Journal of Experimental Botany, Vol. 48, No. 311, pp. 1309-1321, June 1997
Journal of
Experimental
Botany
Long-term effects of C0 2 enrichment and temperature
increase on the carbon balance of a temperate grass
sward
E. Casella and J-F. Soussana1
Fonctionnement et Gestion de I'Ecosysteme Prairial, INRA-Agronomie, 12 Av. du Brezet, F-63039 ClermontFerrand Cedex 2, France
Received 15 May 1996; Accepted 14 February 1997
Abstract
Perennial ryegrass swards were grown in large containers on a soil, at two N fertilizer supplies and were
exposed during two years in highly ventilated plastic
tunnels to elevated (700 ft\ I" 1 [COJ) or ambient atmospheric CO2 concentration at outdoor temperature and
to a 3 C increase in air temperature in elevated C0 2 .
The irrigation was adjusted to obtain a soil water deficit during summer. The daily net C assimilation was
increased in elevated C0 2 by 29 and 36% at the low
and high N supplies, respectively. Canopies grown in
elevated C0 2 for 14 to 27 months photosynthetized
significantly less rapidly, in both elevated and normal
CO2 concentrations, than their counterparts developed
in ambient C0 2 , but the magnitude of this effect was
small (-8% to -13%). Elevated C0 2 resulted in a large
increase in the fructan concentration in the pseudostems and laminae ( + 46% and +189%, respectively).
In elevated CO2, the hexose and sucrose pool
increased by 28% in the laminae, whereas it did not
vary significantly in the pseudo-stems. A 3°C temperature increase in elevated CO2 did not affect significantly the average WSC concentrations in the pseudostems and laminae. The elevated C0 2 effects on the
net C assimilation and on the nocturnal shoot respiration were greater in summer than in spring. On average, a 35% increase in the below-ground respiration
was measured in elevated CO2. At the high N supply,
a 3 °C increase in air temperature led to a decline in
1
the below-ground respiration due to a low soil moisture. The below-ground carbon storage was increased
by 32% and 96% in elevated CO2 at the low and high
N supplies, respectively, with no significant increased
temperature effect. The role for the below-ground
carbon storage of C02-induced changes in the root
fraction of the grass and of temperature-induced
changes in the moisture content of the soil are
discussed.
Key words: Climate change, grassland, gas exchange,
carbohydrates, carbon cycle.
Introduction
The effects of climate change on the role of the biosphere
in the global carbon cycle will depend on an altered
balance between photosynthesis and respiration. On a
global scale, it is estimated that the atmospheric carbon
is recycled by the vegetation every 4-6 years (Schlesinger,
1991). Therefore, an altered balance between CO2 uptake
and release in the biosphere could modify the upward
trend of the increasing atmospheric CO2 concentration
(Keeling et al., 1995). Grassland ecosystems account for
approximately 12% of the total carbon storage in the
continental biosphere (Schlesinger, 1991) and might
therefore contribute significantly to these changes.
In temperate grassland ecosystems, as the shoots
are periodically grazed or clipped, carbon sequestration
occurs mostly below-ground. Under elevated CO2, several
To wtiom correspondence should be addressed. Fax: +33 4 73 62 44 57. E-mail: soussana©clermont.inra.fr
Abbreviations: 350, outdoor climate; 700, outdoor climate + 350 /J \~' [CO2]; 700 + , outdoor climate + 350 /J I " ' [COj] and + 3 °C; N - , low N fertilizer
supply; N + , high N fertilizer supply; A3, gross canopy photosynthesis; An, net canopy photosynthesis; 8 b , carbon balance of the soil, roots and
stubbles; Bc, carbon balance before cutting; DM, dry-matter; Fb, carbon flux to the soil, roots and stubbles; /?„, (root and soil) respiration; Ra, aboveground dark respiration; WSC, water soluble carbohydrates; Vo, carbon above-ground yield.
© Oxford University Press 1997
1310
Casella and Soussana
factors might contribute to enhance this below-ground C
sequestration: (i) a sustained increase in plant productivity (Field et al., 1992; Gifford, 1994); (ii) an increased
allocation of photosynthates below-ground (Luo et al.,
1994); and (iii) a decreased decomposability of roots and
of plant residues due to a larger C:N ratio (Owensby,
1993; Ross et al., 1995). On the other hand, an increased
air temperature may lead to an increase in soil respiration
and in the rate of decomposition of soil organic matter
(Jenkinson et al., 1991).
With isolated plants, it has often been observed that
the initial stimulation in relative growth rate is only
transient, due to negative feedbacks through sink limitations (Arp, 1991) or morphological changes (Fonseca,
1996). Moreover, at the canopy scale, the stimulation of
photosynthesis does not automatically lead to enhanced
formation of new leaf area and plant dry matter (Korner,
1995). It is therefore necessary to study the carbon
balance of plant communities over long time periods,
extending over several years and in realistic conditions.
The carbon balance of pot-grown perennial ryegrass
(Lolium perenne L.) swards was previously studied by Nijs
et al. (1989) and, recently, Schapendonk and Goudriaan
(1995) reported the carbon balance of an established
perennial ryegrass sward over a full year. However, in
these studies the effect of an elevated temperature was
not considered and the water and nitrogen supplies were
non-limiting. The elevated CO2 and increased temperature
effects on the C balance of an established grass sward
grown at two N supply levels during two years, with a
summer water limitation, are discussed here. The results
concerning the productivity, the water balance, the nitrogen budgets and the root fraction of these swards were
previously reported by Casella et al. (1996) and by
Soussana et al. (1996).
It is shown that, during the growing season, the belowground carbon storage is increased in elevated CO2, both
at ambient and + 3 °C air temperature, and that this
partly results from the elevated CO2 effects on the root
fraction of the grass and from the increased temperature
effects on the water status of the soil.
Materials and methods
Climate control facility
A facility consisting of three 70 m2 plastic tunnels, with outdoor
temperature and CO2 concentration tracking devices was used
for the experiment. The climate inside each tunnel is controlled
(Casella et al., 1996) to obtain the following conditions: (350),
outdoor climate; (700), elevated CO2 (+350 ^tl I" 1 [CO2]);
(700+), elevated CO2 and temperature increase (+350^1 I"1
[CO2]and +3°C).
The means per regrowth of the microclimate parameters
(PAR, temperature, vapour pressure deficit of the air and CO2
concentration) outdoor and inside the three tunnels were
previously reported by Casella et al. (1996). A detailed
description of the experimental set-up and of the sward
management can also be found in Casella et al. (1996).
Sward management
In September 1991, 87 swards consisting of perennial ryegrass
(Lolium perenne L., cv. Preference) were sown in 0.5 m2 (45 cm
deep) containers, filled with a well homogenized loamy soil,
and were grown outdoor. In March 1993, 18 months after
sowing, these swards were assigned randomly to the three
highly ventilated plastic tunnels: 350, 700 and 700+ . From
March 1993 to June 1995, the established swards were grown
continuously under the 350, 700 and 700+ climatic conditions.
All swards were cut simultaneously at 4 cm height on five
occasions. Two rates of N fertilizer supply were compared: 160
(N-) and 530 (N+) kg N ha" 1 year"1.
All swards received the same irrigation volumes. From
September to June, the irrigation rate was equal to that supplied
to a (350) N + control sward kept at field capacity. In July
and August, to simulate a summer water deficit, the irrigation
rate was lowered to 70% and 50%, respectively, of that supplied
to the fully irrigated control sward (Casella et al., 1996).
Radiation interception and canopy gas exchanges
In 1994, during each regrowth radiation (PAR) interception
was measured twice weekly, that is before the start and after
the end of the canopy gas exchange measurements. At 12.00 h
(solar time), a ceptometer (Decagon Devices Inc, Pullman, WA,
USA) was placed successively above and below the grass
canopy, along each diagonal of two replicate containers.
The radiation (PAR) extinction profile within the canopy was
measured, before the cuts in May and in October 1994 around
10.00 h solar time with two replicate containers. A ceptometer
was placed at different depths within the canopy (every 5 cm at
the top and thereafter every 10 cm), hence delimiting horizontal
canopy layers. The canopy layers were then clipped with battery
powered shearers and the green lamina area was measured (Li
3100 Area Meter, LiCor, Lincoln, Nebraska, USA). The
extinction coefficient was calculated by fitting an exponential
(Monsi and Saeki, 1953) to the transmitted radiation and
cumulated leaf area data.
Using a computer-controlled open-flow system (Fig. 1),
canopy gas exchange was measured simultaneously in each
tunnel on one N - and one N + grass sward during 2-3 d each
week from April to October in 1993 and 1994. The same
containers were used during one growing season, but different
replicates were used in 1993 and in 1994.
Within each tunnel, the grass canopy was placed in a
transparent (polyethylene film, 30fim) enclosure (0.7 x
0.7 x 0.5 m, approximately 250 1) and the indoor tunnel air
was circulated (800-900 lmin' 1 , that is 3-4 vols min"1) in
the canopy enclosure by a centrifuge fan (Fig. 1). In order to
increase the precision of the dark respiration measurements, the
flow rate was decreased at night time (500-600 lrnin" 1 ) by
automatically adjusting the vent. The flow rate through the
enclosure was determined using the CO2 dilution method
(Daudet, 1987): a small flow of pure CO2 delivered by a mass
flow-meter (Tylan, Torrance, California, USA) was diluted in
the airflow,just after the fan, and the resulting increase in the
CO2 concentration at the inlet of the canopy enclosure was
determined. All CO2 concentrations were measured using an
automated sampler (Siemens, Germany) and an IRGA (Model
Unor 600, Mahiak, Hamburg, Germany) calibrated every 2
weeks with a 700 jll I" 1 [CO2] standard (Messer-Griesheim,
Germany).
The mean air temperature was, on average, 0.3 =C higher
Carbon balance under climate change 1311
Blower
soil. As a result, the values of the below-ground respiration
were equal when measured as described above, or by placing a
rubber stopper at the outlet of the soil container and measuring
after an equilibration period the efflux of CO2 in the shoot
enclosure (data not shown). Moreover, with bare soil, the CO2
concentrations at the inlet and at the outlet of the shoot
enclosure were equal when the pump was on (data not shown).
This indicates that the back-diffusion of CO2 from the soil was
negligible and, hence, that the CO2 production rates in the soil
were correctly estimated.
Interpolation of the carbon fluxes between two successive
measurement penods
The response of canopy net assimilation rate (AJ to the
transmitted radiation inside the tunnel (PAR) was fitted to a
simple rectangular hyperbola, according to the following model:
A^KA^PARyiKn
Soil air flow
Fig. 1. Schematic diagram of an automated open-flow gas exchange
device used to measure continuously CO2 exchange rates (canopy net
assimilation rate and below-ground respiration) of soil-grown perennial
ryegrass swards. The CO2 concentration was measured sequentially
with a first IRGA (calibrated m the range 0-1000 ^11"' [CO2]) at three
points in the main air stream: after the blower (A), at the inlet (B) and
at the outlet (C) of the canopy enclosure. The CO2 concentration was
measured with a second IRGA (calibrated m the range 0-5000 ^1 1"'
[CO2]) at the outlet of the soil (D). The air flow through the canopy
enclosure was measured by diluting a small amount of pure CO2
delivered by a massflow-meter(d) and by measuring the resulting CO2
enrichment m B (see Materials and methods).
during the day and 2 °C lower at night, in the canopy enclosures
compared to the tunnels. With a similar device, Daudet (1987)
also observed a cooling effect of the canopy encosure at night
(due to a negative radiative balance) and a daytime temperature
effect ranging from negative (for well-watered crops) to positive
(for water-limited crops). The mean values of transmitted PAR
in the canopy enclosures and in the tunnels were not significantly
different (data not shown), due to the high transmittance of the
polyethylene film used (Daudet, 1987).
A fraction of the airflowingthrough the shoot enclosure was
pumped (Model Miniport, KNF, Germany) at a constant flow
rate (6.0±0.2 1 min"1) through the soil column of the container
(Fig. 1). Both airflow and CO2 concentration were measured
at the outlet of the soil using, respectively, a flow-meter and
another IRGA (Model Beryl 100, Cosma, France) calibrated
every 2 weeks with a 4500 /^l I"1 [CO2] standard (L'Air Liquide,
France). The difference in response between the two IRGA's
was checked automatically every 30 min with outdoor air. From
the flow rate of air passing through the soil and the difference
between the CO2 concentration of the mixed air in the enclosure
and that at the outlet of the soil, the CO2 production rates in
the soil can be calculated separately from the gas exchanges in
the canopy (Van de Geijn et al., 1994).
A small pressure head (approximately 50 Pa) was maintained
by the open-flow system. This pressure head and the depression
caused by the pump eliminated back-diffusion of CO2 from the
+ PAR)}-!^
(1)
where AmMJL is the maximal photosynthetic rate of the canopy,
Km is the transmitted PAR value at half / l , ^ and R^ is the
canopy dark respiration rate. This model always explained
more than 90% of the total variability of An during one
measurement period. The measured (Amcl) and simulated (A^
mean diurnal values of An were highly correlated, without any
significant bias, since during the growing season: ^4^,, = 0.998
^mei. ^ = 0.97, P< 0.0001. The rectangular hyperbola model
was preferred to that of Acock et al. (1978), also used with
perennial ryegrass by Nijs and Impens (1989), since the Acock
model includes too many parameters, that can not all be fitted
without using further assumptions.
To establish the carbon balance of the swards during one
regrowth, the values of the net assimilation rate in between two
successive measurement periods were interpolated as following.
First, An was fitted using the rectangular hyperbola model for
each measurement period. Secondly, the parameters obtained
during two successive periods were averaged and the values of
An for the missing days were simulated, using the averaged
rectangular hyperbola model and the 15 min average values of
the transmitted PAR. In this way, interpolated daily means of
Aa and of R^ were calculated.
Carbon fluxes per regrowth
The carbon fluxes per regrowth were calculated by cumulating
the daily means of the canopy net assimilation (A^, dark
respiration (i^) and below-ground respiration (Rb). Yc, the
carbon yield of the sward was calculated by multiplying the
harvested biomass of the sward (i.e. the drymatter yield,
previously reported by Casella et al., 1996) by the carbon
concentration of the shoots (on average, 40%, data not shown).
The other carbon fluxes were then calculated from these data.
The carbon balance of the grass swards before cutting (Bc) was
calculated as An-Rb. The carbon flux from the above to the
below-ground carbon compartment (Fb) was calculated as
An-Yc. Finally, the carbon balance of the below-ground
compartment (Bb) was calculated as Fb-Rh. The canopy gross
assimilation {As) was defined as the daily net assimilation plus
the estimated daily shoot respiration. To provide an estimate
of the 24 h shoot respiration (R4), the diurnal and nocturnal
dark shoot respiration rates were assumed to be equal at a
given air temperature. The diurnal respiratory rate (Raa) was
estimated from the measured nocturnal rate (Rao) corrected for
temperature as: R,id = Ran Qi0l{T''~T°"10}
where Ta and
Tn,
respectively, are the daily averages of the diurnal and of the
nocturnal air temperature in a canopy enclosure. With wheat,
for short-term variations of ± 5 °C around the growth temper-
1312
Casella and Soussana
ature, Qw values of 1.8, 1.6, 1.5, and 1.3 at growth temperatures
of 15, 20, 25, and 30 °C, respectively, were reported by Gifford
(1995). Using these results, a distinct Ql0 value was calculated
in each tunnel and for each regrowth. With this procedure, on
average, the respiration rate was increased by 53% during the
day compared to its value at night. It should be noted that, by
using a constant Ql0 value of 2 (Faurie, 1995), a similar result
(+62% increase) would have been obtained.
Acclimation of canopy photosynthesis
Acclimation of photosynthesis was tested in May and October
1994 and in May and June 1995, at the canopy level, by
measuring, at the end of the regrowth period, the photosynthetic
response for a CO2 concentration which differed, or not, from
that experienced during growth. Swards grown at 350 and at
700 /xl I"1 [CO2] were exchanged between tunnels in the evening
and, on the next day, the gas exchanges were measured
according to the procedure described above. Thereby, the short
and the long-term responses of canopy photosynthesis to
elevated CO2 were compared. Since the mean daily irradiance
varied from one day to the next, the radiation response curve
of canopy photosynthesis was fitted to the rectangular hyperbola
model (Eq. 1) and simulation results for two irradiances (500
and 1000 fimo\ photon PAR m~2 s"1) were compared.
Water soluble carbohydrates
In May, July and October 1993 and 1994 and in June 1994,
approximately 1 week after and 1 week before cutting, mature
tillers (sampled from 5 containers, with each replicate consisting
of 4 tillers) were cut with scalpels at the ground level and were
separated into leaf lamina, pseudo-stem (i.e. leaf sheaths, nodes
and internodes) and dead material. In order to avoid diurnal
fluctuations in the carbohydrate content, the samples were
harvested around 12.00 h solar time.
The leaf and pseudo-stem samples were stored at — 20 °C,
freeze-dried and extracted twice during 30 min at 80 °C, first in
2 ml alcohol:water (40/60, v/v)-mixture and then in 2 ml distilled
water. The extract was vacuum dried (Uniequip, Germany),
solubilized in 1 ml distilled water and passed first through a
Sep-Pak C18 cartridge (Millipore, USA), to remove apolar
components and then through ion-exchange columns (extra-sep
columns SAX and SCX-Lida, USA), which were rinsed with
3 ml pure water. The neutral eluant was then vacuum dried,
solubilized with 2 ml pure water and filtered at 0.45 ^m.
Component sugars were quantified by high pressure liquid
chromatography (HPLC), according to Bancal and Gaudillere
(1989), using a HPX-87 P column (Biorad, France) and a
refractometer, against inulin, stachiose, sucrose, glucose,
rhamnose, and fructose as standards.
Statistical analysis
The statistical design is a split-plot with the climate as main
factor and N supply as split factor. A disadvantage is the
absence of true replication (duplicates of the tunnels or
chambers). This can be compensated for by the comparison of
the tunnels (Casella et al., 1996) and chambers (see above)
microclimate. Two replicate containers in each tunnel were
selected at random at the start of the experiment and were used
successively in the two growing seasons for the measurement of
gas exchanges. In each tunnel and for each N supply the annual
dry-matter yield of these containers did not differ significantly
CP>0.95, Student's Mest) from that of the remaining containers
although it was, on average, 8 ±3% lower. An ANOVA was
performed on the data of the annual carbon balance, with the
climate and the N supply as factors and the year (1993 or
1994) as the time variable. A repeated measure procedure is
not needed in this case, since the corresponding gas exchange
data were obtained with different containers in the two years.
An ANOVA was also performed on the water soluble
carbohydrates data, which were obtained by destructive sampling, with the climate, the N supply and the stage of regrowth
as factors. The homogeneity of variance was checked by
applying the variance test ratio. All statistical analysis
were made using the Statgraphics Plus (Manugistics, USA)
statistical package.
Results
Carbon fluxes during the growing season
All carbon fluxes (except R^ and Rb, the above and belowground respirations) in the grassland ecosystem differed
significantly between the two years (Table 1). The N
fertilizer supply effect was highly significant, resulting in
an increase in the gross canopy photosynthesis, in the
dark respiration and in the C balance before cutting of
the grass sward (Table 1).
During the growing season, the cumulated gross assimilation (At) was significantly increased in elevated CO 2 ,
by 28% and 35% at N - and N + , respectively (Table 1).
In the same way, both the above and the below-ground
cumulated respirations were significantly stimulated in
elevated CO 2 . The dark respiration was increased, on
average, by 25-28% for both N fertilizer supplies
(Table 1). The below-ground respiration was increased in
elevated CO 2 by +33% and +36% at N - and N + ,
respectively (Table 1). As a result of these changes, the
C balance before cutting (calculated as An-Rb),
was
enhanced in elevated CO 2 by +28% and +44% at N and N + , respectively.
The carbon flux to the below-ground compartment (Fb)
was calculated as the net assimilation minus the carbon
yield of the cut swards; this flux was also significantly
increased, by 31% and 66% at N - and N + , respectively.
Finally, the carbon balance (Bb) of the below-ground
compartment, which reflects the carbon storage in the
soil and in the non-harvested plant parts (roots and
stubbles) was significantly increased by 30% and 96%, at
N - and N + , respectively (Table 1). On average, from
April to October of each year, a supplemental belowground carbon storage of 1.3 and 2.7 t C ha" 1 occurred
in elevated CO 2 at N - and N + , respectively.
At N-, both in ambient and in elevated CO 2 the above
(Rd) and below-ground (Rb) respirations accounted for
36-37% and 20-21% of the gross assimilation (Ag),
whereas the above (y c ) and below-ground (Bb) carbon
accumulations amounted to 13% and 30% of Ag (Table 1).
Therefore, the fate of the photosynthetic C was similar
in ambient and in elevated CO 2 at N - .
Increasing the N supply resulted in a shift of C accumulation from the below-ground compartment towards the
harvested shoots (5b and Y^ Table 1). At N + , elevated
Carbon balance under climate change
Table 1. (A) Average seasonal carbon fluxes (±1 s.d.) in 1993 and 1994 of perennial
October) and (B) AN OVA with the climate, nitrogen supply and year as factors
ryegrass swards
(tC
1
ha'
from
1313
April
to
Note that, as different replicate containers were used in different years, the interactions between the year and the other factors were
not included in the ANOVA model. Gross canopy photosynthesis (At), above-ground dark respiration (R^), (root and soil)
respiration (Rb), carbon balance before cutting (Bc), carbon above-ground yield (Tc), carbon flux to the soil, roots and stubbles
(Fb) and carbon balance of the soil, roots and stubbles (Bb). The daily values of An, R^ and Rb were measured during 2-3 days
per week and interpolated in between two successive measurements (see Materials and methods). The diurnal respiration rate was
calculated by correcting for temperature and the nocturnal respiration rate, assuming that diurnal and nocturnal shoot respirations
are equal at a given air temperature (see Materials and methods). A%, B^ Fb, and Bb were then calculated as: AD + Ri, An-R^
An— Yc, and An — Rb— Yc, respectively. Yc is the annual carbon yield, calculated from the annual DM yield, previously reported by
DM, N - : 160 kg N ha ~1 y" 1 ; N + : 530 kg N ha - 1 y" 1 ; 350:
CaseUaera/. (1996), and from the C concentration in the harvested
control climate; 700: + 350 f iir 1 [CO 2 ]|;700 + : + 3 5 0 ^ 1 " 1 [CO2] and + 3 ° C .
N+
N-
(A)
350
700
700 +
14.9 ±0.4
5.5±0.3
3.0 ±0.4
6.4 ±0.1
2.0 ±0.1
7.4 ±0.5
4.4 ±0.2
19.0±0.4
6.9±0.3
4.0±0.4
8.2±0.1
2.4±0.1
9.8±0.5
5.8 ±0.2
20.6 ±0.4
8.0±0.3
4.8±0.4
7.8±0.1
(t Cha" 1 from April to October)
\
B-4
Rb
Bc
Yc
fb
A,
2.2±0.1
10.4±0 5
5.6±0.2
(B)
CO 2
Temperature
N
Year
COjxN
Temperature x N
N.S
•**
*••
N.S
N.S
+•
*
N.S
**
N.S
N.S
N.S
N.S
N.S
N.S
N.S
*
350
700
700 +
18.9±0.4
6.9±0.3
3.0 + 0.4
8.5±0.1
5.7 + 0.1
5.8±0.5
2.8±0.2
25.1 ±0.4
8.8 + 0.3
4.1 ±0.4
12.2±0.1
6.7±0.1
9.6±0.5
5.5±0.2
24.4 ±0.4
9.0 ±0.3
3.3 ±0.4
12.1 ±0.1
5.8±O.l
Be
F
N.S
N.S
N.S
**
N.S
N.S
••
•**
N.S
N.S
N.S
( t C h a - 1 1 from April to October)
*••
*••
N.S
9.6±0.5
6.3±0.2
*
•
and ••• denote a significant effect at P<0.05, P<0.0\ and /><0.001, respectively.
CO 2 increased the share of the photosynthetic carbon that
was accumulated below-ground, since Bb accounted for 22%
of At at 700, compared to 15% at 350 (Table 1). Therefore,
during the growing season, the fraction of the assimilated
carbon that was stored below-ground was not modified in
elevated CO 2 at N - and was slightly increased at N + .
A significant interaction between N and elevated CO 2
effects occured with Bc and Bb, as the elevated CO 2 effects
on the C balance before cutting and the below-ground
carbon storage were larger at N + than at N - (Table 1).
A supplemental 3 CC in elevated CO 2 slightly increased
the daily dark respiration (by 16% and 2%, at N - and
N + , respectively), but this temperature effect was not
significant. The increase in air temperature had no significant effect on the below-ground respiration (Rb) at N-,
but reduced it at N + . Finally, the below-ground carbon
storage (2?b) was not significantly affected by the elevated
air temperature (Table 1).
Seasonal changes in the elevated CO2 effects
The seasonal changes in the elevated CO 2 effects on the
carbon fluxes were examined in 1994, as during this year
the soil water content was similar at 350 and at 700
(Casella et al., 1996). Figure 2A and B shows the seasonal
time-course of the differences between the diurnal net C
assimilation rates at 350 and 700^1 I" 1 [CO 2 ]. The
fluctuations are partly due to the small effect of elevated
CO 2 just after each harvest and the gradual increase of it
during the following weeks until the next cut. In the
present study, the percentage increase of the diurnal An
in elevated CO 2 was similar in spring or autumn (+26%,
on average, for the May, June and October regrowths,
data not shown), but larger in summer (+49%, on
average, during the July and September regrowths, data
not shown) (Fig. 2).
The relative rates of change of the above-ground (dark)
and below-ground respirations in elevated CO 2 were
both highly correlated with that of the canopy gross
assimilation (P<0.001, data not shown). However, the
magnitude of the elevated CO 2 effects on the aboveground respiration (up to + l g C m ~ 2 d ~ 1 , Fig. 2C, D)
was much smaller, in 1994, than on the below-ground
respiration (up to + 8 g C m " 2 d ~ 1 , Fig. 2E, F) and the
increase in the above-ground respiration in elevated CO 2
was not significant during the May and June regrowths
(Fig. 2C, D).
Seasonal changes in the temperature increase effects
Figure 3A and B shows the seasonal time-course of the
differences between the diurnal An at +3°C (700+) and
1314
Casella and Soussana
E
O
<
I
1
1
E
U
1
D>
O
m
m
•o
<r
l
o
o
TJ
<r
0
0
-0
E
O
cc
I
.£>
IT
100
150
200
250
100
150
Julian days (1994)
200
250
300
Julian days (1994)
Fig. 2. Differences between elevated (700) and ambient (350) CO 2 for the diumal net canopy carbon assimilation (A, B), the nocturnal dark
respiration (C, D), and the daily below-ground respiration (E, F) during the regrowths ( O B , May; A A , June; V T , July; < > • , August and
September; &-*• , October) in 1994. (A, C, E) N - , low inorganic N supply (160kg N ha" 1 year" 1 ); (B, D, F) N + , high inorganic N supply
(530 kg N ha" 1 year" 1 ). Measured values (•) and simulated values (—) obtained by interpolating (see Materials and methods) Please note the
difference in scale between the figures.
Table 2. Influence of growth and measurement concentration on the net assimilation rate (mg C02m
canopies under 500 (An 50J and 1000 (A, l0OO) iiinol photons m~2s~l
2
min
l
) of perennial ryegrass
The results are the means of four measurements in May and October 1994 and in May and June 1995, with one replicate container at each
occasion. The statistical significance was analysed by ANOVA, with the growth and measurement CO 2 concentrations, the N supply and the
observation date as factors
Growth CO 2
Measurement CO 2 I
NN+
NN+
1
Statistical significance
700
350
'[CO 2 ]
350
700
350
700
44
65
61
94
51
82
80
128
39
61
49
85
51
72
74
110
and ••* denote a significant effect at P<0.05. / > <0.01 and / ) <0.001, respectively.
Growth CO 2
Measurement CO 2 N supply
Carbon balance under climate change
1315
E
O
<
I
E
O
I
O
cc
I
100
150
200
250
Julian days (1994)
100
150
200
250
300
Julian days (1994)
Fig. 3. Differences between + 3°C (700+) and ambient (700) temperatures in elevated CO2 for the diurnal net canopy carbon assimilation (A, B),
the nocturnal dark respiration (C, D), and the daily below-ground respiration (E, F) during the regrowths (DB, May;A • June, V T, July; O • ,
August and September; •&+, October) in 1994. (A, C, E) N-, low inorganic N supply (160 kg N ha" 1 year"1); (B, D, F) N + , high inorganic N
supply (530 kg N ha ' year '). Measured values (•) and simulated values (—) obtained by interpolating (see Materials and methods). Please note
the difference in scale between the figures.
at ambient air temperature (700). A supplemental 3°C
in elevated CO2 increased the photosynthetic response
during the spring and autumn regrowths and, conversely,
decreased it in summer, especially at N +.
The relative rate of change of the above-ground (dark)
respiration at +3°C in elevated CO2 was highly
correlated with that of the canopy gross assimilation
(P<0.0001, data not shown). However, the changes in
the below-ground respiration at + 3 CC were not correlated with that of Ag: at N-, the temperature increase
effect was negative during summer and positive in spring
and autumn but, except for the last cut, a supplemental
3 °C lowered the below-ground respiration of the grass at
N + throughout the growing season (Fig. 3E, F).
Radiation interception and extinction within the canopy
Figure 4 shows the percentage radiation interception at
700 versus 350 ^.1 I" 1 [CO2] in 1994. The percentage
radiation intercepted by the grass canopy was not significantly modified at N - (Fig. 4A). Nevertheless, at N + ,
the grass sward intercepted significantly more radiation
in elevated compared to ambient CO2 (Fig. 4B). At N + ,
the percentage radiation interception at 700 versus 350
followed a highly significant logistic sigmoid model.
According to this model, when the percentage radiation
interception reached 50% in the control, 66% of the
incident radiation was, on average, intercepted in the
elevated CO2 treatment (Fig. 4).
By contrast, an increased air temperature (+3°C) in
1316
Casella and Soussana
o
o
r—
100
•
20
Light(PARI interception 350
(X)
100
60
80
100
Light|PAR| interception 350
(X)
D
o
o
I—
40
N+
80
60
1
TT—
/
/
•j
40
20
0
20
40
60
80
interception 700
(X)
20
40
60
80
100
Light|PAR| interception 700
(X)
Fig. 4. Percentage radiation interception by perennial ryegrass swards, during the five regrowths ( • • , May; A A, June; V T, July; O • , August
and September; •&+, October) in 1994 at 700 versus 350 ^1 I"1 [CO2] (A, B) and at +3 °C versus ambient temperature in elevated CO2 (C, D). (A,
C) N- (open symbols), low inorganic N supply (160 kg N ha" 1 year"1); (B, D) N+ (closed symbols), high inorganic N supply (530kg N ha" 1
year"1). The vertical bars denote the standard error of the mean whenever it was larger than the symbol size. The regression plotted in (A)
is: r=(0.999±0.013) -Tfor /i = 48 and ^ = 0.971. The data in (B) were fitted to a logistic sigmoid: K=100/(l-e"LIX-;r30)), with JT5O = (39.3±0.3)
and £ = (0.063 ±0.002). ^ = 0.989, /><0.0001). The dotted lines depict the confidence interval of the regression at (/><0.05).
elevated CO 2 had variable effects on the percentage
radiation interception of the grass sward. For both N
supply treatments, an increase in radiation interception
was observed in May and in October, but the percentage
radiation interception was lowered at +3°C in July
(Fig.4C, D).
The extinction coefficient (k) within the canopy was
calculated from the data of transmitted PAR and cumulated leaf area before the cuts in May and in October
1994. There was no significant effect of the climate
treatments on the extinction coefficient (data not shown).
Acclimation of canopy photosynthesis
As the swards were exchanged in the evening and were
measured during two successive days, the long-term acclimation was compared with acclimation during 10—48 h.
There was, however, no indication that the net assimilation rate varied between the two successive days of
measurement at a CO 2 concentration differing from the
growth concentration. On average, ryegrass canopies
grown in elevated CO 2 photosynthetized less rapidly, in
both elevated and normal CO 2 concentrations, than their
counterparts developed in ambient CO 2 (Table 2). This
negative effect of the growth CO 2 concentration on the
net assimilation rate was significant both for low and
high radiation conditions (An500 and ^niooo. simulated
values of An under 500 and 1000 ^mol PAR m~2 s" 1 )
and reached —8% and - 1 3 % , on average, with An500
and Anloo0, respectively (Table 2).There were no significant interactions between the effects of the growth and of
the measurement CO 2 concentrations and the N supply
level.
Water soluble carbohydrates
At the time of sampling, around noon, fructans accounted
for 40% and 67% of the water soluble carbohydrates
(WSC) in the laminae and pseudo-stems, respectively
(Table 3). The remaining WSC were mostly sucrose,
Carbon balance under climate change
1317
Table 3. Average concentration of hexose and sucrose (glucose, fructose and sucrose), fructans and total water-soluble carbohydrates
in the lamina and pseudo-stem of perennial ryegrass in 1993 and in 1994
Different letters for a factor in one column indicate a highly significant difference (P<0.01) between treatments (LSD, multiple range test). The
data are the mean of five replicates sampled in May, July and October 1993 and 1994 and in June 1994, approximately one week after and one
week before cutting. 350: control climate; 700. + 350^1 I" 1 [CO 2 ]; 700 + : +350 M l I" 1 [CO2] and + 3°C. N - : 160kg N h a " 1 y ' 1 ; N + : 530kg
' 1
Hexose and sucrose
(mg g" 1 DM)
Fructans
( m g g " 1 DM)
Total water soluble
carbohydrates (mg g" 1 D M )
Lamina
Pseudo-stem
Lamina
Pseudo-stem
Lamina
Pseudo-stem
Climate
350
700
700 +
43 a
55 b
59 b
50 a
47 a
56 b
19 a
55 c
33 b
112 a
164 b
151 b
63 a
111 b
94 b
163 a
212 b
209 b
N supply
NN+
52 a
52 a
47 a
54b
49 a
22 b
164 a
120 b
102 a
76 b
213 a
177 b
Regrowth
Start
End
47 a
58 b
43 a
59 b
10 a
61 b
93 a
191 b
58 a
121 b
137 a
252 b
fructose and glucose. Traces (less than 1% of the WSC)
of stachiose, kestose and DP3 saccharides were also
found.
The total WSC concentration increased significantly
during regrowth and was greater in the pseudo-stems,
compared to the laminae (Table 3). A reduction in the N
supply resulted in a significant increase in the WSC and
fructans concentrations in the pseudo-stems and laminae
(Table 3). However, the average concentration of the
sucrose and hexose pool in the pseudo-stems and laminae
was, respectively, reduced or unchanged at a low compared to a high N supply (Table 3).
Elevated CO2 resulted in a large increase in the fructan
concentration in the pseudo-stems and laminae (+46%
and +189%, respectively). In elevated CO2, the hexose
and sucrose pool increased by 28% in the laminae, whereas
it did not vary significantly in the pseudo-stems (Table 3).
After 1 year exposure to elevated CO2, the total WSC
concentration in the laminae was multiplied, on average,
by a factor of 2 to 3 (Table 3).
A 3 °C temperature increase in elevated CO2 did not
affect significantly the WSC concentrations in the pseudostems and laminae (Table 3). Nevertheless, on average, a
supplemental 3 °C increased significantly the concentration of the hexose and sucrose pool in the pseudo-stems,
while the opposite was observed with the fructan pool in
the laminae (Table 3).
Discussion
Water soluble carbohydrate accumulation
Ample rooting volume was supplied in the present study,
although deep roots (below 45 cm) were not allowed to
develop. Under these experimental conditions, a large
increase in the water soluble carbohydrate pool in elevated
CO2, both in the pseudo-stems and in the laminae
(Table 3) was observed. The starch content in these plant
parts was not measured. However, in a free air CO2
enrichment experiment, the starch content of laminae and
pseudo-stems of perennial ryegrass was always less than
10% of the WSC (M Frehner, personal communication).
Smart et al. (1994) reported that elevated CO2 induced
fructan synthesis in all leaf tissue fractions of vegetative
wheat canopies, although fructan formation was greatest
in the uppermost leaf area. The increase in elevated CO2
of the fructan concentration in the pseudo-stems (+46%,
Table 3) might have contributed to the faster formation
of new leaf area after clipping in the high N treatement
(Fig. 4). However, with the low N supply treatment, leaf
growth was clearly N limited since, despite an increase in
the WSC concentration in the pseudo-stems after cutting,
no enhancement of radiation interception was observed
in elevated CO2 (Fig. 4).
In the laminae, the average share of fructans in the
total WSC increased from 30 to 50% in elevated CO2,
due to a large increase (+190%) in the size of this
vacuolar sugar pool (Table 3). Fructan accumulation
occurs during periods where photosynthetic capacity is in
excess (Eagles, 1967; Pollock and Cairns, 1991). The
large increase in the fructan content of the laminae in
high CO2 might indicate that part of the non-structural
carbohydrates assimilated were not translocated from the
leaves at night, as reported previously by Wong (1990),
with cotton high CO2-grown plants.
Acclimation of the canopy photosynthesis
In the present study, no attempt was made to determine
leaf photosynthetic characteristics, but photosynthetic
acclimation was studied at the canopy level. Any difference between the long-term and short-term responses of
canopy photosynthesis to elevated CO2 may result from
changes in stomatal density or canopy architecture.
1318
Casella and Soussana
Nevertheless, these factors were apparently not modified
in elevated CO2, as: (i) no change in stomatal density
were reported by Ryle and Stanley (1992) and by Gay
and Hauck (1994) with Lolium plants grown in elevated
CO2; (ii) no significant elevated CO2 effect on the extinction coefficient was observed; (iii) the measurements were
made before cutting when radiation was fully intercepted
at N +; (iv) there was no difference in radiation interception between ambient and elevated CO2 at N-.
Due to the low magnitude of the down-regulation of
photosynthesis (-8% to -13%, Table 2) in high-CO2
grown swards, ryegrass plants grown at 700 /nl I" 1 [CO2]
maintained a substantially higher rate of photosynthesis
than those grown at normal ambient CO2 (Table 2).
While photosynthetic down-regulation may be considered
as the rule rather than the exception (Bowes, 1993),
complete down-regulation was not observed with Lolium
perenne (Ryle et al., 1992; Nijs et al., 1995), with a native
C3 saltmarsh sedge, Scirpus olenyi (Jacob et al., 1995),
or with pot grown C3 grasses (Gloser and Bartak, 1994;
Lutze and Gilford, 1995).
Jacob et al. (1995) reported that an elevated CO2
concentration increased carbohydrate concentration and
the ensuing acclimation of the photosynthetic apparatus
(Stitt, 1991) led to a reduction in the protein complement,
especially, Rubisco, which reduced the photosynthetic
capacity. In the present work, there are some clear
indications, of an increase of the soluble carbohydrates
(Table 3) and of a decline in the protein content (Soussana
et al., 1996) of leaves from high CO2-grown plants, but
it was not possible to ascertain whether these changes
contributed, or not, to the down-regulation of canopy
photosynthesis.
Canopy photosynthesis
On average, for the two N supplies, a + 26% increase of
An was observed in spring and autumn in elevated CO2,
in good agreement with Schapendonk and Goudriaan
(1995) who observed a 28% increase of the daily An, in
their experiment with ryegrass canopies without any
nutrient or water limitation. However, the corresponding
increase was larger in summer (+49%, on average for the
July and September regrowths, Fig. 2), presumably due
to the mild water stress conditions which occurred at 350
and 700 during the summer regrowths (Casella et al.,
1996). With field-grown perennial ryegrass, the threshold
for stomatal closure is approximately at —1.0 MPa (Jones
et al., 1980). In these conditions, although the pre-dawn
leaf water potential was apparently always above this
value in the 350 and 700 treatments (Casella et al., 1996),
the light response curves of An displayed an hysteresis
pattern for some of the measurements during the July
and September regrowths (data not shown). The elevated
CO2 effect on the diurnal An was apparently enhanced in
summer due both to these mild water stress conditions
(Andre and Du Cloux, 1993) and to the warmer air
temperatures (Long, 1991).
The average increase of AD in elevated CO2 was smaller
at N - (+29%) compared to N + (+ 36%) (Table 1). From
the daily net assimilation rate and from the daily amount
of radiation intercepted, the radiation use efficiency of
the grass canopy was calculated. On average, elevated
CO2 increased the radiation use efficiency at N - from
0.036 to 0.044mol C mol PAR'1, that is by +22%. At
N +, the corresponding increase in radiation use efficiency
was smaller (from 0.037 to 0.041 mol C mol PAR'1) and
part of the elevated CO2 effect on the net assimilation
rate resulted from the enhancement of radiation interception which occurred at the start of the regrowth in the
high N supply treatment (Fig. 4).
A supplemental 3 °C in elevated CO2 increased the
radiation interception and the photosynthetic response
during the spring and autumn regrowths and, conversely,
decreased it in summer, especially at N + (Figs 3, 4).
When the air temperature (mean of 700 and 700 +) was
below-optimal (below 18.5 ±1°C, Casella et al., 1996),
the enhancement of radiation interception resulted from
the stimulation of the above-ground growth by a supplemental 3°C (Fig. 4). Also, an increased air temperature
favoured the primary effect of CO2 on photosynthesis,
that is a reduction of photorespiration (reviewed by Long,
1991). Nevertheless, during summer, and especially at
N +, a supplemental 3 °C in elevated CO2 had nil or
negative effects on radiation interception (Fig. 4) and
diurnal net assimilation (Fig. 3). This seasonal change
partly resulted from the lower soil moisture at 700+ ,
compared to 700, which reduced the above-ground growth
in this treatment (Casella et al., 1996) and, partly, from
the increased vapour pressure deficit of the air in the
+ 3 °C tunnel, which presumably led (in conjunction with
the soil water deficit) to an increased diurnal stomatal
closure.
Above and below-ground respiration
Nijs et al. (1989) reported a large increase in Rd in highCO2-grown ryegrass canopies. By contrast, Bunce and
Caufield (1991) measured decreased whole-plant respiration rates in elevated CO2 in their experiments with
ryegrass. In the present study, in agreement with Ryle
et al. (1992) and with Schapendonk and Goudriaan
(1995), the elevated CO2 effect on the nocturnal shoot
respiration was small in spring (+6% during the May
and June regrowths). This effect became, however, larger
in summer (+50% during the July and September
regrowths) (Fig. 2). This seasonal change in the CO2
effect on the dark shoot respiration apparently resulted
from the large stimulation of the above-ground drymatter yield in elevated CO2 during summer: +48% for
Carbon balance under climate change
the July and September cuts, compared to +6% during
the rest of the year (Casella et al, 1996).
On average, a supplemental 3 °C in elevated CO2
increased the nocturnal shoot respiration (Fig. 2) by 21%
and 9%, at N - and N +, respectively, despite a decline in
the average shoot mass at cutting date (Table 1). This
indicated, an average increase in the shoot respiration per
unit mass at an increased air temperature in elevated
CO2. Assuming a Q10 value of 2, the shoot maintenance
respiration per unit mass would be increased by 23% with
a 3°C temperature increase in elevated CO2. However,
Gifford (1995) has shown that the shape of the response
of whole plant respiration to growth temperature was
different from that of the short-term response, being a
slanted S-shape, declining at above-optimal air temperatures. According to Gifford (1995), the ratio of shoot
respiration to photosynthesis is approximately constant
with respect to air temperature and CO2 concentration.
Our results support this conclusion, as it can be calculated
from Table 1 that this ratio was comprised in a narrow
range (between 35% and 39%).
By contrast with the report by Schapendonk and
Goudriaan (1995), who observed a small (-7%) decline
in the below-ground respiration, a rather large increase
in this carbon flux was obtained, on average, + 33% and
+ 36% at N - and N + , respectively (Table 1). This
discrepancy can be assigned to differences in the
methodology used.
Pressure differentials inside an open-top chamber,
resulting from the blower operation, lead to a decrease
in soil CO2 concentration (Nakayama and Kimball,
1988). In the present study, this problem was avoided
through the use of the technique developed by Van de
Geijn et al. (1994), resulting in the separate measurement
of the canopy and soil CO2 fluxes. Nevertheless, this
technique, which was also used by Schapendonk and
Goudriaan (1995), was adapted in order to reduce the
air flow through the soil, since these authors reported
that it may lead to an overestimation of the below-ground
respiration, because the soil layers are continuously
flushed with air, which changes the partial pressures of
CO2 and O2 in the soil, resulting in a large increase of
the respiratory activity of the soil microbial biomass
(Dommergues and Mangenot, 1970).
In the present study, only a small fraction (0.25-0.5%)
of the air flow passing through the canopy chamber was
pumped through the soil. Per unit ground area, the air
flow passing through the soil was approximately 27 times
lower (12 versus 320 lmin" 1 m~2) than in the study
reported by Schapendonk and Goudriaan (1995). Hence,
the soil oxygenation was limited, lower values of (root
and soil) respiration (on average, 7.3 versus 30 g CO2
m" 2 d" 1 ) were measured and, by contrast with the report
by Schapendonk and Goudriaan (1995), the soil respiration did not exceed the net CO2 assimilation (Table 1).
1319
The relative rate of change of the below-ground respiration in elevated CO2 closely correlated with that of the
gross assimilation (r2 = 0.648, P<0.001, data not shown),
indicating that the below-ground respiration was partly
under the control of the photo-assimilate supply to the
roots. However, the below-ground respiration was also
positively correlated with the air temperature (data not
shown) and with soil moisture (r2 = 0.13, P< 0.001, data
not shown). The role of the soil moisture can best be
illustrated by the fact that a 3 °C increase in air temperature led to a strong decline in the below-ground respiration at N + (Fig. 3), due to the low soil moisture content
in the 700+ N + treatment (less than 20% of the soil
water holding capacity from June to October, Casella
et al., 1996). This result underlines that the changes in
the carbon and water balances of a grass sward at an
increased air temperature are likely to be strongly
interactive.
Below-ground carbon balance
Due to the large amounts of organic carbon in grassland
soils, changes in carbon pool sizes under elevated CO2
are often hardly detectable through direct measurements
(Ross et al., 1995). By contrast, being based on seasonal
carbon fluxes, the gas-exchange experimental approach
reported herein is not affected by the variability of belowground carbon pool sizes and small differences in carbon
sequestration between treatments can be readily detected
over a growing season (Table 1).
With this technique, it was shown that both the carbon
turnover (below-ground respiration) and the carbon
sequestration in the soil and non-harvested plant parts
are stimulated when a ryegrass sward is exposed to a step
increase in the atmospheric CO2 concentration. In a
previous controlled environment study, the magnitude of
the supplemental carbon storage in elevated CO2 was
shown to be smaller at increasing N limitations (Lutze
and Gifford, 1995). Indeed, under these experimental
conditions, the below-ground C sequestration was less
stimulated in elevated CO2 at the low, compared to the
high N supply (Table 1). This result is partly accounted
for by the lower increase in root phytomass under elevated
CO2 at N - compared to N+ (+1.2 and +1.8 t C ha" 1 ,
respectively, according to Soussana et al., 1996).
However, roots accounted only for a fraction of the total
(2.8 and 5.4 t C ha" 1 during the two growing seasons at
N- and N +, Table 1) supplemental below-ground carbon
storage in elevated CO2.
The stubble mass was measured in May and in October
1993 and 1994 and was previously reported by Soussana
et al. (1996). From these data and from the carbon
concentration in the stubbles (approximately 40%), on
average of the two growing seasons, changes in stubble
mass amounted to a carbon flux comprised between —0.2
1320
Casella and Soussana
and +0.3 t C ha * in 5 months (May to October) (data
not shown). Hence, the contribution of the stubbles to
the sequestration of carbon by the plant and soil ecosystem was negligible, indicating that part of the supplemental below-ground C sequestration also occurred in
the soil organic matter. This conclusion is supported by
the increase in the size of the coarse soil organic matter
fractions for this experiment (Loiseau et cil., 1994; Loiseau
and Soussana, unpublished results) and by the increase
in the soil carbon content, which was previously reported
by Lutze and Gifford (1995) with pot-grown Danthonia
swards grown at low N supplies.
With clipped swards two major factors could contribute
to an increase in the soil carbon accumulation under
elevated CO2: (i) an increase in the production of litter,
especially below-ground through increases in root mass
and length (Rogers et al, 1994; Jongen et al., 1995); (ii)
a decrease in litter quality, due to changes in the C:N
ratio (Owensby, 1993; Gorissen et al., 1995), in the lignin
content and lignin:N ratio (Cotnifo and Ineson, 1996).
In the present experiment, the decline in the N concentration of the stubbles and roots and the large increase in
the root fraction of the grass sward in elevated CO2
(Soussana et al., 1996), respectively increased belowground litter production and reduced litter quality,
thereby favouring the accumulation of carbon in the soil
organic matter pools.
At this point, a link between the nitrogen and carbon
cycles is closed, since Soussana et al. (1996) reported that
a decline in the availability of inorganic N to the soilgrown grass sward contributed to the increased belowground allocation of it's growth. However, changes in
soil C content are not only the result of root litter
production and decomposition; they are also affected by
other factors such as root exsudation and the decomposition of native soil organic matter.
In conclusion, this study shows that, in an established
C3 perennial grass, a relatively large part of the additional
photosynthetic carbon is stored below-ground during the
two first growing seasons after exposure to elevated CO2,
thereby increasing significantly the below-ground carbon
pool. This stimulation of the below-ground carbon
sequestration in temperate grassland soils could exert a
negative feed-back on the current rise of the atmospheric
CO2 concentration, thereby contributing to the terrestrial
sink for carbon (Gifford, 1994; Keeling et al., 1995).
Changes in the nitrogen and water cycles appear to
contribute to this increased below-ground storage
through, respectively, an increased allocation of growth
to the below-ground compartment in elevated CO2
(Soussana et al., 1996) and a soil dessication at +3°C
(Casella et al., 1996), which restricts the below-ground
respiration in the high N treatment.
Part of the extra-carbon stored during the first growing
seasons might be lost during winter or in following years,
especially if carbon turnover in the soil fractions increases
(Ross et al., 1995; Navas et al., 1995). Further studies
on the turnover of carbon in the roots and in the soil
organic matter will therefore be required to gain understanding on the fate of the supplemental carbon that is
accumulated below-ground during the first growing seasons after the exposure of a C3 grass sward to elevated
CO2.
Acknowledgements
This research was funded by EU (contract EV5V CT
920169-CROPCHANGE), INRA and CNRS Programme
Environnement (Comite Ecosystemes). We thank I Chabaux
and E Villeneuve for their technical assistance, R Falcimagne
and M Martignac for the set-up and the maintenance of the
climate change facility and P Loiseau for helpful discussions.
References
Acock B, Charles-Edwards DA, Fitter DJ, Hand DW, Ludwig
LJ, Warren-Wilson J, Withers AC. 1978. The contribution of
leaves from different levels within a tomato crop to canopy
photosynthesis: an experimental examination of two canopy
models. Journal of Experimental Botany 29, 815-27.
Andre M, Du Cloux H. 1993. Interaction of CO 2 enrichment
and water limitations on photosynthesis and water efficiency
in wheat. Plant Physiology and Biochemistry 31, 103-12.
Arp WJ. 1991. Effects of source-sink relations on photosynthetic
acclimation to elevated CO 2 . Plant, Cell and Environment
14, 869-75.
Bancal P, Gaudillere JP. 1989. Oligofructan separation and
quantification by high performance liquid chromatography.
Application to Asparagus officinalis and Triticum aestivum.
Plant Physiology and Biochemistry 27, 745-50.
Bowes G. 1993. Facing the inevitable: plants and increasing
atmospheric CO 2 . Annual Review Plant Physiology and
Molecular Biology 44, 309-32.
Bonce JA, Caulfleld F. 1991. Reduced respiratory carbon
dioxide efflux during growth at elevated carbon dioxide in
three herbaceous perennial species. Annals of Botany 67,
325-30.
Casella E, Soussana JF, Loiseau P. 1996. Long-term effects of
CO 2 enrichment and temperature increase on a temperate
grass sward. I. Productivity and water use. Plant and Soil
182, 83-99.
Cotnifo MF, Ineson P. 1996. Rising CO 2 , decomposition
processes and soil C stores. Biogeochemistry (in press).
Daodet FA. 1987. Un systeme simple pour la mesure in situ des
echanges gazeux de couverts vegetaux de quelques metres
carres de surface foliaire. Agronomie 108, 39-78.
Dommergues Y, Mangenot F. 1970. Ecologie microbienne du sol.
Paris, France: Masson et Cie.
Eagles CF. 1967. Variation in the soluble carbohydrate content
of climatic races of Dactylis glomerata (cocksfoot) at different
temperatures. Annals of Botany 31, 645-51.
Faurie O. 1995. Interactions carbone-azote dans des associations
prairiales graminee (Lolium perenne L.) legumineuse (Trifolium
repens L.). Etude d'associations simulees en conditions
controlees. PhD Thesis, Universite Blaise Pascal, ClermontFerrand II, 203pp.
Carbon balance under climate change
Field CB, Chapin m FS, Matson PA, Mooney H. 1992.
Responses of terrestrial ecosystems to the changing atmosphere. Annual Review of Ecology and Systematic^ 23, 201-35.
Fonseca FG. 1996. The response of herbaceous plants to
elevated CO 2 . Short-term modification of C and N partitioning and growth of Plantago major ssp. pleiosperma. PhD
Thesis, University of Groningen, 99pp.
Gay AP, Hauck B. 1994. Acclimation of Lolium temulentum to
enhanced carbon dioxide concentration. Journal of
Experimental Botany 45, 1133—41.
Gilford RM. 1994. The global carbon cycle: a viewpoint on
the missing sink. Australian Journal of Plant Physiology 21,
1-15.
Gilford RM. 1995. Whole plant respiration and photosynthesis
of wheat under increased CO 2 concentration and temperature.
Long-term versus short-term distinctions for modelling.
Global Change Biology 1, 385-96.
Gloser J, Bartak M. 1994. Net photosynthesis growth rate and
biomass allocation in a rhizomatous grass Calamagrostis
epigejos grown at elevated CO 2 concentration. Photosynthetica
30, 143-50.
Gorissen A, van Ginkel JH, Keurenrjes JJB, van Veen JA. 1995.
Grass root decomposition is retarded when grass has been
grown under elevated CO 2 . Soil Biology Biochemistry 27,
117-20.
Jacob J, Greitner C, Drake BG. 1995. Acclimation of photosynthesis in relation to Rubisco, non-structural carbohydrates
contents and in situ carboxylase activity in Scirpus olneyi
grown at elevated CO 2 in the field. Plant, Cell and Environment
18, 875-84.
Jenkinson DS, Adams DE, Wild A. 1991. Model estimates of
CO 2 emissions from soil in response to global warming.
Nature 351, 304-6.
Jones MB, Leafe EL, Stiles W. 1980. Water stress in fieldgrown prennial ryegrass. n . Its effect on leaf water status,
stomatal resistance and leaf morphology. Annals of Applied
Biology 96, 103-10.
Jongen M, Jones MB, Hebeisen T, Blum H, Hendrey G. 1995.
The effects of elevated CO 2 concentrations on the root growth
of Lolium perenne and Trifolium repens grown in a FACE
system. Global Change Biology 1, 361-71.
Keeling CD, Wborf TP, Wahlen M, van der Plicht J. 1995.
Interannual extremes in the rate of rise of atmospheric carbon
dioxide since 1980. Nature 375, 666-70.
KOmer Ch. 1995. Towards a better experimental basis for
upscaling plant responses to elevated CO 2 and climate
warming. Plant, Cell and Environment 18, 1101-10.
Loiseau P, Soussana JF, Casella E. 1994. Effects of climatic
changes (CO 2 , temperature) on grassland ecosystems. First 5
months experimental results. In: Rousevell MDA, Loveland
PJ, eds, Soil responses to climate change, NATO ASI Series
Vol. 123. Berlin: Springer, 223-8.
Long SP. 1991. Modification of the response of photosynthetic
productivity to rising temperature by atmospheric CO 2
concentrations: has its importance been underestimated?
Plant, Cell and Environment 14, 729-39.
Luo Y, Field CB, Mooney HA. 1994. Predicting responses of
photosynthesis and root fraction to elevated [CO 2 ] a : interactions among carbon, nitrogen, and growth. Plant, Cell and
Environment 17, 1195-204.
Lutze JL, Gilford RM. 1995. Carbon storage and productivity
of a carbon dioxide enriched nitrogen limited grass sward
after one year's growth. Journal of Biogeography 22, 227-33.
1321
Monsi M, Saeki T. 1953. Ober den Lichtfactor in den
Pflanzengesellschaften
und seine Bedeutung fur die
Stoffproduktion. Japanese Journal of Botany 14, 22-52.
Nakayama FS, Kimball BA. 1988. Soil carbon dioxide distribution and flux within the open-top chamber. Agronomy Journal
80, 394-8.
Navas ML, Guillerm JL, Fabreguettes J, Roy J. 1995. The
influence of elevated CO 2 on community structure, biomass
and carbon balance of mediterranean old field microcosms.
Global Change Biology 1, 325-35.
Nijs I, Behaeghe T, Impens I. 1995. Leaf nitrogen content as a
predictor of photosynthetic capacity in ambient and global
change conditions. Journal of Biogeography 12, 177-83.
Nijs I, Impens I, Behaeghe T. 1989. Leaf and canopy responses
of Lolium perenne to long-term elevated atmospheric carbondioxide concentration. Planta 177, 312-20.
Nijs I, Impens I. 1989. Effects of long-term elevated atmospheric
carbon dioxide on Lolium perenne and Trifolium repens using
a simple photosynthesis model. Vegetatio 104, 421-31.
Owensby C. 1993. Potential impact of elevated CO 2 and above
and below-ground litter quality of a tall-grass prairie. Water
Air Soil Pollution 70, 413-24.
Pollock CJ, Cairns AJ. 1991. Fructan metabolism in grasses
and cereals. Annual Review of Plant Physiology and Plant
Molecular Biology 42, 77-101.
Rogers HH, Runion GB, Krupa SV. 1994. Plant responses to
atmospheric CO 2 enrichment with emphasis on roots and the
rhizosphere. Environmental Pollution 83, 155-89.
Ross DJ, Tate KR, Newton PCD. 1995. Elevated CO 2 and
temperature effects on soil carbon and nitrogen cycling in
ryegrass/white clover turves of an Endoquaept soil. Plant and
Soil 176, 37^+9.
Ryle GJA, Powell CE, Tewson V. 1992. Effect of elevated CO 2
on the photosynthesis, respiration and growth of perennial
ryegrass. Journal of Experimental Botany 43, 811-18.
Ryle GJA, Stanley J. 1992. Effect of elevated CO 2 on stomatal
size and distribution in perennial ryegrass. Annals of Botany
6, 563-5.
Schapendonk AHCM, Goudriaan J. 1995. The effects of
increased CO 2 on the annual carbon balance of a Lolium
perenne sward. Proceedings of the Tenth International
Photosynthesis Congress, Montpellier.
Schlesinger WH. 1991. Biogeochemistry: an analysis of global
change. New York: Academic Press.
Smart DR, Chatterton NJ, Bugbee B. 1994. The influence of
elevated CO 2 on non-structural carbohydrate distribution
and fructan accumulation in wheat canopies. Plant, Cell and
Environment Yl, 435—42.
Soussana JF, Casella E, Loiseau P. 1996. Long-term effects of
CO 2 enrichment and temperature increase on a temperate
grass sward. II. Nitrogen budgets and root fraction. Plant
and Soil Ml, 101-14.
Stitt M. 1991. Rising CO 2 levels and their potential significance
for carbon flow in photosynthetic cells. Plant, Cell and
Environment 14, 741-62.
Van de Geijn SC, Groenwold JV, Goudriaan J, Leffelaar PA.
1994. The Wageningen rhizolab—a facility to study soil-rootshoot-atmosphere interactions in crops. Plant and Soil
161, 275-87.
Wong SC. 1990. Elevated atmospheric partial pressure of CO 2
and plant growth. II. Non-structural carbohydrate content
in cotton plants and its effect on growth parameters.
Photosynthesis Research 23, 171-80.