Association between Carbonyl Sulfide Uptake and ¹⁸Δ during Gas Exchange in C₃ and C₄ Leaves^,

Abstract

Carbonyl sulfide (COS) and C¹⁸OO exchange by leaves provide potentially powerful tracers of biosphere-atmosphere CO₂ exchange, and both are assumed to depend on carbonic anhydrase (CA) activity and conductance along the diffusive pathway in leaves. We investigated these links using C₃ and C₄ plants, hypothesizing that the rates of COS and C¹⁸OO exchange by leaves respond in parallel to environmental and biological drivers. Using CA-deficient antisense lines of C₄ and C₃ plants, COS uptake was essentially eliminated and discrimination against C¹⁸OO exchange (¹⁸Δ) greatly reduced, demonstrating CA’s key role in both processes. ¹⁸Δ showed a positive linear correlation with leaf relative uptake (LRU; ratio of COS to CO₂ assimilation rates, A^s/A^c, normalized to their respective ambient concentrations), which reflected the effects of stomatal conductance on both COS and C¹⁸OO exchange. Unexpectedly, a decoupling between A^s and ¹⁸Δ was observed in comparing C₄ and C₃ plants, with a large decrease in ¹⁸Δ but no parallel reduction in A^s in the former. This could be explained by C₄ plants having higher COS concentrations at the CA site (maintaining high A^s with reduced CA) and a high phosphoenolpyruvate carboxylase/CA activity ratio (reducing ¹⁸O exchange efficiency between CO₂ and water, but not A^s). Similar A^s but higher A^c in C₄ versus C₃ plants resulted in lower LRU values in the former (1.16 ± 0.20 and 1.82 ± 0.18 for C₄ and C₃, respectively). LRU was, however, relatively constant in both plant types across a wide range of conditions, except low light (<191 μmol photon m⁻² s⁻¹).

The seasonal cycling of CO₂ concentration measured in the background atmosphere is often taken as evidence of the breathing of Earth and corresponds to about 10 Pg carbon. This is less than 10% of what we estimate to be the total cycling of CO₂ due to respiration and photosynthesis of the terrestrial biosphere (Beer et al., 2010). The discrepancy stems from the fact that most of the cycling CO₂ mixes in the atmosphere near the surface, and we only see the net sum of respiration and photosynthesis, which are often more or less balanced over the seasonal cycle. Modeling and measurements of CO₂ exchange by ecosystems over diurnal cycles have been used to deconvolve the primary biological processes (e.g. Desai et al., 2008). However, our confidence in these approaches disintegrates as we move to larger scales because of the increase in the internal mixing of CO₂ fluxes. Other approaches are needed to quantify the basic physiological processes at these scales.

Recent studies have identified another atmospheric trace gas that may be helpful in this regard—carbonyl sulfide (COS)—a chemical analog of CO₂ that is routinely measured by the National Oceanic and Atmospheric Administration atmospheric sampling program (Montzka et al., 2007). However, its cycling differs from that of CO₂. The major source of COS is the ocean either by direct emission or via CS₂ oxidation (Kettle et al., 2002), and its major sink is the leaves of the terrestrial biosphere, where it is taken up in parallel with photosynthesis. There is no significant source of COS from terrestrial ecosystems; hence, the mixing between sources and sinks occurs on a much larger scale in the atmosphere. Several authors have suggested that COS might be a good tracer for terrestrial gross primary productivity (GPP; Montzka et al., 2007; Blake et al., 2008; Campbell et al., 2008; Suntharalingam et al., 2008; Seibt et al., 2010). The initial development of this tracer was built on theoretical insight into the likely mechanism of COS uptake and a few empirical studies that compared the rates of GPP and COS uptake at the leaf scale (Sandoval-Soto et al., 2005). However, if the goal is to use COS exchange estimates as a proxy for gross photosynthesis (or GPP) on larger scales, then it becomes important to understand the physiological basis for linking these fluxes.

At the leaf scale, correlations of both COS and CO₂ fluxes with photosynthetically active radiation have been interpreted as evidence of stomata-dominated control of COS uptake rates, an interpretation that is supported by the daily pattern of this flux (Bartell et al., 1993; Hofman, 1993; Kuhn et al., 1999; Xu et al., 2002). Stimler et al. (2010a) also recently confirmed that there is no leaf-scale compensation point for COS, supporting the notion of no COS emission on this scale. The deposition velocity (flux normalized to concentration) for COS has been shown to be greater than that for CO₂ in most species, with a mean ratio of the deposition velocities of COS to CO₂ at the leaf level of about 3 (Sandoval-Soto et al., 2005). Recently, Campbell et al. (2008) expressed these ratios as leaf-scale relative uptake (LRU):

where A is the leaf uptake rate, C_a is the ambient concentration, and superscript s and c denote COS and CO₂, respectively. As clearly revealed by Campbell et al. (2008), characterizing LRU and its response to environmental and physiological drivers is key to incorporating COS measurements into flux-partitioning studies.

In vitro studies have established that COS, which is meta stable in water, can be hydrolyzed to H₂S and CO₂ by the enzyme carbonic anhydrase (CA; Kluczewski et al., 1985; Lorimer and Pierce, 1989; Miller et al., 1989; Tiwari et al., 2005; see also Yakir, 2002). This enzyme is ubiquitous in plant leaves and is thought to be the biochemical sink for COS, which is further enhanced as this is an exergonic reaction, making the reverse reaction unfavorable (Liu et al., 2010):

This enzyme is also known to catalyze the exchange of stable isotopes of oxygen between CO₂ and leaf water. This isotopic exchange can be monitored by measuring the change in the ¹⁸O content of CO₂ in gas exchange as well as in large-scale flux studies (Farquhar et al., 1993; Yakir and Wang, 1996; Gillon and Yakir, 2001). For CO₂, CA facilitates the reversible hydration (Lindskog and Coleman, 1973):

This reversible hydration/dehydration of CO₂ provides the opportunity for efficient ¹⁸O exchange between leaf water, normally enriched in ¹⁸O, and CO₂ inside the leaf (Francey and Tans, 1987; Farquhar and Lloyd, 1993; Farquhar et al., 1993; Yakir and Wang, 1996; Gillon and Yakir, 2000, 2001; Seibt et al., 2007).

C¹⁸O¹⁶O isotopic exchange and COS flux can be expected to respond to the same variables. Sufficient CA activity is needed to both bring about full isotopic equilibrium and maintain near-zero COS concentration at the site of CA activity. Furthermore, the same diffusion resistances limit both the inward diffusion of COS flux into leaves and the retrodiffusion of ¹⁸O-labeled CO₂ back into the atmosphere. Note, however, that the diffusion gradient of CO₂ is driven by the photosynthetic flux, while that of COS is independent of it.

Recognizing the potential importance of incorporating COS into gas exchange and flux measurements, we developed a laser-based instrument for the continuous measurement of COS concentration at normal atmospheric levels (approximately 400 parts per trillion, or one-millionth the concentration of CO₂). This instrument enables conducting gas exchange studies analogous to those typically conducted with nondispersive infrared gas analyzers (Stimler et al., 2010a). A subsequent study (Stimler et al., 2010b) examined the stoichiometry of COS and CO₂ fluxes in leaves of C₃ and C₄ species with particular emphasis on the control of these processes by stomatal conductance (g_s). In this article, we focus on the mechanism of COS uptake. We also hypothesize that COS uptake will be closely linked to ¹⁸O labeling of CO₂ during leaf gas exchange, and we take advantage of natural and genetically engineered variations in CA to examine this hypothesis.

RESULTS AND DISCUSSION

Relative COS/CO₂ Uptake in C₃ versus C₄ Plants

The mean rates of net CO₂ assimilation (A^c) and COS uptake (A^s) and the fluxes involved are summarized in Table I and Figure 1, respectively, and are consistent with previous studies (Kesselmeier and Merk, 1993; Sandoval-Soto et al., 2005; Stimler et al., 2010a). Due to the higher A^c but similar A^s values for C₄ compared to C₃ plants, the LRU values were higher in the latter (Table I; Fig. 2; see Kesselmeier and Merk, 1993; Sandoval-Soto et al., 2005; Seibt et al., 2010; Stimler et al., 2010a). Irrespective of these differences, the results showed that, as noted previously for C₃ plants (Stimler et al., 2010a), LRU is relatively constant across a wide range of light intensities in both C₃ and C₄ plants. Large deviations were only observed under low light (<191 μmol photon m⁻² s⁻¹), when the ratio could increase to up to 9.6 in both C₃ and C₄ plants (Fig. 2B). These results support the assumption that both A^c and A^s are influenced to a similar extent by stomatal conductance. Indeed, A^s was linearly correlated to g_s (A^s = 322.0g_s + 11.2; r² = 0.78, P < 0.0001 for C₃, and A^s = 109.0g_s + 3.4; r² = 0.86, P < 0.0001 for C₄). The deviations from the relatively constant LRU under low light probably reflect the light dependency of Rubisco activation and of ribulose 1,5-bisP supply, whereas CA activity is presumably light insensitive (Sage and Sharkey, 1987).

Table I. COS versus CO₂ uptake fluxes; simple (A^s/A^c) or normalized (LRU = A^s/A^c × ([CO₂]/[COS]) ratio of COS and CO₂ uptake rates, and COS total conductance under maximum-intensity light, for C₄ and C₃ plants.

Gas exchange conditions during the light response experiments were T approximately 24°C, RH approximately 70%, [COS] approximately 550 pmol mol⁻¹, and [CO₂] = 400 μmol mol⁻¹; two to five replicates per species; sd is indicated in parentheses.

Plant Species	As/Ac	LRU	As	Ac	gts
	μmol mol−1		pmol m−2 s−1	μmol m−2 s−1	mol m−2 s−1
C4
Maize	1.97 (0.50)	1.05 (0.76)	42.33 (3.65)	16.23 (5.04)	0.17 (0.18)
Sorghum	1.91 (0.39)	1.05 (0.44)	34.10 (13.91)	20.20 (4.18)	0.09 (0.06)
Amaranthus	2.51 (0.85)	1.37 (0.79)	41.54 (10.38)	17.48 (3.72)	0.13 (0.04)
Average	2.13 (0.33)	1.16 (0.18)	39.32 (4.54)	17.97 (2.03)	0.13 (0.04)
C3
Tobacco	2.61 (0.96)	1.63 (0.36)	27.01 (4.89)	9.41 (2.08)	0.1 (0.03)
Sage	2.67 (1.11)	1.82 (0.89)	40.80 (24.38)	13.50 (5.71)	0.05 (0.03)
Hibiscus	2.87 (1.28)	2.02 (0.29)	40.23 (12.68)	9.91 (1.25)	0.07 (0.02)
Average	2.72 (0.14)	1.82 (0.20)	36.01 (13.98)	10.94 (3.01)	0.07 (0.03)

Figure 1.

Comparison of the diffusion pathways of COS and CO₂ from the atmosphere to the site of biochemical reactions in C₃ and C₄ leaves. Concentrations of COS (pmol mol⁻¹) and CO₂ (μmol mol⁻¹) are indicated above and below the lines, respectively. Conductance through resistance steps associated with the boundary layer, stomata, and mesophyll (mol m⁻² s⁻¹) are indicated by g_bl, g_s, and g_m, respectively, based on leaves with net CO₂ flux of 14 or 20 μmol m⁻² s⁻¹ (C₃ and C₄, respectively), and COS uptake flux of approximately 35 pmol m⁻² s⁻¹ (for both plant types). Values of g_m (0.3 mol m⁻² s⁻¹ for both C₄ and C₃ leaves, respectively) were estimated by Stimler et al. (2010a) and incorporate dissolution, liquid-phase diffusion, and the biochemical step (assuming first-order reaction); g_m estimates for CO₂ were based on CO₂ concentration at the hydration site derived from isotopic measurements (Gillon and Yakir, 2001).

Figure 2.

COS uptake rates (pmol m⁻² s⁻¹; A) and LRU values (B) during light response measurements in C₄ plants. Species used were: maize, sorghum, and amaranthus. Conditions during the experiments were: T approximately 24°C, RH approximately 70%, [COS] approximately 550 pmol mol⁻¹, and [CO₂] = 400 μmol mol⁻¹. Symbols without error bars have sd < 2% (in section B only). Inset: LRU (the normalized ratio of COS/CO₂ uptake rates) under high light intensity.

Stomatal conductance is only one of the two major factors regulating leaf COS exchange. Ultimately, COS uptake depends on the reaction rate of COS with CA and water to produce CO₂ and H₂S (Eq. 2; Liu et al., 2010; Stimler et al., 2010a). Here, we confirmed the critical role of CA by carrying out light response experiments with CA-deficient antisense lines of C₃ and C₄ species (Price et al., 1994; von Caemmerer et al., 2004). COS uptake was essentially eliminated in both of these antisense lines (Fig. 3). In the C₃ plants, this was not associated with significant changes in g_s or A^c (Fig. 3, A, C, and E). In C₄ plants, however, antisense lines also showed reductions in A^c and g_s (Fig. 3, B, D, and F). This presumably reflects the C₄ pathway’s dependence on bicarbonate, produced by the CA-facilitated hydration of CO₂.

Figure 3.

Stomatal conductance (g_s; mol m⁻² s⁻¹; A and B), CO₂ uptake (A^c; μmol m⁻² s⁻¹; C and D), and COS uptake (A^s; pmol m⁻² s⁻¹; E and F) in tobacco (C₃) and F. bidentis (C₄) wild-type and antisense lines with different levels of CA activity during the light response experiments. Conditions during the experiments were as in Figure 2. Symbols (black circles, white circles, and gray squares) indicate wild type, moderate (10%, C₃ and C₄ plants), and low (2%, C₃ only) CA activities of the antisense lines, respectively.

Note that in this study, the use of CA-deficient antisense lines did not address the entire range of CA types or enzyme locations in the plants’ leaves (Fabre et al., 2007). Nevertheless, the CA that was affected in these lines clearly dominated the COS and ¹⁸O response.

COS Exchange and ¹⁸Δ

Consistent with previous studies (Farquhar and Lloyd, 1993; Gillon and Yakir, 2001), leaf-scale discrimination against C¹⁸OO (¹⁸Δ) was correlated with C_cs/C_a in all plants (Fig. 4D; where C_cs refers to CO₂ concentration at the chloroplast surface). Moreover, the C₄ plants showed markedly lower ¹⁸Δ than the C₃ plants. This most likely reflects incomplete isotopic equilibrium between CO₂ and water in the C₄ leaves, and accounting for this effect (i.e. θ_eq < 1; Gillon and Yakir, 2000, 2001) revealed consistent relationships between ¹⁸Δ and C_cs/C_a in both C₄ and C₃ plants (Fig. 4).

Figure 4.

Relationships between C¹⁸OO (¹⁸Δ; ‰) and rates of COS uptake (pmol m⁻² s⁻¹; A), CO₂ uptake (μmol m⁻² s⁻¹; B), and LRU (C), observed in C₃ (circles) and C₄ (triangles) plants. Each curve represents the full set of data from a complete light response experiment. In C, r² values for the linear best-fit lines for C₄ and C₃ plants are indicated (P < 0.0001). D, The relationship between ¹⁸Δ and CO₂ concentrations at the hydration site (C_cs) from different light response experiments and plants. Conditions during gas exchange measurements were as in Figure 2. Black circles and crosses represent observed discrimination values; white symbols represent predicted values based on the indicated parameters and as described in Table II, following Gillon and Yakir (2001).

As expected (see above; see also Stimler et al., 2010a), a clear negative correlation was observed between A^s and ¹⁸Δ. This correlation was apparent in all leaves during the light response experiments, for both C₄ and C₃ species (Fig. 4A, showing representative species). Increasing light intensity was generally associated with a decrease in ¹⁸Δ values and increases in A^s. For example, A^s increased from 15 to 43 pmol m⁻² s⁻¹ and from 8 to 24 pmol m⁻² s⁻¹, whereas ¹⁸Δ decreased from 84‰ to 13‰ and from 48‰ to 12‰ in amaranthus (Amaranthus cruentus) and sorghum (Sorghum halepense), respectively. In C₃ species, A^s increased from 20 to 46 pmol m⁻² s⁻¹ and from 20 to 35 pmol m⁻² s⁻¹, while ¹⁸Δ values decreased from 157‰ to 40‰ and from 239‰ to 81‰, for sage (Salvia longispicata × Salvia farinacea) and tobacco (Nicotiana tabacum), respectively. This inverse correlation between A^s and ¹⁸Δ is readily explained by the effects of g_s and C_cs: Increasing g_s with light intensity results in increasing uptake rates of COS (as well as CO₂). At the same time, the increase in photosynthetic capacity with increasing light intensity leads to a decrease in C_cs, resulting in decreased retroflux of ¹⁸O-labeled CO₂ out of the leaf and, consequently, in the observed reduction in¹⁸Δ (see Farquhar and Lloyd, 1993; Gillon and Yakir, 2001).

On average, a reduction of 75‰ in ¹⁸Δ was associated with a decrease in C_cs/C_a, from 0.7 to 0.4, in C₃ plants and from 0.5 to 0.2 in C₄ plants. As the decrease in C_cs reflects a proportionately greater increase in photosynthetic capacity (influencing only A^c and not A^s) relative to stomatal conductance (but g_s influences both A^c and A^s), it can be expected that the LRU will also change under these circumstances. Indeed, linear relationships between LRU and ¹⁸Δ were observed (r² = 0.76, 0.87, and 0.99) for the C₄ species sugarcane (Saccharum officinarum), sorghum, and maize (Zea mays), and r² = 0.89, 0.98, and 0.79 for the C₃ species hibiscus (Rosa sinensis), tobacco, and sage, respectively. Grouping the results for the C₄ and C₃ plants yielded best-fit lines of y = 0.51x − 4.58 with r² = 0.78 (P < 0.0001) for C₄ plants and y = 0.03x − 0.20 with r² = 0.92 (P < 0.0001) for C₃ plants.

The COS-¹⁸Δ Link in C₄ versus C₃ Plants

The correlation between LRU and ¹⁸Δ covered the entire range of both wild-type and CA-deficient antisense lines of C₃ plants (Fig. 5A), but not of C₄ plants (Fig. 5B), or in comparing C₃ to C₄ plants (Fig. 6). Unexpectedly, while ¹⁸Δ was lower in C₄ than in C₃ plants, consistent with previous studies (Gillon and Yakir, 2000, 2001), there was essentially no difference in the rate of COS uptake between the two photosynthetic groups (mean A^s values of 36.0 ± 13.9 and 39.32 ± 4.54 pmol m⁻² s⁻¹ for C₃ and C₄, respectively). Other variables generally showed the expected differences between C₃ and C₄ leaves: A^c was markedly higher, C_i/C_a (CO₂) lower, and g_s (CO₂) similar in C₄ compared to C₃ leaves (Fig. 6; Table I). C_i^s/C_a^s (for COS) was <0.2, as observed previously in C₃ plants (Stimler et al., 2010a), but twice as high in C₄ plants (Fig. 6D), which may reflect differences in internal resistances, as well as the fact that CA is located higher up (toward the atmosphere) in the diffusive pathway in C₄ versus C₃ leaves (see Fig. 1).

Figure 5.

Relationship between LRU and ¹⁸O discrimination (¹⁸Δ) in plants with different levels of CA activity (antisense lines). A, Mean values in C₃ species (tobacco) with moderate and low levels of CA activity (percent of wild type). B, Data of C₄ species (F. bidentis) with moderate CA activity. Values are based on light response experiments under high light intensity (1,889 μmol photon m⁻² s⁻¹) and all other conditions as in Figure 2. C, Schematic summary of the relationships between LRU and ¹⁸Δ values, indicating the possible factors underlying the observed changes based on data from A and B. CA deficiency eliminates COS uptake (LRU approaches zero) and reduces ¹⁸O discrimination (¹⁸Δ) in both C₃ and C₄ plants. Compared to C₃ wild-type plants, the C₄ wild-type (WT) species show a small decrease in LRU (lower A^s/A^c due to high A^c values in these plants) but a large decrease in ¹⁸Δ (PEPC in these plants limits ¹⁸O exchange and CO₂-water isotopic equilibrium).

Figure 6.

Relationships between gas exchange parameters and ¹⁸O discrimination (¹⁸Δ) in C₃ and C₄ plants. A and B, Rates of CO₂ uptake (A^c; A) or COS uptake (A^s; B); C and D, ratio of intercellular to ambient concentrations of CO₂ (C_i^c/C_a^c; C) and COS (C_i^s/C_a^s; D); E and F, total leaf conductance of CO₂ (g_t^c; E) or COS (g_t^s; F). Data are compiled from different light response experiments with C₄ (circles) and C₃ (squares) species. Each data point represents four to six measurements (sd is indicated). Black symbols indicate mean values. C₄ species used were maize, sorghum, and amaranthus, and C₃ species were sage, tobacco, and hibiscus.

Table II. Isotopic and gas exchange values for a C₄ plant (amaranthus) based on light response measurements (compare, Fig. 4D).

¹⁸Δ^m and ¹⁸Δ^p are measured and predicted ¹⁸O discrimination (‰) when θ_eq = 1 and θ_eq = 0.6 (full or partial isotopic equilibrium, respectively), ε = C_cs/(C_a − C_cs) where C_a and C_cs indicate ambient and chloroplast surface CO₂ concentrations, Δ_ea is isotopic discrimination between atmospheric CO₂, δ_a, and that in equilibrium with leaf water, δ_e: Δ_ea = 1,000 × [(δ_e/1,000 + 1)/( δ_a/1,000 + 1) − 1]. Irradiance values are in μmol photon m⁻² s⁻¹. Analysis and definitions are based on Gillon and Yakir (2001).

Irradiance	18Δm	18Δp (θeq = 0.6)	18Δp (θeq = 1)	ε	Δea
171	30.2	30.35	48.87	0.95	35.55
352	31.1	31.58	51.10	1.00	31.25
524	23.35	23.62	36.93	0.72	31.30
1,179	16.95	17.26	25.54	0.45	32.78
1,889	3.77	10.20	14.20	0.18	34.59

The explanation for the unexpected discrepancy between changes in ¹⁸Δ and A^s in going from C₃ to C₄ plants can be based on two mechanisms. These mechanisms are outlined below, but at this stage we do not have sufficient information to accurately quantify their relative contributions.

First, note that as indicated above, low ¹⁸Δ values for C₄ plants result from incomplete isotopic equilibrium between CO₂ and water (Gillon and Yakir, 2000; Eq. 3), which is assumed to be due to reduced CA activity (observed in whole-leaf extracts of C₄ leaves; Gillon and Yakir, 2000). No parallel reduction in COS flux was observed in the C₄ plants examined (Fig. 6B), indicating sufficient CA activity. Note that in going from C₃ to C₄ plants, g_s remains similar but C_i/C_a (CO₂) decreases and C_i^s/C_a^s (COS) increases. This is likely because a reduction in C_i^c (CO₂) in C₄ leaves is associated with higher A^c, whereas an increase in C_i^s (COS) is associated with lower CA activity. The increase in C_i^s (COS), however, helps restore A^s to levels similar to those in C₃ plants. Therefore, given that A^s is a function of both CA and the COS concentration at the enzyme site, the higher C_i^s/C_a^s observed in the C₄ plants (Fig. 6D) could accommodate proportionately lower CA activity while maintaining relatively high A^s values. This could explain, at least in part, the observed discrepancy in A^s versus ¹⁸Δ. Accordingly, estimating total internal conductance to COS, g^s_m (combining dissolution and the biochemical steps, assuming a first-order reaction, as: g_m^s=1/[(C_a^s/A^s) − (1/(1.94/g_s^w) + 1/(1.56/g_bl^w))]; see Stimler et al., 2010a), indicated a mean g_m^s value of approximately 0.3 mol m⁻² s⁻¹ similar to our earlier study and, considering the observed variations (about ±0.5), similar for both C₃ and C₄ plants. This upholds the high internal COS concentration near the biochemical reaction site (Fig. 1).

The second mechanism that could help explain the A^s versus ¹⁸Δ discrepancy (required when, for example, the reduced CA activity accommodated by the first mechanism remains too high to produce previously reported θ_eq values) relies on the fact that in C₄ plants, the primary photosynthetic substrate is bicarbonate consumed by phosphoenolpyruvate carboxylase (PEPC; Cousins et al., 2007). A high ratio of PEPC to CA activity, ρ, must result in reduced bicarbonate concentrations which, in turn, reduce the efficiency of the isotopic exchange between CO₂ and water (the ratio of Rubisco to CA activities is assumed to be low and is neglected here; see Farquhar and Lloyd, 1993). Since PEPC activity directly influences the hydration rate, the apparent CO₂ hydration rate constant, K_H, should not be identical to the CA rate constant, CA_leaf, originally used in estimating θ_eq (Gillon and Yakir, 2001). To account for this effect, we revised the disequilibrium term introduced in Gillon and Yakir (2001), substituting K_H for CA_leaf, where K_H = CA_leaf × (1 − ρ), to obtain:

where CA_leaf is the hydration rate scaled to leaf conditions (CA_leaf = k/F_in, k is CA rate constant, F_in is the gross CO₂ influx rate, μmol m⁻² s⁻¹), and 3 reflects the slower isotopic exchange associated with the three oxygens in bicarbonate.

Limited information is available on the variations in PEPC and CA activities among plant species. Based on our gas exchange data, supplemented with values in the literature (e.g. Cousins et al., 2007), we can assume for a typical C₄ leaf, with A^s of approximately 40 pmol m⁻² s⁻¹ and C_i^s/C_a^s of approximately 0.4, midrange CA activity of approximately 600 μmol m⁻² s⁻¹ and PEPC activity of approximately 100 μmol m⁻² s⁻¹, which yields ρ = 0.17. This, in turn, results in a disequilibrium, θ_eq, value of approximately 0.7, consistent with the observed mean value of approximately 0.6 (Fig. 4D) for our C₄ plants. As expected, θ_eq values approach zero and 1 when ρ values approach 1 and zero, respectively. Note that the above estimates of the ρ effects incorporate both the effect of the relatively high PEPC activities and the high C_i^s in C₄ leaves.

Similar effects of PEPC on ¹⁸Δ have been demonstrated previously using amaranthus (C₄) plants with different levels of PEPC (Cousins et al., 2007). In that study, suppressing the enzyme activity by a factor of approximately 40 enhanced ¹⁸Δ from 16.8‰ to 207‰, consistent with our results, but this was ascribed to errors in estimating CA activities. Note that in C₄ plants, there was no change in ¹⁸Δ between the wild-type and antisense lines (Fig. 5B), which may indicate similar CA/PEPC in both plants or an effect on ¹⁸Δ of CA/PEPC in the wild type similar to the reduced CA in the antisense plants.

Our current, preliminary, perspective of the links between COS uptake rates and ¹⁸Δ, summarized in Figure 5C, highlights three points. First, ¹⁸Δ in C₄ plants is markedly reduced compared to C₃ plants, as has been repeatedly reported. Second, CA activity is key to both ¹⁸O discrimination and COS uptake. Third, ¹⁸Δ and LRU are decoupled in C₄ plants because of the high C_i^s and reduced bicarbonate concentrations associated with high PEPC activity.

CONCLUSION

In this study, we extend our previous results showing that LRU is relatively constant in both C₃ and C₄ plants. We show that the two processes, COS uptake and C¹⁸OO exchange, can be measured simultaneously and that their rates are closely correlated, whether modulated by changing environmental conditions or by manipulating the activity of the key enzyme CA. The unexpected decoupling of COS and C¹⁸OO exchange in comparing C₃ and C₄ plants demonstrated that our understanding of CA reaction and ¹⁸O discrimination in leaves is incomplete. Both COS and C¹⁸ΟΟ can serve as potentially powerful tracers of plant-atmosphere CO₂ fluxes, and as a means to partition these fluxes to their gross components. Using the observed links between COS and ¹⁸Δ could greatly improve their use as tracers and constrain interpretations of the underlying processes.

MATERIALS AND METHODS

Plant Material

C₃ species sage (Salvia longispicata × Salvia farinacea), tobacco (Nicotiana tabacum), and hibiscus (Rosa sinensis) were purchased from local nurseries. C₄ species maize (Zea mays) and sorghum (Sorghum halepense) were grown from seeds in the greenhouse, while amaranthus (Amaranthus cruentus) and sugarcane (Saccharum officinarum) were purchased from nurseries. CA-deficient antisense line and wild-type seeds of tobacco (C₃) and Flaveria bidentis (C₄) were contributed by Prof. S. von Caemmerer (Australian National University) and grown in the greenhouse. Various levels of CA activity were achieved using the suppression methods described by Price et al. (1994) and von Caemmerer et al. (2004). Plants were kept under ambient light and temperature during the experimental period.

Gas Exchange Measurements

Uptake rates of CO₂ and COS were measured in all C₃ and C₄ plants with attached leaves. The experimental system consisted of a flow-through leaf cuvette made of Teflon-coated stainless steel with a magnetically operated fan and a glass window at the top. A whole leaf or branch was sealed in the cuvette (O-ring seal except around the petiole, which was sealed with high-vacuum putty). Measurements were performed under a relative humidity (RH) of approximately 70% and an air temperature of approximately 24°C. Light intensity was 55 to 1,889 μmol photon m⁻² s⁻¹, regulated with layers of miracloth, and filtered through 5 cm of water. Continuous airflow coming out of the cuvette was split into two paths for COS and CO₂ analysis. Following the CO₂ analysis, the sampled air was passed through a magnesium perchlorate drying trap (Sigma-Aldrich) to remove water before collecting the air in 115-mL flasks for analysis of the isotopic composition of CO₂. COS and CO₂ mixing ratios in air entering the leaf cuvette were adjusted to the desired values by mixing purified synthetic air with known gas mixtures produced from a COS-permeation device (VICI Metronics), and a compressed gas mixture of 1% CO₂ in air, followed by a three-stage dilution system. All flow rates were regulated and measured by mass-flow controllers (MKS).

CO₂ and COS Analyses

CO₂ and water-vapor concentrations in the air entering and leaving the leaf cuvette were measured by an infrared gas analyzer (Li-6262; Li-Cor) at a precision better than 0.5 μmol mol⁻¹ for CO₂ and 0.1 mmol mol⁻¹ for water vapor.

COS concentrations were measured using a mid-IR dual quantum cascade laser spectrometer (Aerodyne Research Inc.) at a wavenumber of 2,056 cm⁻¹ with a LN₂-cooled HgCdTe (MCT) detector (Kolmar Technologies) as described by Stimler et al. (2010b). Briefly, direct detection of the absorption spectrum was followed by quantitative spectral fitting combined with the measured pressure, temperature, and path length of the absorption cell and the laser spectral line width using TDL WINTEL software, as described by Nelson et al. (2004). The concentrations of COS and the laser line widths were determined in real time from the spectra through a nonlinear least-squares fitting algorithm that uses spectral parameters from HITRAN (Rothman et al., 2003). The data analysis procedure included pulse normalization reduction of the sample signal and automatic background correction (N₂). Pulse normalization corrects for variations in pulse-to-pulse amplitude in pulsed laser systems, by normalizing the signal pulse train to a reference pulse train. The automatic background correction divides the sample spectra by the spectrum of dry N₂. Corrections were carried out every 300 s.

Isotopic Analysis of CO₂

Oxygen isotope analysis of CO₂ was based on sampling CO₂ in the air entering and exiting the leaf cuvette by passing it through a 115-mL glass flask under atmospheric pressure. The CO₂ in the flask was measured as described by Klein et al. (2005). Briefly, a 1.5-mL aliquot was removed from each flask into a sampling loop and the CO₂ was cryogenically trapped using helium as the carrier gas. It was then passed through a Carbosieve G-packed column at 70°C to separate N₂O, and the eluted CO₂ was analyzed in a Europa 20-20 continuous-flow isotope ratio mass spectrometer (Crewe). Batches of 15 flasks at a time were measured from an automated manifold system, with five flasks of a standard gas being measured for every 10 samples at a precision of 0.2‰. Results are expressed in the small δ notation (δ‰) versus VPDB-CO₂ for ¹⁸O, where δ‰ = (R_sample/R_standard − 1) × 1,000 and R_sample and R_standard are the isotopic ratios of the sample and the appropriate standard, respectively. Instantaneous leaf discriminations, Δ, were calculated as described by Evans et al. (1986) for ¹³C and extended to ¹⁸O (Gillon and Yakir, 2001). The isotopic composition of CO₂ in air supplied to the leaves during the experiments was around δ¹⁸O = −27‰. The δ¹⁸O values of water supplied to the plants was −5‰, and that of leaf water at the evaporating surfaces was estimated using the modified Craig-Gordon model (Craig and Gordon, 1965; Farquhar and Lloyd, 1993) to be 10‰ ± 2‰ (depending on conditions and species).

Leaf discrimination against C¹⁸OO, ¹⁸Δ, and the extent of CO₂-water isotopic equilibrium, θ_eq, were calculated as detailed by Farquhar and Lloyd (1993) and Gillon and Yakir (2000). Calculation of CO₂ concentration at the hydration site (C_cs) was based on the difference between the measured and predicted discrimination against ¹³C (¹³Δ) as described by Gillon and Yakir (2000), based on the approach of Evans et al. (1986). For ¹⁸Δ estimation in C₄ plants, incomplete isotopic equilibrium (θ_eq) was assumed, as measured by Gillon and Yakir (2001).