Nutrient and genotypic effects on CO2-responsiveness: photosynthetic regulation in Leucadendron species of a nutrient-poor environment

Journal of Experimental Botany, Apr 1999

Four South African Leucadendron congenerics with divergent soil N and P preferences were grown as juveniles at contrasting nutrient concentrations at ambient (350 µmol mol−1) and elevated (700 µmol mol−1) atmospheric CO2 levels. Photosynthetic parameters were related to leaf nutrient and carbohydrate status to reveal controls of carbon uptake rate. In all species, elevated CO2 depressed both the maximum Rubisco catalytic activity (Vc,max, by 19–44%) and maximum electron transport rate (Jmax, by 13–39%), indicating significant photosynthetic acclimation of both measures. Even so, all species had increased maximum light-saturated rate of net CO2 uptake (Amax) at the elevated growth CO2 level, due to higher intercellular CO2 concentration (ci). Leaf nitrogen concentration was central to photosynthetic performance, correlating with Amax, Vc,max and Jmax. Vc,max and Jmax were linearly co-correlated, revealing a relatively invariable Jmax:Vc,max ratio, probably due to N resource optimization between light harvesting (RuBP regeneration) and carboxylation. Leaf total non-structural carbohydrate concentration (primarily starch) increased in high CO2, and was correlated with the reduction in Vc,max and Jmax. Apparent feedback control of Vc,max and Jmax was thus surprisingly consistent across all species, and may regulate carbon exchange in response to end-product fluctuation. If so, elevated CO2 may have emulated an excess end-product condition, triggering both Vc,max and Jmax down-regulation. In Leucadendron, a general physiological mechanism seems to control excess carbohydrate formation, and photosynthetic responsiveness to elevated CO2, independently of genotype and nutrient concentration. This mechanism may underlie photosynthetic acclimation to source:sink imbalances resulting from such diverse conditions as elevated CO2, low sink strength, low carbohydrate export, and nutrient limitation.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://jxb.oxfordjournals.org/content/50/333/533.full.pdf

Nutrient and genotypic effects on CO2-responsiveness: photosynthetic regulation in Leucadendron species of a nutrient-poor environment

G.F. Midgley 0 S.J.E. Wand 0 N.W. Pammenter 0 0 Ecology and Conservation, National Botanical Institute , Post Bag X7, Claremont 7735, Cape Town, South Africa 2 Department of Biology, University of Natal , Durban 4041, South Africa Four South African Leucadendron congenerics with divergent soil N and P preferences were grown as juveniles at contrasting nutrient concentrations at ambient (350 mmol mol1) and elevated (700 mmol mol1) atmospheric CO2 levels. Photosynthetic parameters were related to leaf nutrient and carbohydrate status to reveal controls of carbon uptake rate. In all species, elevated CO2 depressed both the maximum Rubisco catalytic activity (Vc,max, by 19-44%) and maximum electron transport rate (Jmax, by 13-39%), indicating significant photosynthetic acclimation of both measures. Even so, all species had increased maximum light-saturated rate of net CO2 uptake (Amax) at the elevated growth CO2 level, due to higher intercellular CO2 concentration (ci). Leaf nitrogen concentration was central to photosynthetic performance, correlating with Amax, Vc,max and Jmax. Vc,max and Jmax were linearly co-correlated, revealing a relatively invariable Jmax:Vc,max ratio, probably due to N resource optimization between light harvesting (RuBP regeneration) and carboxylation. Leaf total non-structural carbohydrate concentration (primarily starch) increased in high CO2, and was correlated with the reduction in Vc,max and Jmax. Apparent feedback control of Vc,max and Jmax was thus surprisingly consistent across all species, and may regulate carbon exchange in response to endproduct fluctuation. If so, elevated CO2 may have emulated an excess end-product condition, triggering both Vc,max and Jmax down-regulation. In Leucadendron, a general physiological mechanism seems to control excess carbohydrate formation, and photosynthetic responsiveness to elevated CO2, independently of genotype and nutrient concentration. This mechanism may underlie photosynthetic acclimation to source:sink imbalances resulting from such diverse conditions as elevated CO2, low sink strength, low carbohydrate export, and nutrient limitation. Introduction The photosynthetic and growth responses of C3 plants to elevated CO2 show a bewildering diversity, ranging from highly positive to neutral and, in rare cases, even negative (Poorter, 1993; Gunderson and Wullschleger, 1994). This greatly complicates the accurate prediction of ecosystem changes as CO2 continues to accumulate in the earths atmosphere. Responses of C3 plants to rising atmospheric CO2 levels are clearly modified by growing conditions (Idso and Idso, 1994), and appear strongly species(Poorter, 1993) and even ecotype- (Norton et al., 1995) and genotype-specific (Curtis et al., 1994; Zhang and Lechowicz, 1995). Because plant growth requires a nutritional balance, that is a balance between carbon uptake above-ground and nutrient uptake below-ground, it is has been suggested that nutrient limitation should constrain plant CO2-responsiveness ( Rastetter et al., 1997 ). However, data suggest that proportional responses to elevated CO2 may be greater in some species under low than high nutrient conditions (Lloyd and Farquhar, 1996). Plant evolutionary history also appears to influence CO -responsiveness, as demonstrated by the relationship 2 between life history strategy (sensu Grime, 1977 ) and CO -responsiveness ( Hunt et al., 1993). The latter view 2 is supported by the finding that biome aYnities are more important than photosynthetic type in predicting CO 2 responsiveness ( Wilsey, 1996 ). It is vital to understand the relative roles of these extrinsic (e.g. nutrient limitation) and intrinsic (adaptive) limitations to CO2-responsiveness, as global change involves changes to resource availability ( Vitousek, 1994), but many ecosystems have evolved under resource-limited conditions. Are the mechanisms which control CO2-responsiveness chiefly intrinsic, species-specific and a function of evolutionary history, or are they externally determined and a function of resource concentration? Photosynthetic acclimation (also termed downregulation) is a crucial component of plant productivity responses to elevated CO (Long et al., 1993 ), and many 2 species show diVerent degrees of photosynthetic acclimation in elevated CO2 ( Harley, 1995). Acclimation appears proximately due to either, or both, a reduction in the concentration or activation state of Rubisco (Sage et al., 1989; Jacob et al., 1995). This response may be due to repartitioning of nitrogen resources within photosynthetic cells (Bowes, 1991 ), feedback regulation by carbohydrate status (Stitt, 1991), or inorganic phosphate limitation (van Oosten et al., 1992). Altered carbohydrate status itself may be due to a combination of factors which lead to an altered source:sink balance in elevated CO2, such as low carbohydrate export rate ( Ko rner et al., 1995 ), low sink demand (Arp, 1991 ) or nutrient supply limitation (Paul and Driscoll, 1997 ). This is a complex chain of linked events, and the process of acclimation is poorly understood (Bowes et al., 1996). There is relatively little information about photosynthetic acclimation in species of nutrient-limited mediterranean-type ecosystems (MTEs). Sclerophylldominated South African MTEs are among the most nutrient-limited in the world ( Kruger et al., 1983). Harley (1995) has proposed that CO2-responsiveness in mediterranean schlerophylls will depend on whether new sinks can be developed to capitalize on increased carbohydrate formation in elevated CO2. Stock and Midgley (1995 ) concluded, from the limited available information, that mediterranean-type species with low growth rates might show muted photosynthetic responses to elevated CO2. Empirical studies of CO2-mediated photosynthetic responses of MTE perennials (Larigauderie et al., 1988; Jenkins, 1993, in Oechel et al., 1995; Miglietta et al., 1995; Bettarini et al., 1995) have produced a mixed bag of results including both down- and up-regulation. At the leaf level, both carbohydrate status and leaf nitrogen content have been clearly shown to influence photosynthetic activity, and both are potentially altered in elevated CO2. Plants must have mechanisms to sense carbohydrate status (van Oosten and Besford, 1996 ); the proven diurnal regulation of carbohydrate production (Geiger and Servaites, 1994) demonstrates that these exist, and may serve to regulate photosynthesis in the medium term. Leaf N content also plays a central role in photosynthesis, and is an important trait that covaries widely with photosynthetic capacity throughout the plant kingdom ( Field and Mooney, 1986 ). Does elevated CO2 aVect these measures consistently across species and nutrient concentration conditions? This study attempts to discern direct eVects of nutrient concentration on the photosynthetic response to elevated CO2, as distinct from species-specific eVects, and to tease apart the relative roles of leaf N and carbohydrate status in modifying photosynthetic rate under elevated CO2 conditions. Materials and methods Species selection Four closely related Leucadendron (Proteaceae) species of congruent growth form, but with inherently diVerent nutrient dependencies, were selected. Leucadendron xanthoconus ( Kuntze) K. Schum. and L. laureolum (Lam.) Fourc. (dystrophic species) are associated with acidic sands of low N and P status, L. coniferum (L.) Meisn. (mesotrophic) is associated with neutral sands of intermediate N and higher P availability, and L. meridianum I. Williams (mesotrophic) is associated with basic sands of higher N and intermediate P status (Richards, 1997a, b; Midgley et al., 1995). Members of the genus Leucadendron do not possess mycorrhizae so common in many fynbos genera (Allsopp and Stock, 1993), thus allowing nutrient concentration to be manipulated hydroponically in a sterile sand/perlite culture medium. Plant material Seeds were collected from at least five plants near Cape Town (L. xanthoconus and L. laureolum) and near Cape Agulhas (L. meridianum and L. coniferum) and stored in sealed containers at room temperature. Seeds were germinated in sterile sand, and 68-week-old seedlings transferred to 0.5 m deep pots (3.3 dm3 volume) containing a sterile sand/perlite mix, and allowed to establish for a further 34 months. Plants were fed monthly with 100 ml of a complete Long Ashton solution diluted to 10% (containing 0.1 mM N as nitrate and ammonium), until cotyledonary reserves were exhausted. Plants were then transferred to open-top chambers, and appropriate CO2 and nutrient treatments initiated. Pots were watered daily, receiving approximately 0.5 l d1 each during cool months, and 1.0 l d1 during warm months. Plants were harvested 6 months later, aged between 11 and 12 months (L. xanthoconus and L. laureolum) and 12 and 13 months (L. coniferum and L. meridianum). Harvesting was conducted during the early morning, and plants were subdivided into leaves, stems and roots. Bulked samples per plant were oven-dried to constant mass at 65 C and milled. Open-top chambers Chambers were hexagonal with 0.38 m long sides and 0.5 m tall, constructed of polycarbonate (1.8 mm thickness) and topped by a removable frustum. They were placed on tables in a polycarbonate-clad greenhouse. Each chamber was ventilated individually by a 12 V DC brushless fan (model FP-108, Commonwealth, Taiwan) which drew air from outside and circulated it in a plenum surrounding the base of the chamber before entering through perforations in the inner plenum wall. For elevated CO2 chambers, pure CO2 was bled into the intake pipes at rates controlled by float metering flowmeters (model DK800, Krohne, Germany). Elevated CO2 chambers were individually calibrated to 700 mmol mol1 using an infra-red gas analyser (LI-6200, Li-Cor, Lincoln, NE, USA), and were generally within 80 mmol mol1 of the target concentration during the first experiment (L. xanthoconus and L. laureolum), and within 50 mmol mol1 for the second experiment (L. coniferum and L. meridianum). CO2 concentrations in the ambient chambers were approximately 350 mmol mol1. The ventilation rate of the open-top chambers was controlled at somewhat more than four air changes per min. Pots were suspended through holes in the table tops, thus preventing pot heating and allowing CO2 fumigation of the soil surface and above-ground plant parts only. Application of nutrient treatments Nutrient treatments comprised a complete Long Ashton solution which was diluted to 20% for the high nutrient treatment (containing 0.20 mM nitrogen as ammonium and nitrate), and diluted a further four times for the low nutrient treatment (5% Long Ashton, containing 0.05 mM nitrogen as ammonium and nitrate). Plants were fed 100 ml each once weekly. Non-structural carbohydrate and nitrogen concentrations Foliar sugar and starch concentrations were analysed using a modified phenol-sulphuric acid method (Buysse and Merckx, 1993). A 50 mg dry sample was extracted overnight in 10 ml 80% ethanol (v/v) and the supernatant analysed for total sugars. The residue was boiled for 3 h in 5 ml 2% HCl (v/v) and the supernatant analysed for starch. Absorbance at 490 nm was measured using a Beckman DU640 spectrophotometer (Beckman Instruments Inc., Fullerton, CA, USA). Total leaf nitrogen concentration was determined using micro-Kjeldahl digestion. Gas exchange Plant gas exchange characteristics were sampled after a minimum of 3 months using leaves which had developed after treatment initiation. All determinations were carried out using an LI-6200 portable photosynthesis system (Li-Cor, Lincoln, Nebraska, USA), configured as a closed system. For L. laureolum and L. xanthoconus, recently mature, fully expanded leaves were positioned singly in a 0.25 dm3 cuvette, and for L. coniferum and L. meridianum the terminal portion of the shoot was enclosed in a 1 dm3 cuvette, because leaves were too short to sample singly. Plants were removed from the open-top chambers and measured in the laboratory on the same day. For the duration of the light-response measurements, the enclosed plant parts were maintained at their respective ambient growing CO2 concentration by periodically injecting pure CO2 to replace CO2 removed by photosynthesis. Using a bank of 50 W tungstenhalogen incandescent lamps (Decostar 51, Osram, Germany), plants were gradually brought to light-saturation point, over a period of 2030 min. This was determined empirically for each treatment (speciesCO2nutrient) and was always above 1000 mmol m2 s1 PFD. At this point, Amax and the reported gs were recorded. Light levels were then reduced in steps of about 300 mmol m2 s1 above 200 mmol m2 s1 PFD and about 50 mmol m2 s1 below 200 mmol m2 s1 PFD, and net CO2 exchange rates measured at each level after suYcient acclimation. To determine dark respiration rate, the cuvette was covered with a black cloth for 5 min. Following the light-response measurements, plants were again exposed to saturating PFD until photosynthetic rate was within 5% of Amax. Thereafter, the CO2 level was decreased in steps of about 150 mmol mol1, allowed to stabilize, and measurements of photosynthetic rate taken to construct an A:ci curve. After the CO2 compensation point had been closely approached or exceeded (a process which took about 4060 min), the CO2 concentration was increased to above 1400 mmol mol1. At this point, stomatal conductance had generally increased due to the depleted CO2 concentration in the cuvette, allowing a rapid and substantial increase in ci, and an accurate estimate of the light and CO2-saturated photosynthetic rate, Jmax. The CO2 concentration was maintained at this level for at least 15 min for full stabilization, and then decreased in steps of about 200 mmol mol1 until the ambient growing CO2 concentration was reached. The photosynthetic rate at this point was again checked to be within 5% of Amax. For gas exchange measurements above 50 PFD, cuvette air temperature was typically maintained at 291 C and air vapour pressure at 202 mb. For gas exchange analysis, three individuals of each species were sampled in each treatment (CO2nutrient combination). Response curves were fitted individually to light- and CO2response data for every leaf or shoot sampled, using iterative non-linear regression ( Unistat 4.51 for Windows, Unistat Ltd., London, UK ). A monomolecular hyperbola (Causton and Dale, 1990) was fitted to light-response data. The function is y=a(1ebcx) where y is the rate of CO2 exchange and x is the independent variable (PFD). The coeYcient a gives the light-saturated rate of CO2 exchange (Amax) and apparent quantum yield (a, the slope, or derivative of the curve at x=0) is given by aceb. These parameters were derived individually for every shoot and leaf sampled, and used in statistical analysis. Carbon dioxide response curves were analysed by fitting the model of Farquhar et al. (1980) to the data for each leaf or shoot sampled, using methods described by Hilbert et al. (1991). Photosynthesis was assumed to be either (a) RuBP saturated, or (b) limited by the light-dependent regeneration of RuBP. In the case of (a) the following holds: A=Vc,max (CC )/(C+k)Rd where Vc,max is the maximum RuBPcase activity, C is the intercellullar partial pressure of CO2, Rd is dark respiration rate, C is the CO2 compensation point, and A=J (CC )/(4.5C+10.5C )Rd k=kC (1+O/kO) where kC and kO are the Michaelis-Menten constants for CO2 and O2, and O is the partial pressure of O2 at the site of carboxylation ( Farquhar et al., 1980). In the case of (b) the following holds: J=JmaxI/(I+2.1Jmax) and I is the instantaneous photosynthetic photon flux density. Iterative non-linear regression analysis was used first to derive Vc,max from each A:ci data set for each leaf or shoot. The conditions for (a) were assumed to be met with a ci of less than 200 mmol mol1. Rd was derived from the light-response curve and substituted into equation 2 (Rd ranged between 0.45 mmol m2 s1 and 1.23 mmol m2 s1 with a mean of 0.74 mmol m2 s1); thus only Vc,max and C were unknowns. Values for Jmax were derived by fitting equation 4 to A5ci data where ci exceeded approximately 300 mmol mol1 (substituting for J according to equation 5 ). Statistical design There were eight open-top chambers in the experimental array, each of which held four individuals of each of two species (eight plants in each chamber, 64 plants in total ). Thus elevated CO2 treatments were replicated four times. Within each opentop chamber, high and low nutrient treatments were replicated twice for each species. The experiments used a split-plot design, giving six degrees of freedom to test for elevated CO2 eVects, and 22 degrees of freedom for nutrient eVects. The allocation of species and treatments within each open-top chamber was formally randomized a priori. Results for each species were analysed separately using split-plot ANOVA. Not all plants could be sampled for gas exchange characteristics due to time constraints, and a subset of three plants per treatment was sampled at random from the open-top chambers. Gas exchange data were tested statistically using standard analysis of variance. Correlations between variables were identified by linearregression, and significant diVerences in regression slopes and intercepts due to elevated CO2 tested using analysis of covariance. All statistical procedures were carried out using Unistat 4.51 for Windows. Gas exchange measurements were carried out above light saturation level (PFD>1000 mmol m2 s1), with cuvette air temperature 291 C, cuvette air vapour pressure 202 mb. L. xanthoconus L. laureolum Results Main CO2 effects and interactions Elevated CO2 had significant negative eVects on Vc,max and Jmax ( Tables 1, 2, 3 ). Elevated CO2 reduced Vc,max in all species under both nutrient treatments, and reduced Jmax in all species at both nutrient concentrations, with two exceptions at the low nutrient concentration (L. xanthoconus and L. coniferum, Fig. 1 ). Elevated CO2 had positive eVects on Amax, PNUE and starch concentration, but did not aVect sugar concentration, leaf N concentration or gs ( Tables 1, 2, 3 ). There was no significant CO nutrient interaction for any measured 2 response ( Table 3 ) and thus the DCO2 values (relative eVect of elevated CO ) presented do not diVerentiate 2 between nutrient concentrations. There was significant CO2species interaction only for gs, as two of the four species showed no gs response to elevated CO2 but L. meridianum decreased gs and L. coniferum increased gs in elevated CO2 ( Tables 1, 2, 3; Fig. 1 ). Nutrientspecies interaction was significant for leaf N and sugar concentrations ( Table 3 ). Nutrient and species effects The higher nutrient concentration significantly increased all measures except starch content, which it significantly decreased ( Tables 1, 2, 3 ). Species diVered significantly in all variables measured ( Table 3 ). Dystrophic species displayed higher Amax, Vc,max, Jmax, and leaf [ N ] ( Table 1 ), and mesotrophic species higher gs, sugar and Nutrient level CO2 level (mmol mol1) (mmol m2 s1) Gas exchange measurements were carried out above light saturation level (PFD>1000 mmol m2s1), with cuvette air temperature 291 C, cuvette air vapour pressure 202 mb. L. meridianum L. coniferum Results are for three-way ANOVA of the combined data sets given in Tables 1 and 2. Nutrient eVect Species eVect CO2nutrient Nutrientspecies CO2species Nutrient level CO2 level (mmol mol1) (mmol m2 s1) **P<0.001; *P<0.01. starch levels ( Table 2 ). Leaf nitrogen levels of dystrophic more responsive to increased concentration than those of mesotrophic species, while sugar levels of mesotrophic species were more responsive nutrient concentration than those of dystrophic species (resulting in significant nutrientspecies interacbut the intercept was significantly increased in elevated CO (P<0.002 ), and the regressions diVered significantly 2 (P<0.005 ). Both J and V were significantlymax c,max positively correlated with leaf [ N ] ( Fig. 3 ), but regressions diVered between CO treatments only for V 2 c,max (P<0.05), due to a significantly increased intercept in elevated CO (P<0.01 ). V and J were significantly 2 c,max max co-correlated in both ambient and elevated CO -grown 2 plants ( Fig. 4 ), but these regressions barely diVered signiCO (P<0.02 ). 2 Both J and V were significantly negatively cor max c,max related with leaf starch concentrations, regardless of CO 2 treatment ( Fig. 5 ). The coeYcient of variation of the V :starch correlation was increased by expressing leaf c,max carbohydrate status on a leaf dry mass basis ( Fig. 5 insert). Both V and J c,max max leaf sugar concentration (data not shown). were poorly correlated with Fig. 2. Relationship between leaf [N ] and Amax in juvenile individuals of four Leucadendron species adapted to soils of diVerent nutrient status, grown under two contrasting soil nutrient regimes and two atmospheric CO2 levels. Each symbol is a two way mean of eight nutrient values and three Amax values, with bars representing standard errors. Solid circles represent means for plants grown at 700 mmol mol1 CO2, and triangles those grown at 350 mmol mol1 CO2. Discussion Nutrient-induced and species differences in assimilation rate and CO2-responsiveness As expected, higher nutrient concentration led to generally increased leaf nutrient status and associated higher photosynthetic rates and stomatal conductance, patterns often reported in the literature ( Field and Mooney, 1986; Evans, 1989). The significant species diVerences for all leaf nutrient and photosynthetic measures were also expected. Unexpectedly, however, dystrophic species had higher foliar N concentrations under both nutrient regimes than did mesotrophic species, and higher total plant N (data not shown). This suggests that the dystrophic species had an inherently higher nutrient uptake capacity than did mesotrophic species. Soil factors, rather than competitive interactions seem to explain species/soil specificity in nature among these and other Proteaceous species ( Richards et al., 1997a). It is possible that diVerences in nutrient uptake capacity may determine these patterns. Even though species photosynthetic characteristics diVered significantly, the photosynthetic response of all four species to both increased nutrient concentration and elevated CO2 was similar, as there were no speciesCO2 and speciesnutrient interactions. Also, the lack of CO2nutrient interaction suggests that nutrient concentration did not alter photosynthetic CO2-responsiveness. If either species-specific characteristics or nutrient concentration were important determinants of CO2-responsiveness, then significant interaction of these factors with CO2 level would be expected. The mechanisms which control photosynthetic CO2responsiveness in these species, therefore, do not seem to be primarily a function of nutrient concentration (i.e. not Fig. 4. Relationship between Vc,max and Jmax in juvenile individuals of four Leucadendron species adapted to soils of diVerent nutrient status, grown under two contrasting soil nutrient regimes and two atmospheric CO2 levels. Each symbol is a two-way mean of three Jmax and Vc,max values, with bars representing standard errors. Solid circles represent means for plants grown at 700 mmol mol1 CO2, and triangles those grown at 350 mmol mol1 CO2. chiefly extrinsic). These mechanisms do not appear to be species-specific either, and so are independent of the recent evolutionary divergence that accompanied the development of their association with distinct soil types. Leaf nitrogen content and photosynthetic capacity Nitrogen is a central determinant of leaf photosynthetic capacity (Amax), and Amax is correlated with leaf [N ] across the plant kingdom ( Field and Mooney, 1986; Evans, 1989; Woodward and Smith, 1994). This study is consistent with that pattern, but reveals a significant increase in the eYciency of nitrogen use in elevated CO2 represented by the increased intercept of the linear leaf Amax:[N ] relationship ( Fig. 2). This study also shows a related and clearly demonstrable increase in PNUE of high CO2-grown plants ( Table 2), and further that the principal determinants of Amax, namely carboxylation eYciency (i.e. Vc,max) and RuBP regeneration capacity (Jmax), were also significantly correlated with leaf [N ] ( Fig. 3). Furthermore, Vc,max and Jmax were consistently co-correlated in the four selected species ( Fig. 4). This pattern is virtually identical to that identified for a broad range of species ( Wullschleger, 1993; Leuning, 1997 ), and is thought to represent optimal distribution of nitrogen between light-harvesting and carboxylation functions ( Lloyd and Farquhar, 1996 ). Thus the dependence of Amax on leaf [N ] seems to be largely due to a consistent partitioning ratio of N resources that may be more or less conserved across the plant kingdom, and the ubiquitous observation of increased PNUE in elevated CO2 (Drake et al., 1997) is the manifest result. It is perhaps surprising that there was no apparent N repartitioning between carboxylation and RuBP regeneration in elevated CO2, given that this would boost plant nitrogen-use eYciency even more than occurs by the reduction in photorespiration alone (Bowes, 1991; Sage, 1994). Nitrogen repartitioning in elevated CO2 has been reported in only very few studies, such as for loblolly pine ( Tissue et al., 1993 ). Nitrogen repartitioning of this type often accompanies the process of shade adaptation ( Woodward, 1990 ), but is probably triggered by a change in light quality, and not carbohydrate availability. Nitrogen allocation at canopy level approximates optimality with respect to carbon assimilation ( Field, 1983; Pons et al., 1993 ), but a lack of repartitioning at leaf level in elevated CO2 suggests that current allocation patterns may not be optimal at future higher CO2 levels (Lloyd and Farquhar, 1996 ). In fact, optimal partitioning of N between carboxylation and light-harvesting functions may be tuned to lower than current ambient atmospheric CO2 levels, found prior to the industrial revolution (~270 mmol mol1 CO2), as appears from a model for Amazonian rainforest (Lloyd et al., 1995). It is possible that species with greater plasticity of nitrogen partitioning in response to carbohydrate availability (such as loblolly pine, Tissue et al., 1993 ), or genotypes with variable partitioning ratios, might be favoured as CO2 continues to rise. Certainly, CO2 responsiveness has been shown to diVer between ecotypes (Norton et al., 1995 ) and may vary, heritably, between genotypes (Curtis et al., 1994). Photosynthetic acclimation in elevated CO2 An understanding of photosynthetic acclimation processes in elevated CO2 remains elusive (van Oosten and Besford, 1996 ). Although responses at leaf level are diverse, some generalizations can be made for woody species (Gunderson and Wullschleger, 1994 ). Net carbon uptake rate measured at growth [CO2] is stimulated under elevated CO2 by roughly 45%, even though the net CO2 uptake rate of elevated CO2-grown plants is 21% lower than that of ambient CO -grown plants when measured 2 at elevated CO2 concentration (Gunderson and Wullschleger, 1994). The results of this study are consistent with these generalizations. The mechanism most commonly implicated in acclimation is feedback regulation of carboxylation by carbohydrate accumulation (Stitt, 1991; van Oosten et al., 1994; Jacob et al., 1995 ). Sugar repression of photosynthesis has been identified as a general trigger for the regulation of photosynthesis in response to changes in sink demand (van Oosten and Besford, 1996). Although relationships have been shown between carbohydrate status and photosynthetic measures in high CO -grown 2 plants, results from elevated CO2 studies are contradictory (Paul and Driscoll, 1997). Photosynthetic acclimation in this study comprised apparently synchronized reductions in both carboxylation and RuBP regeneration capacity, and not only the often-cited reduction in carboxylation capacity. Feedback regulation of RuBP regeneration capacity has not received the same emphasis as short-term photosynthetic acclimation in response to tissue carbohydrate status (van Oosten and Besford, 1996) involving regulation of carboxylation capacity. The consistent negative relationship found in the current study between starch accumulation and both Vc,max and Jmax ( Fig. 5) identifies the central role of carbohydrate accumulation (which responded to both elevated CO2 and nutrient concentration) in photosynthetic regulation. This pattern suggests a general mechanism of photosynthetic regulation in response to both nutrient and carbohydrate concentration, supporting the contention that photosynthetic responses to nutrient deficiency are almost identical to those to elevated CO2 (Paul and Driscoll, 1997 ). This suggests that elevated CO emulates an excess 2 end-product condition, triggering photosynthetic downregulation in a response which may have evolved under conditions of source:sink imbalance, such as periodic nutrient limitation. This response is likely to be particularly well-developed in species subject to periodic episodes of nutrient stress and, therefore, may have an important genetic component (Sage et al., 1989). Plant species diVer in their propensities for accumulating starch relative to sugars, and there is an identified need for a better understanding of how starch- versus sugar- accumulating species respond to elevated CO2 (Bowes et al., 1996). Species studied here showed roughly 5-fold greater starch than sugar concentrations, and starch status seemed more important in photosynthetic feedback regulation than in many other studies. Carbohydrate relations in source leaves are regulated both on a shortterm (instantaneous and diurnal ) basis ( Fondy et al., 1989; Geiger and Servaites, 1994) and on a longer term basis reflected by the baseline (morning) total nonstructural carbohydrate ( TNC ) concentration. Jacob et al. (1995 ) reported increased sugar concentrations in photosynthetically-acclimated high CO2-grown Scirpus olneyi only at midday, whereas both baseline and midday starch concentrations were higher. This would support the suggestion that baseline TNC status, rather than the more ephemeral diurnal sugar fluctuation, is the cue for in vivo photosynthetic downregulation in elevated CO2 in the medium- to long-term. Acknowledgements The World Wildlife Fund (South Africa) is gratefully acknowledged for the funds provided through the Roland and Leta Hill Trust for greenhouse facilities. Nic Bennet, Deryck de Witt, Barry Jagger, and Stanley Snyders provided technical assistance.


This is a preview of a remote PDF: https://jxb.oxfordjournals.org/content/50/333/533.full.pdf

G.F. Midgley, S.J.E. Wand, N.W. Pammenter. Nutrient and genotypic effects on CO2-responsiveness: photosynthetic regulation in Leucadendron species of a nutrient-poor environment, Journal of Experimental Botany, 1999, 533-542, DOI: 10.1093/jxb/50.333.533