Ecology Letters, (2004) 7: 427–440
Joy K. Ward* and John K. Kelly Department of Ecology and Evolutionary Biology, University of Kansas, Room 2041 Haworth Hall, 1200 Sunnyside Avenue, Lawrence, KS 66045-7534, USA *Correspondence: E-mail: [email protected]
Both authors contributed equally to this work.
Scaling up evolutionary responses to elevated CO2: lessons from Arabidopsis Abstract Results from norm of reaction studies and selection experiments indicate that elevated CO2 will act as a selective agent on natural plant populations, especially for C3 species that are most sensitive to changes in atmospheric CO2 concentration. Evolutionary responses to CO2 may alter plant physiology, development rate, growth, and reproduction in ways that cannot be predicted from single generation studies. Moreover, ecological and evolutionary changes in plant communities will have a range of consequences at higher spatial scales and may cause substantial deviations from ecosystem level predictions based on short-term responses to elevated CO2. Therefore, steps need to be taken to identify the plant traits that are most likely to evolve at elevated CO2, and to understand how these changes may affect net primary productivity within ecosystems. These processes may range in scale from molecular and physiological changes that occur among genotypes at the individual and population levels, to changes in community- and ecosystem-level productivity that result from the integrative effects of different plant species evolving simultaneously. In this review, we (1) synthesize recent studies investigating the role of atmospheric CO2 as a selective agent on plants, (2) discuss possible control points during plant development that may change in response to selection at elevated CO2 with an emphasis at the primary molecular level, and (3) provide a quantitative framework for scaling the evolutionary effects of CO2 on plants in order to determine changes in community and ecosystem productivity. Furthermore, this review points out that studies integrating the effects of plant evolution in response to elevated CO2 are lacking, and therefore more attention needs be devoted to this issue among the global change research community. Keywords Arabidopsis, carbon gain, elevated CO2, global change, net primary productivity, photosynthesis, scaling. Ecology Letters (2004) 7: 427–440
It is clear that global changes such as increasing temperatures, changes in precipitation, and altered atmospheric composition will have profound effects on the functioning and productivity of terrestrial ecosystems in the future (IPCC 2001; McCarthy et al. 2001). However, unlike changes in temperature and precipitation that vary across regional scales, rising atmospheric CO2 concentration from fossil fuel combustion and deforestation is occurring at the same rate on a global scale and is producing novel levels of carbon availability for photosynthesis (Schlesinger 1997). Atmospheric CO2 concentration has increased from 270 ppm at the onset of the Industrial Revolution
(125 years ago; Petit et al. 1999) to a current value of 375 ppm CO2 (39% increase). Therefore, it is critical to address the question of whether plants will evolve in response to increasing CO2, and whether these responses will influence community and ecosystem productivity at higher spatial scales (Ward et al. 2000; Kohut 2003). Elevated CO2 directly influences plant populations and communities by altering plant growth and development. CO2 concentration influences the rate of carbon fixation, which subsequently influences growth processes, plant functioning, and reproductive output (Norby et al. 1999; Pritchard et al. 1999; Kinugasa et al. 2003). Elevated CO2 has been shown to increase the growth and reproduction of C3 plants within a single generation of exposure, especially 2004 Blackwell Publishing Ltd/CNRS
428 J. K. Ward and J. K. Kelly
when other resources such as light, water, and nutrients are non-limiting (Ward & Strain 1999). A survey of 156 C3 species grown at elevated CO2 for one generation resulted in an average 41% stimulation in biomass production at 700 relative to 350 ppm CO2, although with considerable variation among species (Poorter 1993). Indirect, longer-term responses to elevated CO2 are also likely. These may be ecological, involving changes in the species composition of communities, or evolutionary, involving changes in the characteristics of individual species. Regarding ecological responses, the prevalence of C3 species relative to C4 species may increase in the future in some ecosystems (Dippery et al. 1995; Tissue et al. 1995; Sage & Kubien 2003), with particular advantages for forest trees (DeLucia et al. 1999). This is because C4 species may benefit less from increasing atmospheric CO2 due to lower saturation levels for photosynthesis. However, C4 plants often exhibit positive responses to CO2 and these are usually manifested through lower stomata conductance and improved water relations, especially in savanna and tallgrass prairie systems (Knapp & Medina 1999). In addition, herbivory of C3 species may increase due to changes in leaf tissue quality at elevated CO2 (Ehleringer et al. 2002; Goverde et al. 2002). Regarding evolutionary responses, elevated CO2 may induce directional selection within C3 plant populations for maximizing the conversion of greater carbon resources to fitness. In this way, evolution is likely to alter the direct physiological, developmental, and growth responses that are observed during a single generation of exposure to elevated CO2 (Thomas & Jasienski 1996; Ward et al. 2000; Klus et al. 2001). There is mounting evidence that changes in atmospheric CO2 over geologic time scales have influenced the evolution of land plants. Beerling et al. (2001) provide evidence for a link between reductions in atmospheric CO2 and the evolutionary advancement of megaphyll leaves during the late Devonian. These leaves displayed higher stomatal densities that increased transpiration (and the ability to cool leaves) beyond the capacity of previous leaf forms. Furthermore, during the Miocene decreasing CO2 concentrations are thought to have been a major selective agent for the evolution of C4 plants (Ehleringer et al. 1997; Cerling et al. 1998) that can maintain high rates of photosynthesis (Tissue et al. 1995) and growth (Dippery et al. 1995) under low CO2 concentrations. During the late Pleistocene (18 000–20 000 years ago), CO2 concentrations ranged between 180 and 200 ppm during glacial maxima (Petit et al. 1999), which are among the lowest values predicted to have occurred during the evolution of land plants (Berner 2003). Modern C3 plants grown under these low CO2 concentrations exhibit a 56–92% reduction in growth, and reproductive failure in some cases (Polley et al. 1993a,b; Dippery et al. 1995; Tissue et al. 1995; Cowling & Sage 2004 Blackwell Publishing Ltd/CNRS
1998). It is likely that these species were relatively more tolerant of low CO2 in the past and their poor performance under low CO2 concentrations reflects evolutionary changes in their physiologies and developmental patterns since the most recent glacial period. Given that CO2 limitation was likely a strong selective agent on C3 plants during recent geologic time scales, it is important to understand how evolutionary responses to low CO2 of the past may have influenced the genetic structure of current plant populations (Ward & Strain 1997). It is also worth noting that the majority of modern plant species have inevitably survived 2 million years of glacial CO2 cycles. Thus, on geological time scales, all currently represented species must have had the necessary genetic resources to adapt, or phenotypic plasticity to cope, with large and abrupt changes in atmospheric CO2 concentration. This review has three objectives. First, we synthesize recent studies investigating the role of atmospheric CO2 as a selective agent on plants. Second, we discuss possible control points during plant development that may change in response to selection at elevated CO2. Third, we develop a framework to consider how evolution of plant traits may interact with ecological processes to determine changes in community and ecosystem productivity. There have been several reviews on the microevolutionary responses of plants to CO2 that have focused on a variety of species including Raphanus raphanistrum and Plantago lanceolata (Curtis et al. 1996), Abutilon theophrasti (Thomas & Jasienski 1996), Prunella vulgaris and P. grandiflora (Schmid et al. 1996) and Brassica rapa (Kohut 2003). We focus on studies involving the model plant Arabidopsis thaliana (hereafter Arabidopsis, a C3 species). Arabidopsis is commonly used for evolutionary CO2 studies and the molecular tools available for this species allow a mechanistic dissection of plant responses to global change factors. Although rarely dominant within any one community, natural genotypes of this species have adapted to a wide range of climatic conditions throughout the world (Pigliucci 1998). Thus, it may be an excellent model system for determining how plants will evolve in response to global change drivers. Furthermore, this annual species is common on disturbed habitats, and therefore adaptive responses are associated with the ability to establish and reproduce rapidly, resulting in completion of the life cycle and high seed production before the onset of the next disturbance event. CO2 AS A SELECTIVE AGENT
A first step in determining whether elevated CO2 may act as a selective agent is to assess the degree of genetic variation within a species for CO2 response. The response of a single genotype to variation in CO2 concentration (or any environmental variable) can be characterized by its
Evolutionary responses to elevated CO2 429
High phenotypic plasticity
Low phenotypic plasticity
Genotype 2 Genotype 1
Genotype–environment interaction Genotype 1
Resource availability Figure 1 The norm of reaction depicts genotypic responses to variation in resource availability. (a) High plasticity genotypes will exhibit a more pronounced change in phenotype across a resource gradient than low plasticity genotypes. (b) Non-parallel reaction norms imply the existence of genotype-by-environment interactions within a population.
Ônorm of reactionÕ (Fig. 1a; see Schlichting & Pigliucci 1995 for a general treatment). In the present context, the reaction norm is the average phenotype produced by a genotype as a function of the atmospheric CO2 concentration in which it is grown. It reflects the plasticity of a genotype. Variation in response is indicated if different genotypes exhibit Ônon-parallelÕ reaction norms (e.g. Fig. 1b). Reactions norms are non-parallel if the magnitude of differences among genotypes changes along an environmental gradient. This phenomenon is denoted Ôgenotypeenvironment interactionÕ (GEI) in the context of evolutionary genetics. Numerous C3 plants have displayed significant GEI in experiments considering both present and future CO2 concentrations. A range of plant characters have been measured including carbon assimilation rate (Curtis et al. 1996; Klus et al. 2001), leaf chemistry (Van Der Kooij et al. 2000; Klus et al. 2001), stomatal regulation (Case et al. 1998), growth (Norton et al. 1995; Zhang & Lechowicz 1995; Schmid et al. 1996; Van Der Kooij et al. 2000), and reproduction (Curtis et al. 1994; Bazzaz et al. 1995;
Farnsworth & Bazzaz 1995). Ward & Strain (1997) estimated reaction norms for field-collected genotypes of Arabidopsis, investigating the effects of CO2 concentration (200–700 ppm) on individual components of fitness (total seed production and proportion of survival) and total fitness (estimated by multiplying survival proportion by total seed production). Genotypes exhibited reductions in estimated fitness ranging from 59 to 87% between 350 (current) and 200 (ice age) ppm CO2 (Fig. 2a), indicating that low CO2 is a stressful agent that exerts a highly variable effect across genotypes. Between 350 and 700 ppm CO2, genotypes showed increases in estimated fitness that ranged between 29 and 84% (Fig. 2a). Furthermore, there was clear evidence of changes in genotypic rank order across CO2 treatments for estimated fitness (Fig. 2a, see bold lines for an example). Such variation in genotypic responses to changes in CO2 concentration, both above and below the current value, indicate that novel CO2 levels may potentially impose strong selective pressure. When individual components of fitness were analysed in Ward & Strain (1997), the statistical analyses revealed that the effects of CO2 on survival were highly variable across genotypes, whereby changes in rank order occurred across CO2 treatments (Fig. 2b, see bold lines for an example). Furthermore, reduced seed production between 350 and 200 ppm CO2 was variable among genotypes, ranging from a 38 to 81% reduction (Fig. 2c). However, variation among genotypes for seed production produced between 350 and 700 ppm CO2 was less pronounced (Fig. 2c). Thus, depending upon the direction of CO2 change, the effects of individual components of fitness on total fitness of Arabidopsis may change, with survival being the dominant factor that differentiates genotypes at elevated CO2, and both survival and seed production differentiating genotypes at low CO2 (Ward & Strain 1997). Taken together, norm of reaction studies support the notion that changing CO2 concentrations are likely to impose selection on natural plant populations. While these experiments suggest that there is potential for evolutionary responses to occur, they may not indicate the full suite of traits that will evolve in response to selection. A more direct method to assess selection responses, and one that may be more relevant in an ecological context, is to maintain a population in the novel environment and allow evolution to occur over multiple generations. To our knowledge, there have been only two studies experimentally characterizing evolutionary responses to controlled selection in novel CO2 environments (Potvin & Tousignant 1996; Ward et al. 2000). In both studies, selection was imposed on total fitness, a trait that integrates the varied responses of individual plant characters. This more closely indicates how specific traits will evolve by natural selection imposed by novel CO2 levels in the field. 2004 Blackwell Publishing Ltd/CNRS
430 J. K. Ward and J. K. Kelly
(a) Fitness (×1000)
(b) 0.9 0.8 0.7 0.6 0.5
Seed number (×1000)
(c) 120 100 80 60 40 20 0 200
CO2 concentration (p.p.m.) Figure 2 The reaction norms of six Arabidopsis genotypes grown at three different CO2 levels (200, 350, and 700 ppm) evaluated for (a) total fitness, (b) proportion of survival, and (c) total seed number.
Tousignant & Potvin (1996) selected Brassica juncea for high silique mass over seven generations (see also Potvin & Tousignant 1996). There were two selection treatments: the first included gradual increases in CO2 and temperature above the current level coupled with heat shocks to simulate future conditions, while the second treatment used the current atmospheric CO2 concentration with constant temperature as a control. With this design, the independent effects of CO2 could not be assessed, although this allowed for measurements of selection with rising CO2 and climate change scenarios. Despite selection for increased silique biomass, the biomass of siliques decreased in both 2004 Blackwell Publishing Ltd/CNRS
treatments during the selection process, and decreased to a greater extent in the future environment than in the control environment. The authors concluded that B. juncea did not show genetic adaptation to the treatment simulating future conditions, and they attributed the lack of a positive response to inbreeding depression that may have been induced by the stressful future environment (Potvin & Tousignant 1996). This study suggests that high temperature may reduce the potential for evolution at elevated CO2. The Ward et al. (2000) study involved a five-generation artificial selection experiment using Arabidopsis selected at both low (200 ppm, glacial) and elevated (700 ppm, future) CO2. The base populations were derived from random crosses between field-collected genotypes (some of which were used in the norm of reaction experiment described above), allowing for selection on variation resulting from recombination and segregation. Selection on experimental populations was for high seed number (25% truncation), which is a major component of fitness, and changes in biomass production and development rate were monitored throughout selection. The experimental design also included control populations that were randomly selected in order to account for possible changes in the growth environment throughout the selection process (see Ward et al. 2000 for a more thorough description of the experimental design and conditions). Ward et al. (2000) found that selected plants had significantly higher seed production than control plants at both 200 and 700 ppm CO2 after only three generations of selection for high seed number, and this pattern persisted throughout the remainder of the selection process (through the fifth and final generation of selection, Fig. 3a,b). Furthermore, in this study, rates of microevolution under the novel CO2 treatments were among the highest observed for a global change factor (Bone & Farres 2001), in part due to the selection design and also due to the capacity of carbon resource availability to serve as a strong selective agent on annual plants. In follow-up to this selection experiment, a reciprocal transplant study was conducted to determine if CO2 was indeed the selective agent that was acting on these plants (Ward et al. 2000). It was found that plants selected for high seed number at 200 ppm CO2 had 30% higher seed number than plants selected at 700 ppm CO2 when both were grown at 200 ppm CO2 (P ¼ 0.0015). Because genetic adaptation can be defined as higher performance under the conditions of selection (Tousignant & Potvin 1996), this result indicated that plants selected at 200 ppm CO2 exhibited true adaptive responses to low CO2. In addition, overall changes in biomass production and development rate occurred in opposite directions during selection at 700 relative to 200 ppm CO2, supporting the possibility that both past and future CO2 concentrations may act as selective agents. Taken together, these results
Evolutionary responses to elevated CO2 431
Days to first flower
Figure 3 The mean values for seed number
indicated that the potential for CO2 to act as a selective agent may be very great, and evolutionary responses to CO2 may occur during time scales that are relevant to ecological studies. In the Ward et al. (2000) selection experiment, plants selected for high seed production at low CO2 (200 ppm) produced on average 35% higher total biomass compared with randomly selected control plants by the last generation of selection (Fig. 3e). The increased biomass was attributed to an extended period of productivity during which plants delayed the onset of reproduction as indicated by a longer period to reach first flower (Fig. 3c) and began senescing later than control plants (data not shown). Furthermore, selection and control plants at 200 ppm CO2 had similar amounts of structural dry matter prior to the initiation of reproduction, suggesting that integrated net carbon uptake was not affected by selection, at least prior to the onset of reproduction (Ward et al. 2000). This study indicated that
(e) Total mass (g)
(a and b), days to first flower (c and d), and total biomass (e and f) for both selected (for high seed number) and control (random selection) populations of Arabidopsis at the fifth and final generation of selection for high seed number. Plants were grown and selected at either 200 ppm (a, c and e) or 700 ppm CO2 (b, d and e). Lines connect mean values for control and selected populations maintained in the same growth chamber. Opened or closed circles within a CO2 treatment designate different growth chambers.
700 p.p.m. CO2
Seed number (X 1000)
Seed number (× 1000)
200 p.p.m. CO2
selection responses may improve plant performance at low CO2 as a result of genetic changes that influence developmental timing. Moreover, these results suggest that ancient plants may have had the potential to adapt to low CO2 conditions that may have influenced primary productivity. Elevated CO2 (700 ppm) also showed indications of acting as a selective agent on Arabidopsis. Overall, seed production increased by 17% for plants selected for high seed production at elevated CO2 relative to randomly selected control plants (in independent populations) by the final generation of selection (Fig. 3b), and significant differences between selection and control plants first occurred after only three generations (Ward et al. 2000). It is also likely that this is a conservative estimate of selection because inadvertent selection is suspected to have occurred in control plants (see Ward et al. 2000). Furthermore, it is worth noting that this increase is comparable with the 2004 Blackwell Publishing Ltd/CNRS
432 J. K. Ward and J. K. Kelly
stimulation in seed production that occurs between plants grown at 350 vs. 700 ppm CO2 (Jablonski et al. 2002), but is in addition to this increase. In contrast to responses at low CO2, plants selected at elevated CO2 exhibited either similar or lower final biomass production relative to randomly selected control plants (Fig. 3f, combined averages from all replicate populations represent a 16% reduction). Lower final biomass in some selection populations was a result of a reduction in the length of the life cycle and a shorter period for accumulation of biomass, as indicated by more rapid initiation of reproduction (Fig. 3d) and earlier senescence (data not shown). However, other populations did not vary for developmental timing or biomass production, and the mechanisms behind increased seed production in those populations are not yet understood. Unlike previous predictions (Schmid et al. 1996), this represents a case where developmental changes associated with evolutionary responses to elevated CO2 may constrain the potential for selection responses to increase biomass production above what is observed during a single generation. If responses that occur in natural field settings are similar to this selection experiment, we would not predict that net primary productivity (NPP) would be enhanced following evolutionary responses, and possibly carbon uptake may even be reduced relative to single generation responses. Kohut (2003) has recently criticized this experiment because (1) Arabidopsis is a self-fertilizing species and thus selection could not act on Ônovel allelic combinationsÕ produced by outcrossing, and (2) the experiment considered only fixed CO2 concentrations (200 and 700 ppm) and did not effectively represent the gradual changes in atmospheric CO2 that are actually occurring. The first criticism is incorrect. While Arabidopsis is predominantly selfing, the Ward et al. (2000) experimental populations were founded by inter-crossing field-collected ecotypes from various regions of the world. This yielded progeny that were heterozygous at all loci differing between the parental lines. Heterozygosity declined over the course of the experiment, but this did not prevent evolutionary change from occurring. More generally, in a primarily selfing species such as Arabidopsis, selection can change a population even in the absence of heterozygotes, so long as all individuals are not homozygous for the same allele (see Kelly 1999a,b for a more detailed treatment of selection and self-fertilization). Kohut’s second criticism is valid, although it should be noted that most ecological experiments use a limited number of discrete experimental treatments to characterize continuous causal variables. However, it does point to an important general consideration about selection imposed by changing atmospheric CO2. Over the next several decades, plant populations will be subject to a continuously changing 2004 Blackwell Publishing Ltd/CNRS
CO2 environment. Short-lived plants may evolve rapidly enough to effectively track these changes. However, species that cannot evolve this rapidly, such as longer-lived tree species, are likely to exhibit a substantial signature of previous selective environments with lower atmospheric CO2. Sage & Cowling (1999) point out that C3 species are sometimes unresponsive to elevated CO2 (Bazzaz 1990; Koch & Mooney 1996) and they argue that these cases may reflect the effects of a past selective environment of lower CO2 on plant physiology. A related issue is that selection on a population at any point along a CO2 gradient will likely affect the entire norm of reaction for many physiological traits. This is because the expressions of a trait at different CO2 levels will be genetically correlated (Falconer & Mackay 1996; Kingsolver et al. 2001). Strong positive correlations are likely for similar environmental values, e.g. plant genotypes that exhibit the highest carbon gain at 300 ppm CO2 are also likely to have relatively high values at 310 ppm. However, we expect lesser correlations, perhaps even negative values with some traits, when comparing environments that are more dissimilar (e.g. 300 vs. 700 ppm). Thus, reaction norms that accurately depict plant performance across a range of CO2 values may prove particularly informative. This is no small feat given that characterizing genotypic performance (and also variation among genotypes) within a single CO2 level requires substantial effort, and because CO2 level varies continuously, there are an infinite number of distinct CO2 levels to be considered. In this regard, global change research may benefit from new analytical techniques that are being developed in quantitative genetics and animal breeding (Kirkpatrick et al. 1990, 1994; Jaffrezic & Pletcher 2000; Kingsolver et al. 2001). The challenge is to interpolate a continuous function (the reaction norm) with data from a limited number of discrete CO2 treatments (see Thomas & Jasienski 1996). A range of techniques, both parametric and non-parametric, have been developed to accomplish this interpolation and to estimate the pattern of variation among reaction norms (Kirkpatrick et al. 1994; Meyer & Hill 1997; Gilbert et al. 1998). If coupled with appropriate measurements of selection, application of these procedures may allow global change researchers to make quantitative predictions of phenotypic evolution across the predicted CO2 gradient. CHANGES RESULTING FROM SELECTION AT ELEVATED CO2
The norm of reaction studies and selection experiments suggest that elevated CO2 will act as a selective agent on natural plant populations, especially for C3 species that are most sensitive to changes in atmospheric CO2 concentration. Therefore, we need to identify the plant traits that
Evolutionary responses to elevated CO2 433
are most likely to evolve at elevated CO2, and to understand how these changes may affect NPP within ecosystems. These processes may range in scale from molecular and physiological changes that occur among genotypes at the individual and population level, to changes in community and ecosystem level processes that result from the integrative effects of different plant species evolving simultaneously, and the effects of these changes on higher trophic levels. Below we provide several examples of possible changes that may occur as a consequence of plant evolutionary responses to changing CO2 concentration. Very little is known about the actual changes that may occur as plants respond to selective pressure from changes in CO2 (low or elevated), and therefore the responses that are presented below are likely candidates to be affected during evolutionary responses to CO2. Physiological and molecular-level responses
Photosynthesis A multitude of studies report that C3 photosynthesis is enhanced at elevated CO2 (Griffin & Seemann 1996; Ward & Strain 1999), and this response often contributes to higher growth and reproduction within a single generation of exposure to elevated CO2 (Curtis & Wang 1998). Increases in net photosynthetic rate are due to greater availability of the CO2 substrate and reductions in the rate of photorespiration (Sage 1994; Ward et al. 1999). However, photosynthetic capacity may decrease following initial exposure to elevated CO2 (also known as photosynthetic down-regulation), and this acclimation response usually occurs when there are limited sinks for utilization of photosynthate (Sims et al. 1999; Tissue et al. 1999). Photosynthetic acclimation most commonly occurs through reduced levels of ribulose-1,5 bisphosphatecarboxylase/oxygenase (hereafter Rubisco) protein content and higher carbohydrate content in leaves (Moore et al. 1999). At the molecular level, the transcription and protein abundance of the Rubisco small subunit (rbcS) appears to be most influential in modulating overall Rubisco protein content, and transcription of this subunit is reduced during acclimation responses (Moore et al. 1999). Sugars have been implicated as signalling molecules (hexokinase sensing) that can repress the transcription of photosynthetic genes under elevated CO2 conditions (Rolland et al. 2002). This response has also been coupled with nitrogen status, because transcription of nitrate reductase (nitrate assimilation) is regulated by a pathway that involves sensitivity of glutamine to sucrose levels (Paul & Foyer 2001). Therefore, the influences of elevated CO2 on carbon and nitrogen uptake are related at the molecular level (Stitt & Krapp 1999).
Photosynthetic acclimation may result in increased fitness, because it reduces excess rubisco production and maintains sugar homeostasis (Rolland et al. 2002), and therefore this response may divert resources to other processes that increase survival and reproductive output (Sage 1994). Cook and co-workers (1998) provided evidence in support of this idea when investigating the physiological properties of Nardus stricta that occurred near and far from CO2 springs in Iceland. N. stricta growing near CO2 springs were exposed to elevated CO2 (c. 650 ppm) for at least several hundred (if not thousands) of years, which may have allowed sufficient time for evolutionary responses to occur. Plants growing near vents showed a 19% reduction in rubisco content and a 35% reduction in rubisco activity. In addition, Saurer et al. (2003) more recently found evidence that Quercus ilex L. trees exhibited evidence of photosynthetic acclimation (down-regulation) under elevated CO2 conditions near a natural CO2 spring, as inferred from carbon isotope ratios. Furthermore, these trees were growing under dry, nutrient-poor conditions, increasing the likelihood of photosynthetic acclimation responses. Photosynthetic acclimation in these studies may involve strict plastic responses that do not involve genetic change. Although given the length of time of elevated CO2 exposure, these responses may have resulted from selection effects that mimic the short-term acclimation response. For CO2 vent (or natural spring) studies, reciprocal transplant experiments and/or common garden studies will be needed to distinguish these alternatives. Under certain scenarios, however, photosynthetic acclimation may not be adaptive based on the availability of other resources (e.g. high nitrogen levels), primarily since a reduction in rubisco content and activity may limit the potential for carbon fixation under elevated CO2 conditions. This would be particularly relevant to non-determinant species that could increase growth indefinitely in response to higher carbon resources. To the author’s knowledge, there have been no studies that directly correlate acclimation of photosynthesis to measurements that are closely tied with fitness (survival, reproduction). Such measurements are necessary for determining whether acclimation of photosynthesis serves as an adaptive response to elevated CO2, and would provide a fuller understanding of how plant productivity may change during evolutionary responses to elevated CO2. Stomatal development and regulation Plants adjust stomatal conductance (g) for decreased water loss when carbon is abundant, and for increased CO2 diffusion into leaves when carbon is limiting (at the expense of greater water loss). Therefore, within a single generation, C3 plants generally exhibit reduced stomatal conductance in response to elevated CO2 (average 21% reduction at twice 2004 Blackwell Publishing Ltd/CNRS
434 J. K. Ward and J. K. Kelly
the current CO2 value among species), resulting in lower transpiration rates and higher water-use-efficiency (Medlyn et al. 2001). In some species, however, stomatal adjustments may be short-lived (Bunce 1992), rendering stomata less responsive to elevated CO2 over time. Leaf stomatal density (number of stomata per leaf area) may also be influenced by elevated CO2 (Woodward et al. 2002). For example, herbarium specimens have shown reductions in stomatal density with rising CO2 since the onset of the industrial revolution (Woodward 1993), and a variety of modern Arabidopsis genotypes exhibited an average 22% reduction in stomata density when grown at 700 ppm (future) compared to 350 ppm (modern) CO2 (Hetherington & Woodward 2003). In addition, fossil leaves have shown higher stomatal densities during periods of low CO2 that occurred within the past 400 million years (Beerling & Woodward 1997), particularly during glacial periods of the late Pleistocene when atmospheric CO2 concentrations were extremely low (180–200 ppm; Beerling et al. 1993; Hetherington & Woodward 2003). Furthermore, it is clear that modern plants show non-linearity for the response of stomatal density to CO2 concentration, with relatively higher sensitivity to CO2 concentrations below the modern value compared with above (Beerling & Royer 2002). Such responses may limit some of the applications for using stomatal density in fossils as an indicator of ancient CO2 concentration (McElwain & Chaloner 1995). The molecular basis for the sensitivity of stomatal density to CO2 is being elucidated in Arabidopsis and is likely associated with the high carbon dioxide (HIC) gene that encodes proteins involved in the formation of long-chain fatty acids in leaves. When this gene is disrupted, very high stomatal densities occur in the leaves of plants grown at elevated CO2 (Gray et al. 2000), and therefore this gene plays some role in the modulation of stomatal number in response to CO2 cues. Furthermore, there is recent evidence suggesting that the detection of ambient CO2 concentration by this signal transduction pathway may involve messages that are transferred from mature to developing leaves where stomatal density is pre-determined at an early stage of development (Lake et al. 2002). Evolutionary modifications in stomatal regulation and development (and the interactions of the two) will likely depend on several coordinated signal transduction pathways. Furthermore, changes in stomatal regulation and development are possible mechanisms, among many, that may alter plant productivity over evolutionary time scales, and may be key to understanding aspects of physiological adaptation to elevated CO2. It is expected that increasing CO2 will invoke selection pressure for reduced stomatal density and lower maximum stomatal conductance, in a fashion that optimizes water conservation relative to carbon fixation for a given microhabitat. It is also important to 2004 Blackwell Publishing Ltd/CNRS
point out that the mechanism of optimization of stomatal regulation may vary, depending on the availability of other resources, particularly water. For example, Woodward et al. (2002) experimented with Arabidopsis genotypes grown under elevated CO2 with both a well-watered and mild drought treatment. They found that a genotype (Col-0) exhibiting relatively low reductions in stomatal density at elevated CO2 had the highest fitness (measured as flower production) under well-watered conditions, whereas a genotype (Ws) exhibiting relatively high reductions in stomatal density at elevated CO2 had the highest fitness advantage under mild drought conditions. Thus, the availability of other resources within the plant microhabitat may play a large role in determining the extent to which changes in stomatal regulation and development will function as adaptive mechanisms for elevated CO2 response. Future physiological and molecular-level studies The completion of the full genome sequence for Arabidopsis (e.g. Salanoubat et al. 2000; Tabata et al. 2000) should greatly facilitate future studies of plant physiology, development, and growth. One application resulting from this endeavor has been the development of a microarray chip that can be used to measure the simultaneous expression of all the c. 28 000 genes of Arabidopsis (e.g. Arabidopsis ATH1 Genome Array, Affymetrix, Santa Clara, CA, USA). This technology has the potential to revolutionize our understanding of plant responses to the environment through the comparison of gene expression profiles. Changes in gene expression may underpin the plastic (physiological) responses that occur within a single plant on the order of days to weeks (e.g. photosynthetic acclimation). This may occur by an increase in the activity of transcription factor(s) as induced directly by the environmental agent or indirectly through a distinct signal transduction pathway. Evolutionary responses may also be driven by changes in gene expression, possibly through gene sequence changes in a promoter region or through genetic changes that alter transcription factors. Microarray technologies have already been applied towards gaining insights into underlying regulatory networks involved in salt and drought stress (Bray 2002; Ozturk et al. 2002), thermal stress (Seki et al. 2002), and ozone stress (Matsuyama et al. 2002). Growth and development
Past studies indicate that elevated CO2 generally enhances the reproductive output of herbaceous species within a single generation by altering flower number, fruit set, and seed production (Ward & Strain 1997; Jablonski et al. 2002). In addition, elevated CO2 often alters the development rate of plants, and in most cases decreases the time required to transition from vegetative to floral stages (LaDeau & Clark
Evolutionary responses to elevated CO2 435
2001). Alterations in flowering time may have large effects on fitness due to a trade-off between accumulation of resources for reproduction and completion of the life cycle before the end of the growing season. In the Ward et al. (2000) selection experiment described above, selection for higher seed production at elevated CO2 (700 ppm) was correlated with faster development rates in a subset of replicate populations (see Fig. 3d, chamber 4 response). This selective response may be accentuated in regions where frost or drought limit reproduction at the end of the growing season. Therefore, there is likely to be a strong inter-play between changes in developmental timing and total fitness under elevated CO2 conditions (although different mechanisms may be equally important as indicated by the selection experiment). The effects of elevated CO2 on plant development need to be understood in more detail, especially at the molecular level. It is unclear how CO2 affects the time to transition from vegetative to floral stages in annual plants. CO2 may influence the rate of maturation in the vegetative phase (e.g. time to shift from juvenile to adult leaf forms; see Simpson et al. 1999), which may ultimately influence the timing of floral induction. Alternatively, elevated CO2 may affect floral induction through changes in carbohydrate status, or may even possibly act as a direct environmental stimulus for floral gene expression, although there is currently no evidence that this is the case. Regardless, the Arabidopsis model system is particularly well suited for determining this mechanism. c. 80 genes in this species have been shown to function in the transition from vegetative to floral stages (Putterill 2001), and it would be fascinating to determine how CO2 cues are incorporated into signal transduction pathways for floral induction. Whole-plant phenotypes such as total biomass, leaf area, and developmental timing are ultimately determined by internal factors such as the regulation of carbon uptake and metabolism. The pattern of character correlations among morphological characters is also a function of these underlying physiological processes. As a consequence, changes in CO2 levels alter not only the mean values for traits such as biomass and leaf area, but also the relationships between these traits (see Fig. 5 of Thomas & Jasienski 1996). This is noteworthy because character correlations are a critical component of evolutionary models and may act as a major ÔconstraintÕ on evolutionary change (e.g. Etterson & Shaw 2001). Microarray studies may provide an understanding of correlations between morphological characters at a more fundamental level. If we can predict changes in character correlations (evolutionary constraints) from responses of the underlying molecular and physiological processes, this will greatly improve the predictive power of evolutionary models.
SCALING UP EVOLUTIONARY RESPONSES TO ELEVATED CO2
Net primary productivity is defined as carbon gain (via photosynthesis) minus carbon loss (via respiration and photorespiration) of plants per unit time. If plant species evolve in response to novel CO2 concentrations, these changes are likely to alter the NPP of entire plant communities and may possibly have cascading effects on the productivity and functioning of whole ecosystems. As a consequence, studies that address evolutionary responses to CO2 become a critical component of global change research, and are necessary for accurately assessing the effects of elevated CO2 concentrations on plant productivity. In this section, we develop a simple conceptual framework to integrate direct (physiological) and indirect (ecological and evolutionary) plant responses to determine changes in community productivity. Net primary productivity is generally considered a property of entire plant communities. However, to consider the consequences of evolutionary changes, it is useful to first focus on a single species (species j). The species level NPP depends on the size of the population, the genetic composition of the population (which influences per capita carbon gain and loss), and the environment. These dependencies can be formally depicted with a simple equation: X NPPSpecies j ½ca ; t ¼ Nj ½t gi ½tKi ½ca ¼ Nj ½tK j ½ca ; t ð1Þ i
where Nj[t] is the number of individuals in species j at time t, gi[t] is the frequency of genotype i at time t, and Ki[ca] is the rate of carbon gain of genotype i per unit time given an atmospheric CO2 level ca. Evaluated over a range of ca values, Ki[ca] represents the norm of reaction of genotype i. As growth is likely to be influenced by many loci, the summation in the central term is taken over all possible multi-locus genotypes. The rightmost term distills the summation over genotypes into a population average of genotypic values, K j ½ca ; t, that we denote the Ôper capita carbon gainÕ of species j. This quantity is essentially the species-level norm of reaction. As a population evolves, changes in genotype frequencies will cause changes in Kj ½ca ; t. Let us now consider species j at a future point in time (t ¢) when the atmospheric CO2 concentration has increased from ca to ca¢. The change in per capita carbon gain is: X gi ½tPi K j ½ca 0 ; t0 K j ½ca ; t ¼ i
Plastic response X þ Dgi Ki ½ca 0
2004 Blackwell Publishing Ltd/CNRS
436 J. K. Ward and J. K. Kelly
where Pi ¼ Ki[ca¢] ) Ki[ca] is the plastic response of genotype i to environmental change and Dgi is the change in the frequency of genotype i from time t to time t ¢. Both of the terms on the right-hand side P of eqn 2 have a simple interpretation. The first term, gi ½tPi , is the direct rei sponse of plants to environmental change via phenotypic plasticity. This term is essentially a weighted average of the plastic responses of individual genotypes. Most experimental studies that evaluate plant performance over a range of CO2 concentrations (e.g. Poorter 1993; Ward & Strain 1997) are estimating this component of change. P The second term in the equation, Dgi Ki ½c0a Pi , reprei sents the effects of evolutionary changes on per capita carbon gain. Several important technical points about the term merit comment. First, selection will change the whole population norm of reaction. This expression evaluates that reaction norm only at the future CO2 concentration ca¢. Second, the summation is taken over all genotypes present in both the current (time t) and the future (time t ¢) populations. Short-term selection experiments (e.g. Potvin & Tousignant 1996; Ward et al. 2000) provide a valuable indication of immediate responses to selection, involving changes in allelic frequencies for loci that are currently polymorphic in the population. Over longer time scales however, mutations at loci that are not currently polymorphic will likely contribute to evolutionary change. These mutations may or may not have the same consequences for NPP as those currently present in natural populations. Current experimental data indicate that the plastic response of plants to increasing CO2 will generally be positive, with greater carbon gain at higher CO2 concentrations, at least among C3 plants (Griffin & Seemann 1996; Curtis & Wang 1998). However, whether the evolutionary response is reinforcing (positive) or antagonistic (negative) to the plastic response depends critically on the types of plant traits that are affected by selection (see previous section). If genotypes with the greatest carbon gain at higher CO2 outcompete genotypes with lower rates of carbon gain, the evolutionary response will reinforce the plastic response. However, several lines of evidence indicate that it is premature to assume that selection will generally act in this way. In the selection experiment of Ward et al. (2000), plant biomass declined or was unaffected as a result of adaptation to higher CO2 levels (Fig. 3f). Also, studies of plant populations occurring near CO2 springs suggest that selection may effectively down-regulate photosynthetic activity (and hence per capita carbon gain) in high CO2 environments (Cook et al. 1998; Saurer et al. 2003). Resources (e.g. nitrogen) that are not used to produce Rubisco can be diverted to increase reproductive output (which is the real target of selection). Finally, the evolutionary changes in the per capita carbon gain of a species
2004 Blackwell Publishing Ltd/CNRS
may be minimal if (1) genetic variation is low, or (2) carbon gain is only loosely related to traits under selection (e.g. Bazzaz et al. 1995), or (3) evolution is limited by antagonistic genetic correlations among traits under selection (e.g. Etterson & Shaw 2001). Responses of plants at the population level will necessarily affect processes at higher spatial scales. It is thus worthwhile to extrapolate these calculations to the next higher level of organization. Let NPPcommunity[ca, t] denote the total NPP of a plant community at time t given atmospheric CO2 concentration ca: X Nj ½tK j ½ca ; t ð3Þ NPPcommunity ½ca ; t ¼ j
where K j ½ca ; t is defined by eqn 1. The change in community NPP from now to a future point in time (t ¢) with CO2 concentration ca¢ is NPPcommunity ½ca 0 ; t 0 NPPcommunity ½ca ; t X Nj ½tfK j ½ca 0 ; t 0 K j ½ca ; tg ¼ j
Evolutionary and plastic response X þ DNj K j ½ca 0 ; t 0
Ecological response where DNj ¼ Nj[t ¢] ) Nj[t] is the change in the number of individuals of species j from time t to time t ¢. As with the change in per capita carbon gain, each of the terms in eqn4 has a simple, qualitative interpretation. The P first term, Nj ½tfK j ½ca 0 ; t 0 K j ½ca ; tg, is a simple j
summation of both plastic and evolutionary responses of individual species to CO2 change [the response of species j (in brackets) is given by eqn 2]. The contribution of each species is weighted by their current population size. Thus, we expect that evolutionary and plastic changes in the most abundant species will have the largest impact on community P NPP. The second term, DNj K j ½ca0 ; t 0 , is the ecological j
response to CO2 change and it reflects the impact of changes in species abundance. This term would also include effects of species extinctions (where Nj[t ¢] ¼ 0) and invasion (where Nj[t] ¼ 0). Thus, the summation must be taken over all species present at either time point. The ecological response implicitly includes changes at both higher and lower trophic levels that impact plant species abundances. The purpose of decomposing changes in NPP into individual components is to clarify how different types of responses (plastic, evolutionary, ecological) combine to change the productivity of plant communities. This exercise indicates how different types of studies (norm of reaction
Evolutionary responses to elevated CO2 437
studies and ecological manipulations) can be combined to predict responses to changing CO2 concentrations. However, the partitioning outlined by eqns 1–4 should not be taken to mean that the underlying processes are independent. As argued earlier, evolution will affect the entire norm of reaction of a species and hence the plastic response to environmental change. Note also that the ecological response term of eqn 4 is directly dependent on plastic and evolutionary responses. The DNj for each species is multiplied by the average per capita carbon gain of future genotypes, K j ½ca 0 ; t 0 . This will differ from the value for current genotypes, K j ½ca ; t because of both evolution and plasticity. Influences from the other direction are also likely; changes in the relative abundance and density of species will alter the growth of individual plants (due to plasticity), and will also affect evolutionary responses by changing competitive regimes (Bazzaz et al. 1995; Andalo et al. 2001). Plants often show the highest plastic growth responses to elevated CO2 when other resources such as nutrients, water, and light are non-limiting (Curtis & Wang 1998). Changes in species densities will clearly affect resource abundances and hence the magnitude of the plastic response term of eqn 2. Changes in species abundances and competitive regimes will affect the nature and strength of natural selection on a range of plant characters. Selection may favour increased plant height via stem elongation if light becomes increasingly limiting with higher plant densities (Dudley 1996; Dudley & Schmitt 1996). Furthermore, the allocation of biomass between above- and belowground structures may become an important component of selection at elevated CO2, with the best allocation pattern dependent on the specific nature of competitive interactions and availability of soil resources. Careful genetic studies may prove particularly valuable here, as some developmental modifications may evolve rapidly, while others may not. For example, the allocation of biomass between roots and shoots is rarely altered in response to elevated CO2 within a single generation, if developmental stage is taken into account (Norby 1994). In addition, variation for allocation of biomass between roots and shoots in response to elevated CO2 was not detected within or among populations of P. lanceolata (Klus et al. 2001). Thus, selection may not be able to rapidly modify the allocation of resources to roots vs. shoots (at a given developmental stage) even if such a modification may be advantageous. FUTURE PERSPECTIVES
Ecological and evolutionary changes in plant communities will have a range of consequences at higher spatial scales
and may cause substantial deviations from ecosystem level predictions based on short-term responses to elevated CO2. The integration of responses at higher spatial scales is closely tied to questions pertaining to carbon cycling and carbon sequestration that have important implications for evaluating whether terrestrial ecosystems will serve as carbon sources or sinks in the future (Pataki et al. 2003). It is therefore critical to incorporate the effects of evolutionary processes when addressing these questions in order to accurately predict the future status of ecosystems. Unfortunately, studies that integrate the effects of plant evolution in response to elevated CO2 at spatial scales larger than the population level are lacking, and therefore more attention needs be devoted to this issue among the global change research community. ACKNOWLEDGEMENTS
We sincerely thank Ben Burgert and Liza Holeski for reviewing a draft of this manuscript. J.K.W. was supported by a grant from the USDA (2003-00791). J.J.K. was supported by an NSF (DEB-9903758) and NIH grant (1 R01 GM60792-01A1). Both authors were supported by a NSF EPSCoR grant (KAN32118). REFERENCES Andalo, C., Goldringer, I. & Godelle, B. (2001). Inter- and intragenotypic competition under elevated carbon dioxide in Arabidopsis thaliana. Ecology, 82, 157–164. Bazzaz, F.A. (1990) The response of natural ecosystems to the rising global CO2 levels. Annu. Rev. Ecol. Syst., 21, 167–196. Bazzaz, F.A., Jasienski, M., Thomas, S.C. & Wayne, P. (1995). Microevolutionary responses in experimental populations of plants to CO2-enriched environments: parallel results from two model systems. Proc. Nat. Acad. Sci. USA, 92, 8161–8165. Beerling, D.J. & Royer, D.L. (2002). Reading a CO2 signal from fossil stomata. New Phytol., 153, 387–397. Beerling, D.J. & Woodward, F.I. (1997). Changes in land plant function over the Phanerozoic: reconstructions based on the fossil record. Bot. J. Linn. Soc., 124, 137–153. Beerling, D.J., Chaloner, W.G., Huntley, B., Pearson, J.A. & Tooley, M.J. (1993). Stomatal density responds to the glacial cycle of environmental change. Proc. R. Soc. Lond. B Biol. Sci., 251, 133–138. Beerling, D.J., Osborne, C.P. & Chaloner, W.G. (2001). Evolution of leaf-form in land plants linked to atmospheric CO2 decline in the Late Palaeozoic era. Nature, 410, 352–354. Berner, R.A. (2003). The long-term carbon cycle, fossil fuels and atmospheric composition. Nature, 426, 323–326. Bone, E. & Farres, A. (2001). Trends and rates of microevolution in plants. Genetica, 112, 165–182. Bray, E.A. (2002). Classification of genes differentially expressed during water-deficit stress in Arabidopsis thaliana: an analysis using microarry and differential expression data. Ann. Bot., 89, 803–811.
2004 Blackwell Publishing Ltd/CNRS
438 J. K. Ward and J. K. Kelly
Bunce, J. (1992). Stomatal conductance, photosynthesis and respiration of temperate deciduous tree seedlings grown outdoors at an elevated concentration of carbon dioxide. Plant Cell Environ., 15, 541–549. Case, A.L., Curtis, P.S. & Snow, A.A. (1998). Heritable variation in stomatal responses to elevated CO2 in wild radish, Raphanus raphanistrum (Brassicaceae). Am. J. Bot., 85, 253– 258. Cerling, T.E., Ehleringer, J.R. & Harris, J.M. (1998). Carbon dioxide starvation, the development of C4 ecosystems, and mammalian evolution. Philos. Trans. R. Soc. Lond. B Biol. Sci., 353, 159–171. Cook, A.C., Tissue, D.T., Roberts, S.W. & Oechel, W.C. (1998). Effects of long-term elevated [CO2] from natural CO2 springs on Nardus stricta: photosynthesis, biochemistry, growth and phenology. Plant Cell Environ., 21, 417–425. Cowling, S.A. & Sage, R.F. (1998). Interactive effects of low atmospheric CO2 and elevated temperature on growth, photosynthesis and respiration in Phaseolus vulgaris. Plant Cell Environ., 21, 427–435. Curtis, P.S. & Wang, X. (1998). A meta-analysis of elevated CO2 effects on woody plant mass, form, and physiology. Oecologia, 113, 299–313. Curtis, P.S., Snow, A.A. & Miller, A.S. (1994). Genotype-specific effects of elevated CO2 on fecundity in wild radish (Raphanus raphanistrum). Oecologia, 97, 100–105. Curtis, P.S., Klus, D.J., Kalisz, S. & Tonsor, S.J. (1996). Intraspecific variation in CO2 responses in Raphanus raphanistrum and Plantago lanceolata: assessing the potential for evolutionary change with rising atmospheric CO2. In: Carbon Dioxide, Populations, and Communities (eds Ko¨rner, C. & Bazzaz, F.A. ). Academic Press, San Diego, pp. 13–22. DeLucia, E.H., Hamilton, J.G., Naidu, S.L., Thomas, R.B., Andrews, J.A., Finzi, A. et al. (1999). Net primary production of a forest ecosystem with experimental CO2 enrichment. Science, 284, 1177–1179. Dippery, J.K., Tissue, D.T., Thomas, R.B. & Strain, B.R. (1995). Effects of low and elevated CO2 on C3 and C4 annuals. I. Growth and biomass allocation. Oecologia, 101, 13–20. Dudley, S.A. (1996). The response to differing selection on plant physiological traits: Evidence for local adaptation. Evolution, 50, 103–110. Dudley, S.A. & Schmitt, J. (1996). Testing the adaptive plasticity hypothesis: density-dependent selection on manipulated stem length in Impatiens capensis. Am. Nat., 147, 445–465. Ehleringer, J.R., Cerling, T.E. & Helliker, B.R. (1997). C4 photosynthesis, atmospheric CO2, and climate. Oecologia, 112, 285–299. Ehleringer, J.R., Cerling, T.E. & Dearing, M.D. (2002). Atmospheric CO2 as a global change driver influencing plant-animal interactions. Integ. Comp. Biol., 42, 424–430. Etterson, J.R. & Shaw, R.G. (2001). Constraint to adaptive evolution in response to global warming. Science, 294, 151– 154. Falconer, D.S. & Mackay, T.F.C. (1996). Introduction to quantitative genetics. Prentice Hall, London. Farnsworth, E.J. & Bazzaz, F.A. (1995). Inter- and intra-generic differences in growth, reproduction, and fitness of nine herbaceous annual species grown in elevated CO2 environments. Oecologia, 104, 454–466.
2004 Blackwell Publishing Ltd/CNRS
Gilbert, P., Moreteau, B., David, J.R. & Scheiner, S.M. (1998). Describing the evolution of reaction norm shape: body pigmentation in Drosophila. Evolution 52, 1501–1506. Goverde, M., Arnone, J.A. & Erhardt, A. (2002). Species-specific reactions to elevated CO2 and nutrient availability in four grass species. Basic Appl. Ecol., 3, 221–227. Gray, J.E., Holroyd, G.H., van der Lee, F.M., Bahrami, A.R., Sijmons, P.C., Woodward, F.I. et al. (2000). The HIC signalling pathway links CO2 perception to stomatal development. Nature, 408, 713–716. Griffin, K.L. & Seemann, J.R. (1996). Plants, CO2 and Photosynthesis in the 21st Century. Chem. Biol., 3, 245–254. Hetherington, A.M. & Woodward, F.I. (2003). The role of stomata in sensing and driving environmental change. Nature, 424, 901–908. IPCC (2001). Climate Change 2001: synthesis report.In: A Contribution of Working Groups I, II, and III to the Third Assessment Report of the Intergovernmental Panel on Climate Change (eds Watson, R.T. & the core writing team). Cambridge University Press, Cambridge, 398p. Jablonski, L.M., Wang, X. & Curtis, P.S. (2002). Plant reproduction under elevated CO2 conditions: a meta-analysis of reports on 79 crop and wild species. New Phytol., 156, 9–26. Jaffrezic, F., Pletcher, S. (2000). Statistical models for estimating the genetic basis of repeated measures and other function-valued traits. Genetics, 156, 913–922. Kelly, J.K. (1999a). Response to selection in partially self fertilizing populations. I. Selection on a single trait. Evolution, 53, 336–349. Kelly, J.K. (1999b). Response to selection in partially self fertilizing populations. II. Selection on multiple traits. Evolution, 53, 350–357. Kingsolver, J., Gomulkiewicz, R. & Carter, P. (2001). Variation, selection and evolution of function-valued traits. Genetica, 112, 87–104. Kinugasa, T., Hikosaka, K. & Hirose, T. (2003). Reproductive allocation of an annual, Xanthium canadense, at an elevated carbon dioxide concentration. Oecologia, 137, 1–9. Kirkpatrick, M., Lofsvold, D. & Bulmer, M. (1990). Analysis of the inheritance, selection, and evolution of growth trajectories. Genetics 124:979–993. Kirkpatrick, M., Hill, W. & Thompson, R. (1994). Estimating the covariance structure of traits during growth and aging, illustrated with lactation in dairy-cattle. Genet. Res., 64, 57–69. Klus, D.J., Kalisz, S., Curtis, P.S., Teeri, J.A. & Tonsor, S.J. (2001). Family- and population-level responses to atmospheric CO2 concentration: gas exchange and the allocation of C, N, and biomass in Plantago Lanceloata (Plantaginaceae). Am. J. Bot., 88, 1080–1087. Knapp, A. & Medina, E. (1999). Success of C4 photosynthesis in the field: Lessons from communities dominated by C4 plants. In: C4 Plant Biology (eds Sage, R.F. & Monson, R.K.). Academic Press, San Diego, pp. 251–283. Koch, G.W. & Mooney, H.A. (1996). Responses of terrestrial ecosystems to elevated CO2: a synthesis and summary. In: Carbon Dioxide and Terrestrial Ecosystems (eds Koch, G.W. & Mooney, H.A.). Academic Press, San Diego, pp. 415–429. Kohut, R. (2003). The long-term effects of carbon dioxide on natural systems: issues and research needs. Environ. Int., 29, 171–180.
Evolutionary responses to elevated CO2 439
LaDeau, S.L. & Clark, J.S. (2001). Rising CO2 levels and the fecundity of forest trees. Science, 292, 95–98. Lake, J.A., Woodward, F.I. & Quick, W.P. (2002). Long-distance CO2 signaling in plants. J. Exp. Bot., 53, 183–193. McCarthy, J.J., Canziani, O.F., Leary, N.A., Dokken, D.J. & White, K.S. (2001). Climate Change 2001: Impacts, Adaptation, and Vulnerability. Cambridge University Press, New York. McElwain, J.C. & Chaloner, W.G. (1995). Stomatal density and index of fossil plants track atmospheric carbon dioxide in the Paleozoic. Ann. Bot., 76, 389–395. Matsuyama, T., Tamaoki, M., Nakajima, N., Aono, M., Kubo, A., Moriya, S. et al. (2002). cDNA microarray assessment for ozonestressed Arabidopsis thaliana. Environ. Pollut., 117, 191–194. Medlyn, B.E., Barton, C.V.M., Broadmeadow, M.S.J., Ceulemans, R., De Angelis, P., Forstreuter, M. et al. (2001). Stomatal conductance of forest species after long-term exposure to elevated CO2 concentration: a synthesis. New Phytol., 149, 247–264. Meyer, K. & Hill, W.G. (1997). Estimation of genetic and phenotypic covariances functions for longitudinal or repeated records by restricted maximum likelihood. Livestock Prod. Sci. 47, 185–200. Moore, B.D., Cheng, S.-H., Sims, D. & Seemann, J.R. (1999). The biochemical and molecular basis for photosynthetic acclimation to elevated atmospheric CO2. Plant Cell Environ., 22, 567–582. Norby, R.J. (1994). Issues and perspectives for investigating root responses to elevated atmospheric carbon dioxide. Plant Soil, 165, 9–20. Norby, R.J., Wullschleger, S.D., Gunderson, C.A., Johnson, D.G. & Ceulemans, R. (1999). Tree responses to rising CO2 in field experiments: implications for the future forest. Plant Cell Environ., 22, 683–714. Norton, L.R., Firbank, L.G. & Watkinson, A.R. (1995). Ecotypic differentiation of response to enhanced CO2 and temperature levels in Arabidopsis thaliana. Oecologia, 104, 394–396. Ozturk, Z.N., Talame´, V., Deyholos, M., Michalowski, C.B., Galbraith, D.W., Gozukirmizi, N. et al. (2002). Monitoring largescale changes in transcript abundance in drought- and saltstressed barley. Plant Mol. Biol., 48, 551–573. Pataki, D.E., Ellsworth, D.W., King, J.S., Leavitt, S.W., Lin, G., Pendall, E. et al. (2003). Tracing changes in ecosystem function under elevated CO2. Bioscience, 53, 805–818. Paul, M.J., Foyer, C.H. (2001). Sink regulation of photosynthesis. J. Exp. Bot., 52, 1383–1400. Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola, J.M., Basile, I. et al. (1999). Climate and atmospheric history of the past420,000 years from the Vostok ice core, Antarctica. Nature, 399, 429–436. Pigliucci, M. (1998). Ecological and evolutionary genetics of Arabidopsis. Trends Plant Sci., 3, 485–489. Polley, H.W., Johnson, H.B., Marino, B.D. & Mayeux, H.S. (1993a). Increase in C3 plant water-use efficiency and biomass over Glacial to present CO2 concentrations. Nature, 361, 61–64. Polley, H.W., Johnson, H.B., Mayeux, H.S. & Malone, S.R. (1993b). Physiology and growth of wheat across a subambient carbon dioxide gradient. Ann. Bot., 71, 347–356. Poorter, H. (1993). Interspecific variation in the growth response of plants to an elevated ambient CO2 concentration. Vegetatio, 104/105, 77–97.
Potvin, C. & Tousignant, D. (1996). Evolutionary consequences of simulated global change: genetic adaptation or adaptive phenotypic plasticity. Oecologia, 108, 683–693. Pritchard, S.G., Rogers, H.H., Prior, S.A. & Peterson, C.M. (1999). Elevated CO2 and plant structure: a review. Global Change Biol., 5, 807–837. Putterill, J. (2001). Flowering in time: genes controlling photoperiodic flowering in Arabidopsis. Philos. Trans. R. Soc. Lond. B Biol. Sci., 356, 1761–1767. Rolland, F., Moore, B. & Sheen, J. (2002). Sugar sensing and signaling in plants. Plant Cell, Supplement 2002, S185–S205. Sage, R.F. (1994). Acclimation of photosynthesis to increasing atmospheric CO2: The gas exchange perspective. Photosynth. Res., 39, 351–368. Sage, R.F. & Cowling, S.A. (1999). Implications of stress in low CO2 atmospheres of the past: Are today’s plants too conservative for a high CO2 world? In: Carbon Dioxide and Environmental Stress (eds Luo, Y. & Mooney, H.A.). Academic Press, San Diego, pp. 289–308. Sage, R.F. & Kubien, D.S. (2003). Quo vadis C-4? An ecophysiological perspective on global change and the future of C4 plants. Photosynth. Res., 77, 209–225. Salanoubat, M., Lemcke, K., Rieger, M., Ansorge, W., Unseld, M., Fartmann, B. et al. (2000). Sequence and analysis of chromosome 3 of the plant Arabidopsis thaliana. Nature, 408, 820–822. Saurer, M., Cherubini, P., Bonani, G. & Siegwolf, R. (2003). Tracing carbon uptake from a natural CO2 spring into tree rings: an isotope approach. Tree Physiol., 23, 997–1004. Schlesinger, W.H. (1997). Biogeochemistry: An Analysis of Global Change, 2nd edn. Academic Press, New York. Schlichting, C.D. & Pigliucci, M. (1995). Gene regulation, quantitative genetics and the evolution of reaction norms. Evol. Ecol., 9, 154–168. Schmid, B., Birrer, A. & Lavigne, C. (1996). Genetic variation in the response of plant populations to elevated CO2 in a nutrientpoor, calcareous grassland. In: Carbon Dioxide, Populations, and Communities (eds Ko¨rner, C. & Bazzaz, F.A.). Academic Press, San Diego, pp. 31–50. Seki, M., Narusaka, M., Ishida, J., Nanjo, T., Fujita, M., Oono, Y. et al. (2002). Monitoring the expression profiles of 7000 Arabidopsis genes under drought, cold and high-salinity stresses using a full-length cDNA microarray. Plant J., 31, 279–292. Simpson, G.G., Gendall, A.R. & Dean, C. (1999). When to switch to flowering. Annu. Rev. Cell Dev. Biol., 99, 519–550. Sims, D.A., Cheng, W.X., Luo, Y.Q. & Seemann, J.R. (1999). Photosynthetic acclimation to elevated CO2 in a sunflower canopy. J. Exp. Bot., 50, 645–653. Stitt, M. & Krapp, A. (1999). The interaction between elevated carbon dioxide and nitrogen nutrition: the physiological and molecular background. Plant Cell Environ., 22, 583–621. Tabata, S., Kaneko, T., Nakamura, Y., Kotani, H., Kato, T., Asamizu, E. et al. (2000). Sequence and analysis of chromosome 5 of the plantArabidopsis thaliana. Nature, 408, 823–826. Thomas, S.C. & Jasienski, M. (1996). Genetic variability and the nature of microevolutionary responses to elevated CO2. In: Carbon Dioxide, Populations, and Communities (eds Ko¨rner, C. & Bazzaz, F.A.). Academic Press, San Diego, pp. 51–81. Tissue, D.T., Griffin, K.L., Thomas, R.B. & Strain, B.R. (1995). Effects of low and elevated CO2 on C3 and C4 annuals. II. Photosynthesis and leaf biochemistry. Oecologia, 101, 21–28.
2004 Blackwell Publishing Ltd/CNRS
440 J. K. Ward and J. K. Kelly
Tissue, D.T., Griffin, K.L., Ball, J.T. (1999). Photosynthetic adjustment in field-grown ponderosa pine trees after six years of exposure to elevated CO2. Tree Physiol., 19, 221–228. Tousignant, D. & Potvin, C. (1996). Selective responses to global change: experimental results on Brassica juncea (L.) Czern. In: Carbon dioxide, Populations, and Communities (eds Ko¨rner, C. & Bazzaz, F.A.). Academic Press, New York, pp. 23–30. Van Der Kooij, T.A.W., De Kok, L.J. & Stulen, I. (2000). Intraspecific variation in the response of Arabidopsis thaliana lines to elevated atmospheric CO2. Phyton (Austria), 40, 125–132. Ward, J.K. & Strain, B.R. (1997). Effects of low and elevated CO2 partial pressure on growth and reproduction of Arabidopsis thaliana from different elevations. Plant Cell Environ., 20, 254– 260. Ward, J.K. & Strain, B.R. (1999). Elevated CO2 studies: past, present and future. Tree Physiol., 19, 211–220. Ward, J.K., Tissue, D.T., Thomas, R.B. & Strain, B.R. (1999). Comparative responses of model C3 and C4 plants to drought in low and elevated CO2. Global Change Biol., 5, 857–867.
2004 Blackwell Publishing Ltd/CNRS
Ward, J.K., Antonovics, J., Thomas, R.B. & Strain, B.R. (2000). Is atmospheric CO2 a selective agent on model C3 annuals? Oecologia, 123, 330–341. Woodward, F.I. (1993). Plant responses to past concentrations of CO2. Vegetatio, 104/105, 145–155. Woodward, F.I., Lake, J.A. & Quick, W.P. (2002). Stomatal development and CO2: ecological consequences. New Phytol., 153, 477–484. Zhang, J. & Lechowicz, M.J. (1995). Responses to CO2 enrichment by two genotypes of Arabidopsis thaliana differing in their sensitivity to nutrient availability. Ann. Bot., 75, 491–499.
Editor, I. F. Woodward Manuscript received 5 December 2003 First decision made 14 January 2004 Manuscript accepted 19 February 2004