The effect of atmospheric CO₂ concentration on carbon isotope fractionation in C₃ land plants

Abstract

Because atmospheric carbon dioxide is the ultimate source of all land-plant carbon, workers have suggested that pCO₂ level may exert control over the amount of ¹³C incorporated into plant tissues. However, experiments growing plants under elevated pCO₂ in both chamber and field settings, as well as meta-analyses of ecological and agricultural data, have yielded a wide range of estimates for the effect of pCO₂ on the net isotopic discrimination (Δδ¹³C_p) between plant tissue (δ¹³C_p) and atmospheric CO₂ (δ¹³C_CO2). Because plant stomata respond sensitively to plant water status and simultaneously alter the concentration of pCO₂ inside the plant (c_i) relative to outside the plant (c_a), any experiment that lacks environmental control over water availability across treatments could result in additional isotopic variation sufficient to mask or cancel the direct influence of pCO₂ on Δδ¹³C_p. We present new data from plant growth chambers featuring enhanced dynamic stabilization of moisture availability and relative humidity, in addition to providing constant light, nutrient, δ¹³C_CO2, and pCO₂ level for up to four weeks of plant growth. Within these chambers, we grew a total of 191 C₃ plants (128 Raphanus sativus plants and 63 Arabidopsis thaliana) across fifteen levels of pCO₂ ranging from 370 to 4200 ppm. Three types of plant tissue were harvested and analyzed for carbon isotope value: above-ground tissues, below-ground tissues, and leaf-extracted nC₃₁-alkanes. We observed strong hyperbolic correlations (R ≥ 0.94) between the pCO₂ level and Δδ¹³C_p for each type of plant tissue analyzed; furthermore the linear relationships previously suggested by experiments across small (10–350 ppm) changes in pCO₂ (e.g., 300–310 ppm or 350–700 ppm) closely agree with the amount of fractionation per ppm increase in pCO₂ calculated from our hyperbolic relationship. In this way, our work is consistent with, and provides a unifying relationship for, previous work on carbon isotopes in C₃ plants at elevated pCO₂. The values for Δδ¹³C_p we determined in our ambient pCO₂ chambers are consistent with the Δδ¹³C_p values measured in large modern datasets of plants growing within the Earth’s wettest environments, suggesting that it may be possible to reconstruct changes in paleo-pCO₂ level from plants that grew in consistently wet environments, if δ¹³C_CO2 value and initial pCO₂ level can be independently quantified. Several implications arise for the reconstruction of water availability and water-use efficiency in both ancient and recent plant Δδ¹³C_p values across periods of changing pCO₂ level. For example, the change in Δδ¹³C_p implied by our relationship for the rise in pCO₂ concentration observed since 1980 is of the same magnitude (= ∼0.7‰) as the isotopic correction for changes in δ¹³C_CO2 required by the input of ¹³C-depleted carbon to the atmosphere. For these reasons, only the portion of the terrestrial isotopic excursion that persists after accounting for changes in pCO₂ concentration should be used for the interpretation of a change in paleo-environmental conditions.

Introduction

Because carbon dioxide is a raw material for photosynthesis, the pCO₂ level of the atmosphere ultimately represents the availability of carbon for land-plant growth. Many studies have demonstrated the influence of increasing pCO₂ on plant growth (Hunt et al., 1991, Hunt et al., 1993, Kimball et al., 1993, Poorter, 1993, Ceulemans and Mousseau, 1994, Wand et al., 1999, Poorter and Navas, 2003, Ainsworth and Long, 2005, Schubert and Jahren, 2011) and on various plant functions, including increased CO₂-assimilation rate (Figure. 1.5 within Fitter and Hay, 2002), short-term decreased photosynthetic rate (Greer et al., 1995), increased seedling growth (Bazzaz, 1974), decreased nitrogen-content in leaves (Figure 14.4 within Bazzaz, 1996), decreased stomatal density (Woodward, 1987, Woodward and Bazzaz, 1988, Woodward and Kelly, 1995, Royer, 2001), and increased intrinsic water-use efficiency (Waterhouse et al., 2004, Gagen et al., 2011). These changes carry implications at the ecosystem level, such that increased pCO₂ implies changes in competition (Hunt et al., 1993), succession of species (Condon et al., 1992, Mousseau and Saugier, 1992, Bazzaz and Miao, 1993), decomposition rates (Melillo et al., 1982, Norby et al., 2001), and reproduction (Downton et al., 1987, Andersson, 1991).

Plant tissues exhibit a ¹³C-depleted isotopic signature relative to the atmosphere because ¹²C is preferentially selected for fixation over ¹³C as CO₂ is converted to sugar within leaves (Park and Epstein, 1960). Within C₃ plants (i.e., plants that employ only the RuBisCO enzyme to catalyze CO₂ fixation), the net isotopic difference between the atmospheric CO₂ and the resultant plant tissue has been modeled according to the following equation (Farquhar et al., 1989):

$$\mathrm{\Delta }\mathrm{\delta }{}^{13}{\text{C}}_{\text{p}}=a+(b-a)(c_i/c_a)$$

where

$$\mathrm{\Delta }\mathrm{\delta }{}^{13}{\text{C}}_{\text{p}}=(\mathrm{\delta }{}^{13}{\text{C}}_{\text{CO}2}-\mathrm{\delta }{}^{13}{\text{C}}_{\text{p}})/(1+\mathrm{\delta }{}^{13}{\text{C}}_{\text{p}}/1000)$$

Within the above, the constants a and b represent the isotopic fractionation due to diffusion through the plant’s stomata and subsequent catalysis by RuBisCO, respectively; δ¹³C_p and δ¹³C_CO2 are the carbon isotope composition of plant tissue and CO₂ in the atmosphere, respectively. Due to the combined action of both diffusion and fixation, the intracellular concentration of CO₂ (c_i) is less than the atmospheric concentration of CO₂ (c_a). Because a land plant’s main mechanism of responding to environmental conditions is by closing or opening stomata, variation in δ¹³C_p is often interpreted as a change in c_i caused by a change in stomatal conductance. Indeed, many environmental characteristics known to directly affect stomatal aperture have been shown to be correlated with δ¹³C_p [e.g., water availability (Warren et al., 2001), precipitation (Diefendorf et al., 2010, Kohn, 2010), relative humidity (Farquhar et al., 1982), and soil moisture content (Ehleringer and Cooper, 1988)]. This has led workers to adopt the interpretation of (c_i/c_a) as a measure of water-use efficiency (Farquhar and Richards, 1984, Farquhar et al., 1988). Other environmental characteristics that may have an indirect affect on stomatal aperture (e.g., through a generalized stress response) have also been observed to correlate with δ¹³C_p [e.g., nutrient availability (Warren et al., 2001), air pollutants (Martin et al., 1988, Savard, 2010), and soil salinity (Lin and Sternberg, 1992)]. The effect of temperature on δ¹³C_p has remained inconsistent, as both positive and negative correlations are commonly reported (Körner et al., 1991, Gröcke, 1998, Schleser et al., 1999, McCarroll and Loader, 2004, Daux et al., 2011).

Because atmospheric pCO₂ has been shown to influence multiple aspects of plant biology, workers have hypothesized that changes in pCO₂ level may directly influence Δδ¹³C_p (Ehleringer and Cerling, 1995, Beerling, 1996, Beerling and Royer, 2002, Tcherkez et al., 2006), but the magnitude of this response is uncertain. While records of Δδ¹³C_p in modern wood have recently been reported to show a positive correlation with increasing pCO₂ over the last 160 years (Gagen et al., 2007, Kirdyanov et al., 2008, Loader et al., 2008, McCarroll et al., 2009, Treydte et al., 2009), early meta-analysis of published data revealed no correlation with pCO₂ (Fig. 1 within Arens et al., 2000). In addition, multiple studies of Δδ¹³C_p have been performed upon plants growing across a gradient of pCO₂ level, either in growth chambers (Beerling and Woodward, 1995, Jahren et al., 2008, Schubert and Jahren, 2011) or in field experiments (Saurer et al., 2003, Sharma and Williams, 2009); other workers have attempted to correlate past pCO₂ level with the Δδ¹³C_p values of sub-fossil tree rings (Feng and Epstein, 1995, Berninger et al., 2000, Hietz et al., 2005, McCarroll et al., 2009, Treydte et al., 2009, Wang et al., 2011) and fossil leaves (Beerling et al., 1993, Beerling and Woodward, 1993, Van de Water et al., 1994, Beerling, 1996, Peñuelas and Estiarte, 1997). The majority of these modern and fossil studies show a positive correlation between Δδ¹³C_p and pCO₂ (Beerling et al., 1993, Van de Water et al., 1994, Beerling and Woodward, 1995, Kürschner et al., 1996, Peñuelas and Estiarte, 1997, Berninger et al., 2000, Saurer et al., 2003, Hietz et al., 2005, Sharma and Williams, 2009, Treydte et al., 2009); however, negative correlation (Beerling and Woodward, 1993, Beerling, 1996) and no correlation (Jahren et al., 2008) have also been reported. Therefore, there exists no clear consensus as to the effect that pCO₂ has on Δδ¹³C_p. However, the above laboratory and field experiments generally lack constant environmental control (especially water availability) across treatments, therefore any direct influence of pCO₂ on Δδ¹³C_p could have been masked or canceled by variation in c_i conferred by the multiple direct and indirect effects of environmental heterogeneity on stomatal conductance. Here, we report new data from laboratory experiments conducted within growth chambers featuring enhanced dynamic stabilization of moisture availability and relative humidity, as well as providing constant light, nutrient, δ¹³C_CO2, and pCO₂ level throughout up to four weeks of plant growth. These experiments, conducted across a wide range of pCO₂ levels while maintaining control over environmental characteristics, yielded new insight into the quantitative effect of pCO₂ level on C₃ land-plant Δδ¹³C_p values.