Oxygen isotope enrichment of organic matter in Ricinus communis during the diel course and as affected by assimilate transport

Authors

  • Arthur Gessler,

    1. Environmental Biology Group; Research School of Biological Sciences, Australian National University, Canberra, Australia;
    2. School of Forest and Ecosystem Science, University of Melbourne, Water Street, Creswick, Victoria, Australia;
    3. Present address: Core Facility Metabolomics, Centre for Systems Biology (ZBSA) and Chair of Tree Physiology, University of Freiburg, Georges-Köhler Allee 53/54, 79085 Freiburg im-Bresigau, Germany
    Search for more papers by this author
  • Andreas D. Peuke,

    1. School of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney NSW 2052 Australia;
    Search for more papers by this author
  • Claudia Keitel,

    1. Environmental Biology Group; Research School of Biological Sciences, Australian National University, Canberra, Australia;
    Search for more papers by this author
  • Graham D. Farquhar

    1. Environmental Biology Group; Research School of Biological Sciences, Australian National University, Canberra, Australia;
    Search for more papers by this author

Author for correspondence: Arthur Gessler Tel: +49 761 2038310 Fax: +49 761 2038302 E-mail: arthur.gessler@sonne.uni-freiburg.de

Summary

  • • The oxygen isotope composition in leaf water and organic compounds in different plant tissues is useful for assessing the physiological performance of plants in their environment, but more information is needed on Δ18O variation during a diel course.
  • • Here, we assessed Δ18O of the organic matter in leaves, phloem and xylem in stem segments, and fine roots of Ricinus communis during a full diel cycle. Enrichment of newly assimilated organic matter in equilibrium with leaf water was calculated by applying a nonsteady-state evaporative enrichment model.
  • • During the light period, Δ18O of the water soluble organic matter pool in leaves and phloem could be explained by a 27‰ enrichment compared with leaf water enrichment. Leaf water enrichment influenced Δ18O of phloem organic matter during the night via daytime starch synthesis and night-time starch remobilization. Phloem transport did not affect Δ18O of phloem organic matter.
  • • Diel variation in Δ18O in organic matter pools can be modeled, and oxygen isotopic information is not biased during transport through the plant. These findings will aid field studies that characterize environmental influences on plant water balance using Δ18O in phloem organic matter or tree rings.

Introduction

The determination of the oxygen isotope composition in leaf water and organic compounds is a promising tool for assessing the physiological performance of plants in their environment (Wang & Yakir, 1995; Barbour et al., 2001, 2002; Cernusak et al., 2005). In particular, the isotopic information in organic matter pools with different turn-over times (e.g. short-lived phloem sugars representing recent assimilates; long-lived structural compounds such as tree ring cellulose) is now widely used – often together with δ13C signatures – to integrate the influence of a range of environmental factors on the water balance of plants (Saurer et al., 1997; Scheidegger et al., 2000; Keitel et al., 2003; Brandes et al., 2006). The oxygen isotope composition of organic matter is mainly influenced by two factors: the δ18O signature of source water and the evaporative enrichment of mean lamina leaf water, which is the reaction water for the newly produced assimilates (Fehri & Letolle, 1977; Cernusak et al., 2003b). Mean lamina leaf water enrichment depends in turn on the diffusion of 18O enriched water from the sites of evaporation back into the mesophyll cells and the convection of unenriched xylem water via the transpiration stream in the opposite direction (Farquhar & Lloyd, 1993; Barbour et al., 2000a). As evaporative enrichment is closely linked to environmental conditions (water vapor pressure in the air, air temperature) and physiological traits (stomatal conductance, transpiration) this measure is of central interest for characterizing water relations in plants.

The theory of evaporative enrichment of leaf water and the imprinting of this signal on newly produced assimilates is well described (Barbour et al., 2000b; Farquhar & Cernusak, 2005). However, there is little information on how the Δ18O signal is altered with time and during transport through the plant. During the day, transitory starch is accumulated in the leaves. At night sugars originating from starch hydrolysis/phosphorolysis (Smith et al., 2003) are loaded into the phloem serving as a carbon source for heterotrophic tissues. It must therefore be assumed that the oxygen isotopic composition of sugars transported in the phloem during the dark period is influenced not only by Δ18O of leaf water during the preceding day (when the starch was produced) but also potentially during the night when starch is broken down since during starch breakdown oxygen from reaction water is introduced into the hexose molecules produced (Smith et al., 2003; Weise et al., 2004). Insights into such effects are important for studies that use Δ18O in different organic matter pools in different plant organs such as phloem sugars (Cernusak et al., 2003a, 2005; Brandes et al., 2006; Keitel et al., 2006) to characterize plant reactions to environmental constraints. Since phloem sugars are the main source of organic matter the stem of trees is supplied with, diel variations might also affect Δ18O of whole wood or cellulose in tree rings (Saurer et al., 1997; Jäggi et al., 2003). Diel variations in the expression of key enzymes of lignin biosynthesis have been described for herbaceous plants (Rogers et al., 2005). There is not much known about a potential circadian control and environmental triggers of cellulose and lignin synthesis in trees but varying carbon fluxes into lignin or cellulose pools during the diel course in combination with variations in Δ18O of phloem sugars would strongly affect the oxygen isotope composition of tree rings.

Post-photosynthetic transformation of organic matter also takes place. Depending on the 18O enrichment of medium water within which these transformations occur, Δ18O of cellulose, lignin, starch and other substances in different organs within the plant will be altered compared with the original organic substrates (Yakir & Deniro, 1990; Farquhar et al., 1998; Roden & Ehleringer, 1999; Roden et al., 2000; Barbour et al., 2004). It is also known that carbohydrate transport in the sieve tubes is highly dynamic and unloading and retrieval of sugars take place in the transport phloem at the same time (van Bel, 2003). In Phaseolus, for example, sieve tubes lose 6% of photosynthate per centimeter of stem, of which about two-thirds is retrieved (Minchin & Thorpe, 1987). Sugars unloaded from the transport phloem might undergo metabolic conversion in the unenriched reaction water of stems before they are reloaded into the sieve tubes. As a consequence, Δ18O of phloem-transported organic matter (Δ18Ophloem-OM; for abbreviations see Table 1) could vary with transport distance and increasingly lose the isotopic information imprinted from leaf water. In addition, Cernusak et al. (2005) concluded that refixation of respired CO2 by photosynthetic bark was responsible for a decrease in Δ18O in bark organic matter in Eucalyptus globulus as bark water within which carbon assimilation was assumed to occur was not subjected to evaporative enrichment. Since R. communis stems are also noticeably green, partial exchange of 18O-depleted bark-assimilated sugars with phloem transported organic matter could also cause basipetal Δ18O gradients within the phloem.

Table 1.  Abbreviations and symbols used in text
Abbreviation or symbolDefinition
ANet assimilation rate
α+Oxygen isotope effect for phase transition between liquid water and vapor
αkOxygen isotope effect during water vapor diffusion from leaf intercellular air spaces to the atmosphere
CMolar concentration of water
ciCO2 concentration in leaf intercellular air spaces
caCO2 concentration in the atmosphere
DDiffusivity of H218O in water
Δ18O18O enrichment of any water or dry matter compared to source water
Δ18OesPredicted steady-state 18O enrichment of water at evaporative sites (without Péclet effect) in leaves compared with source water
inline imagePredicted steady-state 18O enrichment of new photosynthates without Péclet effect (Δ18Oes + ɛwc)
Δ18OL18O enrichment of mean lamina leaf water compared with source water
Δ18OLnPredicted nonsteady-state 18O enrichment of mean lamina leaf water (considering a Péclet effect) compared with source water
inline imagePredicted nonsteady-state 18O enrichment of new photosynthates (Δ18OLn + ɛwc)
Δ18OLsPredicted steady-state 18O enrichment of mean lamina leaf water (considering a Péclet effect)
inline imagePredicted steady-state 18O enrichment of new photosynthates (Δ18OLs + ɛwc)
Δ18Opw18O enrichment of phloem water compared to source water
Δ18Ophloem-OM18O enrichment of organic matter transported in the phloem compared to source water
Δ18Ototal-leaf18O enrichment of total leaf organic matter compared to source water
Δ18Ototal-stem18O enrichment of total organic matter in stem sections compared to source water
Δ18Ov18O enrichment of water vapour compared to source water
Δ18Ows18O enrichment of water soluble organic matter in leaves compared to source water
Δ18Oxylem-OM18O enrichment of organic matter transported in the xylem compared to source water
δ18O18O/16O relative to Vienna standard mean ocean water (VSMOW = 0)
ETranspiration rate
eaVapor pressure in the atmosphere
eiVapor pressure in leaf intercellular air spaces
ɛ+Equilibrium 18O fractionation between liquid water and vapor
ɛkKinetic fractionation for diffusion of H218O from leaf intercellular spaces to the atmosphere
ɛwcEquilibrium 18O fractionation between organic oxygen and medium water
gTotal leaf conductance (stomata + boundary layer conductance) to H2O from leaf intercellular air spaces to the atmosphere
gsStomatal conductance to H2O
LScaled effective path length
pexProportion of oxygen atoms exchanging with medium water during cellulose synthesis
Péclet number
Rs18O/16O of source water
RVSMOW18O/16O of Vienna standard mean ocean water (VSMOW)
rbBoundary layer resistance to water vapour diffusion
rsStomatal resistance to water vapour diffusion
wiMole fraction of water vapour in leaf intercellular air spaces
WLeaf lamina water concentration

About 15 yr ago Jeschke and Pate (1991) showed that considerable amounts of organic matter (approx. 8% of the net C assimilation) are loaded into the xylem in stems and roots of R. communis in order to contribute to the supply of the growing shoot. Depending on the metabolic origin and composition of this organic matter pool it might carry a signature influenced by the evaporative enrichment of leaf water. Basipetal gradients in Δ18O of phloem- or xylem-transported organic compounds could therefore also influence the isotopic enrichment of structural organic matter in different plant parts.

In order to obtain more precise information on potential variations of Δ18O in organic matter pools during the diel course and along the plant axis, we assessed the oxygen isotope enrichment of the water soluble (Δ18Ows) and total organic matter fraction in leaves of R. communis18Ototal-leaf) during a full diel cycle. In addition, we characterized Δ18O in total organic matter of stem segments (Δ18Ototal-stem) at six different positions and in fine roots and determined Δ18Ophloem-OM and Δ18O in xylem-transported organic matter (Δ18Oxylem-OM) at different positions along the plant axis. Barbour et al. (2000a) and Cernusak et al. (2003b) showed that measured or modeled evaporative enrichment of leaf water in R. communis was in good agreement, in the steady state, with phloem organic matter Δ18O collected from the petiole. We extended this to the nonsteady state, by measuring the oxygen isotopic enrichment of newly assimilated organic matter during the dynamic environmental conditions of a diel course, and comparing this with the enrichment of leaf water as calculated from a new nonsteady-state model (Farquhar & Cernusak, 2005) of leaf water enrichment.

In general, no direct exchange between phloem-transported sucrose and phloem water occurs since the sucrose molecule contains no carbonyl groups. However, there are other processes after equilibration with leaf water in the light that may affect organic matter isotopic enrichment. One is starch breakdown, and another is phloem exchange with stem sugars. Therefore we aimed to explore the following notion:

  • 1Δ18Ows and Δ18Ophloem-OM during the light period can be described with the nonsteady-state model taking a 27‰ enrichment between leaf water and organic matter (Yakir & Deniro, 1990) into account.
  • 2The soluble organic matter transported in the phloem during the night and originating from transitory starch is not only influenced by the photosynthesis-weighted leaf water enrichment during day but also by the differing leaf water Δ18O during the night.
  • 3The dynamic efflux to and retrieval of sugars from the stem may alter Δ18O of phloem organic matter in the basipetal direction.

Materials and Methods

Plant material

Seeds of Ricinus communis L. were germinated in vermiculite moistened with 0.5 mm CaSO4. After 13–15 d the plants were transferred to substrate, one plant in a pot of 5 l. The substrate consisted of two parts commercial potting soil (Floradur; Floragard GmbH, Oldenburg Germany) and one part Perlite (Perligran; G, Deutsche Perlite GmbH, Dortmund, Germany). Every third day the pots were well watered with tap water and after the first month the plants were supplied every second week with a commercial fertilizer (0.3% Hakaphos Blau; Compo GmbH, Münster, Germany).

The plants were cultivated for 35–40 d in a glasshouse (26 ± 5°C) with a 16 h photoperiod provided by natural daylight plus mercury-vapour lamps (Osram HQL 400; Osram, Munich, Germany) supplying the plant with a minimum of 300–500 µmol photons m−2 s−1.

Experimental design

The Δ18O of phloem organic matter was determined at six different positions (p-A to p-F; Fig. 1) along the axis at six time points (four in the light (10 : 30, 12 : 00, 16 : 30, 19 : 00 h (± approx. 1.0 h)) and two in the dark period (24 : 00, 03 : 00 h (± approx. 1.0 h)) during a diel course. At each time-point three to four plants were harvested destructively.

Figure 1.

Scheme of the plants used in the experiments with sampling positions. For the analysis of phloem organic matter, phloem sap was obtained from six positions along the stem (p-A to p-F). From the same position stem sections (S-A to S-F) were harvested after phloem sampling. The cotyledons (Cot1 and Cot2) were already shed at the start of the experiments. Stomatal conductance (gs), transpiration and oxygen isotope composition were determined for all fully developed leaves (L1–L7). In addition, at all six time-points fine root samples were harvested. Xylem sap was obtained from cut midribs of the leaves (x-III to x-VII) and from tissue flaps of stem internodes (x-I and x-II) at three time points.

Phloem sap was sampled by cutting the bark with a scalpel according to the procedure described by Jeschke and Pate (1991) transferred with a capillary into a 1.5 ml reaction tube and frozen immediately. The collections of phloem sap from each single plant lasted 45–60 min. After sampling of phloem sap, stem sections with a length of c. 3 cm were collected from the same positions (S-A to S-F).

In addition, all seven fully expanded leaves (L1 to L7) and fine roots were harvested and frozen in liquid nitrogen for the assessment of Δ18O in total and, for leaves only, water soluble organic matter (Fig. 1).

All tissue samples (leaves, stem sections and roots) were homogenized with mortar and pestle in liquid nitrogen. For the extraction of water soluble organic matter from leaves 1.5 ml of deionized water was added to 0.1 g aliquots of plant material. The mixture was agitated for 1 h at 4°C. The extract was heated at 100°C for 1 min to precipitate proteins and centrifuged (12 000 g for 5 min at 4°C). The supernatant is the water soluble (exportable) organic matter fraction consisting mainly of sugars but with some organic acids and amino acids (Brandes et al., 2006).

For all leaves, leaf temperature, transpiration, net CO2 assimilation, and stomatal conductance (gs) were determined at each time point (immediately before harvest) using a portable leaf gas exchange measurement system (LCA 4; ADC BioScientific Ltd, Hoddesdon, UK). The relative humidity (RH) near the lower leaf surface was determined with a temperature and humidity probe (Humicap 113y; Vaisala, Helsinki, Finland). Three to four plants were analysed per time point.

In addition to leaves, stem sections and phloem sap, xylem sap was obtained at three time points during a diel course (09 : 30, 18 : 30, 03 : 00 h (± 2 h)) by applying pneumatic pressure to the root system enclosed in a pressure vessel (Jeschke & Pate, 1991; Peuke, 2000). From each single experimental plant, a series of seven samples of xylem sap was collected first from cut midribs of the leaf and finally from tissue flaps of stem internodes (see Fig. 1, x-VII to x-I). Pressurizing the roots to a level of 0.35 ± 0.04 MPa caused exudation at the various sampling points, enabling sap volumes of sufficient size to be collected for analysis. During the collection procedure sap samples were stored on ice and thereafter kept at −80°C.

Irrigation water on the day of measurement and water vapour (during day and night) of the glasshouse air were collected for oxygen isotope analysis. Water vapour was collected by drawing air though a dry ice–ethanol trap at a flow rate of 1 l min−1 (Cernusak et al., 2003b).

Oxygen isotope theory

Steady-state enrichment of water at the evaporative site of the leaf (Δ18Oes) can be described according to Dongmann et al. (1974)

image(Eqn 1)

+ is the equilibrium fractionation between liquid water and water vapour; ɛk the kinetic fractionation during water vapour diffusion; Δ18Ov is the enrichment of water vapour in the atmosphere above source water; and ea/ei the ratio of ambient to intercellular water vapour concentration).

The ɛ+ value can be calculated according to Bottinga and Craig (1969) from a regression equation relating it to leaf temperature (T in K):

image(Eqn 2)

ɛk can be calculated according to Farquhar et al. (1989):

image(Eqn 3)

where rs and rb refer to stomatal and boundary layer resistance to water vapour (m2 s mol−1) and 32‰ and 21‰ are associated fractionation factors (Farquhar & Cernusak, 2005) based on new determinations of the isotopic effect for diffusion of H218O in air (Cappa et al., 2003).

Average lamina mesophyll water is, however, thought to be less enriched than the water at the evaporative sites because of the influx of xylem water into the leaf (Farquhar & Lloyd, 1993). Steady-state enrichment of mean lamina leaf water (Δ18OLs) depends on the steady-state enrichment at the evaporative site of the leaf (Δ18Oes) and on the lamina radial Péclet number ℘ (Farquhar & Gan, 2003).

image(Eqn 4)

where ℘ is defined as EL/CD, where E is transpiration rate (mol m−2 s−1), L is a scaled effective path length (m), C is the molar concentration of water (mol m−3), and D is the diffusivity of H218O in water (m2 s−1).

Eqn 4 describes mean lamina leaf water enrichment only under steady-state conditions. Although leaves are likely to reach isotopic steady state in the early afternoon, there will be times during the diel course when leaf water enrichment is not at steady state (Wang & Yakir, 1995; Cernusak et al., 2005). Thus, nonsteady-state conditions must be considered in a model that describes leaf water enrichment during the diel course. Nonsteady-state mean lamina leaf water enrichment (Δ18OLn) can be described according to Farquhar and Cernusak (2005) as follows:

image(Eqn 5)

+ is defined as (1 + ɛ+), αk is (1 + ɛk); W is the lamina leaf water concentration (mol m−2); t is time (s), g is total leaf conductance (stomata + boundary layer conductance); and wi is the mole fraction of water vapour in the air spaces inside the leaf).

Newly produced assimilates are assumed to carry the signature of the leaf water at the time when they were produced with an equilibrium fractionation factor (ɛwc) resulting in carbonyl oxygen being c. 27‰ more enriched than water (Sternberg & Deniro, 1983; Yakir & Deniro, 1990). Thus the enrichment in sugar produced during photosynthesis should be given by Δ18OLn + ɛwc.

Isotope measurements and isotopic calculations

Homogenized bulk samples, water extracts, phloem sap and xylem sap were dried at 60°C for 12 h. Samples were combusted in silver capsules (IVA Analysentechnik, Meerbusch, Germany) in a high-temperature conversion/elemental analyser (TC/EA Finnigan MAT GmbH, Bremen, Germany) coupled to an isotope ratio mass spectrometer (Delta Plus; Finnigan MAT GmbH) by a Conflo II interface (Finnigan MAT GmbH). The samples contained c. 300 µg O. One microliter of liquid phloem sap samples, irrigation water and trapped water vapor were injected manually into the TC/EA using a syringe.

The precision of the measurements as determined by repeated measurements of standards was 0.3‰ (1 SD, n = 10). Oxygen isotope ratios (δ18Ο) are presented relative to Vienna standard mean ocean water (VSMOW).

image(Eqn 6)

(Rs and RVSMOW are the isotope ratios (18O/16O) of the plant material and VSMOW, respectively).

Phloem sap water oxygen isotope ratios were determined as described by Cernusak et al. (2002), based on a technique for determination of the water component of a homogeneous mixture of water and dry matter (Gan et al., 2002).

The oxygen isotope composition of organic matter (δ18Oplant material) in leaves, roots, xylem and phloem was expressed as enrichment (Δ18O) above irrigation water (source water; δ18Osource water, −6.5‰):

image(Eqn 7)

Nonsteady-state enrichment (Δ18OLn) of leaf water was calculated according to Eqn 5. From Δ18OLn, 18O enrichment of newly produced organic matter (inline image) in isotopic equilibrium with leaf water was calculated assuming ɛwc = 27‰. For calculating the Péclet number ℘ a scaled effective path length L-value of 11.1 mm according to Cernusak et al. (2003b) was used. We also calculated the expected 18O enrichment of newly produced organic matter (1) taking into account a Péclet effect but assuming steady-state conditions (inline image) and (2) assuming no Péclet effect and steady-state conditions (inline image). The inline image was calculated from steady-state enrichment of mean lamina leaf water (Δ18OLs; Eqn 4) and inline image from steady-state enrichment of water at the evaporative site of the leaf (Δ18Oes; Eqn 1) assuming in both cases assimilates to be enriched by 27‰ compared with water.

In order to integrate calculated inline image, inline image, inline image over the whole canopy we weighted 18O enrichment for assimilation taking into account leaf area (m2) and CO2 assimilation rate (µmol m−2) of leaves L1 to L7. Similarly, observed Δ18O in leaf water soluble (Δ18Ows) and total organic matter (Δ18Ototal-leaf) were weighted by leaf area and oxygen content of the respective organic matter pools per m2 leaf area.

To calculate mean daytime Δ18O values, oxygen isotope enrichment was weighted by CO2 assimilation rate according to Cernusak et al. (2005)

image(Eqn 8)

(∫ A Δ18O · dt is the daily integral of the product of A and Δ18O and ∫ A · dt is the daily integral of photosynthesis).

Statistical analyses

All statistical analyses were performed using NCSS 2004 (Number Cruncher Statistical Software, Kaysville, UT, USA). Differences in Δ18O between different chemical fractions, time-points or different positions were determined using analysis of variance (GLM-anova). For analysis position was nested within time point. Regression lines between Δ18O from different organic matter pools were determined by linear regression analysis. Differences between mean Δ18O during day and night for a given organic matter pool were calculated applying Student's t-test. In order to estimate the uncertainties for the different 18O enrichment models we applied the principles of the Gaussian error propagation. For this calculation we included standard deviations of single measured parameters as errors.

Results

Leaf gas exchange

Figure 2 shows mean leaf temperature (TL) and RH of the air near the leaf of the seven leaves examined as well as transpiration rate (E) and stomatal conductance (gs). Leaf temperature showed a distinct diurnal course with maximum values of 31°C in the afternoon and minimum values of 28.5°C in the morning and at night. Relative humidity displayed an almost inverse pattern with values close to 95% during the night and a minimum of 53% at 16 : 30 h. Mean leaf area-weighted transpiration followed RH inversely and increased from 1.2 mmol m−2 s−1 in the morning to 4.5 mmol m−2 s−1 at 16 : 30 h. Night values were below 0.1 mmol m−2 s−1. Mean leaf area-weighted gs showed a comparable diel pattern with maximum stomatal conductance of approx. 0.25 mol m−2 s−1 in the afternoon. There was no significant difference in gs or E among the seven leaves examined.

Figure 2.

(a) Mean leaf temperature and mean relative humidity (RH) in the air close to the lower leaf surface; (b) transpiration rates E and (c) stomatal conductance (gs) of all seven leaves of the Ricinus communis plants examined during the diel course. Data shown in (b) and (c) are mean values of three to four plants. In addition, the average standard deviation for the mean values is given for each time-point. The bold lines in (b) and (c) refer to the leaf area weighted canopy mean values. The shaded areas denote the dark period. L1 to L7, leaf numbers as indicated in Fig. 1.

Δ18O in leaf organic matter

Observed Δ18O in leaf water-soluble organic matter (Δ18Ows) did not differ significantly among the seven leaves included in the study (Fig. 3a). Mean canopy Δ18Ows (Fig. 3a, thin black line) was approximately constant at 38‰ from 10 : 30 to 12 : 00 h. Towards the afternoon Δ18Ows peaked at >  42‰ and decreased again in the evening. At 24 : 00 h values from the morning were regained and later in the night a diel minimum of 35.7‰ was reached.

Figure 3.

(a) Predicted and observed Δ18O in water-soluble organic matter and (b) observed Δ18O total organic matter in leaves of Ricinus communis during the diel course. The symbols denote observed Δ18Ows (a) and Δ18Ototal (b) in the seven leaves L1 to L7. Data shown are mean values of three to four plants. In addition, the average standard error for the mean values is given. The effects of position (leaf number) and time as calculated with the GLM-anova procedure are given. The thin black lines refer to canopy integrated mean values of observed Δ18Ows (a) and Δ18Ototal-leaf (b). Canopy integrated predicted 18O enrichment of newly assimilated organic matter is indicated by the bold gray (inline image), the bold black dotted (inline image) and the bold black dashed line (inline image). For calculating inline image we assumed nonsteady-state conditions and accounted for a Péclet effect; for inline image we assumed steady-state conditions and accounted for a Péclet effect and for inline image we assumed steady-state conditions and no Péclet effect.

We calculated expected nonsteady-state (inline image) (Fig. 3a, gray bold line) and steady-state (inline image) (Fig. 3a, black dotted bold line) 18O enrichment of newly produced organic matter in isotopic equilibrium with mean lamina leaf water taking into account a Péclet effect as well as assuming steady-state conditions and no Péclet effect (inline image) (Fig. 3a, black dashed bold line).

Assimilation-weighted calculated inline image (the Craig-related enrichment) was consistently higher (up to 2.5‰) than observed Δ18Ows during the whole light period (Fig. 3a). By contrast, there were only slight differences between the observed Δ18Ows and predicted nonsteady-state inline image. Assimilation-weighted predicted steady-state inline image was comparable to nonsteady-state inline image during the light period. The errors for the model calculations amounted to between 0.2 and 0.3‰.

Τhere was no significant difference in Δ18O of total foliar organic matter between different leaves or during the diurnal course (Fig. 3b). Mean leaf area-weighted Δ18Ototal-leaf varied in time between approx. 34 and 37‰.

Δ18O in stem organic matter along the axis

There was no significant spatial variation of the 18O enrichment in stem total organic matter (Δ18Ototal-stem) along the plant axis (Fig. 4a). In the morning there was a general trend for Δ18O to decrease from 10 : 30 h to 12 : 00 h. This tendency was most pronounced in the lowermost segment A. Contrasting with that observation, Δ18Ototal-stem varied thereafter only slightly for the lower segments but a more intensive decrease was observed for the uppermost stem sections at the beginning of the dark period. Δ18Ototal in roots was always less than in the lowermost stem section and showed only slight diel variations with a peak-to-peak variation of approx. 2‰. As with Δ18Ototal-stem, Δ18Ophloem-OM (Fig. 4b) did not differ between different collecting positions along the stem. There was a temporal trend comparable with Δ18Ototal of stem sections during the light period with a decrease in Δ18O from morning to midday and a diel maximum in the afternoon/evening. In contrast to stem total organic matter, a strong decline in Δ18Ophloem-OM was observed in the second part of the night. The maximum variation was up to 7‰. As in phloem organic matter there was no variation in Δ18O of phloem sap water (Δ18Opw) between sampling sites (Fig. 4c). Phloem sap water was slightly enriched compared with source water and showed a diel maximum at 19 : 00 h when calculated leaf water enrichment (Δ18OLn) was close to its maximum.

Figure 4.

Δ18O in total organic matter of (a) stem sections S-A to S-F and fine roots; (b) phloem-transported organic matter and (c) Δ18O in phloem sap water collected at stem positions p-A (base of the stem) to p-F (between leaves L4 and L5) (see Fig. 1) of Ricinus communis during the diel course. Data shown are mean values (n = 3–4). In addition, the average standard deviation for the mean values is given. Effects of position along the axis and of time (excluding the data for fine roots) on Δ18O, as calculated with the GLM-anova procedure, are given. Δ18OLn in (c): calculated and canopy-weighted mean lamina leaf water enrichment; the bold dashed line in (c) refers to the mean value of Δ18Opw from all sampling positions.

The Δ18Oxylem-OM varied between approx. 34.5 and 42‰ and was thus within the range of leaf water soluble and phloem organic matter (Fig. 5). There was no significant difference among sampling positions. A diel maximum in the morning and lower values in the evening and during the night were observed for all positions collected.

Figure 5.

Δ18O in xylem organic matter of Ricinus communis collected at different positions (see Fig. 1) during the diel course. Data shown are mean values (n = 3–9). In addition, the average standard deviation for the mean values is given. Effects of position along the axis and time on Δ18O as calculated with the GLM-anova procedure are given.

Transfer of the isotopic signal from the leaf to the roots

The Δ18Ows of leaves explained 93% of the variation of Δ18O of phloem organic matter during the whole diel course (Fig. 6). The phloem Δ equaled the leaf Δ at night and in the morning, but phloem organic matter was slightly depleted in 18O from midday to evening (Fig. 6a). However, there was no significant correlation between leaf total and phloem organic matter (Fig. 6b). During the whole diel course phloem organic matter was enriched in 18O compared with the bulk leaf fraction.

Figure 6.

Relation in Ricinus communis plants between Δ18O of phloem transported organic matter sample (y-axis) directly below the canopy (Position p-C see. Fig. 1) and (a) leaf area weighted Δ18O of water-soluble matter (x-axis) and of (b) total organic matter of leaves (x-axis). Data shown are means ± SD (n = 3–4). The closed circles are night values. Bold lines denote 1 : 1 lines. The thin black line in (a) is the linear regression line defined by the linear equation given.

We assume that phloem sugar during the night represents mainly soluble sugars remobilized from starch. In order to estimate the extent to which this deviates from an integrative Δ18O signature of the organic matter produced during daylight, we compared the average values from day and night (Table 2). Mean values for day Δ18O were weighted for assimilation rate (Eqn 8), while for the night nonweighted averages were calculated.

Table 2.  Comparison of modeled and measured Δ18O in leaf and phloem organic matter during day and night
 Mean Δ18O dayMean Δ18O night 
  1. Mean Δ18O of newly produced organic matter as calculated from Eqn 5 taking an enrichment of organic matter of 27‰ compared with water into account (inline image) and of leaf water soluble (Δ18Ows), leaf bulk (Δ18Ototal-leaf) and phloem organic matter (Δ18Ophloem-OM) during day and night are shown. Day values were weighted for assimilation rate. Letters a–c indicate homogeneous groups when mean Δ18O of different fraction during day or during night where compared.

inline image40.6 ± 0.6 a  
Δ18Ows40.0 ± 0.5 a36.9 ± 1.0 adifferent at P < 0.01
Δ18Ototal-leaf35.0 ± 0.5 c34.4 ± 2.5 anot different
Δ18Ophloem-OM39.0 ± 1.3 b36.7 ± 1.1 adifferent at P < 0.05

During the day, measured leaf water soluble organic matter (Δ18Ows) matched modeled mean daytime inline image, whereas measured phloem organic matter (Δ18Ophloem-OM) was significantly less enriched (by 1‰) than Δ18Ows. At night, however, phloem and leaf soluble organic matter were equally enriched. Mean diel (day and night) Δ18O did not differ significantly between leaf soluble (38.9 ± 0.7‰) and phloem organic matter (38.2 ± 1.4‰). The Δ18Ows pool differed by 3.1‰ between day and night, while the phloem organic matter pool differed by 2.3‰ between day and night, both being greater during daylight. The enrichment of total leaf organic matter, Δ18Ototal-leaf, did not differ between light and dark period and was relatively depleted in 18O compared with the other pools of organic matter during the day and not significantly different from them during the night.

Total organic matter of the stem sections (S-A to S-F) was depleted in 18O compared with phloem sugars collected from the same positions (p-A to p-F). The regression line (Fig. 7a) calculated with all data from all positions is characterized by the following linear equation:

Figure 7.

Relation in Ricinus communis plants between Δ18O of phloem organic matter and (a) total organic matter of stem sections along the axis (positions A–F) and (b) fine roots. Data shown are means ± SD (n = 3–4). The bold black lines denote the 1 : 1 line. In (b) phloem organic matter obtained from the lowermost sampling position (p-A in Fig. 1) was used for correlation.

Δ18Ototal-stem = 0.605Δ18Ophloem-OM + 8.42 [‰]

At an average Δ18Ophloem-OM of 37‰, the difference between Δ18Ophloem-OM and Δ18Ototal-stem was 6.2‰.

We observed that the fine roots were depleted compared with the assumed source of organic matter (phloem organic matter obtained from the lowermost sampling position (p-A)) (Fig. 7b). The depletion amounted to between 7.4 and 11.4‰. There was no significant correlation between Δ18Ophloem-OM and Δ18O of root organic matter during the diel course.

Discussion

In the present study we assessed the diel variation of water-soluble organic matter pools in the leaves and phloem of R. communis compared with the values calculated from nonsteady-state models of evaporative enrichment of leaf water and organic matter (Cernusak et al., 2005; Farquhar & Cernusak, 2005). Previously, Barbour et al. (2000a) and Cernusak et al. (2003b) showed that Δ18O of phloem organic matter collected from leaf petioles could be reasonably well explained by leaf water enrichment (either measured or calculated taking a Péclet effect into consideration). We calculated enrichment of newly produced assimilates applying a nonsteady-state evaporative enrichment model for leaf water and assuming ɛwc to be +27‰ (inline image) and found good agreement from morning to afternoon between modeled values and measured Δ18O in leaf soluble organic matter (Fig. 3a). A reasonably good agreement was also obtained when steady-state conditions were assumed modeling Δ18O (inline image in Fig. 3a). We calculated inline image and inline image only for the light period when organic matter was newly assimilated. Comparable to our modeling of Δ18O of leaf sugars, nonsteady-state and steady-state models for leaf water evaporative enrichment (which include the Péclet effect) both well describe measured lamina leaf water enrichment during the light period. Mainly during night the prediction of leaf water enrichment is improved by the application of nonsteady-state models (Cernusak et al., 2005). The 18O enrichment of leaf water-soluble organic matter was overestimated when no Péclet effect was taken into account (inline image). The difference between inline image on the one hand and inline image and inline image on the other strongly exceeds the errors calculated for the models.

The soluble pool of organic matter appears to be in, or close to, isotopic equilibrium with lamina leaf water even though environmental and physiological conditions change during the day. However, we do not have data for the period between 12 : 00 h and 17 : 00 h, and it is likely that the peak in leaf water enrichment occurred within this period (i.e. before 17 : 00 h). With this in mind the data might suggest a slight lag, as the measured sugar enrichments are less than the modeled leaf water early in the day, and greater at 17 : 00 h. Assuming a sugar pool size of c. 60 mmol C m−2 (A. Gessler, unpublished), and an assimilation rate of c. 10 µmol m−2 s−1, the time lag would be approx. 2 h. From midday to evening, when the most intensive change of environmental conditions occurred, Δ18O of leaf soluble organic matter differed slightly from phloem sugars collected directly below the canopy whereas both values were comparable during night and in the morning (Fig. 6). Barbour et al. (2000a) showed that it took approx. 3.5 h for phloem sugars harvested at the petiole to reach an isotopic equilibrium. However the phloem equilibration time will exceed the equilibration time of the leaf pool, as it includes the time of transport from the leaf lamina to the petiole.

Thus, even though phloem transport velocity is c. 0.75 m h−1 in R. communis (Peuke et al. 2001), it is likely that phloem sugars isotopically lag the current environmental conditions, as previously observed for trees (Gessler et al., 2004).

However, no obvious time-lag was observed among Δ18O of phloem organic matter collected from different positions along the axis. In addition, there was no significant basipetal gradient in Δ18O of phloem organic matter. There was, however, a diminution of the diel amplitude between leaves and phloem: the leaf soluble sugar enrichments increased by c. 4.5‰ from morning until afternoon and then diminished by c. 8‰ until 03 : 00 h, whereas the corresponding changes in the phloem organic matter were approx 3 and 4‰. Taking into account that the average diel Δ18O did not differ significantly between leaf soluble and phloem organic matter we can conclude that the isotopic signal imprinted on newly assimilated carbohydrates in leaves as integrated over the whole day is conserved in the carbohydrates during phloem transport. However, the diel dynamics are altered during phloem loading. The small but consistent reduction in mean daytime Δ18O in phloem organic matter compared with the soluble organic matter fraction in the leaves (Table 2) might be caused by the different isotopic equilibrium times of the leaf and phloem pools. However, we cannot rule out the possibility that organic compounds newly assimilated in the green tissues in the stems of R. communis contribute partly to phloem organic matter. In the stems, reaction water is not or only slightly enriched and sugars fixed there should have a Δ18O of approx. 27‰ (Cernusak et al., 2005), which is well below the enrichment for carbohydrates assimilated in leaves.

The dynamic Münch mass flow model recently reviewed by van Bel (2003) proposes that while a proportion of the sucrose from sieve tubes is released during phloem transport, carbohydrates are also transported back into the sieve tubes in stem tissues (Minchin & Thorpe, 1987). According to our results the retrieved organic matter is either isotopically similar to the released sugars, or the amount of carbohydrates reloaded is too low to alter the Δ18O of phloem organic matter significantly.

Mean daily assimilation-weighted inline image or foliar Δ18Ows should be representative of the oxygen isotope enrichment of transitory starch laid down during the day. Whereas during daylight triose phosphates are the main export forms leaving the chloroplast, starch breakdown during the night releases maltose as the major export compound (Weise et al., 2004). Assuming that phosphatases break the O–P bond, the number of exchangeable oxygen atoms can be estimated (Sternberg et al., 1986). Thus, during starch hydrolysis one oxygen atom per hexose molecule produced can be exchanged with medium water (when we ignore the α1–6 glycosidic bonds). Assuming a rapid equilibrium between glucose–phosphate and fructose–phosphate (Farquhar et al., 1998), which are the precursors for sucrose synthesis, two further oxygen atoms can be exchanged per hexose molecule. During the formation of the α1–2β glycosidic bond one equilibrated oxygen atom per molecule of sucrose is, however, lost. As a consequence, five out of 11 oxygen atoms in sucrose generated from starch are exchanged. Since calculated mean leaf water Δ18O during the night was 6.1‰ (cf. Fig. 4) and mean daytime Δ18Ows was 40‰ (Table 2), Δ18O of sucrose produced from hydrolysed starch is calculated to amount to c. 36.9‰ (i.e. 6/11 × 40 +5/11 × (27 + 6.1)), a value comparable to the measured 18O enrichment in leaf water soluble (36.9‰) and phloem organic matter (36.7‰) (Table 2). In addition to sucrose, the phloem sap of R. communis also contains amino acids, organic acids and inorganic ions (Hall & Baker, 1972; Jeschke & Pate, 1991). However, sucrose makes up approx. 80–89% of phloem organic matter and Cernusak et al. (2003b) found no strong difference in the oxygen isotope enrichment between bulk phloem material and purified sucrose.

The 18O enrichment of total stem organic matter was considerably lower than that of phloem organic matter. When cellulose and other structural compounds are produced from sugars unloaded from the phloem, exchange with unenriched reaction water of the stem occurs (Barbour & Farquhar, 2000). Cernusak et al. (2005) reviewed the literature for information on the proportion of oxygen atoms in sugar that exchange with water during cellulose synthesis (pex). They observed a high consistency for pex among a wide range of species with a mean value of 0.42 when substrates for tissue synthesis were carbohydrates. Since the bulk organic material of the stem sections analysed here consists (apart from cellulose) of a mixture of various compounds, the portion of oxygen atoms exchanged may differ from 0.42.

Compared with phloem organic matter, but also with bulk stem material, total organic matter in roots is depleted in 18O (Figs 4A and 7B). The mean daily value of enrichment is c. 28.7‰ and thus, not much greater than ɛwc. This might point to the fact that, on a daily average basis, part of the evaporative 18O enrichment imprinted from the leaf water on the newly assimilated organic matter is lost in the roots which is also indicated by the small diel variation in Δ18O of root total organic matter compared with the stem (Fig. 4a). One possible explanation for that observation is associated with CO2 refixation in the roots by phosphoenolpyruvate carboxylase (PEPC). Badeck et al. 2005 observed CO2 fixation by PEPC in roots of Phaseolus vulgaris to exceed net respiration rates. From these findings the authors estimated that PEPC can potentially refix CO2 at rates of several 10% of the gross respiration rate in roots. As a consequence, we suggest that PEPC refixation may influence Δ18O of organic matter in the below ground tissues. As assimilation of organic matter in the roots takes place within unenriched reaction water, and as the carboxyl oxygens of malate – the product of PEPC refixation – are less 18O enriched than carbonyl oxygen (Schmidt et al., 2001), we can explain the relatively low Δ18O values of root organic matter.

Xylem organic matter is more strongly enriched in 18O than is organic matter in the roots. Ranging between 42‰ and 35‰ during the diel course Δ18O is much closer to phloem organic matter along the stem. We consequently conclude that the major part of xylem organic matter is not coming from the roots but originating from phloem-to-xylem exchange, thus representing current evaporative enrichment conditions in leaves. When comparing the diel patterns of Δ18O in phloem and xylem organic matter a substantial time-lag becomes obvious. Whereas maximum Δ18O values were observed in the afternoon for phloem organic matter, 18O enrichment in xylem organic matter was highest during night and in the early morning. The time-lags might be related to the timing of xylem loading during the diel course. Future research on whole-plant cycling of organic matter will have to address the underlying causes in detail.

In conclusion, the 18O isotopic enrichment of the water soluble organic matter pool in leaves during the course of a day can be reasonably well explained by current evaporative enrichment models (working hypothesis (1)). We also showed (working hypothesis (2)) that Δ18Ophloem-OM during the night is influenced by leaf water enrichment during both day and night (i.e. during starch synthesis in the light and remobilization in the dark).

We also showed that 18O enrichment of phloem organic matter reflected that of water-soluble leaf organic matter when averaged over the whole day; during the diel course, however, there was a diminution of the diel amplitude of the oxygen isotope enrichment in phloem as compared to leaf soluble organic matter. Once loaded into the phloem, the diel Δ18O pattern of organic matter did not change during basipetal transport in the sieve tubes. We therefore conclude that the isotopic signal imprinted on newly assimilated organic matter in the leaf can be found in phloem organic matter when averaged over a diel cycle and (working hypothesis (3)) that phloem sugars generally conserve the oxygen isotope signal during transport to the stem base. These findings are of central importance for studies which assess Δ18O in phloem organic matter in the field for characterizing the water balance of plants as affected by environmental conditions (e.g. Cernusak et al., 2003a; Keitel et al., 2006). When calculating or modeling transpiration (Cernusak et al., 2003a) and stomatal conductance (Brandes et al., 2007) from Δ18O in phloem organic matter it is normally assumed that no postphotosynthetic exchange of organic oxygen during phloem loading or transport occurs. We showed here that this assumption is justified when averaging Δ18O over the whole day. However, diel variations influenced by night-time starch remobilization may have to be taken more strongly into account when applying Δ18O approaches in the field. There is a strong need to verify our observation in other species, especially trees, under field conditions.

Acknowledgements

A.G. acknowledges financial support by a research fellowship from the Deutsche Forschungsgemeinschaft (DFG) under contract number GE 1090/4-1 and by a DFG research grant (GE 1090/5–1). G.D.F. acknowledges the Australian Research Council for their support. We thank Matthias Cuntz for helpful comments on the manuscript.

Ancillary