Summary In this review, I focus on modeling studies of tree responses to CO2 enrichment. First, I examine leaf-scale
mod-els of assimilation with respect to the interaction between low temperature and CO2 enrichment. Second, because changes in
allocation within a tree may be significant in determining the growth response of trees to CO2 enrichment and low
tempera-tures, I review models of the control of allocation in plants. Finally, models of stand-scale processes are discussed with respect to their ability to make reliable estimates of likely vegetation responses to predicted climate change. I conclude that our ability to make reliable predictions is hindered by our lack of understanding of several processes, namely: the inter-action between increased atmospheric CO2 concentration and
low temperatures; the control of allocation in plants; and the modeling of stand-scale processes.
Keywords: carbon assimilation, growth, leaf-scale models.
Introduction
Atmospheric CO2 concentrations are increasing at
approxi-mately 1.8% per annum. It is predicted that this increase will have a significant impact on the physiology and growth of trees. The short- and long-term responses to increases in at-mospheric CO2 concentration have been extensively reviewed.
In this review, therefore, I focus on the models that are being developed to study tree responses to CO2 enrichment. I
exam-ine leaf-scale models of assimilation with respect to the inter-action of low temperature with CO2 enrichment. Because
changes in allocation within a tree may be significant in deter-mining the growth response of trees to CO2 enrichment, I
review models of the control of allocation in plants. Finally, models of stand-scale processes are discussed with respect to their ability to make reliable estimates of likely vegetation response to predicted climate change.
Leaf-scale models of assimilation
The response of assimilation to changes in temperature, and the influence of temperature both on the solubilities of CO2 and
O2 and on the specificity of Rubisco to CO2 and O2 are well
known. Similarly, the short-term influence of CO2
concentra-tion on assimilaconcentra-tion is well documented. The response of
maximum assimilation rate to increasing leaf temperature at different CO2 concentrations has also been documented (Long
1991, Eamus et al. 1995). As temperature rises, the stimulation of assimilation resulting from any given increase in CO2
con-centration is increased. For example, in response to a doubling in atmospheric CO2 concentration, CO2 assimilation at
saturat-ing PPFD, Asat , is predicted to increase by 54 and 73% at 20
and 30 °C respectively (Long 1991), and the optimum tem-perature for photosynthesis is also predicted to increase (Ea-mus et al. 1995). However, Rubisco activity declines by 20 to 40% in response to CO2 enrichment (Sage et al. 1989). As the
percentage decline in activity increases, the ratio of Asat at 700
µmol mol−1 to Asat at 350 µmol mol−1 declines, and this affects
the temperature at which CO2 enrichment has a positive effect
on A (Long 1991). Thus, when Rubisco activity declines in response to CO2 enrichment, the temperature threshold at
which CO2 enrichment can have a positive impact on
assimi-lation increases. For example, this temperature increases from about 14 to 24 °C as Rubisco activity declines from 80 to 60% of control values (Long 1991). At temperatures below the threshold, typically less than 16--19 °C, CO2 enrichment has a
negative impact on assimilation rate.
Based on these simulations, it has been concluded that CO2
enrichment will have a larger impact on assimilation at high temperatures than at low temperatures (Jones et al. 1985, Baker et al. 1989). In several studies it has been shown that CO2 enrichment has a negative impact on plant growth at
temperatures below 18 °C (cf. Idso et al. 1987, Kimball et al. 1993, Rogers and Dahlman 1993, Rozema et al. 1993). How-ever, CO2 enrichment does not always have a negative impact
on growth at low temperatures suggesting that C fixation may not be the only process underlying the response of CO2
-en-riched plants to low temperatures (Stitt and Schulze 1994). For example, Sionit et al. (1981) showed that okra was able to grow at lower temperatures when supplied with CO2-enriched air
than when supplied with ambient air. Similarly Potvin and Strain (1985) found a C4 grass was more responsive to CO2
enrichment at low temperatures than at high temperatures. Potvin (1994) observed that CO2 enrichment had a significant
impact on the biomass increment of two C3 species and two C4
species and the effect was not influenced by temperature. In an Arctic tundra study, Grulke et al. (1990) showed that rates of canopy uptake were significantly increased by CO2
enrich-Tree responses to CO
2
enrichment: CO
2
and temperature interactions,
biomass allocation and stand-scale modeling
DEREK EAMUS
School of Biological Sciences, Northern Territory University, P.O. Box 40146, Casuarina, Northern Territory 0811, Australia
Received March 16, 1995
ment, despite the low temperatures of the Arctic tundra. Thus, the published data indicate that the growth response to CO2 enrichment does not increase with increasing
tempera-ture in all species, and some species exhibit a positive growth response to CO2 enrichment at low temperatures. Possible
explanations for these results are that, in some species, Ru-bisco activity or content may not decline in response to ele-vated CO2 (cf. Delgado et al. 1994), or that declines in Rubisco
activity or content are not correlated with a decline in Amax .
Other explanations to account for the increase in growth rate at low temperatures include changes in allocation, or changes in efficiency of use of fixed carbon. Alternatively, the flux control for photosynthesis (Kacser and Burns 1973) may change with temperature and CO2 concentration so that the
degree of limitation imposed by Rubisco on carbon fixation may decline under CO2-enriched conditions at low
tempera-tures (Stitt and Schulze 1994).
Changes in specific leaf area, leaf weight area, shoot weight ratio, carbon content of dry matter, and respiration rates of roots and shoots can all directly influence growth. All of these relationships and processes respond to CO2 enrichment, but
both the magnitude and direction of the response are variable (Poorter 1993, Stulen and den Hertog 1993, Wullschleger et al. 1994). Johnson and Lincoln (1990) concluded that even a decline in the carbon content of plant biomass can result in an increase in growth. Therefore, although the Rubisco model predicts a decrease in the rate of carbon fixation per unit leaf area with decreasing temperature, a change in allocation of carbon within the plant could increase growth rate per plant, if this results in increased photosynthetic area.
Increased efficiency of conversion of fixed carbon to support growth may also contribute to the observed increased growth at low temperatures. Stitt and Schulze (1994) have discussed the concept of wastage of stored carbohydrate based on the observation that changes in response to reduced Rubisco con-tent in Rubisco antisense transformed tobacco cells included a decline in the wastage of stored carbohydrate to compensate for reduced rates of carbon fixation. Wullschleger et al. (1994) concluded that respiration can increase or decrease in response to CO2 enrichment. Maintenance and growth respiration may
respond differently to CO2 enrichment at low temperatures. A
decline in respiration rate in response to CO2 enrichment at low
temperatures and an increase or no change in respiration rate in response to CO2 enrichment at higher temperatures has been
observed (Ziska and Bunce 1994). Thus, at low temperatures, a reduction in respiratory carbon loss could contribute to en-hanced growth (Ziska and Bunce 1994).
I conclude that growth of CO2-enriched plants may increase
at low temperatures despite a decline in the rate of carbon fixation. Changes in allocation, efficiency of use of fixed carbon, and changes in flux control for photosynthesis may all contribute to an increase in growth rate at low temperature. The finding that the growth rate can increase in response to low temperatures even though the Rubisco model predicts a de-cline in the rate of carbon fixation with decreasing temperature implies that either the model is incorrect, or the relationship between carbon fixation rate and growth is complex. Because
the Rubisco model appears to be correct, the latter explanation is favored.
Biomass allocation models
An understanding of the mechanisms underlying the control of biomass allocation in trees is especially relevant to the devel-opment of models for predicting canopy and catchment scale responses to climate change and CO2 enrichment. Models of
the control of allocation in plants can be classified into four groups (Wilson 1988): (a) allometric; (b) functional equilibria; (c) Thornley-type models; and (d) hormone models. Allomet-ric models fit equations to data sets so that the equation is a good description of the data. However, such models are empiri-cal and do not provide insight into mechanisms that may underlie the relationship between shoot and root growth.
Functional equilibria models emphasize the importance of the differential functioning of shoots and roots. A generalized functional equilibrium model can take the form: root mass × root activity = k× shoot mass × shoot activity, where root activity signifies ion (or N) uptake capacity (Givnish 1986, Dewar 1993), shoot activity is the capacity for carbon assimi-lation, and the proportionality constant, k, represents the amount of nitrogen used in plant material per unit amount of carbon (Makela and Sievanen 1987). Although functional equilibria models are a significant development over allomet-ric models, these models only describe a data set and provide no information about the mechanisms that control the alloca-tion of biomass to roots and shoots.
Thornley developed mechanistic models based on three fea-tures. First, root or shoot growth is dependent on the concen-trations of carbon and nitrogen in root and shoot; second, the movement of labile C and N between shoots and roots is dependent on the difference between shoot and root in concen-trations of labile C and N compounds; and third, structural material and labile compounds are treated as separate pools. Of models developed by Thornley, the transport resistance model (Thornley 1972), which views the allocation of C and N as a means of balancing the functional relationships of roots and shoots, appears to be the most robust and has been incorpo-rated into several forest models (Rastetter et al. 1991, Thornley and Cannell 1992).
propor-tional to C and N concentration, the product of C × N can be used to explain the response of allocation to changes in N or C availability.
The weakness of the TR model is the assumption that allo-cation of C to the root and alloallo-cation of N to the shoot is proportional to the amount of root and shoot biomass, respec-tively. That is, the model does not attempt to account for the retranslocation of C and N within the plant. This weakness has been addressed in the mechanistic model proposed by Minchin et al. (1993), which is based on a kinetic analysis of substrate unloading from the phloem in sinks. In this model, which is based on the Michaelis-Menten equation, it was shown that if the Michaelis constant, km, and maximum velocity of phloem
unloading, Vmax , of two sinks differ, and if source loading
declines due to a decline in source activity, then partitioning of photosynthate to the two sinks will differ. Consequently, with appropriate values of km and Vmax for competing sinks, it was
possible to model the observed increase in partitioning to shoots in response to shading of the source leaves without having to invoke changes in allocation of N from the roots (Minchin et al. 1994). An important feature of this model is that the feedback between changes in supply of C from source leaves and allocation of photosynthate between source and sink is explained on the basis of kinetic properties of the sinks, rather than on the basis of shoot and root biomass. The model of Minchin et al. (1993, 1994) deals only with the allocation of C substrates. Whether it can be extended to include the alloca-tion of N remains to be tested. However, uptake of nitrogenous compounds from xylem sap into leaf cytoplasm and loading of nitrogenous compounds into the phloem of leaves for re-translocation to roots with subsequent unloading in roots, provide steps for carrier-mediated transport across mem-branes. Thus, the idea of differences in km and Vmax may be
applicable to the allocation of N in plants. Because up to 50% of the N of roots is derived from N retranslocated from leaves, it is possible that, in the absence of active mechanisms for N unloading from the xylem and phloem, the link between changes in N availability in the soil and changes in allocation of C and N to roots could be mediated via the translocation of C from the shoots.
Dewar (1993) has expanded the TR model based on the work of Givnish (1986) to include a role for water uptake by roots. This root--shoot partitioning model simulates the transport of water, labile N and C, and the interaction between these factors in the control of growth. The model assumes that the growth rate of a site (root/shoot) is determined by the availability of labile C and N at that site, and by its water status. It also assumes that a fraction of the total N taken up by roots is transported to the shoot via the xylem, and the remaining fraction is passed to the root phloem, where it joins the N that is moving down the plant from the shoot.
The model has four important characteristics. First, when the fraction of N that is moved to the shoot is less than or equal to the fraction of total plant biomass in the shoot, then the concentration of labile N in the root increases and is available to support increased root growth. Second, the existence of counter-gradients of labile C and N determine the root and
shoot responses to changes in the supply of C or N. Third, root and shoot water potentials and biomass are explicitly linked in this model. Finally, unlike the TR model, this root--shoot partitioning model can account for the difference in the re-sponse of root/shoot ratio to changes in N supply versus changes in potassium, magnesium and manganese supply.
Luo et al. (1994) have modeled the response of allocation of biomass in plants grown under CO2-enriched conditions on the
basis of specific leaf and plant N concentrations, leaf mass per unit area, and plant N productivity. Generally, N concentra-tions declined and leaf mass per unit area and plant N produc-tivity increased in response to CO2 enrichment. They were also
able to predict up and down regulation of photosynthesis on the basis of the relative changes in leaf mass per unit area and leaf N concentrations. They then compared two regulatory models, a simple functional balance model and a photosynthe-sis--growth model, to explain changes in root and shoot alloca-tion. The photosynthesis--growth model is based on the hypothesis that photosynthesis and RGR are mechanistically related, whereas the functional balance model is based on the hypothesis that allocation and physiological activity of roots and shoots are mechanistically linked. The functional balance model was found to be a poor description of the experimental data; however, the photosynthesis--growth model, which pre-dicted root allocation on the basis of relative sensitivities of photosynthesis and relative growth rate to CO2 enrichment,
more closely matched published results. The model predicted that root allocation increases whenever the rate of photosyn-thesis per unit leaf mass increases more than RGR in response to CO2 enrichment.
It seems likely that changes in allocation of biomass, result-ing from CO2 enrichment influence water use, nutrient and
light capture, and nutrient use efficiency, and hence the func-tional ecology of catchments. However, without a detailed understanding of the mechanisms controlling biomass alloca-tion, our ability to develop models of such responses is seri-ously restricted.
Modeling at the stand scale
A recent development in forest process modeling has been the G’DAY model of Comins and McMurtrie (1993). In this model, photosynthesis, tree productivity, soil carbon--nitrogen dynamics and allocation are included to generate equilibrium responses of tree stands to changes in specific environmental conditions. Three biomass fractions are included (foliage, wood and fine roots). The soil component has four litter com-partments (surface structural, soil structural, surface metabolic and soil metabolic). In addition, the soil has three organic fractions (active, slow and passive). When CO2 concentrations
are doubled, the model predicts a short-term, rapid and large increase in productivity. Consequently, C input to the soil increases and this results in increased sequestration of nutri-ents in the soil and nutrient availability declines rapidly (within 1 year). As a result, plant C/N ratio increases and productivity is substantially reduced.
it incorporates many aspects of tree and forest ecophysiology previously ignored. Thus, N deposition, N release from wood decay, and other processes in the N cycle, as well as the influence of foliar N content on carbon assimilation, are taken into account to a lesser or greater extent. Carbon contents and sequestration into a variety of pools are also explicitly dealt with.
Kirschbaum et al. (1994) tested the effects of varying certain model parameters on model predictions and obtained the fol-lowing information. (1) When the N/C ratio of wood is propor-tional to the N/C ratio of foliage, the impact of CO2 enrichment
on productivity increases from 1.1 (when the ratio is kept constant) to 14.1%, indicating the importance of experimen-tally determining the relationship between wood and foliage N/C ratios for trees exposed to ambient and enriched CO2. (2)
The relationship between foliar N content and maximum as-similation rate can change under CO2 enriched conditions.
Nitrogen use efficiency can increase (Tissue et al. 1993) and decrease (Stitt 1991). Given this uncertainty, the relationship between foliar N and Amax was assumed to be unchanged. (3)
The sensitivity of the whole canopy to CO2 concentration was
sensitive to the choice of photosynthetic parameters and to the choice of the relationship between foliar N content and assimi-lation. More information for mature trees growing in a CO2
-enriched environment is required. (4) If the allocation ratio for foliage, fine roots and woody parts is fixed at 0.2, 0.2 and 0.6, respectively, productivity increases by 14.1% when CO2
con-centration is doubled. When a negative linear response for allocation to roots as a function of foliar N content is incorpo-rated, productivity increases to 14.8%. Interestingly, the sensi-tivity of wood production to a doubling of the CO2
concentration declines from 14.1 to 10.6%, a decline of ap-proximately 25%, as a result of enhanced allocation to fine roots. Changes in specific leaf area substantially influence the response of productivity to CO2 enrichment. When specific
leaf area, SLA, declined by 20%, sensitivity of productivity declined from 14.1 to 8.2%. Thus there is a significant negative feedback between SLA and light interception and responsive-ness to CO2 concentration. (5) The data of Mousseau and
Enoch (1989) show that leaf longevity increases in response to CO2 enrichment. Similarly, there is experimental and
theoreti-cal evidence showing that the light compensation point for photosynthesis is decreased by CO2 enrichment (Long 1991),
and consequently, leaf longevity may increase. However, from considerations of the relationship between foliar nitrogen con-tent and leaf longevity, Kirschbaum et al. (1994) showed that productivity increased from 14.1 to 19.3% when leaf longevity increased. The increase was especially noticeable on soils of low N content. Clearly, leaf longevity can have a substantial impact on canopy responses to CO2 enrichment. (6) The model
showed that nitrogen limitations do not provide an overriding constraint on the ability of trees to respond to CO2 enrichment.
Rather, several interactions between soil, plant and atmosphere determine the final response of trees to CO2 enrichment.
The study by Kirschbaum et al. (1994) revealed the impor-tance of the percentage of N lost from the system through leaching and gaseous emissions. It was the N recycling
con-straint curve that essentially determined the magnitude of the response of productivity to CO2 enrichment.
I conclude that tree-based mechanistic studies of long-term responses to CO2 enrichment are needed. Vertical integration
of these long-term responses at the three major spatial scales, namely leaf scale, whole-plant scale and stand scale, should be possible by the end of the century.
Acknowledgments
I thank Drs. J.D. Barnes and H. Griffiths for their critical review of this manuscript and extensive discussions of the topics covered.
References
Baker, J.T., L.H. Allen, K.J. Boote, P. Jones and J.W. Jones. 1989. Response of soybean to air temperature and CO2 concentration.
Crop Sci. 29:98--105.
Comins, H.N. and R.E. McMurtrie. 1993. Long-term biotic response of nutrient limited forest ecosystems to CO2 enrichment:
equilib-rium behaviour of integrated plant--soil models. Ecol. Appl. 3:666--681.
Delgado, E., R.A.C. Mitchell, M.A.J. Parry, S.P. Driscoll, V.J. Mitchell and D.W. Lawlor. 1994. Interaction effects of CO2 concentration,
temperature and nitrogen supply on the photosynthesis and compo-sition of winter wheat. Plant Cell Environ. 17:1205--1213. Dewar, R.C. 1993. A root--shoot partitioning model based on
carbon--nitrogen--water interaction and Münch phloem flow. Funct. Ecol. 7:356--368.
Eamus, D., C. Berryman and G.D. Duff. 1995. A study of the impact of CO2 enrichment on leaf water potential, transpiration and
hy-draulic conductivity of Maranthes corymbosa and Eucalyptus tetro-donta. Aust. J. Bot. 43:273--282.
Givnish, T.J. 1986. Optimal stomatal conductance, allocation of en-ergy between leaves and roots and the marginal cost of transpira-tion. In On the Economy of Plant Form and Function. Ed. T.J. Givnish. Cambridge University Press, Cambridge, pp 171--213. Grulke, N.E., R.H. Reichers, W.C. Oechel, H. Hjelm and C. Jaeger.
1990. Carbon balance in tussock tundra under ambient and elevated CO2. Oecologia 83:485--494.
Hilbert, D.W., A. Larigauderie and J.F. Reynolds. 1991. The influence of CO2 and daily photon-flux density on optimal leaf nitrogen
concentration and root--shoot ratio. Ann. Bot. 68:365--376. Idso, S.B., B.A. Kimball, M.G. Anderson and J.R. Mauney. 1987.
Effects of atmospheric CO2 enrichment on plant growth: the
inter-active role of air temperature. Agric. Ecosyst. Environ. 20:1--10. Johnson, R.H. and D.E. Lincoln. 1990. Sagebrush and grasshopper
responses to atmospheric CO2. Oecologia 84:103--110.
Jones, P., L.H. Allen and R. Valle. 1985. Photosynthesis and transpira-tion responses of soybean canopies to short-and long-term CO2 treatments. Agron. J. 77:119--126.
Kacser, H. and J.A. Burns. 1973. The control of flux. SEB Symposium 27, Cambridge Univ. Press, Cambridge, U.K., pp 65--107. Kimball, B.A., J.R. Mauney, F.S. Nakayama and S.B. Idso. 1993. In
CO2 and Biosphere. Eds. J. Rozema, H. Lambers, S.C. van de Geijn
and M.L. Cambridge. Adv. Veg. Sci. 14:65--75.
Long, S.P. 1991. Modification of the response of photosynthetic pro-ductivity to rising temperature by atmospheric CO2 concentration:
has its importance been under-estimated? Plant Cell Environ. 14:79--739.
Luo, Y., C.B. Field and H.A. Mooney. 1994. Predicting responses of photosynthesis and root fraction to elevated [CO2]a: interactions
among carbon, nitrogen and growth. Plant Cell Environ. 17:1195--1204.
Makela, A.A. and R.P Sievanen. 1987. Comparison of two shoot--root partitioning models with respect to substrate utilisation and func-tional balance. Ann. Bot. 59:129--140.
Minchin, P.E.H., M.R. Thorpe and J.F. Farrar. 1993. A simple mecha-nistic model of phloem transport which explains sink priority. J. Exp. Bot. 44:947--955.
Minchin, P.E.H., M.R. Thorpe and J.F. Farrar. 1994. Short-term con-trol of root:shoot partitioning. J. Exp. Bot. 45:615--622.
Mousseau, M. and H.Z. Enoch. 1989. CO2 enrichment reduces shoot
growth in sweet chestnut seedlings. Plant Cell Environ. 12:927--934.
Poorter, H. 1993. Interspecific variation in the growth response of plants to an elevated ambient CO2 concentration. In CO2 and
Bio-sphere. Eds. J. Rozema, H. Lambers, S.C. van de Geijn and M.L. Cambridge. Adv. Veg. Sci. 14:77--98.
Potvin, C. 1994. Interactive effects of temperature and atmospheric CO2 on physiology and growth. In Eds. R.B. Alscher and A.R.
Welburn. Chapman and Hall, Cambridge, pp 39--54.
Potvin, C. and B.R. Strain. 1985. Photosynthetic response to growth temperature and CO2 enrichment in 2 species of C4 grasses. Can. J.
Bot. 63:483--487.
Rastetter, E.B., M.G. Ryan, G.R. Shaver, J.R. Mellilo, K.J. Nadelhof-fer, J.E. Hobbie and J.D. Aber. 1991. A general biochemical model describing the responses of the C and N cycles in terrestrial ecosys-tems to changes in CO2 and climate, and N deposition. Tree Physiol.
9:101--126.
Rogers, H.H. and R.C. Dahlman. 1993. Crop responses to CO2
enrich-ment. In CO2 and Biosphere. Eds. J. Rozema, H. Lambers, S.C. van
de Geijn and M.L. Cambridge. Adv. Veg. Sci. 14:117--132.
Rozema, J., H. Lambers, S.C. van de Geijn and M.L. Cambridge. 1993. CO2 and biosphere. Adv. Veg. Sci. 14:1--484.
Sage, R.F., T.D. Sharkey and J.R. Seeman. 1989. Acclimation of photosynthesis to elevated CO2 in 5 C3 species. Plant Physiol.
89:590--596.
Sionit, N., B.R. Strain and H.A. Beckman. 1981. Environmental con-trols on the growth and yield of okra. Effects of temperature and CO2 enrichment at cool temperatures. Crop Sci. 21:885--888.
Stitt, M. 1991. Rising CO2 levels and their potential significance for
carbon flow in photosynthetic cells. Plant Cell Environ. 14:741--762.
Stitt, M. and E.-D. Schulze. 1994. Does Rubisco control the rate of photosynthesis and plant growth? An exercise in molecular eco-physiology. Plant Cell Environ. 17:465--487.
Stulen, I. and J. den Hertog. 1993. Root growth and functioning under atmospheric CO2 enrichment. In CO2 and Biosphere. Eds. J.
Rozema, H. Lambers, S.C. van de Geijn and M.L. Cambridge. Adv. Veg. Sci. 14:99--116.
Thornley, J.H.M. 1972. A model to describe the photosynthate during vegetative plant growth. Ann. Bot. 36:419--430.
Thornley, J.H.M. and M.G.R. Cannell. 1992. Nitrogen relations in a forest plantation--soil organic matter ecosystem model. Ann. Bot. 70:137--151.
Tissue, D.T., R.B. Thomas and B.R. Strain. 1993. Long-term effects of elevated CO2 and nutrients on photosynthesis and Rubisco in
loblolly pine seedlings. Plant Cell Environ. 16:859--865.
Wilson, J.B. 1988. A review of evidence on the control of shoot:root ratio, in relation to models. Ann. Bot. 61:433--449.
Wullschleger, S.D., L.H. Ziska and J.A. Bunce. 1994. Respiration responses of higher plants to atmospheric CO2. Physiol. Plant.
90:221--229.
Ziska, L.H. and J.A. Bunce. 1994. Direct and indirect inhibition of single leaf respiration by elevated CO2 concentrations: interaction
