especially under conditions of declining soil water and rising ewlporative demand (see previous section for details), and the A13C value of all genotypes may not respond in the same
manner or at the same time to these environmental perturbations. For example, the value of A13C inay fall faster in genotypes that use water faster or
that have shallow root systems 01- that have stomata that are more senqitive
to increasing vapor pressure deficit of the air. Among the eight wheat genotypes grown in the study suiumarized in Figure 6.3, genotypic variation in
the slope of the seasonal decline in A1-T was strongly correlated with the
proportion of profile water con{omed at anthe sis by each genotype. The rate
of decline of AI3C was faster for genotypes that had used I-elatively more of
the available soil water at anthesis (Condon et al., 1992).
Broad-sense heritability is :I me:Isure of the extent to which phenotypic
variation in a trait can be attributed to genot>~pic
differences, rather than the
effects of environn~ent,G x E, and sampling. Thus it reflects the extent to
which the range in genotypic v:llues is repeatable, and therefore it also re-
flects the potential for progress in selection during breeding. Estimates of'
broad-sense heritability for A1?Ccan be high. in the order of 90 percent on a
genotype-mean basis and 80 percent on a single-plot basis (Condon and
Richards. 1992), but st~hqtantiallylower values are not uncommon (Ehdaie
and Waines, 1994; Araus et nl., 1998; Merah et al., 2001). Different estimates of broad-scnsc heritability for AI3C may reflect the different tissues
sampled for AI3C measurements. For temperate cereals. values of heritability for Ai3Care often lowest for plant material sampled near anthesis,
such as flag leaves. the rachis. and peduncle. This is when any genotypic
diSSerences in soil-water depletion are likely to be greatest. Herit;lbilitics
may be higher for grain, but this is not always the case across contrasting
soil rnoisturc I-egimes (Condon and Richards. 1992). Heritabilities have
been found to be highest for dry matter laid dow~lbefore or during early
stem elongation, when p1:ints are essentially unstressed and repeatability
greatest (Condon and Richards, 1992).
Values of n;rrrouT-senseheritability lor AI3C of bread wheat reported by
Rebetzke, Condon, et 31. (2002) were also high: 93 percent on a yenotypemean haqis and 63 percent on a single-plot basis. In the same 5tudy. the corresponding n:irro\v-sense heritabililies for grain yield were 55 and 14 percent, respectively. The high values of heritability for A13Cwere obtained by
employing dry matter sanlpling strategies that minimized thc iinpact of potenti;~llylarge G x E for A13C. Measurements o S A I T were done on vegetative leaf material laid down early in the standard growing season when
plants weir effectively unstressed. as recommended by Condon and Richards (1992).
Genetic Corltml of A]-'C
Narrow-sense heritability reflects that portion of the phenotj p'lc variance
that is transmitted from parent to progeny. Tlr~iq,it reflect$ breeding value.
High narrow-sense heritabilities reported by Rebet~ke.Condon, et 211.
(2002) for A13C of field-grown plants were consistent with high narrowsen$e heritabilities for AfiC rneasurcd for progeny from a range of wheat
crosses evaluated in the glasshouse and field (Rebet~keet al., 2006). Furthermore, these high narrow-sense heritabilities were consistent with high
heritabilities for tranqpiration efficiency measured on bread (Malik et al.,
1999) and durum \\/heat (Solomon and Labuschagne, 2004) plants evaluated under water-limited and well-watered conditions in the glasshotlse.
Genetic stildicq in u7henthave shown a piepondel-ance of additive gene effects for A1.'C (Rebelrke et al., 2006) and tr;~nspirationefficiency (M:~lik
et al., 1999; Solon~onand Labuschagne, 2004). Although s~nall,some evi-
5
2.
s."
5
b
.zy
I>.z
3-G
g
ga o%
-
s;".i%
q
a
-.<
W
Vj
I
5996
g
-2& g
u
cog
53..
PF-.'P
y
.
;
g9g
K
s o - g
< g-
3
g-g 5
9
a
sug
g- -2:.2
Sf g
g4
g g c w
yj-rl."
5
0
z zc cy - '
m 3 C.
c.5-z.
2
-
s - s.
E
g'aO6
O Z 5 " c
5
5 3
"
3 6
fqe 2
$ & < $
3 7 2 z
d '
g T .3J. e
g g m
-
J. 0 -2
s e 5 c :
C
PI.
P
-dI
Transpiration efficiency (g.kg-~)
3
t-4
3 E Z @S+
5 V
E w
z. 6.
0
"s
g 05.
2rrG 0
- , 3 <
2
3 CD
5.
EE':g7
-.
55
-8ga
E
g,,
0
0
2 =st<
3C3 Z "
3
Qg.g$
<
3
q
0
d
5.
gz;zp.
- . c 5:
sg$?a
5.5.g
B, w ,
ZFs
r z cn
2
cr
<cpg
-
c.
5
CD
gez-
5
.
d
2
9
'f-5.3
4.
g g <2. 3
v.g-5'55 5
S 3 ~ , 0
" D z V E
slower water use and perhaps consel-vative growth of low-Al3C genotypes
may be adv:intageous in sustainillg higher grain ilumber and grain yield.
Relationships between yield and AfiC may bc particularly variable in Mediterranean-type environments if low Al3C is associated with slow early
growth. This is because in rainfed Mediten-anean-type environments any
difference in transpiration between high-A'" and lo\v-A13C gellotypes
may not be reflected in a similar difference in total crop-water use to
anthesis if soil evaporiition is greater from under the slower-growing canopies of low-Al3C genotypes (Condon et al., 1993; Condon et al.. 2002).
Why Would Low-A13C Cereals Grow More Slowly?
As stated in the previous section, an obvio~lsreason why low A13Ccould
be associated with relatively slow crop-growth rate in cereals is if low A1)C
in the absence of soil-watel- deficit is the result of loiv stornatal conductance. Genotypes with lower sto1nat:ll conductance will tend to have higher
A/T and lower A"C, all else being equal, but higher A/T is likely to be essociated with lower pllotosy~ltheticrate per unit leaf area and consequelltly a
slower rate of crop growth.
Low conductance iuay not be the only reason why low A13Cis associated
with a slow crop-growth rate. Variation in A13C in cereals can also result
froin variation in photocy~ltheticcapacity (Condon et al., 1990; Morgan and
LeCain, 1991 ). If low Al3C is the result of high photosynthetic capacity, the
expectation might be a higher rate of photosyllthesis per unit leaf area and
thus faster crop-growth rate, However, crop growth rate may actually be
slower becaose, in cereals, substantial illcreases in photosynthetic capacity
arc most madily achieved by concentl-ating N into sinallsr leaves that have
greater macs per unit area and that intercept less light per unit N. This may
slow the rate of crop growth ~lntilfull light interception is achieved. IS fill1
light interception is achieved only briefly, as often occurs in drier cropping
environments, or not achieved at all, then high photosynthetic capacity may
not result in greater growth. In fact, the reverse may occur: cereal genotypes
with low photosynthetic capacity may actually achiew faster crop growth.
The eight genotypes shown in the s t ~ ~ ddepicted
y
in Figure 6.5 (Condon
et al., 1993) provide an example. Biomass production to anthcsis was positively correlated with A'" (Figure 6.5a), but the three genotypes with the
highest Al3C values did not have greater sto~natalcoilductallce (Figure
6.5b). Instead, the inference, giver1 they had higher valnec of A13C (ca. 1 per
mil higher) at sinlilar values of conductance, was that these three genotypes
had lower photosynthetic capacity. This inference was confirmed in studies
on glasshouse-grown plants (Condon et al., 1090).
18.5
19.0
19.5
20.0
Discrimination (1
20.5
x
21 0
A'~c)
-
0.35
0 40
0.45
0.50
Conductance (mmol m-2.s-')
FIGURE 6.5. (a) The relationship between aboveground biomass production to
anthesis and leaf Al3C, and (b) the relationship between leaf ~ 1 %and stomatal
conductance, for eight wheat genotypes grown at Moombooldool, SE Australia
in 1985. Values of A13C are average values for leaves from well-watered plants
sampled well before anthesis. Values of stomatal conductance are the average
of data obtained at five sampling times before anthesis. Least-squares linear regressions are shown for statistically-significant relationships (P < 0.05). Open
symbols indicate genotypes with low photosynthetic capacity. Sources:Adapted
from Condon et al., 1990; Condon et al., 1993.
The eight genotypes grown in this study were very diverse, and other factors may have been associated with low photosynthetic capacity and high
AJrC that were also associated with higher anthesis biomass. For example,
just as substantial genotypic variation in dry matter and N partitioning appears to have existed within the leaves, affecting photosynthetic capacity,
genotypic variation in factors iilfluencing within-plant partitioning may
ha\-c nlco existed that could also have influenced crop growth rate. Detailed
{tlrdies on ntclch rnol-e closely relatcd genotypes have been initiated to resolve accociations between factors contributing to variation in Al3C and
factor5 contributing to variation in crop growth rate, factors influericing bioTnase partitioning at leaf and plant levels, and the extent to which such associations are due to pleiotropic effects or genetic linkage.
CASE STUDY: AYI3LIGA7'IC)~VO F CANBONISOTOl'E
ANALYSIS IN VVHEA T BREEL)INrr FOR A UL"iTRALIA
Wheat production in Australia is a ~najorcomponent of a highly inechanized, commodity-driven agricultural system. The vast bulk of the wheat
o
I
P
0
t
G
0
,0
2
2
2
,
Drysdale yield (9/, of check)
Yield advantage (%)
maintain stomata more open after anthesis despite increaving soil and atrnospheric water stress (Condon et al., 1993). Any of these chiiracteristics acsociated with high grain Al3C (fast crop growth rate, ability to remobilize
stored reserves, earlier flowering, better water extraction, stoniatal insensitivity to water deficit) would be ~lsefiilfor cereals in Meditermnean-type
environments.
Carbon lsotape Discrimination in C , Cereals
Can variation in Al3C to breed for iniproved performance of C4 species
in water-limited agriculture be exploited'? This n ~ a ybe possible, but both
theory and available data indicate tliat progrevs with C, species is likely to
be even less straightforward than with C3 species.
S o ~ n evariation iii A13C has been observed within C4 specie? (e.g.,
Hubick et al., 1990; Mortlock and Hammer, 1999). but the extent of this
variation in C, species is inuch less than in C3 species. In addition. the relationship between transpiration efficiency and A13C is different in C, species. This is because of the substaiitial differences between C3 and C, species in the processes of C-assimilation. In C, species C is fixed into 4carbon compounds (malate or aspartate) in ~liesophyllcells by the enzyme
PEP carboxylase. With respect to gaseous CO,, this enryrne actually disin contrast to tlie C, enLynie R~~bisco.
criminates a little in favor of 1".
which has a relatively large discritnination against "CO?. After the initial
carboxylation step in C, pilotosynthesis, tlie 4-carbon co~npoondsai-e then
transported to the bundle-sheath cells surrounding the vascular bundles.
Here. after a decarboxylation step, C 0 2 enters the C, pathway and is fixed
by Rubisco. Tlie bundle sheath is very close to being gas-tight, and this has
two implications: The first is that it allows the C0, concentration to be held
mnuch higher than in the outside air. so c;trhoxylntion by Rubisco is much
more efficient than in C, plants. Second, little opportunity exists for CO, to
escape from the bundle sheath. and so little oppol-tunity exists for any disc~-iminntionagainst ' C O O by
, Rubisco to be expressed. Discrimination by
Rubisco can only be expressed if C 0 2 can escape to the atiiiosphere. In
reality the bundle sheath is iiot absolutely gas-tight, so some discrimination
by iiubisco occurs.
'rhese con~plcxitiesof discl-imination during C, photosynthesis can be
expressed in a simple mathematical description relating Al3C to c/c,,, as follows (Farquhar, 1983; Henderson et al.. 1992: Rrugnoli and Farquhar,
2000):
In this equation, n is the Sriiction:itinn dr~ringdiffusion through the
stomata (4.4 per mil) and b' = [h, i-sP - (b, - s)]. where b4 is the discrimination by PEP carhouyla\c (-5.7 per mil), h, is the cli\crirnination hy Rubivco
(30 per mil), s is dibcrinlination during lcakage fi-om the bundle sheath (1.8
per mil) and the factor @ accounts for the proportion of CO, released by
dccorhoxylation that leak5 out of the hundlc vhcath (Rrugnoli and Farquhar,
2000). It should be noted tliat the value of cD is not constant. It can also be
noted that Equation 6.8 has the same for111as Equation 6.1, which dcscrihed
ditcrimination during C, photosynthesis. with the greater con~plexitiesof
C , photosynthesis captul-ed in the term b' in Equation 6.8, analogous to the
terrrl h in Equation 6.1 .
Compared to C3 species, for C4 species relatively little variation in A T
associated with the prirnary pi-ocecces of C assimilation exists. This is because the value of the term ( b - o) is close to zero, whereas the term ( h -a).
from Eq~iatioti6.1, has a value of ca. 24 per mil. lo fact. for C, photosynthesis, the dependence of A T on c,/c,, [nay be positive or negative depending
on whether the value of 17 5 s greater or legs than then value of a, arid this will
depend an the iirnount of leakage from the bundle sheath. (Using the par;lmeter values cited, h' = a at @ = 0.36, and no dependence of A13C on c,/c,
would exist.) In practice. the value of sP has been usually found to be a little
less than 0.3. Conseqt~ently.the relationship between A'" ;and c,/c, for C,
species is usually negative, but with a small slope (Figure 6.1). For C3plants
tlie relationship between A13Cand c,/cclis positive, with a large slope.
Discriminatioii during C4 photosynthrsis may therefore vary with variation in c/c, or with variation in @ (or both c,/c, and @). In practice, it is diffjcult to distinguish between these two inlluences, and both of'tlienl will be
subject to genetic and environmental effects. Tlie value of c,/c,, would be
expected to respond to changes in stol~iatalconductancc, which is under genetic and environmental intluenccs as found for C3 (pccics. The value of 0
is subject to several poorly understood inlluences. Genetic variation in @
has been indicated fronl the results of comparisons of C4 species that differ
in aspects of their biochemistry and anatomy (Ehleringer and Pearcy,
1983). In soiiie C, specie%,the value of @ has bcen observed to change in response to drought and salinity stresses. These changes have been attributed
to changes in the activity of R~ibiscorelative to PEP carboxylase (Bowman
et al.. 1989; Saliendra et al.. 1996). Hubick et al. (1 990) attsibuted variation
in A13C among soghurn (Soqqiium birolor L.) genotypes to variation in either or both c,/c,, and @, whereas Henderson et al. ( 1992) found cD to be relatively constant at 0.2 and variation in A l T to be related to variation in
c-, /C(,.
Farqli har ( 1 983) concluded that radiation-LIS~
efficiency may be higher if
@ was low. and that this may be reflected i n productivity. Likewise, MI'
Plat leaf area (cm*)
LG
D E
t:
J;
0
c
2
5
0-E
g
s sg
zg g
6.
00:
* m u
m
E
rCD R
CD t,
Z m i T
g
$
q g g
'A
3
oG'
P
g
k5s
02:
g ez
8 6%
rAE c.
g z g
UQ CTC, 3
5. s.
-3-2
CD UQ
Z
Ec2.
R ~
"", + g
.,
0
KrJs
&zg
z.g ;
:53
00%
s d=r,
g+oJ
"
m 0
0
~
.$)plic.utior~cfCc~rt)o~r
Isotope Lh'acri~trir~ntiorl
in Ci~rcrrII~t~~,ror~crr~~r~t
21 7
Henderson. S.A., S. von Caemmerer, and G.D. Farquhar (19922). Short-term measurements of carbon isotope discrimiwition in several C4 species. illirtl-r~li~ir~
Jnrll-nn1 Plant Ph~.riology19: 263-285.
Henderson. S.A., S. von Caemlneser, G.D. Farquhar, L. Wade, and G. Hamn~cr
(1998). Correlation bctwecn carbon isotope discrimilr;ilion and transpiration efin tlrc glasshouse and ihc
ficiency in lincs of ihe C4 species Sol-,qhu,rz bi~*olor
field. Alcstralic111Joiln~nlcfYlunt Y I ~ ~ ~ i o l o25:
g j 1' 11-123.
Herve, D.. F. Fabre. E.F. Rerrios, N. Lerouk, G. A1 Chaarani. C. Planchoo, A.
Sarrafi. and L. Gentzbittel(2001). QTL analysis of photosynthesis and water status traits in sunflower (H~litn.rthusL L ~ ~ L I L I . \ L.) under greenhouse conditions.
.Joi(rr~nlcfE~r~crin~e11t~71
Bofart~l52: 1857- 1863.
Hubick, K.T. and G.D. Farquhar (1989). Genetic variation in carbon isotope discrimination and thc ratio of carbon gained to water lost in barley. Plnrzt. Crll find
En~.irorlrrlerlr12: 795-804.
Huhick. K.T., G.L. Hammer, G.D. Farquhar, L.J. Wadc. S. von Cacmmcrcr. and
S.A. liendcrson (1990). Carbon isotopc discrirnination varies gcnctically in C4
species. Ylnnt Yh?.siolo,q~.
92: 534-537.
Ishinlal-u K., K. Shirota, M. Higa, and Y. Kawa~nitsu(2001). Identification of quantitative trait loci for adaxial and abaxial stornatal frequencies in 01yzcr s a t i ~ ~ c ~ .
Plant Plljsiologj und Rioc.l~c.mi,tr~~
39: 173- 1 77.
Isla. R., R. Aragues. and A. Royo ( 1 998). Validity of variour physiological traitc as
screening criteria for salt tolerance in barley. Ficld Crops Reseur(.lz 5 8 : 97- 107.
James. R.A., A.R. Rivelli. R. Munns, and S. von Caemmerer (2002). Factor\ affectr y growth in salt-stressed durum \vheat. Rmcing C 0 2 assin~ilation,lcnf i n j ~ ~ and
tionnl Plczrrt Biology 29: 1393- 1403.
Johnson. D.A., K.H. Asay, L.T. Tics~cn.J.R. Ehlcringcr, and P.G. Jefferson (1990).
Carbon isotopc discrimination: Potcntial in screening cool-sca\on grasscs for
water-limited environments. Crop Scienc.c 30: 338-313.
Johnson, R.C. and L.hl. Bassett (1991). Carbon isotope discrirnination and water
use efficiency in four cool-season grases. Cr-opScic17i.r3 1 : 157-162.
Jornpuk, C., Y. Fracheboud, P. Stamp, and J. Leipner (2005). Mapping of quantitarntLy\ L,.) seedlings
tive trait loci associated with chilling tolerance in maize (ZPLL
Roioti!~56: 1 151-1 1 h i .
grown under field conditions. Jor~rrlolc?/Er~)ori~iicirial
Jones, H.G. 1993. Drought tolerance and water use efficiency. 111 J.A.C. Smith and
H. GrifYiths (etls.), iVuto- deft'cit,r:Pltrfrt r-c..rl~o/l,rcs
J r o ~ nr-rll to co~nrnrir~ity
(pp.
193-203). Oxford. UK: Bios Scientific Publishers.
Kanp, Y.K., K.A. Richards, and A.G. Condon (1 996). Intcraclions among carbon
isotopc discrimination. watcr use cfficicncy and nilrogcn niilrilion in whcat and
barley. Ko~.eirnJozirnill cjf Crop Scienc*c4 1: 3 18-33 1.
Liang Y.L. and K.A. Richards ( 1 993). Colcoptilc tiller dcvelopmcnt is associated
80: 1 19- 123.
with fdst early vigour in wheat. Ez(ph~,tic.cr
I.ope7-Castaneda, C. and R.A. Richards (1 493). Variation in temperate ce~salsin
rainfed environments. 111. Water ure and water-use efficiency. Fivld Crop, RPseurc.lz 39: 85-98.
Lope7-Castaneda, C.. R.A. Richards, and G.D. Fw-quhar (1995). Variation in early
vigor between wheat and 1,arley. Crop S C ~ ' C I35:
I C (472-179.
~
cv
nlalik. T.A., I). Wright, and D.S. Virk (1999). Tnlier~tanceof net photosynthesis and
nestil'r~r~l
L., under drought.
transpiration efficiency in spring wheat, TI-iticrlr~~
Ylarzt R~-c~etlirrg
11 8 : 93-95.
blartin. H., J. Nicnhiiis, G. King, and A. Schaefcr (1989). Restriction fragment
length polymorphis~nsassociated with water-use efficiency in tomato. Science
243: 1725- 1728.
hlt~hlc.J. and G.D. Farcluhar (1988). Effects of soil strcnglh on the rclalion of wateruse efficiency and growth to carbon isotope discrimination in wheat seedlings.
Ylurzt Phjsiology 86: 3 2-38.
Xlcrah, 0..E. Delccns, and P. Monncvcux (I 999). Grain yiclcl, carbon isotopc discri~rlination,mineral nrrd silicon content in d~lrunlwheat under different precipitalion rcgimcs. Pl~ysiolngiaPlclr~tur~(r?~
107: 387-194.
Merah, O., E. Dclecns, I. Souyris, M. Nachit. and P. Monnevcux (2001 ). Stabilitv of
carbon isotope discrimination and grain yield of durum wheat. Crop S c i e n c ~41 :
677-68 1.
3lo1gan. J.,4. and D.R. LeCain ( 1991). Leaf gas exchange and related leaf traits
amt)ng 15 winler wheal genotypes. Crop Scier2c.t. 3 1 : 443-448.
Mostlock, h1.Y. and G.L. Hammer (1999). Gcnolypc and Lvalcr limitalion effects on
transpiration efficiency in sorgltum. .Tour-nu1 of Clrlp Proll'r(ctio~r2: 265-286.
Pns\iourn, J.R. (1 977). Grain yield, harvest index and water uie of wheat. Jom~,rli~l
qf
tllc Arrvtralian JII stitr(te of ~lgricrlltr~ral
Scierlre 43: 1 17- 120.
Paccioura. J.B. (2004). Water use efficiency in the farmers' fields. In M.A. Bacon
(ed.), \Zfater r t ~ eeffi'cit7ncyiri jtlnrzt biology (pp. 302-321). Oxford. UK: Blackwell.
Peng S., K.G. Cassman, and M.J. Kropff (1995). Relatioriship between leaf photosynthesis and nilrogcn contcnl of ficld-grown rice in tropics. Cl.op Science 35:
1627- 1630.
Preston. T. (1992). The nleasurement of stable isotope natural abundance variations. Plant, C'cll crtrci! Envirorz~rrt.rzt15: 1091- 1097.
Rebetzke. C.J., T.L. Rotwright, C.S. Moore, R.A. Richards, and A.G. Condon
(2004). Genotypic variation in specific lcaf area for genetic improvement o f
early vigour i n whcat. Field Ct-opr Resccrrcl~88: 179-189.
Rebct7ke, G.J., A.G. Condon. R.A. Richards, and G.D. Farquhar (2002). Selection
lor rccluccd carbon-isotope discrimination incrcascs aerial biomass and grain
yield of rainfed bread wheat. Crop Scicrzce 42: 739-745.
Rehet7ke. G.J., A.G. Condon. R.A. Richards. and J.J. Kcad (2001 ). Phenotypic variation and sampling for leaf conductance in wheal (Tritir-11177(ie~tivi~nl
L. brecdinp populations. Euphyticu 121 : 335-341.
Rebet~ke,G.J., R.A. Richards, A.G. Condon, and G.D. Farquhar (2005). Wheat
gel-mplasm with i~nprovcdyield pcrformancc under drought for Australian
breeding programs. Project report. Grains Research and Development Corporation. Unpublished data.
Rebetzke, G.J., R.A. Richards, A.G. Condon. and G.D. Farcluhar (2006). Inhericrestivum L.).
tance of carbon isotope discrimination in bread wheat (7'1-iticr~n1
E~rpllyrira,klarch, pp. 1 - 10.