TY - JOUR
AU - Beerling, David, J
AB - Abstract Background and Aims By the year 2100, atmospheric CO2 concentration ([CO2]a) could reach 800 ppm, having risen from ~200 ppm since the Neogene, beginning ~24 Myr ago. Changing [CO2]a affects plant carbon–water balance, with implications for growth, drought tolerance and vegetation shifts. The evolution of C4 photosynthesis improved plant hydraulic function under low [CO2]a and preluded the establishment of savannahs, characterized by rapid transitions between open C4-dominated grassland with scattered trees and closed forest. Understanding directional vegetation trends in response to environmental change will require modelling. But models are often parameterized with characteristics observed in plants under current climatic conditions, necessitating experimental quantification of the mechanistic underpinnings of plant acclimation to [CO2]a. Methods We measured growth, photosynthesis and plant–water relations, within wetting–drying cycles, of a C3 tree (Vachellia karroo, an acacia) and a C4 grass (Eragrostis curvula) grown at 200, 400 or 800 ppm [CO2]a. We investigated the mechanistic linkages between trait responses to [CO2]a under moderate soil drying, and photosynthetic characteristics. Key results For V. karroo, higher [CO2]a increased assimilation, foliar carbon:nitrogen, biomass and leaf starch, but decreased stomatal conductance and root starch. For Eragrostis, higher [CO2]a decreased C:N, did not affect assimilation, biomass or starch, and markedly decreased stomatal conductance. Together, this meant that C4 advantages in efficient water-use over the tree were maintained with rising [CO2]a. Conclusions Acacia and Eragrostis acclimated differently to [CO2]a, with implications for their respective responses to water limitation and environmental change. Our findings question the carbon-centric focus on factors limiting assimilation with changing [CO2]a, how they are predicted and their role in determining productivity. We emphasize the continuing importance of water-conserving strategies in the assimilation response of savannah plants to rising [CO2]a. Acacia, Vachellia karroo, Eragrostis curvula, C4 photosynthesis, elevated CO2, global change, hydraulics, low CO2, savannah, starch, stomata, thorns Introduction All photosynthetic organisms use the same ancestral C3 photosynthesis pathway, in which Rubisco catalyses separate reactions with CO2 (carboxylation) and O2 (oxygenation). While carboxylation fixes carbon into sugars and produces O2 as a by-product, oxygenation results in photorespiration in which previously fixed carbon is oxidized to CO2. In warm environments, under current atmospheric CO2:O2 concentrations, photorespiration can release 25–30 % of previously fixed CO2 (Bellasio et al., 2014a), necessitating metabolically costly recapture of CO2 and regeneration of toxic by-products (Eckardt, 2005). When atmospheric CO2 concentration ([CO2]a) is low, leaf temperatures are above 25–30 °C and water is limiting, photorespiration can completely offset carbon fixation, impairing growth and making C3 plants increasingly reliant on stored carbon to fuel basic metabolism (Walker et al., 2016). Demand for stored carbon is greatest under low [CO2]a and when stomata close to reduce water losses, thereby reducing CO2 assimilation (A) (Hartmann et al., 2013; Mitchell et al., 2013; Quirk et al., 2013, 2018). The decline in [CO2]a towards a lower plateau of ~200 ppm ~24 Myr ago (Beerling and Royer, 2011) was followed by a climatic shift to greater seasonal aridity driven by mountain uplift, altered monsoons and expansion of Antarctic ice sheets (Zachos et al., 2001, Zhisheng et al., 2001). This provided the ecological setting for low-latitude forest retreat and the rise of flammable savannah grassland ecosystems scattered with fire-suppressed trees (Edwards et al., 2010; Bond and Midgley, 2012). C4 plants evolved an augmentation to C3 photosynthesis consisting of a biochemical CO2-concentrating mechanism. This mechanism concentrates CO2 around Rubisco, allowing C4 plants to operate at lower stomatal conductance (gS) and transpiration (E), giving higher instantaneous photosynthetic water-use efficiency (WUE = A/E) compared with C3 plants, particularly in warm environments and under low [CO2]a (Ward et al., 1999; Anderson et al., 2001; Seibt et al., 2008; Taylor et al., 2011; Cunniff et al., 2016). Selection pressure to reduce photorespiration under high evaporative demand and low [CO2]a promoted the evolution of lineages that subsequently played a role in the expansion of C4-grass-dominated savannahs at the expense of closed woodland 8–10 Myr ago (Cerling et al., 1997; Edwards et al., 2010; Hoetzel et al., 2013). With anthropogenic CO2 emissions, [CO2]a has risen above 400 ppm and is set to reach 600–800 ppm before the year 2100 (IPCC, 2013). Individual plants may respond to rising [CO2]a with plasticity mechanisms (Leakey and Lau, 2012), resulting in acclimation phenotypes (Morgan et al., 2001), which are generally not heritable (but see e.g. Chinnusamy and Zhu, 2009). Over longer timescales, plant populations may evolve heritable responses, known as adaptations, that are not addressed here. Experimental evidence indicates that acclimation may not keep pace with projected [CO2]a rise (Hoffmann and Sgro, 2011), as plants may be approaching their potential capacity to respond to rising [CO2]a (Polley et al., 1993, 2002; Ghannoum et al., 2000, 2010; Gill et al., 2002; Gerhart and Ward, 2010; Gerhart et al., 2011; Medeiros and Ward, 2013; Quirk et al., 2013; Pinto et al., 2014; Cunniff et al., 2016). Nevertheless, acclimation to rising [CO2]a will depend in large part on photosynthetic type, with C3 plants expected to benefit more than C4 from rising [CO2]a through declines in gS, increased WUE and reduced demand for starch accumulation and use (Morgan et al., 2011). However, meta-analyses suggest that relative effects of rising [CO2]a will diminish (Makino and Mae, 1999; Temme et al., 2013) as indirect physiological factors become limiting to photosynthesis and growth (Bellasio et al., 2018), and evidence for the relative effect on C4 photosynthesis is equivocal (Makino and Mae, 1999; Temme et al., 2013). Savannahs are a highly dynamic and globally important biome governed by competitive interactions between trees and C4 grasses. Savannahs evolved on every continent under low [CO2]a and are thought to be highly responsive to [CO2]a-related effects on plant–water relations. Savannahs are prone to disturbance-driven vegetation transitions between open C4-dominated grasslands and closed woodland as [CO2]a, climate, rainfall patterns and water availability change (Bond and Midgley, 2012; Staver et al., 2017). However, the effect of increasing [CO2]a on the relative performance of savannah trees and C4 grasses is uncertain (Midgley and Bond, 2015). This uncertainty has critical implications for our understanding of vegetation dynamics in savannahs and impinges on our capacity to model past and future directional trends in vegetation patterns (Bond, 2008; Hirota et al., 2011; Staver et al., 2011; Moncrieff et al., 2014; Midgley and Bond, 2015). Empirical data characterizing the responses of different savannah plants to water limitation and changing [CO2]a are needed to predict directional trends in savannah vegetation. Presently, no large-scale models include representation of the physiological acclimation to future CO2 level (Rogers et al., 2017). In particular, we need to better understand how savannah species respond to changing [CO2]a and water availability, and model inputs need to be derived with dedicated experiments in which key C3 and C4 species are grown at different [CO2]a levels for extended periods (Makino and Mae, 1999). Here, we aimed to characterize the comparative responses of a nitrogen-fixing C3 tree and a C4 grass to rising [CO2]a by assessing acclimation of photosynthetic characteristics and growth, leaf water status, and leaf and root starch accumulation. To provide initial insights into the question of acclimation responses for two species of different growth form, photosynthetic pathways and life histories (sexual versus clonal) to environmental change, we investigated a representative C3 savannah tree (the acacia Vachellia karroo, a nitrogen-fixing legume) and a C4 savannah grass [Eragrostis curvula, of the NAD-malic enzyme (NAD-ME) C4 subtype] that both evolved under low [CO2]a. Plants were grown at 200, 400 or 800 ppm [CO2]a for 20 months in replicated controlled-environment chambers and subjected to moderate wetting–drying cycles, enabling us to study plant traits without the confounding influence of drought stress responses (e.g. chlorosis, photoinhibition and leaf senescence). We measured growth, starch accumulation, plant physical defences, photosynthesis, plant–water relations and stomatal characteristics. We also performed comprehensive gas exchange experiments to derive parameters of photosynthetic capacity. MATERIALS AND METHODS Plants and growth conditions Seeds of Vachellia karroo obtained from the Desert Legume Program (Tucson, AZ, USA) and Eragrostis curvula obtained from the Germplasm Resources Information Network (United States Department of Agriculture, Washington D.C., USA) were germinated and grown in six controlled-environment growth chambers (BDW40, Conviron, Manitoba, Canada). Growth chambers (two per [CO2]a treatment) were maintained at three [CO2]a levels (200, 400 and 800 ppm) and otherwise constant conditions of 26/17 °C and 70/50 % relative humidity (day/night). A 12-h photoperiod with midday photosynthetic photon flux density (PPFD) at canopy level of >800 μmol m−2 s−1 was provided from a 3:1 mix of 39-W white fluorescent tubes (Master TL5, Philips, Eindhoven, the Netherlands) and 39-W red–blue fluorescent tubes (Grolux T5, Havells-Sylvania, Newhaven, UK), augmented with six 105-W halogen light bulbs (product code 0022114, Havells-Sylvania). Low [CO2]a was achieved by routing the chamber air supply through a scrubbing unit packed with soda lime (Sofnolime, 1–2 mm, Molecular Products, Essex, UK) and [CO2]a was monitored using a CO2 sensor (Carbocap GMP242, Vaisala, Finland) linked to a feedback system regulating the air inlet between the scrubbing unit and ambient air. The 400-ppm [CO2]a treatment was supplied with ambient air, and 800-ppm [CO2]a was achieved by supplementing air with automated injections of CO2. Chamber [CO2]a was periodically checked with an infrared gas analyser (IRGA, LI6400XT, LI-COR, Lincoln, NE, USA). After germination, acacia seedlings and clones of Eragrostis produced by splitting a randomly selected plant were potted-on into 2.5-dm3 pots (n = 8), randomly distributed between the growth chambers, and grown in three-parts commercial loam-free topsoil (Boughton Ltd, Kettering, UK) plus one-part John Innes No. 3 compost (John Innes Manufacturers Association, Reading, UK). Pots were watered gravimetrically to 80 % of pot capacity three times per week after 24–32 photoperiod hours since last watering to maintain a cycle of soil wetting and moderate drying. This meant that at a given point in time between waterings, soil water availability was a function of canopy transpiration and plant water use, which will have varied between species and [CO2]a. Three full sets of measurements and tissue sampling took place over 6 weeks (Supplementary Data Fig. S1) following an initial 18-month growth and acclimation period. After 20 months, plants were harvested and above- and below-ground biomass, and acacia thorns and viable sapwood, were measured. Throughout their growth, plants were rotated weekly within, and monthly between, cabinets, with environmental settings to minimize block effects. Starch content of leaves and roots Starch was determined with the starch-specific protocol of Bellasio et al. (2014b), which minimizes interference from cell wall and matrix components. Roots and leaves were sampled three times over 6 weeks and these triplicate technical replicates were averaged (Supplementary Data Fig. S1). Leaves were randomly selected from the canopy and roots were extracted and rinsed from two soil cores, 2 cm in diameter and 8 cm deep, taken from each pot. Once roots were removed, the extracted soil was used to refill the core. Fresh samples were microwaved immediately to stop enzymatic activity, then oven-dried at 80 ºC for 48 h and ground with a mixer mill and a 50-mm steel jar set (MM 400, Retsch, Haan, Germany). Pulverized samples were first digested within a water bath at 100 ºC for 12 min with pure thermostable α-amylase (E-BLAAM, Megazyme International, Bray, Ireland) in 0.1 m sodium acetate buffer (pH 5.0). Samples were cooled then digested further in a water bath at 50 ºC for 45 min with pure amyloglucosidase from Aspergillus niger (E-AMGDF, Megazyme International), under continuous agitation with an 8-mm magnetic stir-bar added to each tube. The resulting glucose was quantified in aliquots of the supernatant with a peroxidase–glucose oxidase enzyme kit (PGO, Product No. P7119, Merck, Darmstadt, Germany), which quantitatively oxidizes the dye, ortho-dianisidine. The reaction was carried out in a 37 ºC water bath and stopped after 45 min using 75 % sulphuric acid to shift the absorption peak away from plant pigment interference and stabilize colour development (Chow and Landhäusser, 2004), which was read at 530 nm on a spectrophotometer (Cecil 9000, Cecil Instruments, Cambridge, UK). To correct for any day-drift, an internal reference comprising a homogenized mixture of all samples (roots and leaves of both species) was added in quadruplicate to every batch of samples (Bellasio et al., 2014b). Leaf gas exchange Instantaneous leaf gas exchange at midday was measured under operational environmental conditions within the growth chambers after 11–14 photoperiod hours since last watering on young, fully expanded leaves. For Eragrostis, we aligned several leaves without overlapping their edges to fill the cuvette, and for acacia we photographed leaves in situ using a bespoke leaf mount and determined fresh leaf area using ImageJ (NIH, Bethesda, MD, USA), with which the gas exchange data were adjusted during measurement. The IRGA was fitted with a 6-cm2 cuvette and a LI-COR 6400-02B light source. Reference CO2 concentration was 200, 400 or 800 µmol mol−1, block temperature was 26 °C and flow rate was 235 μmol s−1. After stabilization, gas exchange measurements were averaged for 10 s and recorded as a single point measurement. Assimilation responses to [CO2] in the sub-stomatal cavity (Ci) and PPFD (A–Ci and A–PPFD response curves) were determined three times per plant over 6 weeks on the same (acacia) or similar (Eragrostis) leaves as those used for instantaneous, operational measurements. The A response curves were measured at the bench with the IRGA described above after ~11 photoperiod hours since last watering (Supplementary Data Fig. S1). Mass flow leaks (Boesgaard et al., 2013) were monitored with a gas flow meter (Bellasio et al., 2016b) and sealed with non-toxic, water-based adhesive. The IRGA was supplied with humidified ambient air adjusted to 60–70 % relative humidity, whilst CO2 was supplied from cartridges (Liss-Group, Répcelak, Hungary). Block temperature was 26 °C and flow rate was 235 μmol s−1. For the A–Ci curves, PPFD was 1500 μmol m−2 s−1. Leaves were allowed 30–60 min to reach full photosynthetic induction before an automated A–Ci routine was launched, comprising 11 reference CO2 increments between 20 and 1200 ppm with 2–3 min between increments. The sample and reference cells were matched before each measurement. For the A–PPFD curves, reference [CO2] was 200, 400 or 800 μmol mol−1 according to experimental treatment. Leaves were given 30–60 min to reach full photosynthetic induction before an automated routine was launched, comprising ten increments between 1500 and 0 μmol m−2 s−1, with 5–7 min between increments, and matching of the sample and reference IRGAs before each measurement. Primary data were corrected for CO2 diffusion and Ci was recalculated after Bellasio et al. (2016b). Plant–water relations Transpiration (Eop) and gSop were taken from in-cabinet, operational gas exchange measurements (denoted by subscript op) and WUE was calculated as WUE = Aop/Eop (µmol CO2/mmol H2O). Leaf relative water content (RWC), leaf water potential at midday (Ψleaf) and predawn (Ψpd) (a proxy for soil water potential, Ψsoil) and predawn leaf solute potential (Ψsolute) were measured on leaves cut on the day and night indicated by the schedule in Supplementary Data Fig. S1. For RWC, the fresh, turgid and oven-dried (48 h; 60 °C) leaf weights were determined after Ghannoum et al. (2002); turgidity was achieved by submerging cut leaves in distilled water within a sealed environment for 3–4 h (Ghannoum et al., 2002). Leaf water potentials were measured using a pressure chamber (Model 1000, PMS Instrument Company, Albany, OR, USA). Leaf Ψsolute was obtained by freezing leaves then squeezing extracts onto 5-mm-diameter filter paper discs (Whatmann No.1, GE Healthcare, Little Chalfont, UK). Osmolality was measured from the discs using a psychrometer (Psypro with L-51 hygrometers, Wescor, Logan, UT, USA) calibrated with five standard NaCl solutions according to the manufacturer’s instructions, and always allowing a fixed equilibration time of 20 min. Plant hydraulic conductance (Kplant) was calculated as Kplant = Eop∕(Ψpd − Ψleaf) and leaf turgor was calculated as turgor = Ψpd − Ψsolute (Fini et al., 2013). Acacia sapwood staining Acacia branch sections (~7 cm long) were recut underwater to ensure open xylem conduits. The bark was removed from the proximal end and connected to a PVC pipe attached to a reservoir, from which 1 mg L−1 of 0.2-µm-filtered safranin O was delivered from a hydrostatic pressure head of 100 kPa for >16 h (Srebotnik and Messner, 1994). Stained branches were cross-cut into 1-mm-thick slices, and their distal sections were imaged at ×10 magnification (Supplementary Data Fig. S2). The area of stained sapwood was determined using ImageJ, and the mean ratio of stained sapwood to total sapwood area was calculated for slices corresponding to 4, 5 and 6 mm from the point of stain introduction. Thorn measurement The lengths of the five longest thorns on each of the three longest branches per tree were measured along with the number of thorns per unit branch length and total mass of harvested thorns. Stomatal morphology The size and density of stomata on the abaxial side of three acacia and Eragrostis leaves per plant (n = 8) were measured using ImageJ and digital images of epidermal peels (Supplementary Data Figs S3 and S4) obtained at ×200 and ×400 magnification (Franks et al., 2012a). Leaf nitrogen Shoot samples from each plant were pooled, powdered and homogenized. Leaf carbon/nitrogen ratios (C:N) were then measured on duplicate sub-samples (~100 mg) using a C:N analyser (Vario EL Cube, Elementar, Hanau, Germany) in accordance with the manufacturer’s instructions. Following the measurement period, plants were harvested and their biomass components were weighed. Gas exchange data analysis and parameterization Sets of fitted enzyme- and light-limited photosynthetic parameters were derived using A-response curves (Supplementary Data Fig. S5, Table 1). A first set of characteristics was derived through empirical modelling, which has the benefit of not requiring assumptions (Bellasio et al., 2016b). These were respiration in the light, RLIGHT; initial quantum yield for CO2 fixation, Y(CO2)LL; light-saturated rate of gross assimilation, GASAT; carboxylation efficiency, CE; the CO2-saturated rate of assimilation, ASAT; the empirical curvature of the A–PPFD and A–Ci curves, m and ω, respectively; and the Ci–A CO2 compensation point, Γ. The maximum rate of phosphoenolpyruvate (PEP) carboxylation for Eragrostis (VPMAX) and the maximum Rubisco carboxylation rate for acacia (VCMAX) were derived mechanistically within the framework of Bellasio et al. (2016a and b, respectively). We made the following assumptions: (for the enzyme-limited C3 model) gM = 0.2 mol m−2 s−1, KC(1 + O/KO) = 550 µbar (Ethier and Livingston, 2004); (for the enzyme-limited C4 model) α = 0.15, gM = 1 mol m−2 s−1, γ* = 0.000193, RM/RLIGHT = 0.5, KC = 650 µbar, KO = 450 000 µbar, KP = 80 µbar (all from von Caemmerer, 2000) and gBS = 0.0015 mol m−2 s−1 (Bellasio and Griffiths, 2014). Table 1. Fitted photosynthetic parameters derived from A–PPFD and A–Ci response curves. Mean values (s.d.), n = 6–8, for fitted parameters. Values sharing the same letter within each row are not statistically different at α = 0.05 based on two-way ANOVA (one-way for VPMAX) testing for effects of plant (acacia or Eragrostis) and [CO2]a (200, 400 or 800 ppm), with post hoc pairwise comparisons. Main statistical effects are given in Supplementary Data Table S2 Symbol Term Unit Atmospheric CO2 concentration Acacia (C3) Eragrostis (C4) 200 ppm 400 ppm 800 ppm 200 ppm 400 ppm 800 ppm ASAT CO2-saturated assimilation measured in A–Ci curve µmol m−2 s−1 25.4 (2.23) ab 24.0 (8.04) ab 26.7 (3.05) a 21.8 (4.51) ab 17.5 (1.98) b 19.5 (4.03) ab CE Initial slope of the A–Ci curve Dimensionless 0.11 (0.02) c 0.09 (0.04) c 0.09 (0.03) c 0.38 (0.11) a 0.29 (0.07) ab 0.25 (0.05) b Γ Ci–A compensation point, i.e. Ci where A = 0 µmol mol−1 49.1 (1.56) a 52.6 (4.98) a 51.9 (8.95) a 1.49 (1.64) b 3.85 (3.13) b 3.99 (2.77) b ω Empirical curvature of the A–Ci curve Dimensionless 0.40 (0.36) b 0.69 (0.19) ab 0.60 (0.34) ab 0.75 (0.13) ab 0.78 (0.13) ab 0.80 (0.06) a GASAT Light-saturated gross assimilation at [CO2]a of light curve µmol m−2 s−1 7.66 (0.94) c 17.5 (8.17) bc 30.2 (7.67) a 20.4 (5.05) b 20.1 (2.91) b 21.6 (4.69) ab JSAT Light-saturated electron transport rate at [CO2]a of light curve µmol m−2 s−1 122 (45.9) a 128 (40.3) a 139 (33.0) a 157 (39.5) a 153 (68.6) a 154 (43.0) a LCP PPFD–A compensation point, i.e. PPFD where A = 0 µmol m−2 s−1 15.7 (7.07) bc 13.1 (6.82) c 12.1 (5.25) c 25.1 (5.92) ab 30.1 (3.94) a 30.5 (6.25) a m Empirical curvature of the A–PPFD curve Dimensionless 0.60 (0.12) a 0.28 (0.33) a 0.18 (0.20) a 0.75 (0.13) a 0.78 (0.13) a 0.80 (0.06) a RLIGHT Respiration in the light/day >0 µmol m−2 s−1 0.43 (0.09) b 0.53 (0.3) b 0.72 (0.34) b 1.25 (0.26) a 1.37 (0.22) a 1.38 (0.32) a VCMAX CO2-saturated Rubisco carboxylation rate* µmol m−2 s−1 123 (24.8) a 88.4 (11.4) a 80.9 (25.1) a 41.0 (11.7) b 38.2 (17.2) b 38.6 (10.8) b VPMAX CO2-saturated PEPC carboxylation rate (C4 only) µmol m−2 s−1 – – – 52.5 (16.4) a 36.0 (8.97) b 31.7 (8.32) b Y(CO2)LL Initial (or maximum) quantum yield for CO2 fixation Dimensionless 0.03 (0.01) b 0.05 (0.02) ab 0.06 (0.01) a 0.05 (0.01) ab 0.05 (0.002) ab 0.05 (0.01) ab Symbol Term Unit Atmospheric CO2 concentration Acacia (C3) Eragrostis (C4) 200 ppm 400 ppm 800 ppm 200 ppm 400 ppm 800 ppm ASAT CO2-saturated assimilation measured in A–Ci curve µmol m−2 s−1 25.4 (2.23) ab 24.0 (8.04) ab 26.7 (3.05) a 21.8 (4.51) ab 17.5 (1.98) b 19.5 (4.03) ab CE Initial slope of the A–Ci curve Dimensionless 0.11 (0.02) c 0.09 (0.04) c 0.09 (0.03) c 0.38 (0.11) a 0.29 (0.07) ab 0.25 (0.05) b Γ Ci–A compensation point, i.e. Ci where A = 0 µmol mol−1 49.1 (1.56) a 52.6 (4.98) a 51.9 (8.95) a 1.49 (1.64) b 3.85 (3.13) b 3.99 (2.77) b ω Empirical curvature of the A–Ci curve Dimensionless 0.40 (0.36) b 0.69 (0.19) ab 0.60 (0.34) ab 0.75 (0.13) ab 0.78 (0.13) ab 0.80 (0.06) a GASAT Light-saturated gross assimilation at [CO2]a of light curve µmol m−2 s−1 7.66 (0.94) c 17.5 (8.17) bc 30.2 (7.67) a 20.4 (5.05) b 20.1 (2.91) b 21.6 (4.69) ab JSAT Light-saturated electron transport rate at [CO2]a of light curve µmol m−2 s−1 122 (45.9) a 128 (40.3) a 139 (33.0) a 157 (39.5) a 153 (68.6) a 154 (43.0) a LCP PPFD–A compensation point, i.e. PPFD where A = 0 µmol m−2 s−1 15.7 (7.07) bc 13.1 (6.82) c 12.1 (5.25) c 25.1 (5.92) ab 30.1 (3.94) a 30.5 (6.25) a m Empirical curvature of the A–PPFD curve Dimensionless 0.60 (0.12) a 0.28 (0.33) a 0.18 (0.20) a 0.75 (0.13) a 0.78 (0.13) a 0.80 (0.06) a RLIGHT Respiration in the light/day >0 µmol m−2 s−1 0.43 (0.09) b 0.53 (0.3) b 0.72 (0.34) b 1.25 (0.26) a 1.37 (0.22) a 1.38 (0.32) a VCMAX CO2-saturated Rubisco carboxylation rate* µmol m−2 s−1 123 (24.8) a 88.4 (11.4) a 80.9 (25.1) a 41.0 (11.7) b 38.2 (17.2) b 38.6 (10.8) b VPMAX CO2-saturated PEPC carboxylation rate (C4 only) µmol m−2 s−1 – – – 52.5 (16.4) a 36.0 (8.97) b 31.7 (8.32) b Y(CO2)LL Initial (or maximum) quantum yield for CO2 fixation Dimensionless 0.03 (0.01) b 0.05 (0.02) ab 0.06 (0.01) a 0.05 (0.01) ab 0.05 (0.002) ab 0.05 (0.01) ab *Values were log-transformed before statistical analysis to ensure normality of variance; the values given are non-transformed means. LCP, light compensation point. View Large Table 1. Fitted photosynthetic parameters derived from A–PPFD and A–Ci response curves. Mean values (s.d.), n = 6–8, for fitted parameters. Values sharing the same letter within each row are not statistically different at α = 0.05 based on two-way ANOVA (one-way for VPMAX) testing for effects of plant (acacia or Eragrostis) and [CO2]a (200, 400 or 800 ppm), with post hoc pairwise comparisons. Main statistical effects are given in Supplementary Data Table S2 Symbol Term Unit Atmospheric CO2 concentration Acacia (C3) Eragrostis (C4) 200 ppm 400 ppm 800 ppm 200 ppm 400 ppm 800 ppm ASAT CO2-saturated assimilation measured in A–Ci curve µmol m−2 s−1 25.4 (2.23) ab 24.0 (8.04) ab 26.7 (3.05) a 21.8 (4.51) ab 17.5 (1.98) b 19.5 (4.03) ab CE Initial slope of the A–Ci curve Dimensionless 0.11 (0.02) c 0.09 (0.04) c 0.09 (0.03) c 0.38 (0.11) a 0.29 (0.07) ab 0.25 (0.05) b Γ Ci–A compensation point, i.e. Ci where A = 0 µmol mol−1 49.1 (1.56) a 52.6 (4.98) a 51.9 (8.95) a 1.49 (1.64) b 3.85 (3.13) b 3.99 (2.77) b ω Empirical curvature of the A–Ci curve Dimensionless 0.40 (0.36) b 0.69 (0.19) ab 0.60 (0.34) ab 0.75 (0.13) ab 0.78 (0.13) ab 0.80 (0.06) a GASAT Light-saturated gross assimilation at [CO2]a of light curve µmol m−2 s−1 7.66 (0.94) c 17.5 (8.17) bc 30.2 (7.67) a 20.4 (5.05) b 20.1 (2.91) b 21.6 (4.69) ab JSAT Light-saturated electron transport rate at [CO2]a of light curve µmol m−2 s−1 122 (45.9) a 128 (40.3) a 139 (33.0) a 157 (39.5) a 153 (68.6) a 154 (43.0) a LCP PPFD–A compensation point, i.e. PPFD where A = 0 µmol m−2 s−1 15.7 (7.07) bc 13.1 (6.82) c 12.1 (5.25) c 25.1 (5.92) ab 30.1 (3.94) a 30.5 (6.25) a m Empirical curvature of the A–PPFD curve Dimensionless 0.60 (0.12) a 0.28 (0.33) a 0.18 (0.20) a 0.75 (0.13) a 0.78 (0.13) a 0.80 (0.06) a RLIGHT Respiration in the light/day >0 µmol m−2 s−1 0.43 (0.09) b 0.53 (0.3) b 0.72 (0.34) b 1.25 (0.26) a 1.37 (0.22) a 1.38 (0.32) a VCMAX CO2-saturated Rubisco carboxylation rate* µmol m−2 s−1 123 (24.8) a 88.4 (11.4) a 80.9 (25.1) a 41.0 (11.7) b 38.2 (17.2) b 38.6 (10.8) b VPMAX CO2-saturated PEPC carboxylation rate (C4 only) µmol m−2 s−1 – – – 52.5 (16.4) a 36.0 (8.97) b 31.7 (8.32) b Y(CO2)LL Initial (or maximum) quantum yield for CO2 fixation Dimensionless 0.03 (0.01) b 0.05 (0.02) ab 0.06 (0.01) a 0.05 (0.01) ab 0.05 (0.002) ab 0.05 (0.01) ab Symbol Term Unit Atmospheric CO2 concentration Acacia (C3) Eragrostis (C4) 200 ppm 400 ppm 800 ppm 200 ppm 400 ppm 800 ppm ASAT CO2-saturated assimilation measured in A–Ci curve µmol m−2 s−1 25.4 (2.23) ab 24.0 (8.04) ab 26.7 (3.05) a 21.8 (4.51) ab 17.5 (1.98) b 19.5 (4.03) ab CE Initial slope of the A–Ci curve Dimensionless 0.11 (0.02) c 0.09 (0.04) c 0.09 (0.03) c 0.38 (0.11) a 0.29 (0.07) ab 0.25 (0.05) b Γ Ci–A compensation point, i.e. Ci where A = 0 µmol mol−1 49.1 (1.56) a 52.6 (4.98) a 51.9 (8.95) a 1.49 (1.64) b 3.85 (3.13) b 3.99 (2.77) b ω Empirical curvature of the A–Ci curve Dimensionless 0.40 (0.36) b 0.69 (0.19) ab 0.60 (0.34) ab 0.75 (0.13) ab 0.78 (0.13) ab 0.80 (0.06) a GASAT Light-saturated gross assimilation at [CO2]a of light curve µmol m−2 s−1 7.66 (0.94) c 17.5 (8.17) bc 30.2 (7.67) a 20.4 (5.05) b 20.1 (2.91) b 21.6 (4.69) ab JSAT Light-saturated electron transport rate at [CO2]a of light curve µmol m−2 s−1 122 (45.9) a 128 (40.3) a 139 (33.0) a 157 (39.5) a 153 (68.6) a 154 (43.0) a LCP PPFD–A compensation point, i.e. PPFD where A = 0 µmol m−2 s−1 15.7 (7.07) bc 13.1 (6.82) c 12.1 (5.25) c 25.1 (5.92) ab 30.1 (3.94) a 30.5 (6.25) a m Empirical curvature of the A–PPFD curve Dimensionless 0.60 (0.12) a 0.28 (0.33) a 0.18 (0.20) a 0.75 (0.13) a 0.78 (0.13) a 0.80 (0.06) a RLIGHT Respiration in the light/day >0 µmol m−2 s−1 0.43 (0.09) b 0.53 (0.3) b 0.72 (0.34) b 1.25 (0.26) a 1.37 (0.22) a 1.38 (0.32) a VCMAX CO2-saturated Rubisco carboxylation rate* µmol m−2 s−1 123 (24.8) a 88.4 (11.4) a 80.9 (25.1) a 41.0 (11.7) b 38.2 (17.2) b 38.6 (10.8) b VPMAX CO2-saturated PEPC carboxylation rate (C4 only) µmol m−2 s−1 – – – 52.5 (16.4) a 36.0 (8.97) b 31.7 (8.32) b Y(CO2)LL Initial (or maximum) quantum yield for CO2 fixation Dimensionless 0.03 (0.01) b 0.05 (0.02) ab 0.06 (0.01) a 0.05 (0.01) ab 0.05 (0.002) ab 0.05 (0.01) ab *Values were log-transformed before statistical analysis to ensure normality of variance; the values given are non-transformed means. LCP, light compensation point. View Large Statistical analysis Experimental observations of starch accumulation, operational gas exchange, water relations, leaf C:N and stomatal properties and fitted model parameters were subject to two-way ANOVAs. Effects of species (acacia or Eragrostis), [CO2]a (200, 400 or 800 ppm) and their interactions were subject to post hoc Tukey pairwise comparison testing for specific differences between means at the species × [CO2]a level (using Minitab v.17, Minitab, State College, PA, USA), with a significance threshold of 95 % after satisfying assumptions of normality of variance (Supplementary Data Table S2). The biomass and specific starch content of both roots and shoots were analysed using three-way ANOVA for species, [CO2]a and organ (shoot or root) effects, with post hoc pairwise comparison testing for specific differences between means at the species × [CO2]a × organ level. Variables pertaining to acacia only (thorns and viable sapwood area) were subjected to one-way ANOVA for effects of [CO2]a, with post hoc Tukey pairwise comparisons. Technical replicates were averaged before statistical analysis. The level of biological replication was n = 8 unless indicated otherwise. RESULTS Growth, starch and thorns Eragrostis produced 3- to 5-fold more above-ground biomass than acacia over the course of the experiment, with the largest difference occurring at 200 ppm [CO2]a (Fig. 1A). Across both species, above- and below-ground growth was lower at lower [CO2]a (Supplementary Data Table S1), but root growth did not differ between species despite a larger effect of [CO2]a on root growth for acacia, which was 38 % lower at 200 ppm than at 800 ppm [CO2]a (Fig. 1B). Eragrostis accumulated an order of magnitude more starch in leaves than roots, and starch accumulation as a proportion of biomass was not affected by growth [CO2]a (Fig. 1B). In contrast, acacia grown at 200 ppm [CO2]a preferentially accumulated 3-fold more starch as a proportion of biomass in roots than leaves, while at 800 ppm [CO2]a acacia accumulated 6-fold more starch in leaves than roots (Fig. 1B). Acacia trees grown at 800 ppm [CO2]a had 6-fold higher leaf starch, but 65 % lower root starch as a proportion of biomass compared with trees grown at 200 ppm [CO2]a (Fig. 1B, Supplementary Data Table S1). In line with lower shoot growth and starch accumulation in leaves, acacia thorns, which provide defence against herbivores, were half the length and their total mass was ~80 % lower at 200 ppm [CO2]a than at 800 ppm [CO2]a (Fig. 2, Supplementary Data Table S1). Fig. 1. View largeDownload slide Biomass and starch accumulation. Total leaf and root biomass (A) and starch relative content (B) for V. karroo (C3 tree) and E. curvula (C4 grass) at each [CO2]a level. Bars show means ± 1 s.e. (n = 8); bars sharing the same lowercase letter are not statistically different at α = 0.05 (three-way ANOVA testing for effects of species, organ and [CO2]a with post hoc Tukey comparisons). Different uppercase letters in (A) denote statistical differences in total biomass (root plus shoot) between treatments at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Fig. 1. View largeDownload slide Biomass and starch accumulation. Total leaf and root biomass (A) and starch relative content (B) for V. karroo (C3 tree) and E. curvula (C4 grass) at each [CO2]a level. Bars show means ± 1 s.e. (n = 8); bars sharing the same lowercase letter are not statistically different at α = 0.05 (three-way ANOVA testing for effects of species, organ and [CO2]a with post hoc Tukey comparisons). Different uppercase letters in (A) denote statistical differences in total biomass (root plus shoot) between treatments at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Fig. 2. View largeDownload slide Physical defences in acacia (V. karroo). Mean length of the longest five thorns on each of three branches per tree (A) and total thorn mass (B) for plants grown at 200, 400 or 800 ppm [CO2]a. Bars show the mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (one-way ANOVA testing for the effect of [CO2]a with post hoc pairwise comparisons). See Supplementary Data Table S1 for main treatment effects Fig. 2. View largeDownload slide Physical defences in acacia (V. karroo). Mean length of the longest five thorns on each of three branches per tree (A) and total thorn mass (B) for plants grown at 200, 400 or 800 ppm [CO2]a. Bars show the mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (one-way ANOVA testing for the effect of [CO2]a with post hoc pairwise comparisons). See Supplementary Data Table S1 for main treatment effects Gas exchange under operational conditions Operational net CO2 assimilation (Aop) for acacia decreased significantly and progressively with decreasing growth [CO2]a from 800 to 400 ppm (38 % decline) and 200 ppm (66 % decline), whereas Aop for C4Eragrostis was maintained across growth [CO2]a treatments (Fig. 3A; Supplementary Data Table S1). At 800 ppm [CO2]a, Aop for acacia was significantly (40 %) higher than for Eragrostis, while at 200 ppm [CO2]aAop was significantly (56 %) lower for Acacia than Eragrostis (Fig. 3A). Leaf-level operational transpiration (Eop) was higher for acacia than Eragrostis, particularly at higher growth [CO2]a (Supplementary Data Table S1). However, for acacia Eop was only marginally higher at 400 and 200 ppm compared with 800 ppm [CO2]a, whereas declines in Eop with increasing growth [CO2]a were substantial and progressive for Eragrostis (Fig. 3B). The pattern of leaf-level transpiration was determined by operational stomatal conductance, gSop (Fig. 3C). Higher gSop at lower [CO2]a was associated with a marked decrease in WUE for both acacia and Eragrostis (Fig. 3D, Supplementary Data Table S1). WUE was consistently higher for Eragrostis at all [CO2]a levels, with the largest difference occurring at 200 ppm [CO2]a, at which WUE was 2.5-fold higher than for acacia. Decreased WUE at lower growth [CO2]a has been documented previously for NAD-ME C4 grasses, including E. curvula (Wand et al., 2001; Pinto et al., 2014). The beneficial effect of the C4 pathway for WUE and soil water conservation was maintained at progressively higher growth [CO2]a due to the sensitivity of Eragrostis gSop to [CO2]a. Fig. 3. View largeDownload slide Gas exchange under operational conditions. (A) Assimilation (Aop) (B) leaf-level transpiration (Eop), (C), stomatal conductance (gSop) and (D) water-use efficiency (WUE = Aop/Eop) of V. karroo (C3 tree) and E. curvula (C4 grass) leaves measured under operational conditions (i.e. measured in-cabinet with gas-analyser set points for temperature, humidity, Ca (Ca is CO2 concentration outside the leaf, in this particular case CO2 concentration in the IRGA cuvette) and PPFD set at cabinet levels). Bars show the mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Fig. 3. View largeDownload slide Gas exchange under operational conditions. (A) Assimilation (Aop) (B) leaf-level transpiration (Eop), (C), stomatal conductance (gSop) and (D) water-use efficiency (WUE = Aop/Eop) of V. karroo (C3 tree) and E. curvula (C4 grass) leaves measured under operational conditions (i.e. measured in-cabinet with gas-analyser set points for temperature, humidity, Ca (Ca is CO2 concentration outside the leaf, in this particular case CO2 concentration in the IRGA cuvette) and PPFD set at cabinet levels). Bars show the mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Plant–water relations Leaf water potentials at midday (Ψleaf) and predawn (Ψpd) were less negative for Eragrostis than acacia, acacia having lowest Ψleaf at 200 ppm [CO2]a (Fig. 4A, B, Supplementary Data Table S1). For Eragrostis, the highest values of Ψleaf occurred at 800 ppm [CO2]a, and Ψleaf was significantly lower at 200 ppm than at higher [CO2]a. Although Ψpd was generally higher at higher growth [CO2]a, reflecting higher WUE and soil water savings (Supplementary Data Table S1), species-specific decreases at higher growth [CO2]a were less pronounced than for Ψleaf (Fig. 4A). Leaf solute potential (Ψsolute) was significantly lower for acacia than Eragrostis across [CO2]a, which may reflect less hydrated leaves, but was not affected by growth [CO2]a (Fig. 4C, Supplementary Data Table S1). Eragrostis leaf RWC was significantly higher than for acacia, which showed no response of leaf RWC to growth [CO2]a (Supplementary Data Table S1). Moreover, leaf RWC for Eragrostis grown at 400 and 800 ppm [CO2]a was 4 and 10 % lower, respectively, than at 200 ppm (Fig. 4D). Eragrostis Kplant was higher than acacia Kplant (Supplementary Data Table S1), but there was no effect of growth [CO2]a for either species (Fig. 4E). Leaf turgor of acacia was less than half that of Eragrostis (Fig. 4F). Fig. 4. View largeDownload slide Plant hydraulics and water relations for V. karroo (C3 tree) and E. curvula (C4 grass). (A, B) Leaf water potential at midday (A) and predawn (a proxy for soil water potential) (B). (C) Predawn leaf solute potential. (D) Leaf RWC at midday. (E) Plant hydraulic conductance. (F) Leaf turgor pressure (F) Bars show mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Fig. 4. View largeDownload slide Plant hydraulics and water relations for V. karroo (C3 tree) and E. curvula (C4 grass). (A, B) Leaf water potential at midday (A) and predawn (a proxy for soil water potential) (B). (C) Predawn leaf solute potential. (D) Leaf RWC at midday. (E) Plant hydraulic conductance. (F) Leaf turgor pressure (F) Bars show mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc Tukey comparisons). See Supplementary Data Table S1 for main treatment effects. Photosynthetic parameters The A–PPFD response curves show that, at high PPFD, A for acacia decreased significantly and progressively with falling growth [CO2]a, but showed no variation for Eragrostis, which maintained high rates across [CO2]a (Supplementary Data Fig. S5). The A–Ci curves for Eragrostis were steeper than for acacia, especially for plants grown at 200 ppm [CO2]a, but saturated rates were similar for both species across [CO2]a (Supplementary Data Fig. S5). Below Ci of ~200 µmol mol−1, A was highest for plants grown at 200 ppm [CO2]a. The hyperbolic response of gS to PPFD is typical, whereas the concave shape of the response of gS to Ci is due to the speed with which the A–Ci curves were measured (Supplementary Data Fig. S6). We analysed the A–PPFD and A–Ci response curves to derive a suite of photosynthetic parameters (Table 1). The CO2-saturated rate of assimilation (ASAT) was not affected by the [CO2]a at which the plants were grown, but was consistently higher for acacia than Eragrostis (Supplementary Data Table S2). The initial slope of the A–Ci curves (carboxylation efficiency, CE) differed between species and decreased with growth [CO2]a for Eragrostis (Table 1, Supplementary Data Table S2). This was reflected by VCMAX (for acacia) and VPMAX (for Eragrostis), which were derived mechanistically. Eragrostis had 1.5- to 1.7-fold higher VPMAX at 200 ppm [CO2]a than at higher [CO2]a, highlighting upregulation of carboxylation capacity in line with field studies (Anderson et al., 2001). The pattern of Aop between species and [CO2]a was reflected in light-saturated GASAT. The light compensation point (the PPFD at which A is zero) was 52 % lower for acacia than Eragrostis, reflecting 2.4-fold higher daylight respiration (RLIGHT) for Eragrostis, and lower basal metabolism and better adaptation to low light for acacia (Supplementary Data Table S2). However, differences in RLIGHT were not significant across growth [CO2]a. The quantum yield for CO2 fixation [Y(CO2)LL] (a measure of light-use efficiency) was lower for acacia than Eragrostis, and was lowest at 200 ppm [CO2]a for acacia, but showed no variation across [CO2]a for Eragrostis (Table 1, Supplementary Data Table S2). Leaf nitrogen For acacia, leaf C:N was lower at 200 and 400 ppm than at 800 ppm [CO2]a, whereas leaf C:N for Eragrostis was significantly higher at lower growth [CO2]a levels, highlighting different acclimation responses to low [CO2]a (Fig. 5). The differences in leaf C:N across [CO2]a were largely driven by changes in leaf N concentration because leaf C was consistently ~45–46 % across [CO2]a levels. Fig. 5. View largeDownload slide C:N ratios in leaves and the relationship of leaf N to maximum Rubisco and PEPC activities. (A) Leaf C:N. (B, C) Relationship between leaf N and maximal rate of Rubisco carboxylation (VCMAX) in V. karroo (B) and maximal rate of PEPC carboxylation (VPMAX) in E. curvula (C). In (B) and (C), circles represent 200 ppm, squares 450 ppm and triangles 800 ppm growth [CO2]a. Bars show means ± 1 s.e. (n = 6–8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc pairwise comparisons). See Supplementary Data Table S1 for main treatment effects. Fig. 5. View largeDownload slide C:N ratios in leaves and the relationship of leaf N to maximum Rubisco and PEPC activities. (A) Leaf C:N. (B, C) Relationship between leaf N and maximal rate of Rubisco carboxylation (VCMAX) in V. karroo (B) and maximal rate of PEPC carboxylation (VPMAX) in E. curvula (C). In (B) and (C), circles represent 200 ppm, squares 450 ppm and triangles 800 ppm growth [CO2]a. Bars show means ± 1 s.e. (n = 6–8); bars sharing the same letter are not statistically different at α = 0.05 (two-way ANOVA testing for effects of species and [CO2]a with post hoc pairwise comparisons). See Supplementary Data Table S1 for main treatment effects. Stomatal characteristics Acacia and Eragrostis both had higher densities of smaller stomata at 200 ppm compared with leaves grown at 800 ppm [CO2]a (Fig. 6A–C, Supplementary Data Table S2). Fig. 6. View largeDownload slide Stomatal size–density relationships and safranin-stained viable acacia sapwood. Stomatal size measured as guard cell area (A), stomatal density per unit leaf area (B) measured with light microscopy and their relationship (C) indicate the physical constraints on leaf gas exchange with the atmosphere and leaf acclimation to [CO2]a. The percentage of acacia sapwood that became stained (D) as safranin O was drip-fed through cut branch segments provides a proxy measure of plant hydraulic conductance. A higher percentage of staining indicates larger and/or more water-conducting vessels within the woody tissue. Bars and symbols show mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 [two-way ANOVA testing for effects of species (acacia or Eragrostis) and [CO2]a (200, 400 or 800 ppm)]. See Supplementary Data Table S1 for main treatment effects. Fig. 6. View largeDownload slide Stomatal size–density relationships and safranin-stained viable acacia sapwood. Stomatal size measured as guard cell area (A), stomatal density per unit leaf area (B) measured with light microscopy and their relationship (C) indicate the physical constraints on leaf gas exchange with the atmosphere and leaf acclimation to [CO2]a. The percentage of acacia sapwood that became stained (D) as safranin O was drip-fed through cut branch segments provides a proxy measure of plant hydraulic conductance. A higher percentage of staining indicates larger and/or more water-conducting vessels within the woody tissue. Bars and symbols show mean ± 1 s.e. (n = 8); bars sharing the same letter are not statistically different at α = 0.05 [two-way ANOVA testing for effects of species (acacia or Eragrostis) and [CO2]a (200, 400 or 800 ppm)]. See Supplementary Data Table S1 for main treatment effects. Stained sapwood The area of safranin O-stained sapwood relative to the total sapwood area of acacia branch cross-sections was 34% at 200 ppm [CO2]a compared with ~25% at higher [CO2]a (Fig. 6D), indicating development of more and/or larger water-conducting conduits under low [CO2]a. Discussion Most [CO2]a response studies have focused on elevated [CO2]a (reviewed in McLeod and Long, 1999; Ainsworth and Long, 2005; Korner, 2006; Miyagi et al., 2007; Springer and Ward, 2007; Lloyd and Farquhar, 2008), and have often concerned the potential of increased plant growth to boost net primary productivity, offset carbon emissions or survive associated increased warming and/or drought. Fewer studies (reviewed in Sage and Coleman, 2001; Gerhart and Ward, 2010) have focused on plant responses to low [CO2]a, which is critical for understanding long-term plant and ecosystem responses to changing [CO2]a (Leakey and Lau, 2012). We characterized the [CO2]a acclimation responses of two species of different growth forms photosynthetic pathways and life histories (sexual versus clonal) across low to high [CO2]a levels and found marked differences in acclimation to [CO2]a in a range of traits related to growth, carbon–water balance, water relations and photosynthetic rate and capacity. Growth irradiance was lower by a factor of ~2 than in typical savannah field conditions, and this may explain the lower operational rates of A reported here compared with field measurements for the same species (Charles-Dominique et al., 2018). Interpretation of our results therefore assumes that possible light limitation affected all species similarly, with any observed difference being representative of the physiological mechanisms of acclimation to CO2. Acacia trees grown at 400 and 800 ppm [CO2]a produced significantly more biomass than trees grown at lower [CO2]a. Such observations are invoked to link rising [CO2]a over the last century with recent woody encroachment of savannah and pasture land (Hoffmann et al., 2004; Morgan et al., 2007; Staver et al., 2011; Bond and Midgley, 2012; Midgley and Bond, 2015; Nackley et al., 2018). With future [CO2]a increases, higher growth and assimilation could allow more saplings to survive by growing sufficiently tall to protect their crowns from fires and thereby outcompete C4 grasses through shading (Hoffmann et al., 2004; Bond, 2008). Belowground starch reserves in savannah tree roots are considered important for regrowth following crown damage from fire or herbivore disturbance (Bond, 2008; Schutz et al., 2009; Wigley et al., 2009; Kgope et al., 2010; Bond and Midgley, 2012). Acacia trees grown at 800 ppm [CO2]a accumulated significantly more leaf starch than trees grown at lower [CO2]a, in line with Nackley et al. (2018), but starch accumulation in roots decreased at successively higher growth [CO2]a levels. As a result, the apportionment of starch between acacia leaves and roots favoured roots at 200 ppm and leaves at 800 ppm [CO2]a (Fig. 1B), in accordance with the results of Allen et al. (1998) for soybean. In contrast, Kgope et al. (2010) found that root carbohydrate accumulation for two Acacia species increased non-linearly with increasing growth [CO2]a from 200 to 1000 ppm [CO2]a. Differences in the findings between studies may be related to differences in starch analysis protocols (for critical review see Bellasio et al., 2014b); Kgope et al. (2010) used an approach based on acid hydrolysis of carbohydrates that was not starch-specific. Under low [CO2]a, when A is reduced, the acacia trees apparently maintained a higher safety margin for effective regrowth by storing a higher proportion of starch in roots, which would be required for coping with increased browsing pressure above ground. Moreover, under low [CO2]a, acacia grew smaller and fewer thorns compared with higher [CO2]a (Fig. 2, Supplementary Data Table S1), which would result in lower physical defence against herbivores (Charles-Dominique et al., 2016). Lower concentrations of phenolic compounds in foliage (potentially reflected here by lower C:N ratios in acacia leaves; Fig. 5A) under low [CO2]a may also lower resistance to herbivores by increasing the palatability of leaves to browsers (Quirk et al., 2013). Pressure to accumulate below-ground starch reserves under low [CO2]a would be exacerbated where trees and C4 grasses compete for resources (February et al., 2013). Productive C4 grasses promote disturbance from both fire and grazing, which would penalize saplings with lower accumulated root starch that could not re-sprout quickly (Bond, 2008). Starch accumulation in Eragrostis roots and shoots, unlike acacia, was unaffected by growth [CO2]a, probably because A was maintained independently of growth [CO2]a, with no need to adjust safety margins for re-sprouting. Coping with water limitation is critical in semi-arid environments and stomatal factors are increasingly recognized as central to optimizing growth in drying soils (Osborne and Sack, 2012; Bellasio et al., 2017). Fundamentally, evaporation for carbon gain trade-offs is determined by stomatal characteristics and regulation (Buckley et al., 2016). We characterized short- and long-term stomatal responses within wetting–drying cycles, and resolved the effect of soil water potential (Ψpd). Both acacia and Eragrostis showed a characteristic increase in stomatal size and decrease in stomatal density at higher growth [CO2]a, consistent with declines in gS at higher growth [CO2]a (Fig. 3B). At lower growth [CO2]a, stomatal size relative to density appears to decline in acacia. This would reduce gw_maxin sensuFranks et al. (2012b), but we caution against inferring a firm link. The stomatal size–density response we observed to [CO2]a is consistent with Franks et al. (2012b), and aligns with the pattern observed over evolutionary timescales (Beerling and Chaloner, 1993). For acacia, changes in stomatal size and density were accompanied by larger and/or more water-conducting vessels in sapwood, which underlines the higher demand on plant hydraulics and risk of evapotranspirative water losses for trees under low [CO2]a. Smaller stomata are linked with faster stomatal adjustment (Franks and Farquhar, 2007), and help mitigate water limitation under low CO2 (Morgan et al., 2001). Moreover, grass stomata respond faster than tree stomata (Franks and Farquhar, 2007; McAusland et al., 2016) and many grasses, including Eragrostis, can further regulate transpiration through leaf rolling (Kipchirchir et al., 2015). Fast-closing stomata and leaf rolling are important competitive attributes in the persistence of grass cover through their effects on soil water savings where water is periodically limiting (Quirk et al., 2018). Yet, as these advantages are related to grass leaf architecture and hydraulic traits, and are independent of photosynthetic type, C3 grasses may become increasingly competitive under future [CO2]a, particularly if rainfall is limiting and thicker woody canopies create increasingly shaded habitats unfavourable for C4 grasses (Polley et al., 1993, 2002; Ibrahim et al., 2008). The negative response of gSop to growth [CO2]a was more substantial for C4Eragrostis than acacia, resulting in higher WUE with higher growth [CO2]a, in line with Polley et al. (1993). This is thought to have been important in water-limited environments when [CO2]a was low and gS was necessarily high (Osborne and Sack, 2012). In our cabinets, where DS (Leaf-to-atmosphere water mole fraction gradient (a measure of VPD)) was constant, the rate of soil moisture decline was proportional to gS and directly reflected by higher Ψpd (Fig. 4B). We have recently shown that decreasing soil water availability impinges on assimilation relatively more for C4 grasses than for C3 trees (Quirk et al., 2018). However, in monoculture stands like our pots, higher WUE for C4 grasses confers soil water savings, allowing stomata to remain open and extending the duration of assimilation. This has been shown in comparative analyses, in which declines in A were slower for C4 than C3 plants, particularly during the initial stages (2–3 weeks) of drought (Schulze and Hall, 1982; Ripley et al., 2010; Taylor et al., 2014), and is a recognized advantage under fluctuating water availability (Morgan et al., 2011; Ladrón de Guevara et al., 2015; Nie et al., 2018). Conversely, if soil water is shared between C3 and C4 plants, such as in mixed stands where the majority of tree and grass roots occupy upper soil layers (February and Higgins, 2010) and compete primarily for the same resources (Scholes and Archer, 1997), soils are likely to be wetter on average than in C3-only stands, providing a window of opportunity for C3 grasses and trees to colonize stands of C4 grasses. Recent theoretical reasoning suggests that minimizing water use would be less competitive than maximizing net carbon gain. The maximization of carbon gain was shown to be consistent with observed stomatal behaviour in trees (Wolf et al., 2016) and would be an evolutionarily stable strategy. This means that plants could invade monoculture stands characterized by strategies other than carbon gain maximization, but could not be invaded by any other strategy when they are in monoculture (Smith, 1982). Indeed, models have shown that increasing [CO2]a, by relieving pressure for water, would promote coexistence among competing species in mixed stands (Ali et al., 2015). However, only C3 species have been considered so far, and the mathematical basis for competition between photosynthetic types remains undetermined. The implications of continued soil water savings for C4 plants as [CO2]a rises are not well understood, but improved WUE in C4 plants has been observed in cropland under [CO2]a enrichment (Markelz et al., 2011; Reich et al., 2018). In addition, for C4Eragrostis, Kplant and Ψsoil were higher compared with acacia at 800 ppm [CO2]a, suggesting Eragrostis still benefitted from the water-conserving advantages of C4 photosynthesis under high [CO2]a. Although experiments are scarce, there is evidence that C4 plants will continue to benefit from water-saving advantages over C3 plants during seasonal drought as [CO2]a increases (Morgan et al., 2004; Medeiros and Ward, 2013; Gray et al., 2016; Reich et al., 2018). Consequently, in seasonally dry savannah environments, water limitation and drought acclimation are likely to mediate the responses of C4 grasses and savannah trees to rising [CO2]a (Nie et al., 2018). Over multi-year timescales of lower than usual precipitation, such as El Niño–La Niña climatic perturbations, the conservative water use and more rapid responses to soil water inputs of C4 grasses over C3 trees could be influential in tipping the transitional balance between the alternative states of closed forest and open, fire- and/or herbivore-controlled savannah grassland (Hirota et al., 2011; Staver et al., 2011). Mathematical models describing stomatal behaviour coupled to leaf-level biochemical models of C3 and C4 photosynthesis form a critical component of dynamic global vegetation modelling (DGVM) (Bonan et al., 2014; Sato et al., 2015). The outputs from these sub-models are used to estimate fluxes of carbon and water in and out of different plant functional types and predict the responses of ecosystem net primary productivity and water budgets (Ostle et al., 2009; de Boer et al., 2011; Paschalis et al., 2017). When DGVM is used to predict future or reconstruct past scenarios, photosynthetic parameters and stomatal behaviour are generally derived from model plants grown under current [CO2]a conditions (Leuning, 1995; Ball et al., 1987; Buckley et al., 2003; Damour et al., 2010; Way et al., 2011). Consequently, the parameters become less applicable the further the environmental drivers deviate from current conditions (Way et al., 2011). Further, no large-scale models include representation of photosynthetic parameters linking gas exchange characteristics to future [CO2]a (Rogers et al., 2017). For acacia, rising [CO2]a increased the light-saturated rate of gross assimilation, GASAT, and at the same time increased the quantum yield for CO2 fixation, Y(CO2)LL (Table 1, Supplementary Data Table S2), in agreement with Overdieck (1989). Notwithstanding an increase in mitochondrial respiration in the day (RLIGHT), this allowed the light compensation point to decrease with growth [CO2]a, also in accordance with Overdieck (1989). For Eragrostis, A–PPFD characteristics were not affected by [CO2]a (Table 1). The empirical analysis of A–Ci responses revealed that the initial slope of the curve (carboxylating efficiency, CE) responded to growth [CO2]a for both species. For acacia, this was mechanistically explained by the upregulation of VCMAX at low growth [CO2]a, in line with Anderson et al. (2001), and with the downregulation of VCMAX at high growth [CO2]a, consistent with Osborne and Beerling (2003) and Sage (1994). In Eragrostis, differences in CE were explained by higher VPMAX at low growth [CO2]a and vice versa at high growth [CO2]a. These acclimation patterns of carboxylating enzymes can be related to the contrasting effects of [CO2]a on leaf C:N ratios between acacia and Eragrostis (Fig. 5). For acacia, leaf N was positively correlated with VCMAX, while for Eragrostis leaf N was not or was negatively correlated with VPMAX (Fig. 5). This may reflect differences in the composition of the photosynthetic machinery in C3 and C4 plants. For acacia, higher [CO2]a reduced the need for nitrogen investment in Rubisco. Because the N content of Rubisco is greater than that of other photosynthetic enzymes, a reduction in Rubisco content allows C3 plants to grow leaves with lower N concentration, resulting in marginally higher C:N at higher growth [CO2]a. For Eragrostis higher [CO2]a induced downregulation of PEP carboxylase, which has a relatively low N content (Makino et al., 2003). This may allow ATP savings and perhaps a reorganization of the electron transport chain in favour of the NADPH generation required by a higher relative rate of phosphoglycerate (3-phosphoglyceric acid) reduction. However, this upregulation of nitrogen-rich electron transport requires more N, especially if it is accompanied by upregulation of C3 activity at higher growth [CO2]a. On the basis of our findings, we urge the modelling community to account for the effect of long-term acclimation to [CO2]a on model input parameters when simulating plant responses to past and future [CO2]a. Conclusions For C3 savannah-adapted acacia, low to high [CO2]a increased Aop, C:N, biomass and leaf starch accumulation, but decreased gS and root starch accumulation, whereas for C4Eragrostis low to high [CO2]a decreased C:N but did not affect Aop, biomass or starch accumulation, and substantially decreased gS. In both plant types, WUE increased with [CO2]a treatment, associated with less negative Ψpd, increased stomatal size and decreased stomatal density. Overall, with increasing [CO2]a towards 800 ppm our findings point to increased competitiveness of savannah trees like acacia, through benefits that include greater allocation of photosynthate away from root reserves into growth, higher productivity, better herbivore defences and improved WUE. However, the consistently higher WUE found for Eragrostis, together with non-assimilatory hydraulic benefits such as higher KPLANT and faster stomata, would extend the duration of photosynthesis within a given time window. Understanding how these factors will interact to shape the competitive interactions of trees and grasses in semi-arid environments will require integration with soil water reserves and climate in mathematical models. In this endeavour, accounting for the acclimation of photosynthetic characteristics in model parameterization will be critical. SUPPLEMENTARY DATA Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Figure S1: schematic representation of the measurement and sampling strategy. Figure S2: branch cross-sections showing stained sapwood of Vachellia karroo. Figure S3: stomatal peels of Vachellia karroo leaves across [CO2]a treatments. Figure S4: stomatal peels of Eragrostis curvula leaves across [CO2]a treatments. Figure S5: photosynthetic A–PPFD and A–Ci curves. Figure S6: stomatal conductance response to PPFD and Ci. Table S1: ANOVA results for physiological observations of growth, starch accumulation, operational gas exchange, water relations, leaf C:N and stomatal properties. Table S2: ANOVA results for photosynthetic parameters derived from A–PPFD and A–Ci curve-fitting experiments. ACKNOWLEDGEMENTS We thank Dr Loredana Saccone for performing and photographing stomatal peels. The authors acknowledge funding through an ERC advanced grant (CDREG, 322998) awarded to D.J.B. C.B. acknowledges funding through a H2020 MSCA individual fellowship (DILIPHO, ID: 702755). LITERATURE CITED Ainsworth EA , Long SP . 2005 . What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analytic review of the responses of photosynthesis, canopy . New Phytologist 165 : 351 – 371 . Google Scholar Crossref Search ADS PubMed WorldCat Ali AA , Medlyn BE , Aubier TG , Crous KY , Reich PB . 2015 . Elevated carbon dioxide is predicted to promote coexistence among competing species in a trait-based model . Ecology and Evolution 5 : 4717 – 4733 . Google Scholar Crossref Search ADS PubMed WorldCat Allen LH , Bisbal EC , Boote KJ . 1998 . Nonstructural carbohydrates of soybean plants grown in subambient and superambient levels of CO2 . Photosynthesis Research 56 : 143 – 155 . Google Scholar Crossref Search ADS WorldCat Anderson LJ , Maherali H , Johnson HB , Polley HW , Jackson RB . 2001 . Gas exchange and photosynthetic acclimation over subambient to elevated CO2 in a C3–C4 grassland . Global Change Biology 7 : 693 – 707 . Google Scholar Crossref Search ADS WorldCat Ball JT , Woodrow IE , Berry JA . 1987 . A model predicting stomatal conductance and its contribution to the control of photosynthesis under different environmental conditions. In: Biggins J , ed. Progress in photosynthesis research: Volume 4 . Dordrecht : Springer , 221–224. Google Preview WorldCat COPAC Beerling DJ , Chaloner WG . 1993 . Evolutionary responses of stomatal density to global CO2 change . Biological Journal of the Linnean Society 48 : 343 – 353 . Google Scholar Crossref Search ADS WorldCat Beerling DJ , Royer DL . 2011 . Convergent cenozoic CO2 history . Nature Geoscience 4 : 418 – 420 . Google Scholar Crossref Search ADS WorldCat Bellasio C , Griffiths H . 2014 . Acclimation to low light by C4 maize: implications for bundle sheath leakiness . Plant, Cell & Environment 37 : 1046 – 1058 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Burgess SJ , Griffiths H , Hibberd JM . 2014a. A high throughput gas exchange screen for determining rates of photorespiration or regulation of C4 activity . Journal of Experimental Botany 65 : 3769 – 3779 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Fini A , Ferrini F . 2014b. Evaluation of a high throughput starch analysis optimised for wood . PLoS ONE 9 : e86645 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Beerling DJ , Griffiths H . 2016a. Deriving C4 photosynthetic parameters from combined gas exchange and chlorophyll fluorescence using an Excel tool: theory and practice . Plant, Cell & Environment 39 : 1164 – 1179 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Beerling DJ , Griffiths H . 2016b. An Excel tool for deriving key photosynthetic parameters from combined gas exchange and chlorophyll fluorescence: theory and practice . Plant, Cell & Environment 39 : 1180 – 1197 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Quirk J , Buckley TN , Beerling D . 2017 . A dynamic hydro-mechanical and biochemical model of stomatal conductance for C4 photosynthesis . Plant Physiology 175 : 104 – 119 . Google Scholar Crossref Search ADS PubMed WorldCat Bellasio C , Quirk J , Beerling DJ . 2018 . Stomatal and non-stomatal limitations in savannah trees and C4 grasses grown at low, ambient and high atmospheric CO2 . Plant Science 274 : 181 – 192 . Google Scholar Crossref Search ADS PubMed WorldCat de Boer HJ , Lammertsma EI , Wagner-Cremer F , Dilcher DL , Wassen MJ , Dekker SC . 2011 . Climate forcing due to optimization of maximal leaf conductance in subtropical vegetation under rising CO2 . Proceedings of the National Academy of Sciences of the USA 108 : 4041 – 4046 . Google Scholar Crossref Search ADS PubMed WorldCat Boesgaard KS , Mikkelsen TN , Ro-Poulsen H , Ibrom A . 2013 . Reduction of molecular gas diffusion through gaskets in leaf gas exchange cuvettes by leaf-mediated pores . Plant, Cell & Environment 36 : 1352 – 1362 . Google Scholar Crossref Search ADS PubMed WorldCat Bonan GB , Williams M , Fisher RA , Oleson KW . 2014 . Modeling stomatal conductance in the earth system: linking leaf water-use efficiency and water transport along the soil–plant–atmosphere continuum . Geoscientific Model Development 7 : 2193 – 2222 . Google Scholar Crossref Search ADS WorldCat Bond WJ . 2008 . What limits trees in C4 grasslands and savannahs? Annual Review of Ecology Evolution and Systematics 39 : 641 – 659 . Google Scholar Crossref Search ADS WorldCat Bond WJ , Midgley GF . 2012 . Carbon dioxide and the uneasy interactions of trees and savannah grasses . Philosophical Transactions of the Royal Society B: Biological Sciences 367 : 601 – 612 . Google Scholar Crossref Search ADS WorldCat Buckley TN , Mott KA , Farquhar GD . 2003 . A hydromechanical and biochemical model of stomatal conductance . Plant, Cell & Environment 26 : 1767 – 1785 . Google Scholar Crossref Search ADS WorldCat Buckley TN , Sack L , Farquhar GD . 2016 . Optimal plant water economy . Plant, Cell & Environment 40 : 881 – 896 . Google Scholar Crossref Search ADS PubMed WorldCat von Caemmerer S . 2000 . Biochemical models of leaf photosynthesis . Collingwood, Australia : CSIRO Publishing . Google Preview WorldCat COPAC Cerling TE , Harris JM , MacFadden BJ , et al. 1997 . Global vegetation change through the Miocene/Pliocene boundary . Nature 389 : 153 – 158 . Google Scholar Crossref Search ADS WorldCat Charles-Dominique T , Davies TJ , Hempson GP , et al. 2016 . Spiny plants, mammal browsers, and the origin of African savannahs . Proceedings of the National Academy of Sciences of the USA . WorldCat Charles-Dominique T , Midgley GF , Tomlinson KW , Bond WJ . 2018 . Steal the light: shade vs fire adapted vegetation in forest–savannah mosaics . New Phytologist 218 : 1419 – 1429 . Google Scholar Crossref Search ADS PubMed WorldCat Chinnusamy V , Zhu J-K . 2009 . Epigenetic regulation of stress responses in plants . Current Opinion in Plant Biology 12 : 133 – 139 . Google Scholar Crossref Search ADS PubMed WorldCat Chow PS , Landhäusser SM . 2004 . A method for routine measurements of total sugar and starch content in woody plant tissues . Tree Physiology 24 : 1129 – 1136 . Google Scholar Crossref Search ADS PubMed WorldCat Cunniff J , Charles M , Jones G , Osborne CP . 2016 . Reduced plant water status under sub-ambient pCO2 limits plant productivity in the wild progenitors of C3 and C4 cereals . Annals of Botany 118 : 1163 – 1173 . Google Scholar Crossref Search ADS PubMed WorldCat Damour G , Simonneau T , Cochard H , Urban L . 2010 . An overview of models of stomatal conductance at the leaf level . Plant, Cell & Environment 33 : 1419 – 1438 . Google Scholar PubMed WorldCat Eckardt NA . 2005 . Photorespiration revisited . Plant Cell 17 : 2139 – 2141 . Google Scholar Crossref Search ADS WorldCat Edwards EJ , Osborne CP , Strömberg CAE , Smith SA , Consortium CG . 2010 . The origins of C4 grasslands: integrating evolutionary and ecosystem science . Science 328 : 587 – 591 . Google Scholar Crossref Search ADS PubMed WorldCat Ethier GJ , Livingston NJ . 2004 . On the need to incorporate sensitivity to CO2 transfer conductance into the Farquhar-von Caemmerer-Berry leaf photosynthesis model . Plant, Cell & Environment 27 : 137 – 153 . Google Scholar Crossref Search ADS WorldCat February EC , Higgins SI . 2010 . The distribution of tree and grass roots in savannahs in relation to soil nitrogen and water . South African Journal of Botany 76 : 517 – 523 . Google Scholar Crossref Search ADS WorldCat February EC , Higgins SI , Bond WJ , Swemmer L . 2013 . Influence of competition and rainfall manipulation on the growth responses of savannah trees and grasses . Ecology 94 : 1155 – 1164 . Google Scholar Crossref Search ADS PubMed WorldCat Fini A , Bellasio C , Pollastri S , Tattini M , Ferrini F . 2013 . Water relations, growth, and leaf gas exchange as affected by water stress in Jatropha curcas . Journal of Arid Environments 89 : 21 – 29 . Google Scholar Crossref Search ADS WorldCat Franks PJ , Farquhar GD . 2007 . The mechanical diversity of stomata and its significance in gas-exchange control . Plant Physiology 143 : 78 – 87 . Google Scholar Crossref Search ADS PubMed WorldCat Franks PJ , Freckleton RP , Beaulieu JM , Leitch IJ , Beerling DJ . 2012a. Megacycles of atmospheric carbon dioxide concentration correlate with fossil plant genome size . Philosophical Transactions of the Royal Society B: Biological Sciences 367 : 556 – 564 . Google Scholar Crossref Search ADS WorldCat Franks PJ , Leitch IJ , Ruszala EM , Hetherington AM , Beerling DJ . 2012b. Physiological framework for adaptation of stomata to CO2 from glacial to future concentrations . Philosophical Transactions of the Royal Society B: Biological Sciences 367 : 537 – 546 . Google Scholar Crossref Search ADS WorldCat Gerhart LM , Ward JK . 2010 . Plant responses to low [CO2] of the past . New Phytologist 188 : 674 – 695 . Google Scholar Crossref Search ADS PubMed WorldCat Gerhart LM , Harris JM , Nippert JB , Sandquist DR , Ward JK . 2011 . Glacial trees from the La Brea tar pits show physiological constraints of low CO2 . New Phytologist 193 : 1049 – 1063 . Google Scholar PubMed WorldCat Ghannoum O , Caemmerer SV , Ziska LH , Conroy JP . 2000 . The growth response of C4 plants to rising atmospheric CO2 partial pressure: a reassessment . Plant, Cell & Environment 23 : 931 – 942 . Google Scholar Crossref Search ADS WorldCat Ghannoum O , Caemmerer Sv , Conroy JP . 2002 . The effect of drought on plant water use efficiency of nine NAD-ME and nine NADP-ME Australian C4 grasses . Functional Plant Biology 29 : 1337 – 1348 . Google Scholar Crossref Search ADS WorldCat Ghannoum O , Phillips NG , Conroy JP , et al. 2010 . Exposure to preindustrial, current and future atmospheric CO2 and temperature differentially affects growth and photosynthesis in Eucalyptus . Global Change Biology 16 : 303 – 319 . Google Scholar Crossref Search ADS WorldCat Gill RA , Polley HW , Johnson HB , Anderson LJ , Maherali H , Jackson RB . 2002 . Nonlinear grassland responses to past and future atmospheric CO2 . Nature 417 : 279 – 282 . Google Scholar Crossref Search ADS PubMed WorldCat Gray SB , Dermody O , Klein SP , et al. 2016 . Intensifying drought eliminates the expected benefits of elevated carbon dioxide for soybean . Nature Plants 2 : 16132 . Google Scholar Crossref Search ADS PubMed WorldCat Hartmann H , Ziegler W , Kolle O , Trumbore S . 2013 . Thirst beats hunger – declining hydration during drought prevents carbon starvation in Norway spruce saplings . New Phytologist 200 : 340 – 349 . Google Scholar Crossref Search ADS PubMed WorldCat Hirota M , Holmgren M , Van Nes EH , Scheffer M . 2011 . Global resilience of tropical forest and savannah to critical transitions . Science 334 : 232 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat Hoetzel S , Dupont L , Schefusz E , Rommerskirchen F , Wefer G . 2013 . The role of fire in Miocene to Pliocene C4 grassland and ecosystem evolution . Nature Geoscience 6 : 1027 – 1030 . Google Scholar Crossref Search ADS WorldCat Hoffmann AA , Sgro CM . 2011 . Climate change and evolutionary adaptation . Nature 470 : 479 – 485 . Google Scholar Crossref Search ADS PubMed WorldCat Hoffmann WA , Orthen B , Franco AC . 2004 . Constraints to seedling success of savannah and forest trees across the savannah-forest boundary . Oecologia 140 : 252 – 260 . Google Scholar Crossref Search ADS PubMed WorldCat Ibrahim DG , Gilbert ME , Ripley BS , Osborne CP . 2008 . Seasonal differences in photosynthesis between the C3 and C4 subspecies of Alloteropsis semialata are offset by frost and drought . Plant, Cell & Environment 31 : 1038 – 1050 . Google Scholar Crossref Search ADS PubMed WorldCat IPCC . 2013 . www.ipcc.chaccessed. COPAC Kgope BS , Bond WJ , Midgley GF . 2010 . Growth responses of African savannah trees implicate atmospheric [CO2] as a driver of past and current changes in savannah tree cover . Austral Ecology 35 : 451 – 463 . Google Scholar Crossref Search ADS WorldCat Kipchirchir KO , Ngugi KR , Mwangi MS , Njomo KG , Raphael W . 2015 . Water stress tolerance of six rangeland grasses in the Kenyan semi-arid rangelands . American Journal of Agriculture and Forestry 3 : 222 – 229 . Google Scholar Crossref Search ADS WorldCat Korner C . 2006 . Plant CO2 responses: an issue of definition, time and resource supply . New Phytologist 172 : 393 – 411 . Google Scholar Crossref Search ADS PubMed WorldCat Ladrón de Guevara M , Lázaro R , Arnau-Rosalén E , Domingo F , Molina-Sanchis I , Mora JL . 2015 . Climate change effects in a semiarid grassland: physiological responses to shifts in rain patterns . Acta Oecologica 69 : 9 – 20 . Google Scholar Crossref Search ADS WorldCat Leakey ADB , Lau JA . 2012 . Evolutionary context for understanding and manipulating plant responses to past, present and future atmospheric [CO2] . Philosophical Transactions of the Royal Society B: Biological Sciences 367 : 613 – 629 . Google Scholar Crossref Search ADS WorldCat Leuning R . 1995 . A critical appraisal of a combined stomatal-photosynthesis model for C3 plants . Plant, Cell & Environment 18 : 339 – 355 . Google Scholar Crossref Search ADS WorldCat Lloyd J , Farquhar GD . 2008 . Effects of rising temperatures and [CO2] on the physiology of tropical forest trees . Philosophical Transactions of the Royal Society B: Biological Sciences 363 : 1811 – 1817 . Google Scholar Crossref Search ADS WorldCat Makino A , Mae T . 1999 . Photosynthesis and plant growth at elevated levels of CO2 . Plant and Cell Physiology 40 : 999 – 1006 . Google Scholar Crossref Search ADS WorldCat Makino A , Sakuma H , Sudo E , Mae T . 2003 . Differences between maize and rice in N-use efficiency for photosynthesis and protein allocation . Plant and Cell Physiology 44 : 952 – 956 . Google Scholar Crossref Search ADS PubMed WorldCat Markelz RJC , Strellner RS , Leakey ADB . 2011 . Impairment of C4 photosynthesis by drought is exacerbated by limiting nitrogen and ameliorated by elevated [CO2] in maize . Journal of Experimental Botany 62 : 3235 – 3246 . Google Scholar Crossref Search ADS PubMed WorldCat McAusland L , Vialet-Chabrand S , Davey P , Baker NR , Brendel O , Lawson T . 2016 . Effects of kinetics of light-induced stomatal responses on photosynthesis and water-use efficiency . New Phytologist 211 : 1209 – 1220 . Google Scholar Crossref Search ADS PubMed WorldCat McLeod AR , Long SP . 1999 . Free-air carbon dioxide enrichment (FACE) in global change research: a review . Advances in Ecological Research 28 : 1 – 56 . Google Scholar Crossref Search ADS WorldCat Medeiros JS , Ward JK . 2013 . Increasing atmospheric [CO2] from glacial to future concentrations affects drought tolerance via impacts on leaves, xylem and their integrated function . New Phytologist 199 : 738 – 748 . Google Scholar Crossref Search ADS PubMed WorldCat Midgley GF , Bond WJ . 2015 . Future of African terrestrial biodiversity and ecosystems under anthropogenic climate change . Nature Climate Change 5 : 823 – 829 . Google Scholar Crossref Search ADS WorldCat Mitchell PJ , O’Grady AP , Tissue DT , White DA , Ottenschlaeger ML , Pinkard EA . 2013 . Drought response strategies define the relative contributions of hydraulic dysfunction and carbohydrate depletion during tree mortality . New Phytologist 197 : 862 – 872 . Google Scholar Crossref Search ADS PubMed WorldCat Miyagi KM , Kinugasa T , Hikosaka K , Hirose T . 2007 . Elevated CO2 concentration, nitrogen use, and seed production in annual plants . Global Change Biology 13 : 2161 – 2170 . Google Scholar Crossref Search ADS WorldCat Moncrieff GR , Scheiter S , Bond WJ , Higgins SI . 2014 . Increasing atmospheric CO2 overrides the historical legacy of multiple stable biome states in Africa . New Phytologist 201 : 908 – 915 . Google Scholar Crossref Search ADS PubMed WorldCat Morgan JA , Lecain DR , Mosier AR , Milchunas DG . 2001 . Elevated CO2 enhances water relations and productivity and affects gas exchange in C3 and C4 grasses of the Colorado shortgrass steppe . Global Change Biology 7 : 451 – 466 . Google Scholar Crossref Search ADS WorldCat Morgan JA , Mosier AR , Milchunas DG , LeCain DR , Nelson JA , Parton WJ . 2004 . CO2 enhances productivity, alters species composition, and reduces digestibility of shortgrass steppe vegetation . Ecological Applications 14 : 208 – 219 . Google Scholar Crossref Search ADS WorldCat Morgan JA , Milchunas DG , LeCain DR , West M , Mosier AR . 2007 . Carbon dioxide enrichment alters plant community structure and accelerates shrub growth in the shortgrass steppe . Proceedings of the National Academy of Sciences of the USA 104 : 14724 – 14729 . Google Scholar Crossref Search ADS PubMed WorldCat Morgan JA , LeCain DR , Pendall E , et al. 2011 . C4 grasses prosper as carbon dioxide eliminates desiccation in warmed semi-arid grassland . Nature 476 : 202 . Google Scholar Crossref Search ADS PubMed WorldCat Nackley LL , Betzelberger A , Skowno A , et al. 2018 . CO2 enrichment does not entirely ameliorate Vachellia karroo drought inhibition: a missing mechanism explaining savannah bush encroachment . Environmental and Experimental Botany 155 : 98 – 106 . Google Scholar Crossref Search ADS WorldCat Nie M , Zou J , Xu X , Liang C , Fang C , Li B . 2018 . Comment on “Unexpected reversal of C3 versus C4 grass response to elevated CO2 during a 20-year field experiment” . Science 361 : eaau3016 . Google Scholar Crossref Search ADS PubMed WorldCat Osborne CP , Beerling DJ . 2003 . The penalty of a long, hot summer. Photosynthetic acclimation to high CO2 and continuous light in “living fossil” conifers . Plant Physiology 133 : 803 – 812 . Google Scholar Crossref Search ADS PubMed WorldCat Osborne CP , Sack L . 2012 . Evolution of C4 plants: a new hypothesis for an interaction of CO2 and water relations mediated by plant hydraulics . Philosophical Transactions of the Royal Society B: Biological Sciences 367 : 583 – 600 . Google Scholar Crossref Search ADS WorldCat Ostle NJ , Smith P , Fisher R , et al. 2009 . Integrating plant–soil interactions into global carbon cycle models . Journal of Ecology 97 : 851 – 863 . Google Scholar Crossref Search ADS WorldCat Overdieck D . 1989 . The effects of preindustrial and predicted future atmospheric CO2 concentration on Lyonia mariana (L.) D. Don . Functional Ecology 3 : 569 – 576 . Google Scholar Crossref Search ADS WorldCat Paschalis A , Katul GG , Fatichi S , Palmroth S , Way D . 2017 . On the variability of the ecosystem response to elevated atmospheric CO2 across spatial and temporal scales at the Duke Forest FACE experiment . Agricultural and Forest Meteorology 232 : 367 – 383 . Google Scholar Crossref Search ADS WorldCat Pinto H , Sharwood RE , Tissue DT , Ghannoum O . 2014 . Photosynthesis of C3, C3–C4, and C4 grasses at glacial CO2 . Journal of Experimental Botany 65 : 3669 – 3681 . Google Scholar Crossref Search ADS PubMed WorldCat Polley HW , Johnson HB , Marino BD , Mayeux HS . 1993 . Increase in C3 plant water-use efficiency and biomass over glacial to present CO2 concentrations . Nature 361 : 61 – 64 . Google Scholar Crossref Search ADS WorldCat Polley HW , Johnson HB , Derner JD . 2002 . Soil- and plant-water dynamics in a C3/C4 grassland exposed to a subambient to superambient CO2 gradient . Global Change Biology 8 : 1118 – 1129 . Google Scholar Crossref Search ADS WorldCat Quirk J , McDowell NG , Leake JR , Hudson PJ , Beerling DJ . 2013 . Increased susceptibility to drought-induced mortality in Sequoia sempervirens (Cupressaceae) trees under Cenozoic atmospheric carbon dioxide starvation . American Journal of Botany 100 : 582 – 591 . Google Scholar Crossref Search ADS PubMed WorldCat Quirk J , Bellasio C , Johnson DA , Osborne CP , Beerling DJ . 2018 . C4 savannah grasses fail to maintain assimilation in drying soil under low CO2 compared with C3 trees despite lower leaf water demand . Functional Ecology 33 : 388 – 398 . Google Scholar Crossref Search ADS WorldCat Reich PB , Hobbie SE , Lee TD , Pastore MA . 2018 . Unexpected reversal of C3 versus C4 grass response to elevated CO2 during a 20-year field experiment . Science 360 : 317 – 320 . Google Scholar Crossref Search ADS PubMed WorldCat Ripley B , Frole K , Gilbert M . 2010 . Differences in drought sensitivities and photosynthetic limitations between co-occurring C3 and C4 (NADP-ME) panicoid grasses . Annals of Botany 105 : 493 – 503 . Google Scholar Crossref Search ADS PubMed WorldCat Rogers A , Medlyn BE , Dukes JS , et al. 2017 . A roadmap for improving the representation of photosynthesis in Earth system models . New Phytologist 213 : 22 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat Sage RF . 1994 . Acclimation of photosynthesis to increasing atmospheric CO2 – the gas-exchange perspective . Photosynthesis Research 39 : 351 – 368 . Google Scholar Crossref Search ADS PubMed WorldCat Sage RF , Coleman JR . 2001 . Effects of low atmospheric CO(2) on plants: more than a thing of the past . Trends in Plant Science 6 : 18 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat Sato H , Kumagai T , Takahashi A , Katul GG . 2015 . Effects of different representations of stomatal conductance response to humidity across the African continent under warmer CO2-enriched climate conditions . Journal of Geophysical Research: Biogeosciences 120 : 979 – 988 . Google Scholar Crossref Search ADS WorldCat Scholes RJ , Archer SR . 1997 . Tree-grass interactions in savannahs . Annual Review of Ecology and Systematics 28 : 517 – 544 . Google Scholar Crossref Search ADS WorldCat Schulze E-D , Hall AE . 1982 . Stomatal responses, water loss and CO2 assimilation rates of plants in contrasting environments. In: Lange OL , Nobel PS , Osmond CB , Ziegler H , eds. Physiological plant ecology II: water relations and carbon assimilation . Berlin : Springer , 181–230. Google Preview WorldCat COPAC Schutz AEN , Bond WJ , Cramer MD . 2009 . Juggling carbon: allocation patterns of a dominant tree in a fire-prone savannah . Oecologia 160 : 235 . Google Scholar Crossref Search ADS PubMed WorldCat Seibt U , Rajabi A , Griffiths H , Berry JA . 2008 . Carbon isotopes and water use efficiency: sense and sensitivity . Oecologia 155 : 441 – 454 . Google Scholar Crossref Search ADS PubMed WorldCat Smith JM . 1982 . Evolution and the theory of games . Cambridge, UK: Cambridge University Press . Google Preview WorldCat COPAC Springer CJ , Ward JK . 2007 . Flowering time and elevated atmospheric CO2 . New Phytologist 176 : 243 – 255 . Google Scholar Crossref Search ADS PubMed WorldCat Srebotnik E , Messner K . 1994 . A simple method that uses differential staining and light microscopy to assess the selectivity of wood delignification by white rot fungi . Applied and Environmental Microbiology 60 : 1383 – 1386 . Google Scholar PubMed WorldCat Staver AC , Archibald S , Levin SA . 2011 . The global extent and determinants of savannah and forest as alternative biome states . Science 334 : 230 – 232 . Google Scholar Crossref Search ADS PubMed WorldCat Staver AC , Botha J , Hedin L . 2017 . Soils and fire jointly determine vegetation structure in an African savannah . New Phytologist 216 : 1151 – 1160 . Google Scholar Crossref Search ADS PubMed WorldCat Taylor SH , Ripley BS , Woodward FI , Osborne CP . 2011 . Drought limitation of photosynthesis differs between C3 and C4 grass species in a comparative experiment . Plant, Cell & Environment 34 : 65 – 75 . Google Scholar Crossref Search ADS PubMed WorldCat Taylor SH , Ripley BS , Martin T , De-Wet L-A , Woodward FI , Osborne CP . 2014 . Physiological advantages of C4 grasses in the field: a comparative experiment demonstrating the importance of drought . Global Change Biology 20 : 1992 – 2003 . Google Scholar Crossref Search ADS PubMed WorldCat Temme AA , Cornwell WK , Cornelissen JHC , Aerts R . 2013 . Meta-analysis reveals profound responses of plant traits to glacial CO2 levels . Ecology and Evolution 3 : 4525 – 4535 . Google Scholar Crossref Search ADS PubMed WorldCat Walker BJ , VanLoocke A , Bernacchi CJ , Ort DR . 2016 . The costs of photorespiration to food production now and in the future . Annual Review of Plant Biology 67 : 107 – 129 . Google Scholar Crossref Search ADS PubMed WorldCat Wand SJE , Midgley GF , Stock WD . 2001 . Growth responses to elevated CO2 in NADP-ME, NAD-ME and PCK C4 grasses and a C3 grass from South Africa . Functional Plant Biology 28 : 13 – 25 . Google Scholar Crossref Search ADS WorldCat Ward JK , Tissue DT , Thomas RB , Strain BR . 1999 . Comparative responses of model C3 and C4 plants to drought in low and elevated CO2 . Global Change Biology 5 : 857 – 867 . Google Scholar Crossref Search ADS WorldCat Way DA , Oren R , Kim HS , Katul GG . 2011 . How well do stomatal conductance models perform on closing plant carbon budgets? A test using seedlings grown under current and elevated air temperatures . Journal of Geophysical Research: Biogeosciences 116 : doi: https://doi.org/10.1029/2011JG001808 . WorldCat Wigley B , Cramer M , Bond W . 2009 . Sapling survival in a frequently burnt savannah: mobilisation of carbon reserves in Acacia karroo . Plant Ecology 203 : 1 . Google Scholar Crossref Search ADS WorldCat Wolf A , Anderegg WR , Pacala SW . 2016 . Optimal stomatal behavior with competition for water and risk of hydraulic impairment . Proceedings of the National Academy of Sciences of the USA 113 : E7222 – E7230 . Google Scholar Crossref Search ADS PubMed WorldCat Zachos J , Pagani M , Sloan L , Thomas E , Billups K . 2001 . Trends, rhythms, and aberrations in global climate 65 Ma to present . Science 292 : 686 – 693 . Google Scholar Crossref Search ADS PubMed WorldCat Zhisheng A , Kutzbach JE , Prell WL , Porter SC . 2001 . Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times . Nature 411 : 62 – 66 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes These authors contributed equally to this work. © The Author(s) 2019. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Response of photosynthesis, growth and water relations of a savannah-adapted tree and grass grown across high to low CO2
JF - Annals of Botany
DO - 10.1093/aob/mcz048
DA - 2019-08-02
UR - https://www.deepdyve.com/lp/oxford-university-press/response-of-photosynthesis-growth-and-water-relations-of-a-savannah-T0fK5VZY3y
SP - 77
VL - 124
IS - 1
DP - DeepDyve
ER -