Environmental and physiological controls on the carbon isotope composition of CO2 respired by leaves and roots of a C3 woody legume (Prosopis velutina) and a C4 perennial grass (Sporobolus wrightii)

Authors

  • WEI SUN,

    Corresponding author
    1. Department of Renewable Resources
    Search for more papers by this author
    • Present address: Institute of Grassland Science, Northeast Normal University, Key Laboratory of Vegetation Ecology, Ministry of Education, Changchun, Jilin 130024, China.

  • VÍCTOR RESCO,

    1. Department of Renewable Resources
    Search for more papers by this author
    • Present address: Hawkesbury Institute for the Environment, University of Western Sydney, Richmond, NSW 2753, Australia.

  • DAVID G. WILLIAMS

    1. Department of Renewable Resources
    2. Department of Botany
    3. Program in Ecology, University of Wyoming, Laramie, WY 82071, USA
    Search for more papers by this author

W. Sun. Fax: +86 431 8569 5065; e-mail: sunwei8868@gmail.com; sunwei@nenu.edu.cn

ABSTRACT

Accurate estimates of the δ13C value of CO2 respired from roots (δ13CR_root) and leaves (δ13CR_leaf) are important for tracing and understanding changes in C fluxes at the ecosystem scale. Yet the mechanisms underlying temporal variation in these isotopic signals are not fully resolved. We measured δ13CR_leaf, δ13CR_root, and the δ13C values and concentrations of glucose and sucrose in leaves and roots in the C4 grass Sporobolus wrightii and the C3 tree Prosopis velutina in a savanna ecosystem in southeastern Arizona, USA. Night-time variation in δ13CR_leaf of up to 4.6 ± 0.6‰ in S. wrightii and 3.0 ± 0.6‰ in P. velutina were correlated with shifts in leaf sucrose concentration, but not with changes in δ13C values of these respiratory substrates. Strong positive correlations between δ13CR_root and root glucose δ13C values in P. velutina suggest large diel changes in δ13CR_root (were up to 3.9‰) influenced by short-term changes in δ13C of leaf-derived phloem C. No diel variation in δ13CR_root was observed in S. wrightii. Our findings show that short-term changes in δ13CR_leaf and δ13CR_root were both related to substrate isotope composition and concentration. Changes in substrate limitation or demand for biosynthesis may largely control short-term variation in the δ13C of respired CO2 in these species.

INTRODUCTION

The two major components of terrestrial ecosystem respiration, plant autotrophic respiration and soil microbial respiration, respond differently to environmental changes, making the interpretation and prediction of ecosystem respiration challenging (Falge et al. 2002). Measurements of the stable isotope ratio of carbon (δ13C) in CO2 exchanged between ecosystems and the atmosphere are useful for partitioning ecosystem respiration into autotrophic and heterotrophic components (Griffis, Baker & Zhang 2005) and detecting environmental effects on ecosystem-level photosynthetic processes (Bowling et al. 2002; Pataki et al. 2003). Such applications require understanding of processes controlling δ13C values of respiration sources at appropriate temporal and spatial scales (Barbour et al. 2011). Currently, our ability to model and predict changes in the δ13C value of autotrophic respiration at the ecosystem level is constrained by poor understanding of the short-term environmental and physiological controls on the δ13C value of CO2 respired by plant roots and leaves (Bowling, Pataki & Randerson 2008).

CO2 evolved from leaves and roots accounts for a large fraction of ecosystem autotrophic respiration (Ryan & Waring 1992). The apparent respiratory 13C/12C fractionation (ΔR) and the δ13C value of respiratory substrates partitioned between leaves and roots can differ substantially among species and in response to environmental stress (Bowling et al. 2008; Cernusak et al. 2009). CO2 evolved from dark-acclimated leaves is typically 13C enriched relative to bulk leaf biomass and putative respiratory substrates (Duranceau et al. 1999; Ghashghaie, Duranceau & Badeck 2001; Xu et al. 2004), which has been attributed mainly to the heterogeneous distribution of 13C across different carbon atoms within hexose molecules and the molecular fragmentation in the TCA cycle (Rossmann, Butzenlechner & Schmidt 1991; Tcherkez et al. 2003). Because of intramolecular 13C/12C differences, CO2 derived from pyruvate decarboxylation is 13C enriched compared with that from oxidation of acetyl-CoA in the TCA cycle (Melzer & Schmidt 1987). Therefore, incomplete oxidation of pyruvate in the TCA cycle associated with allocation of acetyl-CoA to biosynthetic pathways could lead to 13C enrichment in leaf-respired CO2 relative to respiratory substrates. The δ13C value of CO2 evolved from leaves can vary over diel periods by up to 8‰ (Hymus et al. 2005; Prater, Behzad & Jeffery 2006; Priault, Wegener & Werner 2009; Sun, Resco & Williams 2009), and differs among plant functional groups (Werner et al. 2007, 2009; Priault et al. 2009; Sun, Resco & Williams 2010).

In contrast, root-respired CO2 is generally 13C depleted relative to biomass or potential respiratory substrates (Ghashghaie, Badeck & Lanigan 2003; Bathellier et al. 2008, 2009; Bowling et al. 2008; Gessler et al. 2009), which may partially explain 13C enrichment in root biomass relative to leaf biomass (Cernusak et al. 2009). However, 13C enrichment of CO2 evolved from stems (Damesin & Lelarge 2003; Brandes et al. 2006), and roots (Gessler et al. 2007) relative to potential respiratory substrates have also been observed. Shifts in the δ13C value of root-respired CO2 (δ13CR_root) at hourly timescales have been observed in some species (Klumpp et al. 2005; Gessler et al. 2009; Unger et al. 2010), but not in others (Wegener, Beyschlag & Werner 2010).

Short-term changes in the δ13C value of CO2 respired by dark-acclimated leaves (δ13CR_leaf) and roots may be attributed partially to shifts in respiratory substrate use. A shift in respiratory substrate use from 13C-enriched carbohydrates to 13C-depleted lipids has been observed in Phaseolus vulgaris leaves maintained in darkness for several days (Tcherkez et al. 2003). However, the same phenomenon was not observed in P. vulgaris roots (Bathellier et al. 2009). Shifts in substrate use are unlikely to occur for plants growing under normal light/dark cycles unless severe carbon limitation arises under stressful environmental conditions (McDowell et al. 2008). Changes in the contribution of CO2 derived from light-enhanced dark respiration in light-acclimated leaves can potentially lead to variation in daytime δ13C values of leaf-respired CO2 (Barbour et al. 2007). However, night-time shifts in δ13CR_leaf are more likely associated with changes in ΔR and/or shifts in the δ13C value of respiratory substrates.

Leaf respiration is fuelled by recently assimilated photosynthates (Nogués et al. 2004), which are likely to vary in their carbon isotope composition over short timescales. As a result of short-term (hourly) changes in environmental conditions, leaf photosynthetic 13C/12C discrimination (ΔP) is expected to change substantially during the light period (Sun et al. 2009). This large variation in ΔP may directly affect δ13C of leaf-exported phloem sugars, which are the primary substrates of root respiration. Carbohydrates derived from transitory starch degradation are the substrates of night-time leaf respiration and the sources of night-time leaf-exported phloem sugars. Transitory starch is probably enriched in 13C because of an isotope effect of the aldose reaction (Gleixner & Schmidt 1997; Gleixner et al. 1998), therefore carbohydrates derived from transitory starch degradation are likely enriched in 13C relative to triose phosphate, the initial products of the Calvin cycle. Thus, the isotope effect associated with transitory starch formation could also lead to a diel shift in the δ13C of primary respiratory substrates and leaf-exported phloem sugars (Brandes et al. 2006; Gessler et al. 2008). The effects of short-term changes in ΔP and isotope effects of transitory starch assimilation and remobilization on nocturnal shifts in δ13CR_leaf and diel variation in δ13CR_root can be tested by frequent sampling and measurement of the δ13C of respiratory substrates (e.g. glucose and sucrose).

ΔR is likely not constant at diel timescales as partitioning of metabolic intermediates, such as acetyl-CoA, between different pathways is sensitive to changes in substrate availability or requirement for precursors for biosynthesis. Hence, even when there is no shift in the δ13C value of respiratory substrates, changes in ΔR can lead to variation in δ13CR_leaf and δ13CR_root. The magnitude of short-term changes in ΔR and its physiological and environmental controls under natural field conditions remains mostly unresolved.

The objectives of this study were to examine and understand underlying mechanisms associated with short-term changes in δ13CR_leaf and δ13CR_root in two important savanna species. We examined diel variation in δ13CR_root, nocturnal variation in δ13CR_leaf, and associated changes in δ13C values and concentrations of leaf and root glucose and sucrose in naturally occurring populations of a C3 woody legume (Prosopis velutina Woot.) and a C4 perennial grass (Sporobolus wrightii Munro ex Scribn.) during the dry pre-monsoon and wet monsoon seasons in southeastern Arizona, USA. The different seasonal conditions associated with the annual summer monsoon allowed us to examine short-term variation in δ13CR_leaf and δ13CR_root across contrasting moisture regimes, which is likely to induce changes in the δ13C value of respiratory substrates and the proportion of substrate diverted to biosynthetic precursors. We hypothesized that: (1) nocturnal shifts in the δ13C value of primary respiratory substrates partially determine variation in δ13CR_leaf and δ13CR_root and (2) variation in ΔR results from changes in substrate availability or requirement for precursors for biosynthesis, which would further influence patterns of δ13CR_leaf and δ13CR_root.

MATERIALS AND METHODS

Study site

The study site (31°39.8′N, 110°10.7′W, 1200 m elevation) was located on a remnant floodplain terrace of the San Pedro River, about 16 km southwest of the Tombstone, Arizona, USA. The overstory vegetation was dominated by velvet mesquite (P. velutina). The understory was covered by sacaton (S. wrightii), a perennial C4 grass, and other herbaceous annual and perennial dicots. Long-term (1893–2007) mean maximum and minimum temperatures are 25.2 and 9.7 °C, respectively, and the mean annual precipitation is 351 mm in Tombstone, Arizona (Western Regional Climate Center; http://www.wrcc.dri.edu). About 60% of the total annual precipitation occurs during the summer monsoon between July and September. Soils at this site consist mainly of gravelly sandy loam layers interspersed with clay and gravel lenses (Scott et al. 2004).

Field sampling of leaf and root dark-respired CO2

Leaf and root dark-respired CO2 was collected in the C3 legume tree species P. velutina and the C4 grass S. wrightii during the dry pre-monsoon (26 June) and rainy monsoon (2 August) seasons. Large-bore (60 mL) syringes were used to collect dark-respired CO2 samples for both leaves and roots using an incubation method modified from Werner et al. (2007). Root dark-respired CO2 was collected at 3 h intervals over a 24 h period; whereas leaf dark-respired CO2 was collected only during the night-time period (2100, 0000 and 0300 h) to avoid the impacts of light-enhanced dark respiration on δ13CR_leaf values. The starting times for the pre-monsoon and monsoon sampling of CO2 respired by roots were 0600 and 0900 h, respectively. Several young, fully expanded leaves from P. velutina and S. wrightii were collected and placed into the syringe for CO2 sampling. Coarse roots from each of the three P. velutina trees and fine roots from each of the three S. wrightii plants that were used for leaf dark-respired CO2 sampling were collected. After soil on the surface of the roots was carefully removed, the roots were cut into segments with length approximately 2 cm and placed into the large-bore syringe. The syringe bore volume was flushed completely with CO2-free air five times by actuating the syringe plunger. CO2 from leaf and root respiration in the barrel was then allowed to build for 15 and 30 min, respectively, and then 5 mL of air in the barrel was injected into septa-capped helium-flushed vials. The incubation time was determined by a pre-experiment to ensure that the CO2 concentration in the sample vials was no less than 2000 ppm. Vials were stored in the field in dry, insulated coolers and shipped overnight to the University of Wyoming Stable Isotope Facility for isotopic analysis.

Leaf gas exchange and water potential measurements

Leaf photosynthesis (at 0600, 0900, 1200, 1500 and 1800 h) and respiration (at 2100, 0000 and 0300 h) were measured using a LI-6400 portable photosynthesis system (Li-Cor Biosciences, Lincoln, NE, USA) on two replicate leaves of each of the three P. velutina and S. wrightii plants used for respiration sampling. Young, fully expanded leaves at the top of the canopy were used for these measurements. For each measurement, environmental conditions inside the leaf chamber (i.e. photosynthetically active radiation, chamber block temperature, relative humidity and CO2 concentration) were set to match ambient conditions. Pre-dawn leaf water potential (Ψpd) was measured on two P. velutina twigs for each of the three trees and one S. wrightii leaf for each of the three S. wrightii plants between 0300 and 0500 h using a pressure chamber (PMS Instruments, Corvallis, OR, USA).

Lipid and starch extractions

Leaves and roots comparable with those used for CO2 sampling were collected at each sampling interval over the 24 h measurement periods. Leaves and roots were immediately flash frozen in liquid nitrogen, stored on dry ice and then freeze dried using a Labconco freeze dryer (Labconco, Kansas City, MO, USA). Leaf and root samples were freeze dried and then ground into fine powder using a ball mill. Lipid and starch extractions were then performed following the procedures described by Wanek, Heintel & Richter (2001) and Göttlicher et al. (2006), respectively. Briefly, 100 mg powdered dry leaf or root material was extracted with 1.5 mL methanol/chloroform/water (MCW, 12:5:3, v/v/v) for 30 min at 70 °C. After centrifuging at 10 000 g for 2 min, 0.65 mL of the supernatant was transferred and phases were separated by adding 0.2 mL chloroform and 0.7 mL water. After centrifugation, 50 µL of the chloroform phase with the isolated lipids was transferred into a tin capsule and air dried in a fume hood and then analysed on an isotope ratio mass spectrometer (described next). The pellet was washed with distilled and deionized water and oven dried after re-extraction with MCW three times. Starch in the oven-dried pellet was gelatinized at 100 °C in a water bath for 15 min. Gelatinized starch was hydrolyzed by heat stable α-amylase (Sigma-Aldrich, St. Louis, MO, USA) at 85 °C for 120 min. The aqueous phase, containing sugars derived from starch, was separated from the pellet through centrifugation at 10 000 g. The aqueous phase was purified in a centrifugal filter device (Microcon YM-10, Millipore, Bedford, MA, USA). Fifty µL of the filtrate was transferred to a tin capsule and oven dried at 65 °C, and then analysed on an isotope ratio mass spectrometer (described next). We extracted a starch standard (Wheat starch, Sigma-Aldrich) using the method described previously and measured its stable carbon isotope composition. No significant isotopic fractionation was detected during extraction of this standard starch (data not shown).

Glucose and sucrose extractions and separations

Extractions and separations of glucose and sucrose in leaf and root samples were performed following the procedures described by Duranceau et al. (1999). Briefly, 50 mg of leaf or root sample was suspended with 1 mL of distilled and deionized water and maintained for 20 min at 4 °C. The resulting extract was centrifuged at 12 000 g for 10 min. The supernatant was boiled for 3 min and centrifuged at 12 000 g for 10 min to denature and precipitate soluble proteins. The water-soluble fraction was filtered with a 0.45 µm centrifugal filter device (Millipore, Milford, MA, USA). Quantification and purification of glucose and sucrose in filtered extracts was carried out by high-performance liquid chromatography (HPLC) analysis on 200 µL aliquots applied to a Sugar-Pak1 column (6.5 mm diameter and 300 mm length; Waters, Milford, MA, USA). The flow rate and the temperature of the column were maintained at 0.5 mL min−1 and 85 °C, respectively. The sugar peaks were detected by a Waters 410 differential refractometer (Millipore, Milford, MA, USA). Sucrose and glucose fractions from HPLC were collected and oven dried at 65 °C. The oven-dried sucrose and glucose were re-dissolved in 50 µL of deionized water. The solution was transferred to a tin capsule and oven dried at 65 °C and then analysed on an isotope ratio mass spectrometer (described next). No significant isotopic fractionation was observed with HPLC purification processes after similar analyses were applied to analytical grade sucrose and glucose solutions (data not shown).

Isotope ratio mass spectrometry

All isotope analyses were performed at the University of Wyoming Stable Isotope Facility. Leaf and root dark-respired CO2 was analysed within 5 d of field collection using a GasBench II attached to a Finnigan Delta+XP continuous flow inlet isotope ratio mass spectrometer (Thermo Finnigan, Bremen, Germany). Repeated measurements with laboratory CO2-in-air working standards had a precision of <0.1‰. The δ13C values of CO2 in respiration samples were corrected to the international Vienna Pee Dee Belemnite (VPDB) standard using CO2-in-air working standards calibrated with CO2-in-air standards obtained from NOAA-CMDL. Carbon isotope ratios of leaf and root biomass, starch, lipids, glucose and sucrose were determined by a Finnigan Delta+XP continuous flow inlet isotope ratio mass spectrometer coupled to a Costech elemental analyzer (Costech Analytical Technologies, Valencia, CA, USA). Precision of repeated measurements of laboratory standards was <0.1‰. Carbon isotope ratios are reported in parts per thousand relative to VPDB as: δ13C (‰) = (Rsample / Rstandard – 1) × 1000‰, where R is the molar 13C/12C ratio.

Flux-weighted δ13C value of CO2 respired during night-time

The flux weighted δ13C value of night-time leaf-respired CO2 was calculated as:

image(1)

where Ri and δ13CR_leaf,i are the respiration rate and the δ13C value of leaf-respired CO2 at time i (2100, 0000 and 0300 h), respectively. As we did not measure root respiration rate, we report the average δ13C value of root-respired CO2 recorded at 2100, 0000 and 0300 h (defined here as ‘δ13CR_root_w’) for comparison.

Respiratory apparent 13C/12C fractionation

Respiratory apparent fractionation relative to a particular substrate X (e.g. biomass, lipids, starch, glucose or sucrose) was calculated as:

image(2)

where δ13CX represents δ13C signature of substrate X; and δ13CR represents δ13C signature of leaf or root dark-respired CO2.

Statistical analysis

Simple linear regression analysis was conducted to evaluate the relationships between δ13C of dark-respired CO2 and δ13C of sucrose and glucose; sucrose concentrations and δ13C of respired CO2, and respiratory apparent 13C/12C fractionation and sucrose concentration. One-way analysis of variance (anova) was conducted to assess nocturnal variation in the δ13C of leaf-respired CO2, leaf glucose and sucrose, and diel variation in the δ13C values of root-respired CO2, root glucose and root sucrose. Two-way anova was conducted to evaluate differences in flux-weighted respiratory apparent 13C/12C fractionation relative to biomass, starch, lipids, glucose and sucrose between species and organs. All statistical analyses were carried out using SAS version 9.0 (SAS Institute Inc., Cary, NC, USA). Average values are reported as the arithmetic mean ± 1 standard error.

RESULTS

Water potential and leaf gas exchange

As a result of differences in both soil and atmospheric drought, S. wrightii and P. velutina expressed contrasting physiological activities and states during the dry pre-monsoon and wet monsoon sampling periods. Pre-dawn water potential (Ψpd) in S. wrightii increased from −2.5 ± 0.4 MPa during the pre-monsoon sampling period to −0.05 ± 0.01 MPa during the monsoon sampling period, whereas Ψpd in P. velutina was not influenced by monsoon precipitation inputs and remained constant at approximately −0.9 MPa. As a consequence of increases in the atmospheric water vapour pressure, maximum daytime leaf vapour pressure deficit (Dl) in S. wrightii and P. velutina declined from 7.6 ± 0.3 and 7.2 ± 0.3 kPa during the dry pre-monsoon sampling period to 2.5 ± 0.1 and 2.3 ± 0.1 kPa, respectively, during the wet monsoon sampling period. Stomatal conductance (gs) and net assimilation rate (A) in S. wrightii and P. velutina during the monsoon sampling period were higher than those observed during the pre-monsoon period; however, the magnitude of seasonal shifts in maximum gs and A in S. wrightii (from 0.08 ± 0.01 to 0.51 ± 0.03 mol m−2 s−1 and 10.3 ± 0.8 to 62.7 ± 3.5 µmol m−2 s−1 for gs and A, respectively) was greater than in P. velutina (from 0.24 ± 0.03 to 0.28 ± 0.03 mol m−2 s−1 and 17.1 ± 2.9 to 19.1 ± 3.3 µmol m−2 s−1 for gs and A, respectively).

Diel variation in sucrose and glucose concentration

Large diel variation in leaf sucrose concentration was observed in the C4 grass S. wrightii and the C3 tree P. velutina during both pre-monsoon and monsoon sampling periods; leaf sucrose concentration increased during the daytime, reaching its maximum values at late afternoon, and decreased thereafter during the night-time (Fig. 1). However, the magnitude of diel variation in leaf sucrose concentration was greater in S. wrightii (from 1.3 ± 0.4 to 17.9 ± 0.3 mg g−1 dry biomass and from 1.3 ± 0.5 to 17.4 ± 1.8 mg g−1 dry biomass for the pre-monsoon and monsoon sampling periods, respectively) than in P. velutina (from 3.0 ± 0.4 to 11.1 ± 1.8 mg g−1 dry biomass and 10.1 ± 0.7 to 21.2 ± 1.2 mg g−1 dry biomass for the pre-monsoon and monsoon sampling periods, respectively). Systematic diel variation in leaf glucose concentration was observed in S. wrightii (from 2.2 ± 0.1 to 12.4 ± 1.1 mg g−1 dry biomass and 7.2 ± 0.4 to 14.8 ± 0.7 mg g−1 dry biomass for the pre-monsoon and monsoon sampling periods, respectively), but not in P. velutina, during both the pre-monsoon and monsoon sampling periods (Fig. 1). Leaf glucose concentration values in P. velutina were greater than in S. wrightii.

Figure 1.

Diurnal variation in sucrose and glucose concentrations (mg g−1 dry biomass) in leaves and roots of S. wrightii (circle) and P. velutina (triangle) during the pre-monsoon (solid) and monsoon (open) sampling periods. Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols).

Systematic diel variation in root sucrose concentration was observed in the C3 tree P. velutina and, to a lesser extent, also in the C4 grass S. wrightii during the pre-monsoon (Fig. 1). Root sucrose concentration in P. velutina decreased during the daytime, reaching its minimum value at 1500 h (14.0 ± 2.5 and 13.0 ± 0.9 mg g−1 dry biomass for the pre-monsoon and monsoon sampling periods, respectively), and increased during the night-time, reaching its maximum value at pre-dawn (19.7 ± 2.0 and 25.8 ± 2.3 mg g−1 dry biomass for the pre-monsoon and monsoon sampling periods, respectively). Root sucrose concentration in S. wrightii during the pre-monsoon sampling period was higher than that observed during the monsoon period, whereas root sucrose concentration in P. velutina during the pre-monsoon sampling period was lower than that observed during the monsoon period (Fig. 1). No apparent diel or seasonal changes in root glucose concentration were detected in S. wrightii and P. velutina (Fig. 1). In general, root glucose concentration was lower than root sucrose concentration except in S. wrightii during the monsoon sampling period (Fig. 1).

δ13C of leaf and root dark-respired CO2

Large nocturnal shifts in δ13CR_leaf were observed in S. wrightii and P. velutina during both the pre-monsoon and monsoon sampling periods (Fig. 2; Table 1). The δ13CR_leaf value gradually decreased during the night-time in both species. Systematic diel variation in δ13CR_root was detected in P. velutina, but not in S. wrightii during both the dry pre-monsoon and wet monsoon sampling periods (Fig. 2; Table 1). The magnitude of diel variation in δ13CR_root in P. velutina was 3.8 ± 0.2 and 3.9 ± 0.1‰ for the pre-monsoon and monsoon sampling periods, respectively. The δ13CR_root value in P. velutina was not significantly different between the pre-monsoon and monsoon sampling periods; however, root-respired CO2 in S. wrightii during the pre-monsoon sampling period was 13C enriched by 2.6 ± 0.4‰ compared with that during the monsoon period.

Figure 2.

Nocturnal variation in the carbon isotope composition (‰) of CO2 evolved from leaves (δ13CR_leaf) and diel variation in the carbon isotope composition of root-respired CO2 (δ13CR_root) in S. wrightii (circle) and P. velutina (triangle) during the pre-monsoon (solid) and monsoon (open) sampling periods. Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols). Nocturnal δ13CR_leaf values in S. wrightii and P. velutina have been reported in Sun et al. (2010).

Table 1.  Degrees of freedom (d.f.) and P values from one-way anova of the nocturnal shift in the carbon isotope composition of leaf dark-respired CO2, leaf glucose and leaf sucrose, as well as of the diel variation in the carbon isotope composition of root dark-respired CO2, root glucose and root sucrose
SpeciesSeasonOrganδ13CRδ13Cglucoseδ13Csucrose
d.f.Pd.f.Pd.f.P
S. wrightiiPre-monsoonLeaf2<0.0120.6220.58
Root80.1380.3480.50
MonsoonLeaf2<0.0120.2720.51
Root80.3080.1680.30
P. velutinaPre-monsoonLeaf20.0320.3420.11
Root8<0.018<0.0180.03
MonsoonLeaf20.0420.2720.06
Root8<0.0180.0180.12

δ13C of glucose and sucrose

We observed no systematic diel variation in the carbon isotope composition of leaf sucrose and glucose in S. wrightii and P. velutina during the pre-monsoon and monsoon sampling periods (Fig. 3; Table 1). The magnitude of diel changes in δ13C value of leaf glucose and sucrose during the monsoon season [glucose: 2.2 ± 0.1‰ (S. wrightii) and 2.4 ± 0.2‰ (P. velutina); sucrose: 1.9 ± 0.2‰ (S. wrightii) and 2.4 ± 0.2‰ (P. velutina)] was greater than that during the dry pre-monsoon season [glucose: 0.8 ± 0.2‰ (S. wrightii) and 1.0 ± 0.1‰ (P. velutina); sucrose: 1.6 ± 0.2‰ (S. wrightii) and 1.5 ± 0.1‰ (P. velutina)]. Leaf sucrose in P. velutina during the pre-monsoon season was slightly 13C-enriched compared with that during the monsoon period; whereas the δ13C value of leaf sucrose in S. wrightii showed no apparent seasonal differences. Systematic diel variation in the carbon isotope composition of root glucose, rather than root sucrose, was detected in P. velutina (Fig. 3).

Figure 3.

Diel variation in the carbon isotope composition (‰) of sucrose and glucose extracted from S. wrightii (circle) and P. velutina (triangle) leaves and roots. The carbon isotope composition of sucrose and glucose was measured during both the pre-monsoon (solid) and monsoon (open) sampling periods Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols).

The observed diel variation in the δ13C value of root glucose in P. velutina was statistically significant (Table 1), with δ13C values decreasing during the morning period, reaching their minimum at 1200 h, and then increasing thereafter and reaching their maximum at pre-dawn. No apparent seasonal differences were detected in the δ13C value of root sucrose and glucose in P. velutina. The δ13C value of root sucrose and glucose in S. wrightii showed no systematic diel changes (Fig. 3; Table 1). However, root sucrose in S. wrightii during the pre-monsoon sampling period was 13C-enriched compared with that during the monsoon period.

Relationships between δ13C of leaf and root-respired CO2 and δ13C of glucose and sucrose

No significant correlation was detected between night-time values of δ13CR_leaf and leaf δ13Csucrose or between δ13CR_root and root δ13Csucrose in S. wrightii and P. velutina (Fig. 4). Leaf dark-respired CO2 in S. wrightii and P. velutina was 13C enriched compared with leaf sucrose during most of the night-time period, whereas root-respired CO2 had similar δ13C values relative to root sucrose (Fig. 4). Strong positive correlations were detected between diel values of δ13CR_root and root δ13Cglucose in P. velutina during both the pre-monsoon (r2 = 0.48; P < 0.05) and monsoon (r2 = 0.84; P < 0.01) sampling periods, but not between night-time δ13CR_leaf and leaf δ13Cglucose in both species, or between δ13CR_root and root δ13Cglucose in S. wrightii. Root-respired CO2 in P. velutina was 13C enriched relative to root glucose (Fig. 4d).

Figure 4.

Relationships between the carbon isotope composition (‰) of CO2 evolved from darkened leaves (δ13CR_leaf) and roots (δ13CR_root) and the carbon isotope composition of sucrose (δ13Csucrose) and glucose (δ13Cglucose) in S. wrightii (circle) and P. velutina (triangle) during the pre-monsoon (solid) and monsoon (open) sampling periods. Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols). r2 and P values are provided.

Flux-weighted respiratory apparent 13C/12C fractionation

We calculated respiratory apparent fractionation in leaf- and root-respired CO2 during the night-time with respect to biomass (ΔR_biomass), starch (ΔR_starch), lipids (ΔR_lipid), sucrose (ΔR_sucrose) and glucose (ΔR_glucose) as putative respiratory substrates. The flux-weighted carbon isotope composition of night-time leaf- and root-respired CO2 and the carbon isotope composition of potential respiratory substrates are provided in Table 2. We observed significant differences in flux-weighted ΔR_biomass (P = 0.04), ΔR_starch (P < 0.01), ΔR_lipid (P < 0.001), ΔR_sucrose (P = 0.02) and ΔR_glucose (P < 0.01) between leaves and roots. Flux-weighted ΔR_biomass, ΔR_lipid, ΔR_sucrose and ΔR_glucose in roots were greater than those in leaves, but ΔR_starch in roots was less than that in leaves (Fig. 5). Significant species differences in respiratory apparent fractionation were detected in flux-weighted ΔR_biomass (P < 0.01), ΔR_sucrose (P < 0.01) and ΔR_glucose (P < 0.01), but not in ΔR_starch (P = 0.9) and ΔR_lipid (P = 0.06). Flux-weighted ΔR_biomass, ΔR_sucrose and ΔR_glucose in the C4 grass S. wrightii were greater than those in the C3 tree P. velutina (Fig. 5).

Table 2.  The carbon isotope composition (‰) of respired CO2 (δ13CR), biomass (δ13Cbiomass), starch (δ13Cstarch), lipids (δ13Clipid), glucose (δ13Cglucose) and sucrose (δ13Csucrose) in S. wrightii and P. velutina during the pre-monsoon and monsoon sampling periods
Speciesδ13CRδ13Cbiomassδ13Cstarchδ13Clipidδ13Cglucoseδ13Csucrose
  1. The carbon isotope ratios of biomass, starch and lipids are the average of samples collected at 0600 and 1800h from three replicate plants of each species. The carbon isotope composition of glucose and sucrose are the average of samples collected at 2100, 0000 and 0300 h. δ13CR in leaves is respiration-rate weighted night-time (2100, 0000 and 0300 h) δ13C of leaf-respired CO2. δ13CR in roots is the average of values measured at 2100, 0000 and 0300 h. Data are means ± 1 SE (n = 3).

S. wrightii      
 Pre-monsoon      
  Leaf−13.7 ± 0.3−15.2 ± 0.1−14.8 ± 0.1−23.4 ± 0.3−15.5 ± 0.2−15.1 ± 0.2
  Root−14.7 ± 0.1−13.9 ± 0.3−16.8 ± 0.2−20.1 ± 0.5−14.9 ± 0.3−13.8 ± 0.3
 Monsoon      
  Leaf−14.5 ± 0.4−14.8 ± 0.1−15.9 ± 0.3−24.9 ± 0.3−13.9 ± 0.2−14.8 ± 0.4
  Root−17.3 ± 0.3−13.5 ± 0.2−18.4 ± 0.4−22.7 ± 0.2−18.0 ± 0.5−16.5 ± 0.5
P. velutina      
 Pre-monsoon      
  Leaf−22.7 ± 0.6−26.5 ± 0.4−23.1 ± 0.4−29.7 ± 0.3−26.7 ± 0.2−24.4 ± 0.4
  Root−22.2 ± 0.2−26.4 ± 0.2−24.7 ± 0.4−29.5 ± 0.4−26.9 ± 0.5−24.3 ± 0.3
 Monsoon      
  Leaf−23.9 ± 0.5−26.7 ± 0.2−24.0 ± 0.1−30.2 ± 0.2−30.1 ± 0.5−25.3 ± 0.4
  Root−23.0 ± 0.2−25.7 ± 0.2−23.9 ± 0.1−29.2 ± 0.2−26.5 ± 0.4−23.9 ± 0.6
Figure 5.

Flux-weighted respiratory apparent 13C/12C fractionation (‰) in night-time (2100, 0000 and 0300 h) leaf- and root-respired CO2 relative to biomass, starch, lipids, sucrose and glucose in S. wrightii and P. velutina during the pre-monsoon (solid) and monsoon (open) sampling periods. The respiration-rate weighted carbon isotope composition of leaf night-time (2100, 0000 and 0300 h)-respired CO2 was used for the calculation of leaf respiratory apparent 13C/12C fractionation. The averaged carbon isotope composition value at 2100, 0000 and 0300 h in root-respired CO2 was used for the calculation of root respiratory apparent 13C/12C fractionation. Data are means ± 1 SE (n= 3).

Correlations between sucrose concentration, δ13CR_leaf, δ13CR_root and ΔR_sucrose

Strong positive correlations were detected between night-time δ13CR_leaf and sucrose concentration in S. wrightii and P. velutina leaves, as well as between δ13CR_root and sucrose concentration in P. velutina roots (Fig. 6). Significant negative linear correlations were observed between ΔR_sucrose and sucrose concentration in S. wrightii leaf and in P. velutina leaf and root (Fig. 7). However, correlations between ΔR_glucose and glucose concentration were not significant for leaves or roots in the two species (data not shown).

Figure 6.

Relationships between the carbon isotope composition (‰) of CO2 evolved from darkened leaves (δ13CR_leaf) and roots (δ13CR_root) and sucrose concentration (mg g−1 dry biomass) in S. wrightii (circle) and P. velutina (triangle) during the pre-monsoon (solid) and monsoon (open) sampling periods. Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols). r2 and P values are provided.

Figure 7.

Relationships between the respiratory apparent 13C/12C fractionation (‰) relative to sucrose (ΔR_sucrose) in leaf and root dark-respired CO2 and sucrose concentration (mg g−1 dry biomass) in S. wrightii (circle) and P. velutina (triangle) during the pre-monsoon (solid) and monsoon (open) sampling periods. Data are the means of three replicate trees or grasses (standard error bars are shown when larger than symbols). r2 and P values are provided.

DISCUSSION

The carbon isotope composition of root-respired CO2

As a major component of ecosystem respiration, diel variation in δ13CR_root has the potential to affect the estimation of the carbon isotope composition of ecosystem-respired CO2 (δ13CR_eco). However, the extent that short-term variation in δ13CR_root will alter δ13CR_eco and the estimation of autotrophic versus heterotrophic respiration rates may depend on the functional type composition of vegetation and the environmental conditions of growth. We observed large (approximately 4‰) and systematic diel variation in δ13CR_root in the C3 woody species P. velutina, but not in the C4 grass S. wrightii (Fig. 2). Short-term shifts in δ13CR_root have been reported in the C3 herbs Helianthus annuus (Klumpp et al. 2005), Ricinus communis (Gessler et al. 2009), Melissa officinalis and Salvia officinalis (Wegener et al. 2010) growing under controlled environment conditions, as well as in a C3 herb Tuberaria guttata (Unger et al. 2010) under field conditions. However, no shifts in δ13CR_root at hourly timescales were reported in a C3 shrub Halimium halimifolium and a C3 non-aromatic herb Oxalis triangularis in a greenhouse environment (Wegener et al. 2010).

Root-respired CO2 in the C4 grass S. wrightii during the night-time was slightly 13C depleted compared with biomass and sucrose (Fig. 5; Table 2) which is consistent with observations from previous studies (Badeck et al. 2005; Klumpp et al. 2005; Gessler et al. 2009). 13C depletion in root-respired CO2 has been attributed to isotope fractionations associated with the re-fixation of respiratory CO2 in stems and roots catalyzed by phosphoenolpyruvate carboxylase (PEPc) (Klumpp et al. 2005; Gessler et al. 2009). PEPc discriminates against 13C during fixation of the bicarbonate (HCO3-) substrate, but the initial step of CO2 hydration favours 13C, such that overall CO2 re-fixation favours 13CO2, favouring preferential loss of 13C-depleted CO2. A recent study suggests that the pentose phosphate pathway (PPP) may also play a role in the observed 13C depletion in root-respired CO2 (Bathellier et al. 2009) because in the PPP cycle the CO2 released originates from the 13C-depleted C-1 of glucose (Rossmann et al. 1991). However, we observed that the root-respired CO2 in P. velutina was 13C enriched relative to biomass, glucose and sucrose (Fig. 5). 13C enrichment in root- and stem-respired CO2 relative to putative organic carbon sources in tree species growing under field conditions has been reported elsewhere (Damesin & Lelarge 2003; Brandes et al. 2006; Gessler et al. 2007). These results highlight potential functional group (C3 versus C4) differences in root ΔR that may be associated with differences in PEPc activity. C3 species have relatively low PEPc activity compared with C4 species (Raven & Farquhar 1990), which may explain to some extent the differences observed between P. velutina, a C3 shrub, and S. wrightii, a C4 grass. Functional group differences in the magnitude of diel variation in δ13CR_root as well as root ΔR need to be carefully studied as they may influence δ13CR_eco and the estimation of carbon fluxes from heterotrophic and autotrophic sources.

Effects of δ13C of primary respiratory substrates on δ13CR_leaf and δ13CR_root

Large diel variation in photosynthetic discrimination (ΔP), especially in the C3 tree P. velutina (data not shown), has been hypothesized to affect δ13CR_leaf by altering δ13C of primary respiratory substrates (Sun et al. 2009). However, we observed no apparent nocturnal variation in leaf δ13Cglucose or δ13Csucrose (Fig. 3), or in leaf δ13Cstarch or δ13Clipids (data not shown). Lack of short-term (diel) changes in δ13C of primary respiratory substrates has been reported elsewhere (Hymus et al. 2005; Göttlicher et al. 2006; Sun et al. 2009). Therefore, the hypothesis that short-term variation in ΔP can affect night-time δ13CR_leaf by altering the δ13C of primary respiratory substrates lacks support from our current observations. Indeed, results of a recent study also suggested that variation in δ13C in glucose, sucrose and fructose was not sufficient to explain diel shifts in δ13CR_leaf (Werner et al. 2009). The impacts of variation in ΔP on δ13Cglucose and δ13Csucrose are likely reduced by the large internal glucose and sucrose pools with relatively low turnover rates.

Significant positive correlations between δ13CR_root and root δ13Cglucose (Fig. 4d) support our hypothesis that diel changes in δ13CR_root in P. velutina resulted, to some extent, from changes in the carbon isotope composition of root glucose. Root respiration is fuelled by leaf-derived phloem C, which is likely to vary in its δ13C values at diel timescales. Indeed, diel variation in the δ13C of carbohydrates being transported from leaf to root has been observed in Eucalyptus delegatensis (Gessler et al. 2007), Pinus sylvestris (Brandes et al. 2006) and R. communis (Gessler et al. 2008, 2009). Short-term changes in the δ13C of phloem sugars may have resulted from variation in the δ13C of leaf sugars or isotope effects of the remobilization of stored carbohydrates along the phloem transport pathway (Gessler, Rennenberg & Keitel 2004), or both. Daytime phloem sugars are derived from recently assimilated triose phosphate, whereas night-time phloem sugars originate from transitory starch degradation. The 13C enrichment (up to 4‰) in starch relative to triose phosphate (Gleixner & Schmidt 1997; Gleixner et al. 1998) may lead to diel variation in the δ13C of phloem sugars (Gessler et al. 2008, 2009). Moreover, short-term changes in ΔP (Hymus et al. 2005; Sun et al. 2009) may also impact the δ13C values of phloem sugars by directly altering δ13C values of triose phosphate. However, the magnitude of diel shifts in the δ13C value of leaf glucose and sucrose in P. velutina (approximately 2‰) was not large enough to explain associated changes in root δ13Cglucose. Thus, the observed diel variation in root δ13Cglucose (Fig. 3) may have resulted from changes in the δ13C of leaf sugars and isotope effects associated with transitory starch formation and degradation.

Effects of substrate availability on respiratory apparent 13C/12C fractionation

The observed strong positive correlations between sucrose concentration and δ13C of respired CO2 (Fig. 6), as well as negative correlations between sucrose concentration and respiratory apparent 13C/12C fractionation relative to sucrose (ΔR_sucrose) (Fig. 7), support our second hypothesis that substrate availability impacts δ13CR_leaf and δ13CR_root by altering ΔR. Changes in ΔR may have resulted from molecular fragmentations at metabolic branch points (Tcherkez et al. 2003), as the 13C distribution within hexoses is not homogenous. The C-3 and C-4 carbon atoms are 13C enriched relative to other carbon atoms (Rossmann et al. 1991; Hobbie & Werner 2004). Therefore, CO2 evolved from decarboxylation of pyruvate is 13C enriched compared with CO2 from oxidation of acetyl-CoA in the TCA cycle. The increase in the allocation of acetyl-CoA to energy production may have resulted from reduced substrate availability or from a diminishing requirement for precursors for biosynthesis. For leaves, a greater proportion of acetyl-CoA is likely to be diverted to biosynthesis at the beginning of the night-time period to meet the requirement of plant growth (Walter et al. 2005; Sadok et al. 2007) and cellular repair and maintenance. The allocation of acetyl-CoA to biosynthesis is likely to cause respired CO2 to be 13C enriched relative to the respiratory substrate (more negative in ΔR). Throughout the night-time, progressively more acetyl-CoA may be oxidized in the TCA cycle to generate ATP leading to an increase in ΔR (less negative in ΔR) and a decrease in δ13CR_leaf. Although our observations (Figs 6 & 7) support the hypothesis that substrate availability controls δ13CR_leaf and δ13CR_root by altering ΔR, we cannot rule out the effects of diminishing requirement for precursors for biosynthesis on ΔR through altering the allocation of acetyl-CoA. Moreover, diminishing requirement for precursors for biosynthesis throughout night-time may down-regulate transitory starch degradation and subsequently reduce substrate availability. Others have argued that variation in ΔR can be associated with changes in substrate availability and produce short-term changes in the δ13C of leaf-respired CO2 (Hymus et al. 2005; Sun et al. 2010). Indeed, our current observations suggest that it may also control, to some extent, diel variation in δ13CR_root.

For C3 species, the theoretical maximum isotopic variation between 0 and 100% commitment of acetyl-CoA to the TCA cycle is 4‰ (Werner 2010), which is sufficient to explain the less than 4‰ nocturnal shift in δ13CR_leaf and diel shift in δ13CR_root observed in P. velutina (Fig. 2). Because of greater intramolecular 13C homogeneity in C4 plants (Rossmann et al. 1991), the theoretical maximum isotopic variation associated with changes in the commitment of acetyl-CoA to the TCA cycle is 2.2‰, which is not enough to explain the observed 4.6‰ nocturnal shift in the C4 grass S. wrightii (Fig. 2). Rather, kinetic isotope effects of respiratory decarboxylating enzymes, such as pyruvate dehydrogenase, isocitrate dehydrogenase and 2-oxoglutarate dehydrogenase (Tcherkez & Farquhar 2005; Werner et al. 2011) may increase nocturnal variation in δ13CR_leaf. Indeed, Werner (2010) suggests that the combined isotope effects of respiratory decarboxylating enzymes, relative carbon flux changes through pyruvate dehydrogenase and the TCA cycle may theoretically induce more than a 9‰ shift in δ13C of respired CO2.

CONCLUSIONS

Large nocturnal variation in δ13CR_leaf was observed in the C3 woody species P. velutina and the C4 grass S. wrightii; however, systematic nocturnal and diel variations in δ13CR_root were observed only in P. velutina. Negative correlations between respiratory apparent 13C/12C fractionation relative to sucrose and leaf sucrose concentrations suggest that night-time changes in δ13CR_leaf may be controlled by short-term changes in ΔR associated with variation in substrate availability, or potentially by changes in substrate demand for biosynthesis. Strong, positive correlations between δ13CR_root and δ13C of root glucose suggest that diel shifts in δ13CR_root in P. velutina were controlled to some extent by short-term changes in δ13C of phloem sugars resulting from changes in the δ13C of leaf sugars or isotope effects associated with transitory starch formation and degradation. The magnitude of short-term changes in δ13C of respired CO2 and night-time ΔR differed significantly between species and organs. These patterns need to be considered when δ13C measurements are used to partition net CO2 exchange between ecosystems and the atmosphere into heterotrophic and autotrophic components.

ACKNOWLEDGMENTS

This material is based upon work supported by the National Science Foundation under Grant no. 0414680. S. Sharma and W. Cable helped with carbon isotope analysis. J-Y. Ma and J. Jones helped with glucose and sucrose purification. R. Scott, T. Huxman, S. Chen and G. Barron-Gafford supported the field work.

Ancillary