- Top of page
- Materials and Methods
- Supporting Information
Increasing atmospheric carbon dioxide concentrations and corresponding climate change have increased the demand for a better process-based understanding of carbon (C) exchange processes (i.e. photosynthesis and respiration) at individual plant and ecosystem scales. Stable C isotopes are a sensitive tool for disentangling C fluxes from the leaf to the ecosystem level (Yakir & Sternberg, 2000; Bowling et al., 2008) and for analysing biophysical and biochemical processes in photosynthetic pathways. The marked discrimination against the heavier isotope (13C) during photosynthesis has been well characterized (Farquhar et al., 1982, 1989). Similarly, fractionation during photorespiration is thought to be significant (Sharkey, 1988; Gillon & Griffiths, 1997), while apparent fractionation during dark respiration has long been considered negligible (Lin & Ehleringer, 1997). However, there is now evidence for substantial apparent fractionation leading to differences between the C isotope composition of leaf dark-respired CO2 (δ13Cres) and its putative substrates in many C3 species (Ghashghaie et al., 2003). Therefore, δ13Cres has become the subject of several recent studies (Schnyder et al., 2003; Tcherkez et al., 2003; Nogués et al., 2004; Hymus et al., 2005; Klumpp et al., 2005; Prater et al., 2006; Bathellier et al., 2008). In general, foliar δ13Cres has been found to be 13C-enriched compared with a wide variety of metabolites, for example, an enrichment of 9‰ in dark-respired CO2 relative to plant organic material was found in Nicotiana sylvestris (Ghashghaie et al., 2001). This apparent fractionation effect is highly variable, changing with species and environmental factors (for a review, see Ghashghaie et al., 2003). Based on these laboratory experiments there is now increasing knowledge on the mechanisms accounting for fractionation occurring during dark respiration at the leaf level (Tcherkez et al., 2003; for recent reviews see Ghashghaie et al., 2003; Werner et al., 2007a). Indeed, based on the heterogeneous C isotope distribution in hexose molecules (DeNiro & Epstein, 1977; Rossmann et al., 1991; Gleixner et al., 1998), Ghashghaie et al. (2001) indicated two metabolic origins for the respired CO2 (oxidation of pyruvate releases 13C-enriched CO2 relative to substrate while the acetyl-CoA decarboxylated through the Krebs cycle is depleted) possibly accounting for the 13C-enrichment of the overall respired CO2 compared with respiratory substrates. Recent work has also shown that CO2 respired by tree trunks is in general 13C-enriched, while that of roots is 13C-depleted compared with their respective bulk organic matter or carbohydrates (Badeck et al., 2005; Klumpp et al., 2005; Gessler et al., 2007; Maunoury et al., 2007; Bathellier et al., 2008).
However, only a few studies have focused on potential diurnal short-term variations of δ13C of dark-respired CO2. Indeed, assuming that the pool of fresh assimilates carries C of variable isotope composition resulting from changes in photosynthetic discrimination, it can be expected that the isotope ratio of dark-respired CO2 may also change during the light period, even without involving any fractionation by the process of respiration itself. In field studies, Hymus et al. (2005) and Prater et al. (2006) found a pronounced enrichment of respired CO2 along a light period up to 5–10‰ when compared with the respired δ13CO2 measured during the dark period. This 13C enrichment was correlated with the concomitant cumulative CO2 assimilation (Prater et al., 2006). Similarly, rapid dynamics, though with smaller magnitudes, have been shown in other ecosystem compartments, for example, at the trunk and soil levels (Maunoury et al., 2007; Kodama et al., 2008).
The isotopic signature of ecosystem-respired CO2 (δ13CR) is a complex response of different respiratory sources, including respiration by autotrophic and heterotrophic organisms. Ecosystem respiration is still poorly understood even though it is a major component of the global C balance (Valentini et al., 2000; Reichstein et al., 2002; Davidson et al., 2006). Understanding the driving environmental factors of δ13CR is therefore important for applications of isotope-based models of the global C budget. So far, the short-term dynamics of the C isotopic composition of respired CO2 have been disregarded in most studies despite their potential implications, for example, for the sampling protocols used to collect nocturnal Keeling plots. There is increasing evidence of rapid dynamics (minutes to hours) of δ13CR (i.e. 4‰ and 6‰ during one night; Bowling et al., 2003; Werner et al., 2006, respectively). However, data on this topic are scarce and the understanding and identification of the isotopic effects during dark-respiration defined by Tcherkez et al. (2003) are far from being resolved (Tcherkez & Farquhar, 2005). Taking advantage of the rapid in-tube incubation method (Werner et al., 2007b), this paper aims to investigate the different metabolic processes influencing the isotope composition of respired CO2 by analysing diurnal variation in dark-respired δ13CO2 in a wide range of ecotypes and species.
Most studies on fractionation during dark respiration have been performed on fast-growing herbaceous species under laboratory conditions. Werner et al. (2007b) were the first to show pronounced differences in two different functional plant types under standardized conditions: no significant diurnal variation in δ13Cres occurred in a fast-growing herb, while a pronounced δ13CLight–Dark amplitude of 8‰ occurred in a Mediterranean oak. Here, we explore the hypothesis that the extent of diurnal increase in δ13Cres varies between plant functional types. We present the first species survey to characterize different functional groups in relation to structural and metabolic features such as leaf thickness, C : N ratios and photosynthesis. Further, we use pyruvate positional labelling experiments, which provide the first direct evidence of the importance of apparent fractionation processes in respiratory pathways for the observed functional differences in diurnal variation of plant δ13Cres.
- Top of page
- Materials and Methods
- Supporting Information
The primary objective of this study was to investigate species-specific differences in the magnitude of diurnal variation in δ13C of leaf dark-respired CO2 (δ13Cres). The diurnal amplitude in δ13Cres varied from 0 to 8‰ among the species examined. Such differences could be attributed to distinct functional plant groups: all evergreen, slow-growing or aromatic species studied exhibited large variation in δ13Cres with a common diel pattern. The δ13Cres continuously increased during the light period compared with the morning values, followed by a decrease during the dark period. Conversely, herbaceous and fast-growing plants did not exhibit marked temporal variations. These results precisely match previous data obtained for T. barbata (Werner et al., 2007b), for one Pinus species (Pinus elliottii; Prater et al., 2006) and for Q. ilex (Hymus et al., 2005; Werner et al., 2007b) in natural conditions, being a strong indication that measured δ13Cres on glasshouse-grown species were fully representative. Nevertheless, changes in δ13Cres diurnal patterns could occur under natural conditions with, for example, drought or temperature.
There has long been evidence for enriched δ13C signals in leaf-respired CO2 relative to leaf organic matter or respiratory substrates (Park & Epstein, 1961; Duranceau et al., 1999; Ghashghaie et al., 2001; Tcherkez et al., 2003; Klumpp et al., 2005), which has been explained by apparent fractionation processes in the respiratory pathways (Ghashghaie et al., 2001, 2003; Tcherkez et al., 2003). Our positional labelling experiments (Fig. 6) provide first evidence, that these processes are also involved in the diurnal changes in δ13Cres. In short, enrichment in dark leaf-respired δ13CO2 is mainly attributed to the heterogeneous 13C-distribution within hexose molecules (Fig. 6), where C3 and C4 are 13C-enriched compared with other positions because of fractionation in the aldolase reaction (Rossmann et al., 1991; Gleixner & Schmidt, 1997). During glycolysis, C1 of pyruvate derived from enriched C3 and C4 of glucose molecules is decarboxylated by pyruvate dehydrogenase (PDH), releasing 13C-enriched CO2, while the lighter carbon atoms are incorporated in acetyl-CoA and decarboxylated in the Krebs cycle. Acetyl-CoA molecules are partially deviated to the biosynthesis of metabolites, for example, fatty acids and secondary compounds, well known to be 13C-depleted compared with carbohydrates (Park & Epstein, 1961). Accordingly, Ghashghaie et al. (2001) proposed that if the carbohydrate molecule is fully consumed during dark respiration no apparent fractionation will be observed (i.e. the overall CO2 released by dark respiration carries the isotopic signature of the substrate). By contrast, in the case of a deviation of light carbon (acetyl-CoA) into biosynthetic pathways, the overall respired CO2 is 13C-enriched.
The pyruvate positional 13C-labelling provides direct evidence that changes in the relative activity of the PDH-reaction (decarboxylation of 13C1-labelled pyruvate) and Krebs cycle (decarboxylation of 13C2-3-labelled pyruvate) do occur (Fig. 6). Moreover, it indicates the importance of changes in relative C flux rates through both respiratory pathways: the diurnal increase in δ13Cres was caused by a marked increase in the C flux through PDH into secondary metabolism relative to the Krebs cycle activity, which remained constant throughout the day (for H. halimifolium, Fig. 6c). Hence, diurnal variations in δ13Cres are related to an increased metabolic activity of the PDH, which exceeded the C flow into Krebs cycle by several times. By contrast, a stable low activity of both pathways was observed in the herb O. triangularis (Fig. 6d), which is consistent with the lack of diurnal variation in δ13Cres in herbaceous, fast-growing species. Nevertheless, the amount of CO2 released from C1 by PDH exceeded the CO2 released in the Krebs cycle (13C2-3-labelled pyruvate, Fig. 6d) even in this species. This indicates that not all the glucose molecules are fully respired even in fast-growing species, but that some acetyl-CoA molecules and/or intermediates of the Krebs cycle, which are precursors for multiple anabolic and catabolic reactions, may always be allocated into other pathways.
Effect on the balance between carbon supply and demand on δ13Cres
Differences between functional plant groups may be attributed to marked differences in magnitude of C supply through photosynthesis (‘supply function’) relative to the respiratory demand for growth and maintenance respiration (‘demand function’).
The demand function is sustained by distinct respiratory energy demand of the functional groups: slow-growing, evergreen woody species, with high diel variation in δ13Cres, generally exhibit a low demand in respiratory substrates in contrast to actively growing species with a high respiratory energy demand (e.g. herbs). The respiratory demand of a species can be broken up into growth and maintenance respiration. The latter contributes to a larger proportion of total plant respired CO2 in slow-growing as opposed to fast-growing plants (Amthor, 1984). Given the differences in leaf structure such as higher lignin content in woody compared with herbaceous plants, Bowling et al. (2008) estimated that growth respiration would produce twice as enriched δ13Cres in woody than in herbaceous plants.
The demand function is further supported by the reduced diurnal δ13Cres variation in growing leaves, with a higher respiratory energy demand, compared with mature leaves from the same evergreen plants (Fig. 5). Most evergreen species exhibit a flush-like growth during a short period (Werner et al., 1999) and woody species can sustain new growth from C reserves of the previous year (Damesin & Lelarge, 2003). By contrast, in fast-growing species, the growing shoots may provide a strong C sink for the whole plant. This is further supported by Ocheltree & Marshall (2004) who found that the enrichment in δ13Cres of Helianthus annuus relative to soluble sugars was negatively correlated to its relative growth rate. It supports the importance of maintenance vs growth respiration to explain isotopic increase in δ13Cres.
However, differences in respiratory energy demand between functional groups are not expected to change on a diurnal time-scale. Hence, what seems to be of greater importance is the balance between the respiratory energy demand and the C supply rate, which can indeed exhibit marked diurnal changes. The supply function is reinforced by the fact that the increase in δ13Cres is linearly related to the cumulative CO2 fixation during the light period (Fig. 4), as already reported for Q. ilex leaves (Hymus et al., 2005). Moreover, by impairing the potential CO2 accumulation through a decreased light intensity or by interrupting the light period, the diurnal increase in δ13Cres declined almost proportionally. The positive linear relationship between the increase in δ13Cres and the cumulative CO2 uptake may indicate that with the diurnal accumulation of metabolites in excess of the respiratory demand a larger proportion can be diverted into secondary metabolism, as shown through the pyruvate-labelling experiments (Fig. 6c). This is supported by findings of Prater et al. (2006) who induced less enriched δ13Cres on P. elliottii needles by artificial shading of leaves. Thus, the diurnal increase in δ13Cres can be attributed to the increasing flux into secondary metabolism with increasing C supply during the day when the sugar pools are filled and the respiratory demand is met.
Influence of fractionation of enzymes and respiratory substrates on δ13Cres
The observed diurnal variation in δ13Cres of up to 8‰ exceeds the variation that can be expected from the heterogeneous intramolecular 13C distribution of a glucose molecule (6‰, Rossmann et al., 1991; Hobbie & Werner, 2004). However, isotope effects of the respiratory decarboxylating enzymes could increase the difference between CO2 evolved via PDH reaction and the Krebs cycle. Tcherkez & Farquhar (2005) assumed that PDH fractionates (Melzer & Schmidt, 1987) but they also suggested that this fractionation would not be evident in case of full decarboxylation of pyruvate molecules (Tcherkez & Farquhar, 2005). Nevertheless, they have shown with quantum chemical calculations, that the enzyme citrate synthase, which catalyses the first step of the Krebs cycle, has an isotope effect of 23‰ (Tcherkez & Farquhar, 2005). Calculations of the overall isotope effects revealed that the Krebs cycle is a source of 13C depletion, both in organic acids that are intermediates in the cycle and in the respired CO2 (Tcherkez & Farquhar, 2005). Hence, this process could increase the difference in δ13Cres above the expected difference originating from heterogeneous distribution within the glucose molecule.
The extent to which respiratory processes fractionate is further dependent on the pool sizes of the substrates (i.e. with increasing pool sizes during the day there is a higher probability for fractionation to occur). Further, the observed increase in δ13Cres could be caused by an increase in δ13C of the respiratory substrates through either (i) a diurnal decrease in photosynthetic discrimination via temporal variation of stomatal and internal conductance and Rubisco activity; or (ii) a shift to respiratory sources with more enriched isotopic signatures or (iii) a change in the relative flux rates from respiratory substrates with different δ13C.
Leaf respiration uses several C sources, including soluble sugars, starch, lipids or amino acids with a rapid turnover and different isotopic characteristics and residence times (Schnyder et al., 2003; Nogués et al., 2004). Further, a change in isotopic signature of stored vs fresh assimilates, that can contribute to up to 50% of respiration, can account for variation in δ13Cres (Schnyder et al., 2003; Nogués et al., 2004). The δ13Cres has been shown to vary over the course of various days in darkness under constant environmental conditions. Those changes were associated with the depletion of different substrate pools and/or shifts in the relative contributions of dark-respiratory pathways (Tcherkez et al., 2003). Could these processes also occur in the light?
We did not observe diurnal changes in either the pool sizes or δ13C of different sugars (data not shown) or leaf organic matter (Fig. 2), in agreement with other recent works that reported little diurnal variation in δ13C of different respiratory substrates, despite marked diurnal variations in δ13Cres (Hymus et al., 2005; Göttlicher et al., 2006). Furthermore, marked diurnal changes in photosynthetic discrimination and thus, in δ13C of fresh assimilates are unlikely under the controlled conditions in the climate chamber. Vapour pressure deficit and temperature were kept constant in the glasshouses, thus, changes in those parameters can also be ruled out as a potential source of variations. Further, as these plants were grown for a prolonged period under constant conditions, there was presumably no substantial difference in the isotopic signature of old and new C reserves. However, there are further potential fractionation processes, such as transitory starch accumulation and remobilization, which have been found to govern the diel rhythm of δ13Cres in short-term turnover pools of soluble sugars in leaves and phloem-transported organic matter (Tcherkez et al., 2004; Gessler et al., 2007).
Light-enhanced dark respiration
Upon darkening of a leaf an immediate increase followed by a subsequent decrease in δ13Cres occurs (Werner et al., 2007b), and hence the time of dark-incubation is important. It has been argued that this transient peak is related to light-enhanced dark respiration (LEDR; Barbour et al., 2007) that can be observed as a post-illuminatory respiration pulse (Atkin et al., 1998). Although the metabolic origin of such an effect is not well known (Atkin et al., 1998), organic acids might be the respiratory substrates during this peak (Cornic, 1973), which might have a different δ13C signature from glucose. Barbour et al. (2007) have suggested that malate could be a substrate for respiration by the NAD+ malic enzyme during the LEDR peak, which would produce enriched CO2. Our on-going work does not provide a strong support for this hypothesis. First, the transient decrease in δ13Cres typically lasts for 30–60 min (up to 120 min, Werner et al., 2007b) which is longer than the time frame of LEDR and second, we observed this transient peak during the decarboxylation of 13C1-labelled pyruvate, which may not be expected if the major source of LEDR is the decarboxylation of (unlabelled) organic acid via the NAD+ malic enzyme. An alternative explanation could be a rapid increase in PDH activity, which is downregulated during the light period (Tcherkez et al., 2005), and subsequent rapid decarboxylation of the available sugar pools, reflecting the pool size and C flow rates. However, more research is needed to clarify the underlying processes. Nevertheless, the diurnal increase in δ13Cres has similar amplitude when measured after 30 min darkening (Q. ilex, data not shown) and the reported values are consistent with data from Hymus et al. (2005) who incubated leaves for approx. 15–30 min. This indicates, that even if the LEDR has an influence on δ13Cres immediately upon darkening, the diurnal pattern observed in this study will be maintained.
Differences in δ13Cres between functional plant groups
Our data cover a broad spectrum of leaf-respired CO2 isotopic signatures, ranging from −16 to −32‰, which may be partially attributed to differences in photosynthetic discrimination between functional groups, reflected in differences in δ13C of leaf organic matter (Fig. 2). Differences in leaf structure, such as SLA, leaf water content and C : N ratios (Fig. 3) may also play a role through its effect on stomatal and mesophyll CO2 conductance. Evergreen species have a lower internal CO2 conductance compared with deciduous trees and herbaceous plants (e.g. 0.1 and 0.24 for Q. ilex and Q. petraea, respectively; Roupsard et al., 1996; for references on mesophyll conductance see Ethier & Livingston, 2004). Mesophyll conductance can vary rapidly with incident light, temperature, CO2 and humidity (Piel et al., 2002; Warren & Dreyer, 2006; Flexas et al., 2007), however, to our knowledge there are no reports on diurnal changes in mesophyll conductance. Stomatal conductance remained constant over the day in species with marked increase in δ13Cres (such as Q. ilex, H. halimifolium and A. unedo; data not shown). Further, a diurnal δ13Cres increase was observed for all leaf types investigated (including needles, sclerophyllous and mesophyllous leaves; see Figs 1 and 3) and in M. piperita, a herbaceous aromatic plant with larger SLA and (most probably) high mesophyll conductance. Thus, while species might be clustered into functional groups by leaf morphology, this is not likely to account for the observed diurnal patterns.
One common function of species with a high diurnal increase in δ13Cres seems to be an increased biosynthesis of secondary metabolites for defence, stress avoidance or aromatic compounds. For 10 species of this group, including the aromatic herb M. piperita (Nogués et al., 2006), we found a literature reference for volatile compound emission, particularly isoprene (no data were available for the remaining species).
In Q. cilex and P. pinea, for example, monoterpene emissions were found to be highly variable at the diurnal timescale, and markedly dependent on the incident light level (Staudt et al., 1997; Sabillon & Cremades, 2001). Isoprene emission is closely linked to photosynthesis and can exhibit a marked diurnal increase (Rapparini et al., 2004). Although several pathways can be involved in isoprene synthesis, Affek & Yakir (2003) showed that 72–91% of emitted isoprene was derived from recently fixed C. An inverse relationship between dark respiration rate and isoprene emission was found in several studies and it has been hypothesized that these two processes compete for the same substrate (Rosenstiel et al., 2003, 2004), which would be in agreement with our results. Loreto et al. (2007) confirmed this relationship for young leaves only, showing that the decrease in respiratory demand when leaves mature is accompanied by a progressive and rapid increase in isoprene emission during leaf development. This is in agreement with the reduced diurnal increase in δ13Cres in young leaves (Fig. 5). Further, several species with marked diurnal variation in δ13Cres are aromatic species (e.g. R. officinalis and M. piperita), which do allocate C into secondary compound synthesis of aromatic volatile molecules.
Hence, the diurnal increase in δ13Cres may be attributed to the increasing flux into secondary metabolism with increasing C supply during the day when the sugar pools are filled and the respiratory demand is met.
It will be important to evaluate the impact of these dynamics in δ13Cres at other spatial scales, from organs (e.g. shoots, stems and roots) to the plant level and for other ecosystem compartments (e.g. soil) to identify, how variations in δ13Cres of different ecosystem compartments will influence the integrated signal of ecosystem respiration (δ13CR). There is now increasing evidence that significant diurnal cycles in δ13Cres also occur in many other respiratory sources in the ecosystem, such as trunk (Maunoury et al., 2007; Kodama et al., 2008), soil (Kodama et al., 2008), roots (S. Unger et al., unpublished) and ecosystem (Bowling et al., 2003; Knohl et al., 2005; Werner et al., 2006; Kodama et al., 2008). Such knowledge is of major importance as it may affect the reliability of our estimates of ecosystem respiration, which is used in many modelling approaches to partition ecosystem C fluxes (Yakir & Wang, 1996; Bowling et al., 2001; Knohl & Buchmann, 2005) and may thus affect our predictions on ecosystem response to environmental changes.
This is the first large species survey on short-term δ13Cres variations that could be attributed to marked apparent fractionation processes in the respiratory pathways. Our results support the hypothesis that the diurnal increase in δ13Cres is enhanced during the light period in species with a high investment in secondary metabolism, whereas fast-growing herbs and grasses with a high respiratory energy demand do not show this diurnal pattern. Pyruvate positional 13C-labelling provided the first direct evidence that diurnal variations in δ13Cres are related to increased metabolic activity of PDH at low constant Krebs cycle activity, and point out the importance of changes in relative flux rates between both pathways. Differences between functional groups may be attributed to marked differences in the balance between the C supply through the amount of fresh assimilates during photosynthesis versus the respiratory demand for growth and maintenance respiration. Hence, C isotope composition of plant-respired CO2 contains information on the fate of respiratory substrates, and may, therefore, provide a nonintrusive way to identify changes in C allocation patterns. These short-term variations in δ13Cres have marked implications at larger scales, particularly for isotope partitioning studies at the ecosystem level.