Planta
Planta (1989) 177:312-320
9 Springer-Verlag 1989
Leaf and canopy responses of Lolium perenne to long-term elevated atmospheric carbon-dioxide concentration I. Nijs 1, I. Impens 1, and T. Behaeghe 2 1 Department of Biology, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, and 2 Faculty of Agricultural Sciences, University of Ghent, Coupure Links 653, B-9000 Gent, Belgium
Abstract. The relationship between leaf photosynthetic capacity (P,,max), net canopy COz- and H20exchange rate (NCER and Et, respectively) and canopy dry-matter production was examined in Lolium perenne L. cv. Vigor in ambient (363 -t- 30 gl. 1- a) and elevated (631 + 43 gl. 1-1) CO2 concentrations. An open system for continuous and simultaneous regulation of atmospheric CO2 concentration and N C E R and E~ measurement was designed and used over an entire growth cycle to calculate a carbon and a water balance. While NCER~,ax of full-grown canopies was 49% higher at elevated CO2 level, stimulation of p . . . . . was only 46% (in spite of a 50% rise in one-sided stomatal resistance for water-vapour diffusion), clearly indicating the effect of a higher leaf-area index under high CO2 (approx. 10% in one growing period examined). A larger amount of CO2-deficient leaves resulted in higher canopy dark-respiration rates and higher canopy light compensation points. The structural component of the high-CO2 effect was therefore a disadvantage at low irradiance, but a far greater benefit at high irradiance. Higher canopy darkrespiration rates under elevated CO2 level and low irradiance during the growing period are the primary causes for the increase in dry-matter production (19%) being much lower than expected merely based on the NCERm,x difference. While total water use was the same under high and low CO2 levels, water-use efficiency increased 25% on the canopy level and 87% on a leaf basis. In the course Abbreviations and symbols: C350 = ambient CO2, 363 +_30 Ixl. 1-1 ; C600 = high CO2, 631 + 43 Ixl. 1-1 ; co = atmospheric COz level; ci = CO2 concentration in the intracellular spaces of the l e a f ; E t = canopy evapotranspiration; Io = canopy light compensation point; N C E R = c a n o p y CO/-exchange rate; p, = l e a f photosynthetic rate; PPFD = photosynthetic photon flux density; ra = leaf boundary-layer resistance; R D = canopy dark-respiration rate; rs=stomatal resistance; W U E = w a t e r use efficiency
of canopy development, allocation towards the root system became greater, while stimulation of shoot dry-matter accumulation was inversely affected. Over an entire growing season the root/shoot production ratio was 22% higher under high CO2 concentration.
Key words: Carbon dioxide and water balance Carbon dioxide concentration (elevated) - Dry matter production - Evapotranspiration - Lolium (high-CO2 response) Photosynthesis (leaf, canopy) Water use efficiency
Introduction Plant photosynthetic responses to increased atmospheric CO2 can be examined in various experimental set-ups: many combinations of measurements on leaves, plants or canopies, over short and long periods of time and with different degrees of environmental control are possible. Not all of them, however, offer the possibility to define the relationship between photosynthetic rate (and its increase with elevated CO2 concentration) and changes in canopy dry-matter production. Even canopy photosynthetic measurements under well-controlled and constant laboratory conditions cannot always provide the basis for a reliable estimate, since it was shown in an earlier paper (Nijs et al. 1988) that canopy photosynthetic enhancement under optimal conditions (77% over several growth stages) was not a good indicator for dry-matter increase (43% on average over the growing season). The mechanism by which canopy photosynthesis affects final yield can be examined best when the former is continuously monitored as it fluctuates in response to environmental changes during a growing period. Therefore, we have built an open sy-
I. Nijs et al. : Leaf and canopy responses of Lolium to CO2
stem in which the stimulation of a high-CO2 atmosphere and the measurement of canopy gas exchange could take place simultaneously and continuously over a long period of time. Drake and Read (1981) succesfully used a similar system on salt-marsh plant communities. Our experimental design has several advantages: the vegetation is grown under near-natural variations of temperature, humidity and solar radiation (which is high compared with phytotrons), measurements can be highly automated and computerized, estimation of the carbon and water balance is possible (Drake et al. 1985). With a proper internal mixture, the absence of moisture and CO2 profiles is a further plus-point. A disadvantage is the absence of true replication (duplicates of the chambers), which can be compensated for by introducing internal replicates (within the chambers) and by scrutinous comparison of the chambers' microclimate. Part of the environmental control is lost compared with indoor controlled-environment chambers, but on the other hand it is difficult to extrapolate results from the latter to the natural environment. A second objective of this study was the investigation of leaf responses (gas exchange and stomatal behaviour) to elevated atmospheric COs concentration, to elaborate how this "primary effect" of CO2 enirchment affects canopy photosynthetic rate and productivity. Although it is possible to calculate the photosynthetic rate of the leaf from that of canopy and vice versa (Boote and Jones 1986), direct measurement is less complicated since a great deal of peripheral information is requested for this procedure. Stomatal behaviour, in particular, is examined here because it is the major barrier for both CO2 and water-vapour transport, as well as an important determinant of leaf water-use efficiency (WUE) (Louwerse 1980). Material and methods Lotium perenne L. cv. Vigor seeds (caryopses) were planted in steamed loamy soil in plastic containers in September 1986 at a density of 60 kg.ha 1. The containers were 13 cm deep and had a 121-cm z surface area. They remained in the open field during the following winter (vernalization). On March 3, 1987, some of the containers were moved to transparent acrylic growth-chambers (ground area=0.36 m 2, height=65 cm) that were part of a system for continuous and simuItaneous simulation of a high or ambient CO2 atmosphere, and canopy gasexchange measurement. The remaining containers were moved to similar growth-chambers with only the CO2 regulation and without the extension for gas-exchange measurements. In the latter, the stands were used for measurement of leaf photosynthetic capacity (/7. . . . . ), stomatal resistance to water-vapour diffusion (r~) and for destructive root- and shoot-production as-
313 sessment. In the former, the stands were untouched in order not to disturb the air flow in the open system during measurement, the only routine operation here was clipping the swards every three weeks, which was a standard procedure for all plant material used. After clipping, the harvested material was removed and the remaining 3-cm-high stubble layer was refertilized with a standard NPK mixture. Inside the chambers the containers were arranged without interspace so that the small area of vegetation in each them was assembled to a closed canopy. Container border effects were minimized by hemming them with gauze. There was no additional illumination so the stands were cultivated under normal daily and seasonal light :fluctuations. Photosynthetic photon fluence rate (PPFD), air temperature and relative humidity in the chambers were measured with quantum sensors (LI-190S; Li-Cor, Lincoln, Neb., USA), copperconstantan thermocouples (type $1121; Honeywell, 'Washington, DC, USA) and capacitive moisture sensors (type 1514; Vaisala, Helsinki, Finland) respectively. Relative humidity and PPFD were comparable with outside values (PPFD was on average 25 35 ~ lower than in the field), while air temperature in the chambers was 5 10% higher than in the qpen field on warm bright days. The average CO2 concentration was 363__ 30 gl.1-1 in the C350 treatment and 631 _+43 gl.1-1 in the C600 treatment.
Integrated units for gas-exchange rneasurement and C02-concentration regulation (Fig. 1). The system consists of four growthchambers (two of which were used in our experiments) that serve as cuvettes in an open system with continuous flow of CO2-enriched or ambient air. Sampling the air for CO2 concentration and water vapour before and after it passes the chambers allows calculation of COz- and water-vapour-exchange rates. The flow rate, manometrically determined with venturitubes calibrated with a standard gasmeter (type G I 6 ; Contigea, Brussels, Belgium), was approx. 0.4m3.min -1. This rate of air flow through the enclosures kept CO2 differentials within reasonable limits (max. 40 ~tI-l-1), but appeared insufficient to prevent chamber air temperature rising above 30 ~ C in early summer, so the air was cooled with a standard air-conditioner before entering the chambers. To avoid contamination from exhaust fumes the ventilation system inlet was placed at 5 m altitude. Injection of COz was regulated with a standard flow meter (SHO-rate purgemeter, R2-15-AA; Brooks, Veenendaal, The Netherlands). Since only one air sample could be analysed at a time, using a time-controlled six-point switching unit with a 2-rain valve-scanning interval (WA-161-WK3; ADC, Hoddesdon, Herts., UK) and an absolute infra-red gas analyser (IRGA; type SB-322; ADC), and since CO2 concentration was usually rapidly fluctuating, true differential measurement Was impossible. Instead, correct measurements were obtained over 1-h periods by placing buffertanks in each polypropylene air-sampling duct, yielding averages per hour of photosynthetic and evapotranspiration rates. The time lag for detecting changes in photosynthetic rate was about 25 min for a 100% change. All signals (from the IRGA, moisture and quantum sensors and thermocouples) were stored in a multi-channel data logger (Delta Logger; Delta-T devices, Burwell, Camb., UK) with a direct line to a personal computer (Olivetti M24) for further data processing. To avoid disturbing the measurements by opening the chambers, containers were irrigated with a trickle system (one drip tube per container) every night at 24.00 h. This way measurement of canopy evapotranspiration (E~) during the day was unaffected.
314
I. Nijs et al. : Leaf and canopy responses of Lolium to COz
8
11
12
.5 13
20
I Canopy gas-exchange rates. The net CO2-exchange rate of the soil-vegetation complex was calculated as N C E R = ( C - J ) / A , where C = t h e CO2 concentration difference before and after the growth-chamber/cuvette, J = flow rate of air through the cuvette and A = ground area. Evapotranspiration rate E, was calculated analogously. The N C E R includes canopy photosynthesis, canopy respiration, root respiration and soil respiration ; Et includes canopy transpiration and soil evaporation. A rectangular hyperbola
Y=A+B/(X+C) (X is the independent variable, Y is the dependent variable) was used to fit observed values on N C E R - P P F D response diagrams. This way N C E R could be calculated at every moment of the day during the growing period. On days lacking data (due to system breakdown or when unloading the data logger), N C E R was estimated as the mean of the values produced by
I
Fig. 1. Integrated system for simultaneous and continuous simulation of a high and an ambient CO2 atmosphere and for measurement of canopy gas-exchange rate. Parts: 1 ventilation air inlet at 5 m altitude, 2 air-conditioner, 3 growth-chamber = cuvette, 4 carbon dioxide injection, 5 blower (turbine), 6 venturimeter, 7 water manometer, 8 outlet for chamber air, 9 copper-constantan thermocouple and quantum sensor, 10 axial fan, l l airsampling lines (polypropylene), 12 buffertank, 13 time-controlled six-point switching unit, 14 pump for continuous flushing of air-sampling lines, 15 capacitive moisture sensor, 16 drying agent (silica gel), 17 rotameter, 18 absolute infra-red gas analyser, 19 data logger, 20 personal computer
the curves of the surrounding days. Eventually the total amount of C02 absorbed in photosynthesis and released in respiration was calculated and compared with dry-matter production over the same time interval. From the Et values a water balance can be calculated. Dividing total production by the amount of water lost, yields W U E over the interval. Maximum leaf photosynthetic rate and stomatal resistance. Maximum leaf photosynthetic rate (p ..... ) was determined at three one-week intervals during the growing periods by placing the containers in a separate small chamber, composed of mylar plastic over a metal frame. By leading air from the larger growth chambers through this small measuring unit, a similar atmosphere was created here. A 400-W sodium lamp was mounted on top of it, resulting in a PPFD of 800 ktmol.m-Z.s-1 at the leaf surface, which was found sufficient to saturate pho-
I. Nijs et al. : Leaf and canopy responses of Lolium to CO2 tosynthesis. Inside, maximum leaf photosynthetic rate p ..... of a single leaf could be determined by placing it in a small acrylic leaf chamber that was part of a closed-loop COz-measuring device, as described by Ceulemans et al. (1986). Since the device was designed for quick, non-destructive and extensive measurements, a large number of repetitions was possible. In the same small measuring chamber, stomatal resistance (r~) was determined with a portable diffusion porometer (Mk-II ; Delta-T Devices). In this context, an interesting parameter is ce (COz concentration in the intercellular spaces of the leaf), since its change with varying climate conditions can indicate changes in W U E (Louwerse 1980). It is calculated as c i = c , - ( p , , r z . c o ) (Gates 1980), where rz,co~= leaf resistance for COz diffusion. The latter was obtained by combining the upper and the lower leaf side resistances in parallel: 1/rl,co~= (1/r~pp~,co) +(l/rt ...... co). On either side the resistance was composed of a stomatal fraction, r,, and a boundary layer fraction, r,, e.g. : rupper,CO a = rs,upper,CO 2 + ra,upper,CO2-
In the above formula, r c o = 1.56.rH~o (Noble 1983). The boundary-layer resistance r, was calculated from evaporation q and " l e a f " to air vapour pressure difference VPD of a damp filter paper in the leaf chamber, using dew point mirrors (type TS 3C, Walz, Effeltrich, FRG). For this simulation of a leaf without stomatal barrier, r~ equals VPD/q.
315 PPFD
C3':iO
CE,O0
PPFD tmmolm-2 s-I] 3
HL-'ER(rag m-2 s--D 11.2
,.,.I
("
z
3
\
.......
LB "3,
'LY,
,"--7%,'
10 13 1G time of day (Gffr+ll
1.4
19
Fig. 2. Time courses of N C E R and PPFD on 2 May 1987, of a Lolium perenne sward under ambient (C350; 363 • 30 gl-1-1) and high (C600; 631 ___43 gl.1-1) CO2 concentrations (canopy age = 11 d)
PPFD
C350
C&O0
PPFO Immolm-2 s-l) 3
Et (rag"m-2 s-.12 20Q
Production assessment. Dry-matter production of some of the
stands in the growth chambers without a CO2-measurement facility was measured one, two and three weeks after clipping. In one, three-week growth period root production was determined. Data from a previous growing season, from April 1986 to July 1986, were used to calculate changes in root/shoot production ratio induced by high COz levels.
/./'='~'~. %.
2
it
0
Results
Canopy measurements. Canopy net CO2-exchange rate, NCER, and evapotranspiration rate, E , were monitored during several growing periods. The period 22 Aprit i987 to I3 May 1987 was selected for further analysis. Typical daily courses for N C E R and E, under ambient (C350, 363 + 30 gl1-1) and high (C600, 631 +_43 gl.1-1) CO2 are represented in Figs. 2 and 3. Under C600, N C E R is higher than under C350 at high P P F D only, while at low irradiance N C E R becomes considerably less under high CO2. Values of Et, however, are highly similar, irrespective of light intensity. On all days examined, considerable hysteresis effects occurred, for N C E R as well as for Et, although in opposite directions when comparing A.M. with P.M. values (Figs. 4, 5). Linear regression in N C E R - P P F D diagrams yielded higher determination coefficients r 2 for the C600 treatment (0.919 versus 0.792), so the hysteresis effect on photosynthesis was more intense at C350. The characteristics from N C E R - P P F D curves derived from diagrams like those in Figs. 4 and 5 are represented in Table 1. Values of N C E R at saturating P P F D were in the same order of magni-
7
,,,,-~X
I , I i0 13 15 time of: da~ (fclr+11
!0 19
Fig. 3. Time courses of E~ and PPFD on 2 May 1987, of a Lolium perenne sward under ambient (C350) and high (C600) CO2 concentrations (canopy age = 11 d)
tude as those obtained by Nijs et al. (1988) under laboratory conditions with the same species. Fullgrown canopies showed a 49% higher NCERm,x under C600, compared with C350. In earlier developmental stages this increase was even higher. Quantum efficiency, canopy dark-respiration rate (RD) and canopy light compensation point (Io) were also much higher under the C600 treatment, the differences were in fact enormous. It is not clear what caused the extremely high RD under C600, although Drake and Read (1981) me.asured RD values of a similar order of magnitude in a mixed-species plant community, also in an open system (RD amounted to up to 30% of maximum NCER). Calculating the carbon balance using these data showed that a large fraction of the CO21 absorbed during the day is released again during ':the night, especially under the C600 treatment (Table 2). Because of this effect, the increase in net CO,, input
316
I. Nijs et al. : Leaf and canopy responses of Lolium to CO 2 1.4 1.31.2-
11
1.1-
lO
1-
~ ~ .'~/'~
-
0.9-
/
0.8-
Ttt
0.70.6-
o4 I
0.5-
E
0.40.3-
m
E
13
/
....-- ""
0.2.
J ,0
0.1 0
z
8
-0.1 -0.2
lg
-
/
--0.3 -0.4-0.5 -0.6
-
-0.7
.
/
/
/
/
~/7
14
/
/ / ig !
0
I
I
I
0.2
I
0.4
I
I
I
0.6
0.8
I
I
1
t
1.2
PPIrD Cmmo! m--Z s.1) rl
C350
A
C600
Fig. 4. Hysteresis effects on N C E R in a Lolium perenne sward under ambient (C350) and high (C600) CO2 concentrations on 2 May 1987 (canopy age= 11 d). Data label=hour of day
caused by higher C O 2 becomes much smaller than expected merely based on the NCERmax difference. Total production is only a fraction of net CO2 uptake, since COz is converted into dry matter with a certain efficiency, depending on chemical composition and energy content of the final product. According to Sestak et al. (197J), 1 g of CO2 is converted into 0.5 g of protein, 0.6 g of starch or 0.4 g of lipids. De Wit (1978) gives similar values. The transformation efficiency in our experiments was about 0.4. The C350/C600 ratio for net CO2 input (0.73=566/711) was in the same order of magnitude as the ratio of total dry-matter production (0.79) between these treatments. Total water use was virtually the same for both treatments. The increase in W U E was 25%, and was solely the consequence of an increase in N C E R (Table 3).
Leaf measurements. At every developmental stage (1, 2 and 3 weeks after clipping), p . . . . . was higher under the C600 treatment. Since there was hardly any change with leaf age, mean values were calculated for the entire growth cycle: p . . . . . at C600
(1.067_+0.121 m g . m - 2 . s -1) was 46% higher than at C350 (0.732_+0.109 m g . m -2 .s-a). On average, rs was 50% higher at C600 than at C350 one-sidedly (1.599_+0.387 s-cm -1 versus 1.067_+0.180s-cm -1) and 46% higher for the whole leaf. There was a slight decrease with leaf age. At only one of three sampling dates was the difference statistically significant, almost certainly as a result of the instability of the diffusion porometer with changing temperature (there was no rigorous temperature conditioning in the measuring unit). The boundary-layer resistance for water-vapour diffusion was 0.79 s. c m - 1 one-sidedly. A summary of most of the important parameters describing stomatal behaviour is presented in Table 4. Although c~ rises substantially with elevated CO2 level, c~/c, remains fairly constant. In spite of the considerable stomatal closure, the extra COz can still result in a 46% rise in photosynthetic capacity. Assuming leaf transpiration rate (q) to be approximately inversely proportional to (r~+r,), an estimate can be made for the change in W U E of the leaf caused by high CO2 in saturating light: since p . . . . . rises by a factor of 1.458 and (rs + r.)
I. Nijs et al. : Leaf and canopy responses of Lolium to
317
CO 2
200 190 180 170 160 150 140 -
12 1
130
3
12
~
"~'-"
,~11
120 110 1 O0 90 80 70 60 50 40 30 20
8
7
100
I
I
0
I
I
0.2
I
I
0.4. Q
I
0.6
I
O.B
I
I
Table 1. Canopy CO2-exchange rate (NCERmax; mg. m - 2. s - 1 ) , RD (mg.m-2.s 1), Io (10 .6 mol'm-2"s -1) and quantum efficiency (c~; g CO2-(mol photons) 1) of the NCER-PPFD response curves for Lolium perenne grown at ambient (C350) and high (C600) CO2 concentrations during the growth cycle 22 April 1987 to 13 May 1987. Values are given for 4-, 13and 20-d-old swards. Canopy dark-respiration rate (RD) was directly calculated from RD values during the night (without fitting). Since asymptotic values were not always approached in full sunlight, NCER at 2 mmoi. m - 2. s- 1 is given in parentheses Development
Treatment
stage (d)
C350
C600
NCERm.x
4 13 20
1.03 (0.64) 1.91 (1.24) 2.93 (1.83)
2.39 (1.51) 2.99 (1.93) 4.37 (2.53)
c~
4 13 20
0.84 1.87 2.93
2.58 3.44 4.37
RD
4 13 20
--0.018 --0.056 --0.068
-0.386 --0.505 --0.716
Io
4 13 20
21 30 27
1.2
PPFD ~mmol m - 2 s - l ) r A C600
Fig. 5. Hysteresis effects o n E t in a Lolium perenne sward under ambient (C350) and high (C600) 1987 (canopy age = 11 d). Data label = hour of day
Characteristic
I
1
150 147 185
CO 2
concentrations on 2 May
Table 2. Carbon balance of Lolium perenne at ambient (C350) and high (C600) CO2 concentrations over the growing period from 22 April 1987 to 13 May 1987. Root production was estimated as a fraction of shoot production. All units are in g- m - 2 Treatment
c o z uptake during the day CO2 loss at night Net CO2 input Shoot production (measured) Root production (estimated) Total production
C350
C600
603.3 37.3 566.0 191.7 43.2 234.9
1186.2 415.2 771.0 228.9 66.1 295.0
i n c r e a s e s b y a f a c t o r o f 1.280 ( c a u s i n g q to d e c r e a s e b y a f a c t o r o f 0.781), W U E will i n c r e a s e b y a f a c t o r o f 1.458/0.781, w h i c h is a 8 7 % i m p r o v e m e n t .
D r y - m a t t e r production. The increase in shoot drym a t t e r p r o d u c t i o n o v e r the g r o w i n g p e r i o d 22 M a r c h 1987 to 13 M a y 1987 w a s 1 9 % . The investigation of the e v o l u t i o n of shoot drym a t t e r p r o d u c t i o n w i t h age s h o w e d t h a t the inc-
I. Nijs et al. : Leaf and canopy responses of Lolium to
318
Table3. Water use (kg.m -2) and W U E ( . 1 0 - S g CO2.(g H20) a) of Lolium perenne vegetations grown at ambient (C350) and high (C600) COz concentrations over the period from 22 April 1987 to 13 May 1987 Treatment
Water use WUE
C350
C600
98.2 239
98.4 300
Table 4. p . . . . . ( g g - m - 2. s- 1), r, (s. c m - 1), cl (gl. 1-1), c, (gl. 1-1) and cf/co of Lolium perenne leaves. Data are mean values over all growth stages (one, two and three weeks after clipping). ro for H 2 0 diffusion was 1.870 s-cm-~ one-sidedly Parameter
Treatment C350
r~,n2o,uppc~ l~af ~iae rs,tf2o,l. . . . l~f ~id, p,, .... ca ci cJc~
1.067 4.437 732 367 282 0.768
C600 1.599 5.800 1067 620 462 0.745
Table 5. Percentage increase of shoot and root dry-matter (DW) production caused by high (C600) CO2 concentration at three stages in the development of Lolium perenne. Growing period = 4 May 1987 to 25 May 1987 Development stage
Shoot D W
Root D W
Day 7 Day 14 Day 21
+34.2 +23.0 + 12.0
--1.9 + 16.0 + 28.0
rease caused by high CO2 became smaller during the course of canopy development towards full cover, while the stimulation of root dry-matter accumulation became stronger (Table 5). Over an entire growing season (April 1986 to July 1986), the root/shoot production ratio was 22% higher under C600 compared with C350. Discussion
Leaf responses to elevated COz level can be explained best with the classical equation F - - C / R ( F = CO2 flux, C = C O 2 concentration difference between the atmosphere and the site of carboxylation and R = resistance to CO2 diffusion). In high CO2, R is increased through stomatal closure, but this
CO 2
is overcompensated for by the increase of C. As a consequence, p, will rise and q will decrease, causing W U E to go up drastically. It appears however, that the leaf response is not the most economical one possible, because p, would rise even more if the stomata were unaffected, and leaf transpiration rate (q) would be similar to its value under ambient CO2 concentration. Obviously the CO2 benefit is only partly used for production purposes and partly for maintainance of survival purposes (through reduction of water use). It is clear that the individual leaf responds in relation to the plant or canopy, instead of as an autonomic entity. Possibly, the leaf anticipates the higher risk for water stress because of the larger transpiring surface under C600, though on the other hand only an equal amount of water is consumed compared with the C350 treatment. Another explanation is an imminent saturation of CO2 processing in the chloroplast or a saturation of carbohydrate export for allocation to other sinks in the plant, as a consequence of the higher COz supply if stomatal closure fails to occur. Even if photosynthesis is not saturated, keeping the stomata open will probably mean departure from optimal stomatal behaviour (dq/dp,=constant, Farquhar and Sharkey 1982), no longer minimizing water loss for extra carbon gain. The stomatal behaviour of Lolium perenne was similar to that of sunflower and barley (Louwerse 1980): ci was proportional to ca with a ratio of _+0.75. In these experiments it was pointed out that another type of response, with stabilization of ci with increasing % is more advantageous when water is limiting. Keeping a constant Ci/Ca ratio is intermediate between this type of behaviour and the situation where stomata do not react at all to rising ca. An increase in Pn,ma• is of course translated in a rise in NCERm,x. The latter is somewhat larger (49%) compared with the former (46%), which demonstrates the effect of CO2 on total leaf area. Calculating backwards from dry matter to leaf area, using allometric relations derived by Nijs et al. (1988) yielded a 9.9% shoot length increase in high CO2 for full-grown stands. The final increase in dry-matter production is only 19%. This is partly a consequence of the relatively low P P F D values during the growing period selected for gasexchange analysis, because the effect of higher CO2 is strongest at high PPFD. At low irradiance the high leaf-area index of the canopy grown at C600 is even a disadvantage, because light extinction is so fast that the vegetation already becomes CO2 deficient at 100 to 200 izmol.m - 2 . s - 1 . A large
I. Nijs et al. : Leaf and canopy responses of Lolium to COz
number of respiring leaves and a high RD and Io are the result. The structural effect of C600 is therefore a disadvantage at low, but a much stronger benefit at high PPFD. Since light can become the limiting factor, the reasoning is confirmed that high CO2 is only a serious benefit if it is the limiting factor. Brown and Higginbotham (1986) found analogous results with Picea glauca: elevated CO2 level increased leaf and total biomass only at high N-availability. In other species, however, production was stimulated on unfertilized soil and in spite of severe N-deficiency (Norby et al. 1986). The influence of limited factors may be speciesspecific. Gas-exchange measurements on the canopy level demonstrated considerable hysteresis effects on CO2 and H 2 0 : N C E R showed a significant drop in the afternoon, while Et rose drastically. The first phenomenon was probably caused by source-sink effects (sink saturation), while the second was almost certainly driven by environmental factors like a higher temperature and a lower relative humidity. The effect of elevated COz is rather positive than negative, because hysteresis is less intense at C600, indicating there was no more sink-saturation compared with C350. An interesting strategy for the vegetation - especially with limited water availability - would be to open the stomata in the first half of the day only, when carbohydrate sinks are not yet saturated, and Et is low because of the relatively low air temperature and vapour-pressure deficit between leaf and atmosphere. A variation of this behaviour is observed in sugarcane: high yields in this species have been empirically correlated with maximal stomatal opening occurring in the morning and evening, when vapour-pressure deficit is low, and reduced conductance during midday when it was high, rather than with continuous opening during the daylight hours (Grantz et al. 1987). It is worth mentioning that the data presented on canopy responses are representative for late spring growth only. An earlier study (Nijs et al. 1988) has shown that gas exchange in summer can deviate somewhat from this general pattern, while some other differences between the results from this earlier study and the present one are also striking, e.g. the dry-matter increase here was only 19%, while it was 46% over the 1986 growing season. Radiation-input differences are obviously responsible for this. The W U E figures are fairly out of line as well, but these are not comparable because the integrated W U E value calculated here over a three-week growing period is a totally different parameter from the W U E figures determined
319
in a fixed laboratory environment. On the other hand the previously measured 77% enhancement of maximum canopy photosynthetic rate was confirmed in this study, which is not remarkable for it is an intrinsic plant characteristic rather than one depending on environmental change. Production assessment showed that at the beginning of the growing period, immediately after clipping, the extra carbon from C600 is invested relatively more into shoots and less into root material. In the course of canopy development this allocation pattern is reversed, which is not illogical since a stand recently mown has a relative surplus of roots. This is in agreement with the earlier observation from Nijs et al. (1988) that an important fraction of the extra dry matter produced in high CO2 originates from the first few days of regrowth after the vegetation is mown. Data from an entire growing season showed that the higher CO2 uptake under the C600 regime was more intensely allocated to the root system than to the shoots. Concluding that Lolium perenne is "wasting" part of the extra COz by converting it into no~L-harvestable biomass, is however not correct; since enhanced root growth or stimulation of storage into roots is a benefit after clipping: the allocated material is directly or indirectly used for establishing a new canopy. Thus it still ends up in the harvestable material, increasing the harvest index. This study was supported by the "Instituut tot Aanmoediging van bet Wetenschappetijk Onderzoek in Nijverheid en Landbouw" (Brussels, Belgium).
References Boote, K.J., Jones, J.W. (1986) Equations to define canopy photosynthesis from quantum efficiency, maximum leaf rate, light extinction, leaf area index and photon flux density. Proc. 7th Congr. on Photosynthesis, Providence, Rhode Island), Biggins, J., ed. Martinus Nijhoff, D'ordrecht, The Netherlands Brown, K., Higginbotham, K.O. (1986) Effect of carbon dioxide enrichment and nitrogen supply on growth of boreal tree seedlings. Tree Physiol. 2, 223-232 Ceulemans, R., Kockelbergh, F., Impens, I. (1986) A fast, low cost and low power requiring device for improving closed loop CO2 measuring systems. J. Exp. Bot. 37, ~234~1244 De Wit, C.T. (1978) Stimulation of assimilation, respiration and transpiration of crops. Centre for Agricultural Publishing and Documentation, Wageningen Drake, B.G., Read, M. (1981) Carbon dioxide assimilation, photosynthetic efficiency and respiration of Chesapeake Bay salt march. J. Ecol. 69, 405-423 Drake, B.G., Rogers, H.H., Allen, L.H. (1985)Methods of exposing plants to elevated carbon dioxide. In: Direct effects of increasing carbon dioxide on vegetation (Rep. DOE/ ER-239, Dept. of Energy, Washington D.C.), pp. 11-35, Strain, B.R., Cure, J.D., eds. NTIS, Springfield, Va., USA
320 Farquhar, D.G., Sharkey, T.D. (1982) Stomatal conductance and photosynthesis. Annu. Rev. Plant Physiol. 33, 317-345 Gates, D.M. (1980) Biophysical ecology. Springer, New York Heidelberg Berlin Grantz, D.A., Moore, P.H., Zeiger, E. (1987) Stomatal responses to light and humidity in sugarcane: prediction of daily time courses and identification of potential selection criteria. Plant Cell Environ. 10, 197504 Louwerse, W. (1980) Effects of CO2 concentration and irradiance on the stomatal behaviour of maize, barley and sunflower plants in the field. Plant Cell Environ. 3, 391 398 Nijs, I., Impens, I., Behaeghe, T.J. (1988) Effects of rising at-
I. Nijs et al. : Leaf and canopy responses of Lolium to
CO 2
mospheric carbon dioxide on gas exchange and growth of perennial ryegrass. Photosynthetica 22, 44-50 Nobel, P.S. (1983) Biophysical plant physiology and ecology. Freeman, San Francisco Norby, R.J., Pastor, K., Melillo, J.M. (1986) Carbon-nitrogen interactions in COz-enriched white oak: physiological and long-term perspectives. Tree Physiol. 2, 243 259 Sestak, Z., Catsky, J., Jarvis, P.G. (1971) Plant photosynthetic production: manual of methods. Junk, The Hague Received 29 December 1987; accepted 21 October 1988
Planta (1989) 177:312-320
9 Springer-Verlag 1989
Leaf and canopy responses of Lolium perenne to long-term elevated atmospheric carbon-dioxide concentration I. Nijs 1, I. Impens 1, and T. Behaeghe 2 1 Department of Biology, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, and 2 Faculty of Agricultural Sciences, University of Ghent, Coupure Links 653, B-9000 Gent, Belgium
Abstract. The relationship between leaf photosynthetic capacity (P,,max), net canopy COz- and H20exchange rate (NCER and Et, respectively) and canopy dry-matter production was examined in Lolium perenne L. cv. Vigor in ambient (363 -t- 30 gl. 1- a) and elevated (631 + 43 gl. 1-1) CO2 concentrations. An open system for continuous and simultaneous regulation of atmospheric CO2 concentration and N C E R and E~ measurement was designed and used over an entire growth cycle to calculate a carbon and a water balance. While NCER~,ax of full-grown canopies was 49% higher at elevated CO2 level, stimulation of p . . . . . was only 46% (in spite of a 50% rise in one-sided stomatal resistance for water-vapour diffusion), clearly indicating the effect of a higher leaf-area index under high CO2 (approx. 10% in one growing period examined). A larger amount of CO2-deficient leaves resulted in higher canopy dark-respiration rates and higher canopy light compensation points. The structural component of the high-CO2 effect was therefore a disadvantage at low irradiance, but a far greater benefit at high irradiance. Higher canopy darkrespiration rates under elevated CO2 level and low irradiance during the growing period are the primary causes for the increase in dry-matter production (19%) being much lower than expected merely based on the NCERm,x difference. While total water use was the same under high and low CO2 levels, water-use efficiency increased 25% on the canopy level and 87% on a leaf basis. In the course Abbreviations and symbols: C350 = ambient CO2, 363 +_30 Ixl. 1-1 ; C600 = high CO2, 631 + 43 Ixl. 1-1 ; co = atmospheric COz level; ci = CO2 concentration in the intracellular spaces of the l e a f ; E t = canopy evapotranspiration; Io = canopy light compensation point; N C E R = c a n o p y CO/-exchange rate; p, = l e a f photosynthetic rate; PPFD = photosynthetic photon flux density; ra = leaf boundary-layer resistance; R D = canopy dark-respiration rate; rs=stomatal resistance; W U E = w a t e r use efficiency
of canopy development, allocation towards the root system became greater, while stimulation of shoot dry-matter accumulation was inversely affected. Over an entire growing season the root/shoot production ratio was 22% higher under high CO2 concentration.
Key words: Carbon dioxide and water balance Carbon dioxide concentration (elevated) - Dry matter production - Evapotranspiration - Lolium (high-CO2 response) Photosynthesis (leaf, canopy) Water use efficiency
Introduction Plant photosynthetic responses to increased atmospheric CO2 can be examined in various experimental set-ups: many combinations of measurements on leaves, plants or canopies, over short and long periods of time and with different degrees of environmental control are possible. Not all of them, however, offer the possibility to define the relationship between photosynthetic rate (and its increase with elevated CO2 concentration) and changes in canopy dry-matter production. Even canopy photosynthetic measurements under well-controlled and constant laboratory conditions cannot always provide the basis for a reliable estimate, since it was shown in an earlier paper (Nijs et al. 1988) that canopy photosynthetic enhancement under optimal conditions (77% over several growth stages) was not a good indicator for dry-matter increase (43% on average over the growing season). The mechanism by which canopy photosynthesis affects final yield can be examined best when the former is continuously monitored as it fluctuates in response to environmental changes during a growing period. Therefore, we have built an open sy-
I. Nijs et al. : Leaf and canopy responses of Lolium to CO2
stem in which the stimulation of a high-CO2 atmosphere and the measurement of canopy gas exchange could take place simultaneously and continuously over a long period of time. Drake and Read (1981) succesfully used a similar system on salt-marsh plant communities. Our experimental design has several advantages: the vegetation is grown under near-natural variations of temperature, humidity and solar radiation (which is high compared with phytotrons), measurements can be highly automated and computerized, estimation of the carbon and water balance is possible (Drake et al. 1985). With a proper internal mixture, the absence of moisture and CO2 profiles is a further plus-point. A disadvantage is the absence of true replication (duplicates of the chambers), which can be compensated for by introducing internal replicates (within the chambers) and by scrutinous comparison of the chambers' microclimate. Part of the environmental control is lost compared with indoor controlled-environment chambers, but on the other hand it is difficult to extrapolate results from the latter to the natural environment. A second objective of this study was the investigation of leaf responses (gas exchange and stomatal behaviour) to elevated atmospheric COs concentration, to elaborate how this "primary effect" of CO2 enirchment affects canopy photosynthetic rate and productivity. Although it is possible to calculate the photosynthetic rate of the leaf from that of canopy and vice versa (Boote and Jones 1986), direct measurement is less complicated since a great deal of peripheral information is requested for this procedure. Stomatal behaviour, in particular, is examined here because it is the major barrier for both CO2 and water-vapour transport, as well as an important determinant of leaf water-use efficiency (WUE) (Louwerse 1980). Material and methods Lotium perenne L. cv. Vigor seeds (caryopses) were planted in steamed loamy soil in plastic containers in September 1986 at a density of 60 kg.ha 1. The containers were 13 cm deep and had a 121-cm z surface area. They remained in the open field during the following winter (vernalization). On March 3, 1987, some of the containers were moved to transparent acrylic growth-chambers (ground area=0.36 m 2, height=65 cm) that were part of a system for continuous and simuItaneous simulation of a high or ambient CO2 atmosphere, and canopy gasexchange measurement. The remaining containers were moved to similar growth-chambers with only the CO2 regulation and without the extension for gas-exchange measurements. In the latter, the stands were used for measurement of leaf photosynthetic capacity (/7. . . . . ), stomatal resistance to water-vapour diffusion (r~) and for destructive root- and shoot-production as-
313 sessment. In the former, the stands were untouched in order not to disturb the air flow in the open system during measurement, the only routine operation here was clipping the swards every three weeks, which was a standard procedure for all plant material used. After clipping, the harvested material was removed and the remaining 3-cm-high stubble layer was refertilized with a standard NPK mixture. Inside the chambers the containers were arranged without interspace so that the small area of vegetation in each them was assembled to a closed canopy. Container border effects were minimized by hemming them with gauze. There was no additional illumination so the stands were cultivated under normal daily and seasonal light :fluctuations. Photosynthetic photon fluence rate (PPFD), air temperature and relative humidity in the chambers were measured with quantum sensors (LI-190S; Li-Cor, Lincoln, Neb., USA), copperconstantan thermocouples (type $1121; Honeywell, 'Washington, DC, USA) and capacitive moisture sensors (type 1514; Vaisala, Helsinki, Finland) respectively. Relative humidity and PPFD were comparable with outside values (PPFD was on average 25 35 ~ lower than in the field), while air temperature in the chambers was 5 10% higher than in the qpen field on warm bright days. The average CO2 concentration was 363__ 30 gl.1-1 in the C350 treatment and 631 _+43 gl.1-1 in the C600 treatment.
Integrated units for gas-exchange rneasurement and C02-concentration regulation (Fig. 1). The system consists of four growthchambers (two of which were used in our experiments) that serve as cuvettes in an open system with continuous flow of CO2-enriched or ambient air. Sampling the air for CO2 concentration and water vapour before and after it passes the chambers allows calculation of COz- and water-vapour-exchange rates. The flow rate, manometrically determined with venturitubes calibrated with a standard gasmeter (type G I 6 ; Contigea, Brussels, Belgium), was approx. 0.4m3.min -1. This rate of air flow through the enclosures kept CO2 differentials within reasonable limits (max. 40 ~tI-l-1), but appeared insufficient to prevent chamber air temperature rising above 30 ~ C in early summer, so the air was cooled with a standard air-conditioner before entering the chambers. To avoid contamination from exhaust fumes the ventilation system inlet was placed at 5 m altitude. Injection of COz was regulated with a standard flow meter (SHO-rate purgemeter, R2-15-AA; Brooks, Veenendaal, The Netherlands). Since only one air sample could be analysed at a time, using a time-controlled six-point switching unit with a 2-rain valve-scanning interval (WA-161-WK3; ADC, Hoddesdon, Herts., UK) and an absolute infra-red gas analyser (IRGA; type SB-322; ADC), and since CO2 concentration was usually rapidly fluctuating, true differential measurement Was impossible. Instead, correct measurements were obtained over 1-h periods by placing buffertanks in each polypropylene air-sampling duct, yielding averages per hour of photosynthetic and evapotranspiration rates. The time lag for detecting changes in photosynthetic rate was about 25 min for a 100% change. All signals (from the IRGA, moisture and quantum sensors and thermocouples) were stored in a multi-channel data logger (Delta Logger; Delta-T devices, Burwell, Camb., UK) with a direct line to a personal computer (Olivetti M24) for further data processing. To avoid disturbing the measurements by opening the chambers, containers were irrigated with a trickle system (one drip tube per container) every night at 24.00 h. This way measurement of canopy evapotranspiration (E~) during the day was unaffected.
314
I. Nijs et al. : Leaf and canopy responses of Lolium to COz
8
11
12
.5 13
20
I Canopy gas-exchange rates. The net CO2-exchange rate of the soil-vegetation complex was calculated as N C E R = ( C - J ) / A , where C = t h e CO2 concentration difference before and after the growth-chamber/cuvette, J = flow rate of air through the cuvette and A = ground area. Evapotranspiration rate E, was calculated analogously. The N C E R includes canopy photosynthesis, canopy respiration, root respiration and soil respiration ; Et includes canopy transpiration and soil evaporation. A rectangular hyperbola
Y=A+B/(X+C) (X is the independent variable, Y is the dependent variable) was used to fit observed values on N C E R - P P F D response diagrams. This way N C E R could be calculated at every moment of the day during the growing period. On days lacking data (due to system breakdown or when unloading the data logger), N C E R was estimated as the mean of the values produced by
I
Fig. 1. Integrated system for simultaneous and continuous simulation of a high and an ambient CO2 atmosphere and for measurement of canopy gas-exchange rate. Parts: 1 ventilation air inlet at 5 m altitude, 2 air-conditioner, 3 growth-chamber = cuvette, 4 carbon dioxide injection, 5 blower (turbine), 6 venturimeter, 7 water manometer, 8 outlet for chamber air, 9 copper-constantan thermocouple and quantum sensor, 10 axial fan, l l airsampling lines (polypropylene), 12 buffertank, 13 time-controlled six-point switching unit, 14 pump for continuous flushing of air-sampling lines, 15 capacitive moisture sensor, 16 drying agent (silica gel), 17 rotameter, 18 absolute infra-red gas analyser, 19 data logger, 20 personal computer
the curves of the surrounding days. Eventually the total amount of C02 absorbed in photosynthesis and released in respiration was calculated and compared with dry-matter production over the same time interval. From the Et values a water balance can be calculated. Dividing total production by the amount of water lost, yields W U E over the interval. Maximum leaf photosynthetic rate and stomatal resistance. Maximum leaf photosynthetic rate (p ..... ) was determined at three one-week intervals during the growing periods by placing the containers in a separate small chamber, composed of mylar plastic over a metal frame. By leading air from the larger growth chambers through this small measuring unit, a similar atmosphere was created here. A 400-W sodium lamp was mounted on top of it, resulting in a PPFD of 800 ktmol.m-Z.s-1 at the leaf surface, which was found sufficient to saturate pho-
I. Nijs et al. : Leaf and canopy responses of Lolium to CO2 tosynthesis. Inside, maximum leaf photosynthetic rate p ..... of a single leaf could be determined by placing it in a small acrylic leaf chamber that was part of a closed-loop COz-measuring device, as described by Ceulemans et al. (1986). Since the device was designed for quick, non-destructive and extensive measurements, a large number of repetitions was possible. In the same small measuring chamber, stomatal resistance (r~) was determined with a portable diffusion porometer (Mk-II ; Delta-T Devices). In this context, an interesting parameter is ce (COz concentration in the intercellular spaces of the leaf), since its change with varying climate conditions can indicate changes in W U E (Louwerse 1980). It is calculated as c i = c , - ( p , , r z . c o ) (Gates 1980), where rz,co~= leaf resistance for COz diffusion. The latter was obtained by combining the upper and the lower leaf side resistances in parallel: 1/rl,co~= (1/r~pp~,co) +(l/rt ...... co). On either side the resistance was composed of a stomatal fraction, r,, and a boundary layer fraction, r,, e.g. : rupper,CO a = rs,upper,CO 2 + ra,upper,CO2-
In the above formula, r c o = 1.56.rH~o (Noble 1983). The boundary-layer resistance r, was calculated from evaporation q and " l e a f " to air vapour pressure difference VPD of a damp filter paper in the leaf chamber, using dew point mirrors (type TS 3C, Walz, Effeltrich, FRG). For this simulation of a leaf without stomatal barrier, r~ equals VPD/q.
315 PPFD
C3':iO
CE,O0
PPFD tmmolm-2 s-I] 3
HL-'ER(rag m-2 s--D 11.2
,.,.I
("
z
3
\
.......
LB "3,
'LY,
,"--7%,'
10 13 1G time of day (Gffr+ll
1.4
19
Fig. 2. Time courses of N C E R and PPFD on 2 May 1987, of a Lolium perenne sward under ambient (C350; 363 • 30 gl-1-1) and high (C600; 631 ___43 gl.1-1) CO2 concentrations (canopy age = 11 d)
PPFD
C350
C&O0
PPFO Immolm-2 s-l) 3
Et (rag"m-2 s-.12 20Q
Production assessment. Dry-matter production of some of the
stands in the growth chambers without a CO2-measurement facility was measured one, two and three weeks after clipping. In one, three-week growth period root production was determined. Data from a previous growing season, from April 1986 to July 1986, were used to calculate changes in root/shoot production ratio induced by high COz levels.
/./'='~'~. %.
2
it
0
Results
Canopy measurements. Canopy net CO2-exchange rate, NCER, and evapotranspiration rate, E , were monitored during several growing periods. The period 22 Aprit i987 to I3 May 1987 was selected for further analysis. Typical daily courses for N C E R and E, under ambient (C350, 363 + 30 gl1-1) and high (C600, 631 +_43 gl.1-1) CO2 are represented in Figs. 2 and 3. Under C600, N C E R is higher than under C350 at high P P F D only, while at low irradiance N C E R becomes considerably less under high CO2. Values of Et, however, are highly similar, irrespective of light intensity. On all days examined, considerable hysteresis effects occurred, for N C E R as well as for Et, although in opposite directions when comparing A.M. with P.M. values (Figs. 4, 5). Linear regression in N C E R - P P F D diagrams yielded higher determination coefficients r 2 for the C600 treatment (0.919 versus 0.792), so the hysteresis effect on photosynthesis was more intense at C350. The characteristics from N C E R - P P F D curves derived from diagrams like those in Figs. 4 and 5 are represented in Table 1. Values of N C E R at saturating P P F D were in the same order of magni-
7
,,,,-~X
I , I i0 13 15 time of: da~ (fclr+11
!0 19
Fig. 3. Time courses of E~ and PPFD on 2 May 1987, of a Lolium perenne sward under ambient (C350) and high (C600) CO2 concentrations (canopy age = 11 d)
tude as those obtained by Nijs et al. (1988) under laboratory conditions with the same species. Fullgrown canopies showed a 49% higher NCERm,x under C600, compared with C350. In earlier developmental stages this increase was even higher. Quantum efficiency, canopy dark-respiration rate (RD) and canopy light compensation point (Io) were also much higher under the C600 treatment, the differences were in fact enormous. It is not clear what caused the extremely high RD under C600, although Drake and Read (1981) me.asured RD values of a similar order of magnitude in a mixed-species plant community, also in an open system (RD amounted to up to 30% of maximum NCER). Calculating the carbon balance using these data showed that a large fraction of the CO21 absorbed during the day is released again during ':the night, especially under the C600 treatment (Table 2). Because of this effect, the increase in net CO,, input
316
I. Nijs et al. : Leaf and canopy responses of Lolium to CO 2 1.4 1.31.2-
11
1.1-
lO
1-
~ ~ .'~/'~
-
0.9-
/
0.8-
Ttt
0.70.6-
o4 I
0.5-
E
0.40.3-
m
E
13
/
....-- ""
0.2.
J ,0
0.1 0
z
8
-0.1 -0.2
lg
-
/
--0.3 -0.4-0.5 -0.6
-
-0.7
.
/
/
/
/
~/7
14
/
/ / ig !
0
I
I
I
0.2
I
0.4
I
I
I
0.6
0.8
I
I
1
t
1.2
PPIrD Cmmo! m--Z s.1) rl
C350
A
C600
Fig. 4. Hysteresis effects on N C E R in a Lolium perenne sward under ambient (C350) and high (C600) CO2 concentrations on 2 May 1987 (canopy age= 11 d). Data label=hour of day
caused by higher C O 2 becomes much smaller than expected merely based on the NCERmax difference. Total production is only a fraction of net CO2 uptake, since COz is converted into dry matter with a certain efficiency, depending on chemical composition and energy content of the final product. According to Sestak et al. (197J), 1 g of CO2 is converted into 0.5 g of protein, 0.6 g of starch or 0.4 g of lipids. De Wit (1978) gives similar values. The transformation efficiency in our experiments was about 0.4. The C350/C600 ratio for net CO2 input (0.73=566/711) was in the same order of magnitude as the ratio of total dry-matter production (0.79) between these treatments. Total water use was virtually the same for both treatments. The increase in W U E was 25%, and was solely the consequence of an increase in N C E R (Table 3).
Leaf measurements. At every developmental stage (1, 2 and 3 weeks after clipping), p . . . . . was higher under the C600 treatment. Since there was hardly any change with leaf age, mean values were calculated for the entire growth cycle: p . . . . . at C600
(1.067_+0.121 m g . m - 2 . s -1) was 46% higher than at C350 (0.732_+0.109 m g . m -2 .s-a). On average, rs was 50% higher at C600 than at C350 one-sidedly (1.599_+0.387 s-cm -1 versus 1.067_+0.180s-cm -1) and 46% higher for the whole leaf. There was a slight decrease with leaf age. At only one of three sampling dates was the difference statistically significant, almost certainly as a result of the instability of the diffusion porometer with changing temperature (there was no rigorous temperature conditioning in the measuring unit). The boundary-layer resistance for water-vapour diffusion was 0.79 s. c m - 1 one-sidedly. A summary of most of the important parameters describing stomatal behaviour is presented in Table 4. Although c~ rises substantially with elevated CO2 level, c~/c, remains fairly constant. In spite of the considerable stomatal closure, the extra COz can still result in a 46% rise in photosynthetic capacity. Assuming leaf transpiration rate (q) to be approximately inversely proportional to (r~+r,), an estimate can be made for the change in W U E of the leaf caused by high CO2 in saturating light: since p . . . . . rises by a factor of 1.458 and (rs + r.)
I. Nijs et al. : Leaf and canopy responses of Lolium to
317
CO 2
200 190 180 170 160 150 140 -
12 1
130
3
12
~
"~'-"
,~11
120 110 1 O0 90 80 70 60 50 40 30 20
8
7
100
I
I
0
I
I
0.2
I
I
0.4. Q
I
0.6
I
O.B
I
I
Table 1. Canopy CO2-exchange rate (NCERmax; mg. m - 2. s - 1 ) , RD (mg.m-2.s 1), Io (10 .6 mol'm-2"s -1) and quantum efficiency (c~; g CO2-(mol photons) 1) of the NCER-PPFD response curves for Lolium perenne grown at ambient (C350) and high (C600) CO2 concentrations during the growth cycle 22 April 1987 to 13 May 1987. Values are given for 4-, 13and 20-d-old swards. Canopy dark-respiration rate (RD) was directly calculated from RD values during the night (without fitting). Since asymptotic values were not always approached in full sunlight, NCER at 2 mmoi. m - 2. s- 1 is given in parentheses Development
Treatment
stage (d)
C350
C600
NCERm.x
4 13 20
1.03 (0.64) 1.91 (1.24) 2.93 (1.83)
2.39 (1.51) 2.99 (1.93) 4.37 (2.53)
c~
4 13 20
0.84 1.87 2.93
2.58 3.44 4.37
RD
4 13 20
--0.018 --0.056 --0.068
-0.386 --0.505 --0.716
Io
4 13 20
21 30 27
1.2
PPFD ~mmol m - 2 s - l ) r A C600
Fig. 5. Hysteresis effects o n E t in a Lolium perenne sward under ambient (C350) and high (C600) 1987 (canopy age = 11 d). Data label = hour of day
Characteristic
I
1
150 147 185
CO 2
concentrations on 2 May
Table 2. Carbon balance of Lolium perenne at ambient (C350) and high (C600) CO2 concentrations over the growing period from 22 April 1987 to 13 May 1987. Root production was estimated as a fraction of shoot production. All units are in g- m - 2 Treatment
c o z uptake during the day CO2 loss at night Net CO2 input Shoot production (measured) Root production (estimated) Total production
C350
C600
603.3 37.3 566.0 191.7 43.2 234.9
1186.2 415.2 771.0 228.9 66.1 295.0
i n c r e a s e s b y a f a c t o r o f 1.280 ( c a u s i n g q to d e c r e a s e b y a f a c t o r o f 0.781), W U E will i n c r e a s e b y a f a c t o r o f 1.458/0.781, w h i c h is a 8 7 % i m p r o v e m e n t .
D r y - m a t t e r production. The increase in shoot drym a t t e r p r o d u c t i o n o v e r the g r o w i n g p e r i o d 22 M a r c h 1987 to 13 M a y 1987 w a s 1 9 % . The investigation of the e v o l u t i o n of shoot drym a t t e r p r o d u c t i o n w i t h age s h o w e d t h a t the inc-
I. Nijs et al. : Leaf and canopy responses of Lolium to
318
Table3. Water use (kg.m -2) and W U E ( . 1 0 - S g CO2.(g H20) a) of Lolium perenne vegetations grown at ambient (C350) and high (C600) COz concentrations over the period from 22 April 1987 to 13 May 1987 Treatment
Water use WUE
C350
C600
98.2 239
98.4 300
Table 4. p . . . . . ( g g - m - 2. s- 1), r, (s. c m - 1), cl (gl. 1-1), c, (gl. 1-1) and cf/co of Lolium perenne leaves. Data are mean values over all growth stages (one, two and three weeks after clipping). ro for H 2 0 diffusion was 1.870 s-cm-~ one-sidedly Parameter
Treatment C350
r~,n2o,uppc~ l~af ~iae rs,tf2o,l. . . . l~f ~id, p,, .... ca ci cJc~
1.067 4.437 732 367 282 0.768
C600 1.599 5.800 1067 620 462 0.745
Table 5. Percentage increase of shoot and root dry-matter (DW) production caused by high (C600) CO2 concentration at three stages in the development of Lolium perenne. Growing period = 4 May 1987 to 25 May 1987 Development stage
Shoot D W
Root D W
Day 7 Day 14 Day 21
+34.2 +23.0 + 12.0
--1.9 + 16.0 + 28.0
rease caused by high CO2 became smaller during the course of canopy development towards full cover, while the stimulation of root dry-matter accumulation became stronger (Table 5). Over an entire growing season (April 1986 to July 1986), the root/shoot production ratio was 22% higher under C600 compared with C350. Discussion
Leaf responses to elevated COz level can be explained best with the classical equation F - - C / R ( F = CO2 flux, C = C O 2 concentration difference between the atmosphere and the site of carboxylation and R = resistance to CO2 diffusion). In high CO2, R is increased through stomatal closure, but this
CO 2
is overcompensated for by the increase of C. As a consequence, p, will rise and q will decrease, causing W U E to go up drastically. It appears however, that the leaf response is not the most economical one possible, because p, would rise even more if the stomata were unaffected, and leaf transpiration rate (q) would be similar to its value under ambient CO2 concentration. Obviously the CO2 benefit is only partly used for production purposes and partly for maintainance of survival purposes (through reduction of water use). It is clear that the individual leaf responds in relation to the plant or canopy, instead of as an autonomic entity. Possibly, the leaf anticipates the higher risk for water stress because of the larger transpiring surface under C600, though on the other hand only an equal amount of water is consumed compared with the C350 treatment. Another explanation is an imminent saturation of CO2 processing in the chloroplast or a saturation of carbohydrate export for allocation to other sinks in the plant, as a consequence of the higher COz supply if stomatal closure fails to occur. Even if photosynthesis is not saturated, keeping the stomata open will probably mean departure from optimal stomatal behaviour (dq/dp,=constant, Farquhar and Sharkey 1982), no longer minimizing water loss for extra carbon gain. The stomatal behaviour of Lolium perenne was similar to that of sunflower and barley (Louwerse 1980): ci was proportional to ca with a ratio of _+0.75. In these experiments it was pointed out that another type of response, with stabilization of ci with increasing % is more advantageous when water is limiting. Keeping a constant Ci/Ca ratio is intermediate between this type of behaviour and the situation where stomata do not react at all to rising ca. An increase in Pn,ma• is of course translated in a rise in NCERm,x. The latter is somewhat larger (49%) compared with the former (46%), which demonstrates the effect of CO2 on total leaf area. Calculating backwards from dry matter to leaf area, using allometric relations derived by Nijs et al. (1988) yielded a 9.9% shoot length increase in high CO2 for full-grown stands. The final increase in dry-matter production is only 19%. This is partly a consequence of the relatively low P P F D values during the growing period selected for gasexchange analysis, because the effect of higher CO2 is strongest at high PPFD. At low irradiance the high leaf-area index of the canopy grown at C600 is even a disadvantage, because light extinction is so fast that the vegetation already becomes CO2 deficient at 100 to 200 izmol.m - 2 . s - 1 . A large
I. Nijs et al. : Leaf and canopy responses of Lolium to COz
number of respiring leaves and a high RD and Io are the result. The structural effect of C600 is therefore a disadvantage at low, but a much stronger benefit at high PPFD. Since light can become the limiting factor, the reasoning is confirmed that high CO2 is only a serious benefit if it is the limiting factor. Brown and Higginbotham (1986) found analogous results with Picea glauca: elevated CO2 level increased leaf and total biomass only at high N-availability. In other species, however, production was stimulated on unfertilized soil and in spite of severe N-deficiency (Norby et al. 1986). The influence of limited factors may be speciesspecific. Gas-exchange measurements on the canopy level demonstrated considerable hysteresis effects on CO2 and H 2 0 : N C E R showed a significant drop in the afternoon, while Et rose drastically. The first phenomenon was probably caused by source-sink effects (sink saturation), while the second was almost certainly driven by environmental factors like a higher temperature and a lower relative humidity. The effect of elevated COz is rather positive than negative, because hysteresis is less intense at C600, indicating there was no more sink-saturation compared with C350. An interesting strategy for the vegetation - especially with limited water availability - would be to open the stomata in the first half of the day only, when carbohydrate sinks are not yet saturated, and Et is low because of the relatively low air temperature and vapour-pressure deficit between leaf and atmosphere. A variation of this behaviour is observed in sugarcane: high yields in this species have been empirically correlated with maximal stomatal opening occurring in the morning and evening, when vapour-pressure deficit is low, and reduced conductance during midday when it was high, rather than with continuous opening during the daylight hours (Grantz et al. 1987). It is worth mentioning that the data presented on canopy responses are representative for late spring growth only. An earlier study (Nijs et al. 1988) has shown that gas exchange in summer can deviate somewhat from this general pattern, while some other differences between the results from this earlier study and the present one are also striking, e.g. the dry-matter increase here was only 19%, while it was 46% over the 1986 growing season. Radiation-input differences are obviously responsible for this. The W U E figures are fairly out of line as well, but these are not comparable because the integrated W U E value calculated here over a three-week growing period is a totally different parameter from the W U E figures determined
319
in a fixed laboratory environment. On the other hand the previously measured 77% enhancement of maximum canopy photosynthetic rate was confirmed in this study, which is not remarkable for it is an intrinsic plant characteristic rather than one depending on environmental change. Production assessment showed that at the beginning of the growing period, immediately after clipping, the extra carbon from C600 is invested relatively more into shoots and less into root material. In the course of canopy development this allocation pattern is reversed, which is not illogical since a stand recently mown has a relative surplus of roots. This is in agreement with the earlier observation from Nijs et al. (1988) that an important fraction of the extra dry matter produced in high CO2 originates from the first few days of regrowth after the vegetation is mown. Data from an entire growing season showed that the higher CO2 uptake under the C600 regime was more intensely allocated to the root system than to the shoots. Concluding that Lolium perenne is "wasting" part of the extra COz by converting it into no~L-harvestable biomass, is however not correct; since enhanced root growth or stimulation of storage into roots is a benefit after clipping: the allocated material is directly or indirectly used for establishing a new canopy. Thus it still ends up in the harvestable material, increasing the harvest index. This study was supported by the "Instituut tot Aanmoediging van bet Wetenschappetijk Onderzoek in Nijverheid en Landbouw" (Brussels, Belgium).
References Boote, K.J., Jones, J.W. (1986) Equations to define canopy photosynthesis from quantum efficiency, maximum leaf rate, light extinction, leaf area index and photon flux density. Proc. 7th Congr. on Photosynthesis, Providence, Rhode Island), Biggins, J., ed. Martinus Nijhoff, D'ordrecht, The Netherlands Brown, K., Higginbotham, K.O. (1986) Effect of carbon dioxide enrichment and nitrogen supply on growth of boreal tree seedlings. Tree Physiol. 2, 223-232 Ceulemans, R., Kockelbergh, F., Impens, I. (1986) A fast, low cost and low power requiring device for improving closed loop CO2 measuring systems. J. Exp. Bot. 37, ~234~1244 De Wit, C.T. (1978) Stimulation of assimilation, respiration and transpiration of crops. Centre for Agricultural Publishing and Documentation, Wageningen Drake, B.G., Read, M. (1981) Carbon dioxide assimilation, photosynthetic efficiency and respiration of Chesapeake Bay salt march. J. Ecol. 69, 405-423 Drake, B.G., Rogers, H.H., Allen, L.H. (1985)Methods of exposing plants to elevated carbon dioxide. In: Direct effects of increasing carbon dioxide on vegetation (Rep. DOE/ ER-239, Dept. of Energy, Washington D.C.), pp. 11-35, Strain, B.R., Cure, J.D., eds. NTIS, Springfield, Va., USA
320 Farquhar, D.G., Sharkey, T.D. (1982) Stomatal conductance and photosynthesis. Annu. Rev. Plant Physiol. 33, 317-345 Gates, D.M. (1980) Biophysical ecology. Springer, New York Heidelberg Berlin Grantz, D.A., Moore, P.H., Zeiger, E. (1987) Stomatal responses to light and humidity in sugarcane: prediction of daily time courses and identification of potential selection criteria. Plant Cell Environ. 10, 197504 Louwerse, W. (1980) Effects of CO2 concentration and irradiance on the stomatal behaviour of maize, barley and sunflower plants in the field. Plant Cell Environ. 3, 391 398 Nijs, I., Impens, I., Behaeghe, T.J. (1988) Effects of rising at-
I. Nijs et al. : Leaf and canopy responses of Lolium to
CO 2
mospheric carbon dioxide on gas exchange and growth of perennial ryegrass. Photosynthetica 22, 44-50 Nobel, P.S. (1983) Biophysical plant physiology and ecology. Freeman, San Francisco Norby, R.J., Pastor, K., Melillo, J.M. (1986) Carbon-nitrogen interactions in COz-enriched white oak: physiological and long-term perspectives. Tree Physiol. 2, 243 259 Sestak, Z., Catsky, J., Jarvis, P.G. (1971) Plant photosynthetic production: manual of methods. Junk, The Hague Received 29 December 1987; accepted 21 October 1988
