SEARCH

SEARCH BY CITATION

Keywords:

  • carbon-13 discrimination;
  • ecosystem respiration;
  • Keeling plot;
  • Nothofagus;
  • phloem sugar δ13C

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References
  • Day-to-day variability in the carbon isotope composition of phloem sap (δ13Chd) and ecosystem respiratory CO213CR) were measured to assess the tightness of coupling between canopy photosynthesis (δ13Chd) and ecosystem respiration (δ13CR) in two mature Nothofagus solandri (Hook. f.) forests in New Zealand.
  • Abundant phloem-tapping scale insects allowed repeated, nondestructive access to stem phloem sap 1–2 m above ground. δ13Chd was compared with δ13C predicted by an environmentally driven, process-based canopy photosynthesis model. Keeling plots of within-canopy CO2 were used to estimate δ13CR.
  • By including a lag of 3 d, there was good agreement in the timing and direction of variation in δ13Chd and predictions by the canopy photosynthesis model, suggesting that δ13Chd represents a photosynthesis-weighted, integrative record of canopy photosynthesis and conductance.
  • Significant day-to-day variability in δ13CR was recorded at one of the two forests. At this site, δ13CR reflected variability in δ13Chd only on days with <2 mm rain. We conclude that the degree of coupling between canopy photosynthesis and ecosystem respiration varies between sites, and with environmental conditions at a single site.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

An understanding of the exchange of energy, carbon and water between the atmosphere and the biosphere is vital for successful predictions of ecosystem response to environmental change. Ecosystem carbon exchange is commonly measured using micrometeorological techniques such as eddy covariance. However, carbon-exchange measurements are unable to separate photosynthetic and respiratory fluxes, and cannot distinguish autotrophic from heterotrophic respiration. The stable isotope composition of CO2 is of particular interest, as it provides information about the components of gross carbon exchange (Yakir & Sternberg, 2000). Further, models using mass balance of stable isotopes of carbon and oxygen are an important tool in constraining the global carbon budget, and determining the size and location of terrestrial carbon sources and sinks (Cias et al., 1995; Battle et al., 2000; Canadell et al., 2000; Randerson et al., 2002).

Carbon isotope discrimination during C3 photosynthesis is reflected in the isotopic composition of plant organic material (δ13Cp, expressed relative to the Vienna Pee Dee Belemnite standard, VPDB), and provides an integrative record of CO2 supply relative to demand. When this plant material is later respired, the δ13C of CO2 released is expected to reflect the isotope composition resulting from photosynthetic discrimination, at least to some extent (Duranceau et al., 1999; Xu et al., 2004). While the processes regulating δ13Cp are generally well understood at the leaf level (Farquhar et al., 1989), and scaling procedures allow leaf-level processes to be extrapolated to canopy level (Hanson et al., 2004), knowledge of the processes determining the carbon isotope composition of ecosystem-respired CO213CR) is lacking.

By invoking conservation of mass, it follows that the isotope composition of carbon respired by an ecosystem should reflect the isotope composition of carbon fixed by photosynthesis (Pataki et al., 2003) over the entire life of the ecosystem. δ13CR has been shown to vary consistently with changes in δ13Cp between sites with trees of differing age (Fessenden & Ehleringer, 2002). Variation in δ13CR has also been found to reflect environmental variability, such as changes in air saturation deficit (D) during a growing season, at single sites (Ekblad & Högberg, 2001; Bowling et al., 2002). The relationship between D and δ13CR is suggested to be driven by the influence of D on stomatal conductance, and hence on δ13C of recently fixed carbohydrate (McDowell et al., 2004b), coupled with the observation that current photosynthesis strongly drives ecosystem respiration (Högberg et al., 2001). Recent work (Scartazza et al., 2004) demonstrates a strong link between the carbon isotope composition of photosynthetic products and δ13CR during a growth season in a northern hemisphere beech forest. In contrast, other studies have shown that δ13CR varies over only a small range during a season, suggesting that environmentally driven variation in δ13Cp is not reflected in δ13CR (Flanagan et al., 1996; Buchmann et al., 1998). Decoupling of δ13CR from δ13Cp can also occur over daily timescales. For example, when δ13CR was measured on consecutive nights over 2 wk, McDowell et al. (2004a) found little variability in δ13CR, despite large changes in D which caused a wide range in canopy conductance (gc).

The decoupling of δ13CR from δ13Cp may be caused by variable time lags between carbon fixation and respiration by different carbon pools (Bowling et al., 2002). McDowell et al. (2004a) list a number of factors that could contribute to variability in the time lag between fixation and respiration, including: distance from source leaves to the respiring tissue, phloem temperature, source and sink strengths, size of ecosystem carbon pools, soil microbial turnover rates, fungal transport rates, nutrient availability, and soil moisture and temperature effects on soil microbial respiration rates. CO2 respired by an ecosystem is derived from a number of sources. If stem and branch respiration is assumed to be negligible, the dominant sources are leaves, roots, litter decomposition and soil microbial respiration. Respiration by microbes in the rhizosphere is assumed to be dependent on plant-derived carbon, so may be isotopically indistinguishable from root respiration (Formanek & Ambus, 2004). The time taken from carbon fixation to CO2 release during respiration is expected to vary considerably between leaves (days) and soil (months to years). Ecosystems in which heterotrophic respiration forms a large proportion of total ecosystem respiration may be expected to display loose coupling between δ13Cp and δ13CR if heterotrophs do not use recently fixed carbon as a substrate.

It is also possible that decoupling of δ13CR from δ13Cp results from isotopic fractionation between carbon fixation and respiration. CO2 respired by isolated protoplasts showed no shift in isotopic composition compared with source carbon (Lin & Ehleringer, 1997), but CO2 respired by whole leaves has been found to be consistently less depleted than that of carbon sources (Duranceau et al., 1999; Ghashghaie et al., 2001; Xu et al., 2004). Further, isotopic fractionation has been demonstrated during litter decomposition (Fernandez et al., 2003) and during uptake of carbon by fungi, resulting in CO2 released during fungal respiration being 1–8‰ more enriched than the carbon supplied to the fungi (Henn & Chapela, 2001).

The low variability in δ13CR observed by McDowell et al. (2004a) may result simply from a lack of variation in δ13Cp (not measured in their experiment). Applying mechanistic models (Farquhar et al., 1989), variation in gc may not have resulted in variation in δ13Cp if this was accompanied by changes in canopy photosynthetic rate (Acan) that maintained a constant ratio of leaf intercellular to ambient CO2 concentration (ci/ca). Simultaneous measurement of δ13C of recently fixed carbon and δ13CR has allowed this possibility to be tested. Scartazza et al. (2004) observed extremely tight coupling between δ13C of phloem sap and δ13CR over a seasonal timescale in an Italian beech forest. The carbon isotope composition of carbohydrate is known to reflect ci/ca (Brugnoli et al., 1988, 1998); and more recently, variability in the stable isotope composition of phloem sap bled from mature trees has been shown to reflect relative rates of canopy photosynthesis and conductance (Pate & Arthur, 1998; Keitel et al., 2003; Cernusak et al., 2005).

Current understanding suggests that photosynthetic discrimination varies considerably from day to day as environmental conditions vary. Insofar as plant respiration uses recently fixed carbon as a substrate, δ13C of CO2 respired by plants is expected to be quite dynamic at timescales of a few weeks, although root respiration may be out of phase with leaf respiration because of the time taken to transport photosynthate from leaves to roots. δ13C of litter- and soil-respired CO2 may be less variable than plant-respired CO2 over a few weeks if soil carbon substrates remain constant. The contribution of respiration by each component to total ecosystem respiration is likely to vary with environmental conditions (e.g. temperature and water availability, for which there is some evidence: McDowell et al., 2004b), suggesting that the tightness of coupling between canopy photosynthesis and ecosystem respiration may also vary in time.

In this paper we investigate coupling between canopy photosynthesis and ecosystem respiration in two mature Nothofagus solandri (Hook. f.) forests on a daily timescale. Carbon isotope discrimination during photosynthesis (δ13Cp) was modelled with an environmentally driven process-based model of canopy photosynthesis. Carbon in phloem sap was sampled for δ13C analysis using the honeydew scale insect (Ultracoelostoma spp.) common on Nothofagus trees in New Zealand and abundant on the sampled trees. The phloem-tapping scale insects allowed repeated, nondestructive access to current photosynthate via excretion of excess carbohydrates at the end of long anal threads (honeydew). By comparing modelled δ13Cp with measured δ13C in honeydew, we test the hypothesis that day-to-day variation in δ13Cp (driven by changes in canopy photosynthesis and conductance) is reflected in δ13C of stem phloem sap, and estimate the time taken for carbon fixed in the canopy to move to the sampling point on the stem 1.5 m above the ground. By comparing δ13C of honeydew (δ13Chd) with δ13C of within-canopy CO2, we test the hypothesis that δ13CR is related to variation in δ13C of recent photosynthate. Finally, results from two forests are compared to test the hypothesis that the degree of coupling between discrimination during carbon fixation and ecosystem respiration varies with environmental conditions (specifically, rainfall before sampling).

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

Description of sites

Samples were collected at two sites in the South Island of New Zealand with contrasting long-term environmental conditions. The first site is at Craigieburn Forest Park (CFP), c. 100 km west of Christchurch at 740 m elevation (43.1° S, 171.4° E). A meteorological station within 6 km of the site recorded average annual rainfall of 1470 mm over the past 40 yr. Mean annual air temperature at the same station is 8.0°C, with an average seasonal range of 11.8°C and an average daily range of 10.7°C. The forest is dominated by Nothofagus solandri[var. cliffortioides (Hook. f.) Poole: mountain beech] trees 15–20 m tall, with basal areas between 0.016 and 0.126 m2. The second site is within Ashley Forest Park (AFP), c. 60 km north of Christchurch (43.1° S, 172.3° E) and at 400 m elevation. Meteorological measurements made at the site during 2003 recorded 1170 mm of rainfall. Average temperature during 2003 was 11.2°C with a seasonal range of 8.6°C and an average daily range of 9.1°C. The forest is dominated by Nothofagus solandri[var. solandri (Hook. f.) Oerst.: black beech] trees 15–25 m tall, with basal areas between 0.016 and 0.229 m2. Forest structure at both sites is similar, with a canopy dominated by beech and a very sparse understorey consisting mainly of small shrubs of the genus Coprosma. Both sites were on moderately steep slopes.

Half-hourly average air temperature, relative humidity, incident irradiance and total rainfall were measured in clearings and recorded using standard techniques and data loggers (CR10X and 21X; Campbell Scientific, Logan, UT, USA) at the existing meteorological station near the CFP site, and at a station installed in December 2002 at the AFP site.

Phloem sap sampling and measurement of carbon isotope composition

Nothofagus forests in the central and northern South Island of New Zealand support large populations of the sooty beech scale insect (Ultracoelostoma spp., Hemiptera: Margarodidae) that live within the bark and tap the phloem sap. Excess carbohydrates are excreted by the insect from long waxy anal threads that extend from the bark, producing droplets of sugar-rich honeydew (Morales, 1991). The rate of honeydew production measured in winter at both sites was 2 mg sugar m−2 bark h−1, but was up to 80 mg sugar m−2 bark h−1 at AFP in summer (Dungan & Kelly, 2003; RJD, unpublished data). We found that ≈25% of trees at CFP and 60% of trees at AFP were heavily infested with scale insects. Preliminary photosynthesis measurements in January and April at AFP suggest that upper-canopy leaves of infested and uninfested trees were not significantly different in ci/ca. (RJD and MHT, unpublished data). This suggests that the δ13C of honeydew (δ13Chd) from the sampled trees is representative of carbon fixed recently by all canopy trees within the forest. The extra sink for carbon created by the presence of the insects does not appear to have a negative impact on the carbon balance of the tree (RJD, unpublished data), with neighbouring, similarly sized, uninfested and highly infested trees showing no differences in rates of growth (Chew, 2003) or health (Wardle, 1984).

Honeydew was sampled using 10 µl capillaries, either by drawing the liquid up into the tube, or by allowing more concentrated and viscous honeydew to adhere to the end of the tube (the viscosity of honeydew is related to air temperature, D, and wind speed; Dungan & Kelly, 2003). Honeydew was then placed in weighed tin capsules, returned to the laboratory and evaporated at 60°C for 24 h. Samples ranged in volume from 0.5 to 10 µl and yielded between 0.5 and 2.5 mg C, depending on honeydew concentration. Samples were divided into 0.5 mg C mass classes and analysed for δ13C composition with standards of similar mass. Carbon isotope analysis was performed on an isotope ratio mass spectrometer (Europa Scientific 20/20) interfaced to a Dumas elemental analyser (Europa Scientific ANCA-SL, Europa Scientific Ltd., Crewe, UK). Isotope ratios are presented in parts per thousand in delta notation as:

  • image(Eqn 1)

where R is the isotope ratio (13C/12C), and the standard used is CO2 from VPDB. The standard deviation for the repeated analysis of an internal standard, commercial sugar, was ±0.14‰. Calibration vs VPDB was achieved using a certified secondary standard from CSIRO, Canberra, Australia.

Honeydew from Ultracoelostoma feeding on Nothofagus has been found to consist of 42% fructose, 23% sucrose, 1% glucose and 33% oligosaccharides (probably tetrasaccharides) with trace amounts (<0.05%) of protein (Grant & Beggs, 1989). Nothofagus species are thought to transport sucrose (Zimmerman & Ziegler, 1975; Grant & Beggs, 1989), so isotopic fractionation during conversion of sucrose into fructose and oligosaccharides by the insect is possible. To be certain that changes in δ13C of honeydew represent changes in δ13C of phloem sap, it was important to establish if the insects change the isotopic ratio of sampled sugars. Insects were removed from the bark, taking care to leave their phloem-tapping stylets in place. Phloem sap continued to bleed from the stylet for 1–3 d, allowing phloem sap to be sampled and compared with honeydew excreted by adjacent insects. Honeydew and phloem sap samples were taken for δ13C analysis over a range of environmental conditions. A paired t-test showed no significant difference between δ13C of phloem sap and honeydew (δ13Chd = δ13C sap; P = 0.004), confirming that there was no significant fractionation of 13C through the insect (Fig. 1).

image

Figure 1. Relationship between the carbon isotope composition of phloem sap and honeydew excreted by adjacent phloem-tapping scale insects on Nothofagus solandri trees. Line represents 1 : 1 relationship.

Download figure to PowerPoint

Another methodological issue to resolve was the degree of variation in δ13C of honeydew over time, i.e. variation between δ13C of ‘standing’ honeydew droplets and those freshly formed within an hour of clearing all droplets from a defined area. In November 2003, δ13C of fresh droplets was compared with δ13C of standing honeydew at both sites, samples being collected between 00:30 and 04:30 h (New Zealand standard time, NZST). No significant difference was found between standing and fresh honeydew δ13C (Fig. 2). In subsequent sampling, only standing droplets were collected.

image

Figure 2. Relationship between the carbon isotope composition of standing and freshly formed honeydew droplets. Line represents 1 : 1 relationship.

Download figure to PowerPoint

At each site three or four trees were selected with naturally high levels of insect infestation and honeydew production. Basal area of sampled trees ranged between 0.038 and 0.096 m2 at CFP and 0.024 and 0.118 m2 at AFP. Three replicate honeydew samples were collected from insects on the stem between 1 and 2 m above the ground (a vertical displacement 5–15 m from source leaves) either between 00:30 and 01:40 h NZST over four nights at AFP in November, or between 1 and 2 h after sunset over six nights at CFP in January and AFP in March. Samples were collected on consecutive nights in summer between 27 January and 1 February at CFP (at 21:30 h NZST), and on alternate nights between 20 and 27 November (at approximately 01:00 h NZST), and 8 and 17 March at AFP (at 20:30 h NZST).

Five replicate samples of soil, from within the upper 0.1 m, and leaf litter were collected at each site in March and dried at 80°C for 48 h. At the same time, leaves from three trees were sampled from 2.5 m above the ground at both sites, and from the sunlit upper canopy at AFP (≈17 m above the ground, with access via a cherry picker not available at CFP). Also sampled at AFP was the abundant black sooty fungus (a number of species of the genera Trichopeltheca, Capnocybe and Capnodium are present; Wardle, 1984) occurring as thick layers coating tree stems and branches. The fungus was also present at the higher-elevation site at CFP, but was not as thick or abundant. Sieved soil (<500 µm) and leaves, litter and fungus were ground and analysed for δ13C as described above for honeydew.

Within-canopy air sampling and estimation of δ13CR

Keeling (1958, 1961) suggested use of a simple mixing model to calculate the δ13C of CO2 respired by the ecosystem (δ13CR):

  • image(Eqn 2)

where c is CO2 concentration, δ13C is the stable carbon isotope ratio of CO2, and the subscripts a and f refer to CO2 in the atmosphere above and within the canopy, respectively. From equation 2 it can be seen that a plot of 1/cf vs δ13Cf gives a straight line with an intercept δ13CR. Estimates of δ13CR were obtained from geometric mean linear regressions as described by Pataki et al. (2003). We assumed no change in δ13CR and δ13Ca during sampling for each Keeling plot.

Within-canopy air was sampled into pre-evacuated 60 ml glass flasks between 0.01 and 7 m above the soil surface from within a 4 m2 plot at each site on the same nights as honeydew collection. Samples were taken during the night to avoid confounding effects of photosynthesis on the carbon isotope composition of the sampled air. Air was pumped through a magnesium perchlorate trap before sampling to remove water vapour. The gas intake was not fixed to a solid structure, as described for most studies using the Keeling plot technique (e.g. Fessenden & Ehleringer, 2002; McDowell et al., 2004a; Scartazza et al., 2004). Rather, the intake tube was moved between heights within the sampling plot to ensure a sufficient range in CO2 concentration (>75 µmol mol−1; Pataki et al., 2003) to minimize errors in δ13CR estimation. An infrared gas analyser (LiCor 6262, Lincoln, NE, USA) was placed in the sampling line to check the CO2 concentration before sampling, but this concentration was not used in Keeling plot analysis. Keeling plots were constructed with the CO2 concentration measured during isotopic analysis, as described below. Care was taken to avoid sample contamination with human breath. Five samples were taken, with the lowest sample intake and highest CO2 concentration taken at 0.02 m below the surface of the litter and each set of samples taking c. 15 min to collect. The range in CO2 concentration for a single Keeling plot varied between 60 and 382 µmol mol−1.

Sampling air within the litter layer raises the possibility that the CO2 within the sampled footprint was very localized and dominated by the δ13C of below-ground respiration. Unfortunately this is a limitation of the Keeling plot approach; a wide range in CO2 concentration is required for Keeling plot construction, and high CO2 concentrations are only found very close to the forest floor. We acknowledge that the value of δ13CR calculated here is unlikely to be a true representation of whole-ecosystem δ13CR, and likely to be biased by below-ground respiration. To partially assess this issue, we constructed Keeling plots with samples collected only above the litter layer, and compared intercepts with those estimated from all samples. The two intercepts were not significantly different for any sampling night (P > 0.05, Student's t-test; Bailey, 1981), suggesting that the sampled footprint within the litter layer was large enough to be somewhat integrative.

Nothofagus cover was continuous within the small watersheds and for at least 2 km uphill of the sampling locations at both sites, and observations of air movement at night suggest that the sampled footprint was dominantly upslope within the Nothofagus forests. Vegetation cover changed beyond the edge of the sampled watersheds (to conifer plantation at AFP and to pasture at CFP), and may have contributed to the sampled CO2. However, the highest sample was collected at 7 m (one-third to one-half canopy height), and 80% of samples were taken within 100 mm of the soil surface, suggesting that the sampled footprint was small and well within the Nothofagus forests. Further, Keeling plots would deviate significantly from linearity if CO2 was sourced from ecosystems with differing δ13C of respiratory CO2, and this was not the case.

Air samples were analysed for CO2 concentration and δ13C using a stable isotope mass spectrometer (MAT 252 IRMS, Finnigan, Bremen, Germany) interfaced to a gas chromatograph (HP5890 series II, Hewlett Packard, Avondale, PA, USA) as described by Ferretti et al. (2000), at the National Institute of Water and Atmospheric Research, Wellington, New Zealand. Isotope ratios are also presented relative to the VPDB standard. The standard deviation for the repeated analysis of two internal standards (clean air samples collected at Baring Head, New Zealand) were ±0.03 and ±0.02‰ for δ13C, and ±0.09 and ±0.13 ppm for CO2 concentration. Calibration of internal standards to VPDB was achieved by measurement of internal standards at CSIRO Atmospheric Research, Melbourne, Australia. δ13C values were assigned using the CSIRO CG2003 reference scale, which is calibrated to PDB via the National Bureau of Standards reference scale (NBS-19).

Keeling plots constructed from within-canopy air samples generally had very high r2 values (between 0.994 and 0.999 at CFP and between 0.979 and 0.999 at AFP), producing estimates of δ13CR with low standard errors (between 0.2 and 0.5‰ at CFP; between 0.1 and 1.3‰ at AFP), confirming our assumption of constant δ13CR and δ13Ca during sampling.

Modelled canopy ci/caandδ13Cp

We estimated canopy intercellular CO2 concentration using a one-dimensional multilayer canopy model described by Leuning (1995) and applied by Whitehead et al. (2002) and Richardson et al. (2005). The model scales leaf-level measurements of photosynthesis, respiration and stomatal conductance to the canopy level, using submodels for radiative transfer, leaf energy balance, evaporation and photosynthesis. The canopy is divided into 12 layers based on assumed vertical distribution of cumulative canopy leaf area index. Values for the parameters required by the model are presented in Table 1. Leaf energy balance and coupled photosynthesis, stomatal conductance and leaf ci (Leuning, 1995) were calculated simultaneously for sunlit and shaded foliage in each layer using biochemical models of photosynthesis (Farquhar et al., 1980). Canopy ci is calculated as a photosynthesis-weighted average of sunlit and shaded leaves from each canopy layer.

Table 1.  Values for parameters in the integrated canopy gas model used to estimate canopy ci for Nothofagus solandri at Craigieburn Forest Park (CFP) and Ashley Forest Park (AFP), New Zealand
ParameterDefinitionReferenceCFPAFPUnits
  1. 1, Hollinger (1989); 2, Richardson et al. (2004); 3, R.J.D. and co-workers (unpublished data); 4, Benecke & Nordmeyer (1982).

  2. Values for photosynthesis parameters are estimated from measurements made at 20°C.

VcmaxMaximum rate of carboxylation at top of canopy1, 2, 3 48.1 78.1µmol m−2 s−1
JmaxMaximum rate of electron transport at top of canopy1, 2, 3 90.6154.5µmol quanta m−2 s−1
RdLight-independent rate of respiration4  0.416  0.416µmol m−2 s−1
αQuantum yield of electron transport2  0.11  0.11mol mol quanta−1
τConvexity of light response curve2  0.66  0.66
a1Parameter related to intercellular CO2 concentration2  6.3  6.3
gsc0Residual stomatal conductance to CO2 transfer2  0.0025  0.0025mol m−2 s−1
Ds0Sensitivity of stomatal conductance to air saturation deficit2817.6817.6Pa
DsminMinimum value of air saturation deficit for decreasing gsc2250250Pa
LLeaf area index3  3.5  3.5m2 m−2

Carbon isotope discrimination (Δ13C, expressed relative to δ13C of atmospheric CO2) was calculated from daily photosynthesis-weighted canopy ci/ca by (Farquhar et al., 1989):

  • image(Eqn 3)

where a is diffusional fractionation (4.4‰); b is biochemical fractionation (27‰ after taking wall resistance effects into account, Brugnoli & Farquhar, 2000) during carbon fixation; and ca is assumed to be 370 µmol mol−1. Modelled Δ13C values were compared with the carbon isotope composition of honeydew (δ13Chd) by conversion to the VPDB scale:

  • image(Eqn 4)

assuming δ13C of source CO2 for carbon fixation is −7.8‰ and invariable during the season (Walcroft et al., 1997). Equation 3 does not include biochemical processes downstream of carbon fixation that could alter the δ13C of carbohydrates transported in the phloem (Jäggi et al., 2002). For example, sucrose is often more enriched than the initial products of photosynthesis, but less enriched than starch (Schmidt & Gleixner, 1998). Recent work (Gessler et al., 2004) suggests that observed differences between leaf and phloem sucrose may be related to temporal variation in discrimination during photosynthesis, rather than fractionation during transport, as has been suggested (Damesin & Lelarge, 2003; Hobbie & Werner, 2004).

Statistical analyses

Correlation analysis was used to test our first two hypotheses, that canopy ci/ca is reflected in δ13Chd and δ13CR. Pearson correlation coefficients were calculated using Microsoft excel. The distance between the location of canopy assimilation and respiring tissue suggests that correlations between δ13Cp, δ13Chd, and δ13CR are likely to be lagged in time. The time lag between assimilation and transport in the stem (δ13Cp and δ13Chd) has been found to be ≈2 d for European beech trees (Keitel et al., 2003), while the time lag between assimilation and respiration (δ13Cp and δ13CR) varies between 1 d (McDowell et al., 2004a) and 5–10 d (Bowling et al., 2002). Accordingly, we calculated Acan-weighted average δ13Cp for each individual day as well as Acan-weighted average δ13Cp for two successive days, then conducted correlation analyses between modelled δ13Cp and measured δ13Chd and δ13CR with lag times of 1–10 d (a subset of these results are presented). Analysis of variance in δ13Chd was conducted in genstat version 6.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

Weather conditions and modelled δ13Cp

A cold front brought 20.9 mm of rain to AFP on 17 and 18 November (days 321 and 322; Fig. 3) before the start of sampling. After the passage of the front, conditions were increasingly warm, dry and sunny until a low pressure system brought cooler, overcast conditions with patchy rain from 22 November (day 326) to the end of the sampling period. A significant rain event (15.7 mm) was recorded on the day before the last samples being taken. Modelled Acan varied between 0.54 and 1.27 mol m−2 d−1 (average 0.92 mol m−2 d−1), driven by variation in irradiance. Coupled with stomatal response to variation in D, variation in Acan resulted in more negative modelled δ13Cp on days 322 and 323 (18 and 19 November) and less negative modelled δ13Cp on 325 (21 November).

image

Figure 3. Variation in weather variables and in modelled canopy photosynthetic rate (Acan) and modelled photosynthesis-weighted carbon isotope discrimination (δ13Cp) for Ashley Forest Park (New Zealand) between 13 and 27 November (days 217–231). (a) daily rainfall; (b) incident irradiance (400–700 nm; Q); (c) air temperature (Ta); (d) vapour pressure deficit (D); (e) Acan; (f) modelled δ13Cp.

Download figure to PowerPoint

During the January sampling at CFP, two cold fronts moved across the site on 26 and 31 January, bringing 25.8 and 31.0 mm of rain, respectively (days 26 and 31, Fig. 4). Days between 22 and 25 January were warm and sunny with moderately high D. After the first front had passed (28–30 January), conditions remained partly cloudy, with lower temperatures and values of D than those before the front. Variation in irradiance resulted in a range in modelled total canopy assimilation rate (Acan) of between 0.31 and 0.78 mol m−2 d−1, with an average of 0.65 mol m−2 d−1. Variation in Acan and D resulted in modelled δ13Cp being less negative in the warm, sunny conditions before the first front, more negative during the passage of both fronts, and intermediate on the partly cloudy days between the two fronts (Fig. 4).

image

Figure 4. Variation in weather variables and in modelled canopy photosynthetic rate (Acan) and modelled photosynthesis-weighted carbon isotope discrimination (δ13Cp) for Craigieburn Forest Park (New Zealand) between 20 January and 1 February (days 20–32). (a) daily rainfall; (b) incident irradiance (400–700 nm; Q); (c) air temperature (Ta); (d) vapour pressure deficit (D); (e) Acan; (f) modelled δ13Cp.

Download figure to PowerPoint

Two cold fronts also moved across the AFP site during the March sampling period. The first brought 38.5 mm of rain to AFP between 1 and 3 March, and the second 22.3 mm on 16 and 17 March (days 61 and 76, respectively, Fig. 5). Between the fronts the days were mostly clear and sunny. Variation in irradiance and D resulted in a range in modelled Acan of between 0.41 and 1.05 mol m−2 d−1, with an average of 0.76 mol m−2 d−1. Air temperature and air saturation deficit were quite variable during this time, with no clear diurnal pattern, but generally higher than in January at CFP. This wide variability in D probably results from differences in wind direction. Hot, dry föhn winds from the northwest are often interspersed with cool, moist easterlies from the Pacific Ocean east of the Southern Alps in the South Island. Variable D resulted in considerable variation in modelled δ13Cp. Modelled δ13Cp reflected changes in D, with less negative δ13Cp associated with warm, dry conditions and more negative δ13Cp with cool, moist conditions (Fig. 5).

image

Figure 5. Variation in weather variables and in modelled canopy photosynthetic rate (Acan) and modelled photosynthesis-weighted carbon isotope discrimination (δ13Cp) for Ashley Forest Park (New Zealand) between 1 and 17 March (days 61–77). (a) daily rainfall; (b) incident irradiance (400–700 nm; Q); (c) air temperature (Ta); (d) vapour pressure deficit (D); (e) Acan; (f) modelled δ13Cp.

Download figure to PowerPoint

Values of δ13Chd and δ13CR

At AFP in November, there was a general decline in the carbon isotope composition of honeydew (δ13Chd) over consecutive nights, from −26.0 to −26.9‰. Two trees (trees 1 and 3) followed similar trends in δ13Chd, while the third tree samples (tree 4) showed little variation (Fig. 6a). anova revealed highly significant (P < 0.001) differences between trees and between nights and a highly significant (P < 0.001) tree × night interaction. Significant variation in δ13CR (between −23.5 and −26.0‰) was observed between nights at AFP in November (Fig. 6b).

image

Figure 6. Temporal variation in the carbon isotope composition of honeydew from four Nothofagus solandri trees (a,c,d) and ecosystem-respired CO2 (b,d,f), at Craigieburn Forest Park (c,d) and Ashley Forest Park in November (a,b) and in March (e,f). Error bars represent (a,c,e) SE of the mean value; (b,d,f) SE of the Keeling plot geometric mean regression.

Download figure to PowerPoint

A general decline in δ13C of honeydew with time was also found at CFP in January, from −25.4 to −26.5‰. Considerable variation in δ13Chd was observed between trees on the same night, but variation between replicate samples from a single tree was generally small. Three of the four trees showed similar patterns in δ13Chd, but the fourth tree showed very little variation in δ13Chd between nights (Fig. 6c). anova revealed highly significant (P < 0.001) differences between trees and between nights and a significant (P = 0.017) tree × night interaction. Values of δ13CR varied between −24.1 and −25.1‰, although differences between nights were not statistically significant (Fig. 6d).

At AFP in March, average values of δ13C of honeydew varied between −27.0 and −27.8‰. Variation between nights for a single tree was larger, up to 1.7‰. As at CFP, three of the four trees showed similar patterns in δ13Chd (Fig. 6e). anova revealed significant (P < 0.01) tree and night effects, and a highly significant (P < 0.001) tree × night interaction. Significant variation in δ13C of ecosystem respiration was observed between nights at AFP in March, with a range in δ13CR of 3.3‰ (Fig. 6f).

Relationships between modelled δ13Cp, and measured δ13Chd and δ13CR

Significant positive correlations were found between modelled δ13Cp (averaged over a single day) and δ13Chd for all three trees sampled at AFP in November when δ13Chd at 1.5 m above the ground lagged 3 d (trees 1 and 3) or 5 d (tree 4) behind canopy carbon fixation (Fig. 7a). Average δ13Chd across all trees was most strongly correlated with modelled δ13Cp with a 3 d lag time and photosynthesis-weighted average δ13Cp over a single day (Fig. 7b). No significant correlation was found between δ13CR and modelled δ13Cp with any time lag between 0 and 10 d (Fig. 7b). Correlation coefficients were not improved significantly for δ13Chd of individual trees or the average across all trees, or δ13CR, when a photosynthesis-weighted 2 d average modelled δ13Cp was used (data not shown).

image

Figure 7. Pearson product-moment correlation coefficients for linear regressions of the carbon isotope composition of honeydew from individual Nothofagus solandri trees (a,c,e); site average honeydew and ecosystem-respired CO2 (b,d,f) vs photosynthesis-weighted daily modelled canopy carbon isotope discrimination time-lagged by 0–7 d for two forests (Craigieburn Forest Park, CFP; Ashley Forest Park, AFP, New Zealand). A 0 d shift corresponds to the photosynthesis-weighted average modelled canopy carbon isotope composition for the day immediately preceding sampling. Dashed line represents a statistically significant positive relationship (P < 0.05; n = 4 (a,b); n = 6 (c–f)).

Download figure to PowerPoint

At the CFP site, significant positive correlations were found between modelled δ13Cp (averaged over a single day) and δ13Chd for three of the four trees when δ13Chd lagged 3 d (trees 1 and 2) or 4 d (tree 4) behind canopy carbon fixation. No significant correlation was found between δ13Chd from tree 3 and modelled δ13C with any time lag between 0 and 10 d, where δ13Cp was averaged over either 1 or 2 d (a subset of correlation coefficients is presented in Fig. 7c). Average δ13Chd across all trees was most strongly correlated with modelled δ13Cp with a 3 d lag time, and photosynthesis-weighted average δ13Cp over a single day (Fig. 7d). No significant correlation was found between δ13CR and modelled δ13Cp with any time lag between 0 and 10 d (Fig. 7d). Correlation coefficients were not significantly improved for δ13Chd of individual trees or the average across all trees, or δ13CR, when a photosynthesis-weighted 2 d average modelled δ13Cp was used (data not shown). The relationship between modelled δ13Cp, with a 3 d lag time, and measured δ13Chd is presented in Fig. 8a and shows that, while the timing and direction of variation in δ13Chd was well predicted, absolute values and the magnitude of day-to-day variation were poorly predicted (the modelled range was 4.2‰, while the measured range in δ13Chd was just 1.3‰).

image

Figure 8. Relationships between measured and modelled carbon isotope composition of honeydew at two forests (Craigieburn Forest Park, CFP; Ashley Forest Park, AFP, New Zealand), including a 3 d lag time at both sites. Dashed line represents linear regression; solid line the 1 : 1 relationship. (a) Measured δ13C = 0.24 × modelled δ13C − 18.98, r2 = 0.63, P = 0.058. (b) Measured δ13C = 0.25 × modelled δ13C − 19.84, r2 = 0.98; P < 0.0001. Error bars represent SE of mean values (n = 12).

Download figure to PowerPoint

At AFP in March, significant positive correlations were also found between modelled δ13Cp and δ13Chd for three of the four trees when δ13Cp lagged 3 d (trees 1 and 4) or 6 d (tree 3) behind canopy photosynthesis. No significant correlation was found between δ13Chd from tree 2 and modelled δ13Cp with any time lag between 0 and 10 d, although the highest correlation coefficient was found with a 3 d lag (Fig. 7e). When δ13Chd was averaged across all trees, a highly significant correlation was observed with modelled δ13Cp including a 3 d lag time (Fig. 7f). No significant correlation was found between δ13CR and modelled δ13Cp at AFP, but the highest correlation coefficient was found with a 3 d lag (Fig. 7f). Correlation coefficients were not significantly improved when a photosynthesis weighted 2 d average δ13Cp was used (data not shown). As at CFP, the canopy photosynthesis model accurately predicted the timing and direction of changes in δ13Chd (with a 3 d lag), but did not predict the absolute values and amplitude of changes in δ13Chd (the modelled range in δ13Cp over both November and March sampling periods was 6.3‰, while the measured range in δ13Chd was just 1.8‰; Fig. 8b).

Coupling between δ13Chd and δ13CR

At both sites there was no significant relationship between δ13Chd and δ13CR when all days were included in the analysis. As suggested in the Introduction, the degree of coupling between δ13Cp and δ13CR may vary with environmental conditions. With this in mind, the sampling days were divided into ‘wet days’ (>2 mm rain fell within 12 h of sampling) and ‘dry days’ (all other days). When wet days were excluded from the analysis, δ13CR tended to increase with increasing δ13Chd at both sites and both sampling times at AFP, although the relationships were not statistically significant (Fig. 9). Further, significant seasonal variation in δ13Chd between spring and autumn sampling periods at AFP was not reflected in δ13CR (Fig. 9; Table 2).

image

Figure 9. Relationships between the carbon isotope composition of honeydew and ecosystem-respired CO2 at Craigieburn Forest Park (CFP, a) and Ashley Forest Park (AFP, b) (New Zealand) for wet (open symbols) and dry (closed symbols) nights. Dashed lines represent linear regressions for dry nights (a, P = 0.06; b, P = 0.19 for November data, P = 0.06 for March data).

Download figure to PowerPoint

Table 2.  Carbon isotopic composition of ecosystem components at Craigieburn Forest Park (CFP) and Ashley Forest Park (AFP), New Zealand
Componentδ13C (‰)
AFP (Nov)CFPAFP (Mar)
  1. Mean ± SE; nd, not determined. Leaves were sampled either 17 or 2.5 m above ground.

Sunlit leaves (17 m)ndnd−26.6 ± 0.2
Leaves (2.5 m)nd−28.8 ± 1.1−29.4 ± 0.7
Honeydew (overall average)−26.5 ± 0.1−26.0 ± 0.1 (Jan)−27.3 ± 0.1
Sooty mouldndnd−27.9 ± 0.2
Leaf litternd−27.4 ± 0.1−28.9 ± 0.1
Soil (0–100 mm)nd−26.5 ± 0.2−27.1 ± 0.2
Ecosystem-respired CO2 (all samples on a single plot)−24.5 ± 0.3−24.7 ± 0.1 (Jan)−23.5 ± 0.3

Respired CO2 was more enriched than recently fixed carbon at both sites, by 1.4, 2.0 and 3.3‰ on average at CFP and AFP in November and March, respectively. At AFP in March, the difference between δ13CR and δ13Chd tended to decrease during the sampling campaign (P = 0.016), from 4.4 to 1.1‰. Respired CO2 was also more enriched than carbon from any measured pool at both sites, and the difference was greater at AFP than at CFP (Table 2). Carbon in ecosystem pools was less depleted in 13C at CFP than at AFP, by between 0.6‰ for soil and 1.5‰ for leaf litter, suggesting that leaves of trees at CFP had lower ci/ca over the long term than did those at AFP.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

The range in weather conditions over the sampling periods provided an ideal opportunity to test the hypotheses that short-term changes in canopy δ13Cp determine the isotopic composition of phloem sap, and influence δ13C of ecosystem-respired CO213CR, but likely to be biased towards below-ground respiration caused by the CO2 sampling strategy). Isotope analysis of phloem sap, sampled via the honeydew scale insect, provided evidence of significant day-to-day variation in canopy δ13Cp, the timing and direction of which were predicted with an integrated canopy photosynthesis model. On dry days (< 2 mm rain within 12 h of sampling), δ13CR tended to increase with increasing δ13C of phloem sap at both sites, although the relationships were not statistically significant (probably because of the small number of samples). Overall, the data suggest that δ13CR was influenced, to some extent, by variation in discrimination during carbon fixation. However, the relationships between δ13C of phloem sap and δ13CR are much weaker than those reported by Scartazza et al. (2004) in a northern hemisphere beech forest, suggesting that the tightness of coupling between canopy photosynthesis and ecosystem respiration varies considerably between ecosystems. The data also provide support for the third hypothesis: that environmental conditions influence the degree of coupling between canopy δ13Cp and δ13CR.

Time lags between canopy carbon fixation, phloem transport and ecosystem respiration

An integrated canopy model accurately predicted the direction of changes in δ13Chd and, with a 3 d lag time, the timing of changes in δ13Chd was also well predicted. A lag of 3 d between canopy photosynthesis and δ13C of phloem sugars collected near the ground fits well with previous observations in European beech trees (Keitel et al., 2003). While the timing and direction of changes in δ13Chd was well predicted by the canopy photosynthesis model, the magnitude of the changes was poorly predicted. Modelled variation in δ13Cp of carbon fixed by leaves was 3.2 and 3.5 times higher than measured variation at CFP and AFP, respectively. This may be caused by incorrect values for model parameters (no attempt was made to ‘tune’ the model), but may also be caused by mixing of carbon fixed by different source leaves. For example, if we assume that carbohydrates are loaded into the phloem immediately after fixation, and phloem transport velocities are constant throughout the tree (to simplify the calculation) at 0.75 m h−1 (Zimmermann & Braun, 1971), we can estimate the time taken for carbon to reach the base of the stem. Carbon fixed by the highest leaves of a 17 m tall tree would reach phloem at 1.5 m above the ground (total path length is 15.5 m, with a transport time of 20.7 h), 13.4 h after carbon fixed at the end of a 3 m long branch that was 4 m above the ground (total path length is 5.5 m, with a transport time of 7.3 h). Further mixing of carbon fixed at different times is likely to be caused by variation in the timing of phloem loading resulting from variation in the soluble carbohydrate content of leaves (i.e. rate of phloem loading is related to leaf carbohydrate concentration; Moing et al., 1994). Laboratory experiments (Grodzinski et al., 1998) and modelling simulations (Thompson & Holbrook, 2004) suggest that phloem transport velocity and assimilate supply are tightly coupled, via osmotic pressure/concentration waves that move much more quickly than the solution itself, so phloem transport velocity (and so lag times) may vary considerably between trees, and could even vary widely over a day within a single tree.

Interestingly, differences in lag times between November and March at AFP were observed for two of the three trees sampled at both times. Tree 1, a dominant tree in an exposed canopy position, showed a 3 d lag for both spring and autumn sampling times. In contrast, trees 3 and 4 (both smaller, subdominant trees) had variable lag times: 3 and 6 d for tree 3 and 5, and 3 d for tree 4, in November and March, respectively. The limited data do not allow firm conclusions to be drawn, but it seems possible that seasonal differences in carbon source and sink strengths could change the rate of carbon transport to the roots. Future work with the Nothofagus–honeydew system will include sampling at higher temporal resolution (both on daily and seasonal scales), more replication of dominant and subdominant trees and at a number of positions along the stem.

There was no evidence of an additional lag between phloem sugars sampled 1.5 m above the ground and ecosystem respiration. A possible explanation for this observation is that the honeydew insect provides a ‘shortcut’ for carbon between the phloem and the microbial community in the litter and soil, via honeydew droplets falling or being washed from branches and stem. The amount of honeydew-derived carbon arriving at the forest floor has not been quantified, so the importance of this shortcut is unknown. A 3 d lag between environmental drivers of variation in δ13Cp and δ13CR compares well with previously published data in a number of forests. For example, Horwath et al. (1994) found a peak in 14CO2 respired by the soil 2 d after Populus trees were radiolabelled; and Mikan et al. (2000) found a lag of 3–4 d in a similar experiment. A lag of 2–4 d in δ13CR following a change in D was found in a boreal forest by Ekblad & Högberg (2001), while Bowling et al. (2002) found peaks in the correlation coefficients between D and δ13CR with a lag of 5–10 d in several coniferous forests, and McDowell et al. (2004b) found a lag of 0–3 d between δ13CR and meteorological and physiological variables at two coniferous sites.

Coupling between δ13Cp and δ13CR

Scartazza et al. (2004) report very tight coupling (r2 = 0.99) between seasonal variability in canopy discrimination and δ13CR in a northern hemisphere beech forest. Such tight coupling was not found on a daily temporal scale in our study. Further, seasonal variability in canopy discrimination was not reflected in δ13CR at AFP. This suggests that the degree of coupling between canopy photosynthesis and ecosystem respiration varies between ecosystems. It seems possible that, while recently fixed carbon was the main substrate for respiration in the northern hemisphere beech forest, other carbon pools with less temporally variable δ13C were the main substrate in the Nothofagus forests studied here. A trend of increasing δ13CR with increasing δ13C of phloem sugar (as reported by Scartazza et al., 2004) was observed for all three sampling periods in our study when wet days (for which >2 mm of rain was recorded) were excluded, although the relationships were not significant statistically. This suggests that the degree of coupling between δ13Cp and δ13CR may vary with environmental conditions at a single site.

Differences between δ13CR and δ13C of respiratory sources

Ecosystem-respired carbon was found to be significantly more enriched in 13C than any measured pool of carbon in the ecosystem (Table 2). A discrepancy between δ13C of carbon sources within the ecosystem and δ13C of respired CO2 has been reported in a number of ecosystems (Pataki et al., 2003), and δ13CR has been reported to be both less depleted (Bowling et al., 2003a), and more depleted (Scartazza et al., 2004), than ecosystem carbon pools. These disequilibria are still poorly understood. Clearly, conservation of mass dictates that fractionation between input (photosynthates) and output (respired CO2) carbon cannot be sustained over the entire life of the ecosystem (Pataki et al., 2003). Further, Lin & Ehleringer (1997) have shown that no 13C fractionation occurs during mitochondrial respiration in isolated leaf protoplasts. However, there is growing evidence of isotopic fractionation during biochemical transformation of carbohydrates before respiration (Duranceau et al., 1999; Ghashghaie et al., 2001; Tcherkez et al., 2003); during respiration itself (Fernandez et al., 2003; Xu et al., 2004); and during carbon uptake by microbes (Henn & Chapela, 2001). Leaf-respired CO2 is often more enriched in 13C than bulk leaf carbon under normal (well watered and not carbon-limited; Tcherkez et al., 2003) conditions because the carbohydrate substrates for respiration are enriched compared with total leaf carbon (Schmidt & Gleixner, 1998), and because the carbon released as CO2 comes from atoms in positions 3 and 4 of the hexose phosphate substrates, which are more enriched than the average δ13C for the molecule (Tcherkez et al., 2003). That is, the nonstatistical (nonrandom) distribution of 13C in the hexose phosphate substrates results in fragmentation fractionation during respiration (Tcherkez et al., 2004). Future studies into variation in δ13CR should quantify the carbon isotope composition of CO2 respired by the various components, rather than the source carbon (although even when this is done, there are problems with interpretation; McDowell et al., 2004a).

Conclusions and recommendations

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

Data presented here suggest that the degree of coupling between δ13Cp and δ13CR varies between ecosystems and with environmental conditions at a single site. A likely explanation for variation in the tightness of the link between environmentally driven variability in δ13Cp and δ13CR found in this and previous studies (Bowling et al., 2003a; McDowell et al., 2004a, 2004b; Scartazza et al., 2004) is variation in the proportional contribution to δ13CR by recently fixed carbon substrates both in time at a single site, and between ecosystems. We suggest that future studies either compare ecosystems with very different plant vs soil respiratory contributions, or compare times at a single site when the proportional contribution of different ecosystem components are widely and predictably variable. For example, respiration from an ecosystem with very low soil organic carbon content should be strongly plant-derived, so δ13CR may be highly temporally variable and closely linked to δ13Cp, while an older ecosystem with high soil organic carbon content may display loose coupling between short-term variation in δ13Cp and δ13CR. An environmental effect that should produce a clear δ13CR signal is the rapid and transient increase in litter decomposition following rewetting of dry litter. Eddy covariance measurements in a number of ecosystems have revealed a large, short-lived increase in ecosystem respiration immediately after the break of a drought (CO2 exchange rates nearly an order of magnitude higher than those before rain; Xu & Baldocchi, 2004), attributed to microbial litter decomposition (Kelliher et al., 2004; Lee et al., 2004). In this scenario, δ13C of litter decomposition will dominate δ13CR immediately after rain, then become increasingly less important as the litter dries. The development of ‘real-time’ methods to measure the δ13C of CO2 using tunable diode laser absorption spectroscopy (Bowling et al., 2003b; Griffis et al., 2004) will greatly increase the temporal resolution of δ13CR estimation, and should allow significant advances in mechanistic understanding of the link between δ13Cp and δ13CR.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References

We thank the Department of Conservation and Carter Holt Harvey for providing access to the sites, N.G. McDowell for discussion and comments on an earlier draft of the manuscript, B.J. Bond and an anonymous reviewer for helpful comments, G.N.D. Rogers for technical assistance in the field, and A. Rajendram, Waikato Stable Isotope Unit, for isotopic analysis of organic samples. This work was funded by the Royal Society of New Zealand Marsden Fund (MMB, JEH, GDF and DW, contract LCR201; MHT and RJD, contract UOC103).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions and recommendations
  8. Acknowledgements
  9. References
  • Bailey NTJ. 1981. Statistical Methods in Biology, 2nd edn. New York, USA: John Wiley and Sons.
  • Battle M, Bender ML, Tans PP, White JWC, Ellis JT, Conway T, Francey RJ. 2000. Global carbon sinks and their variability inferred from atmospheric O2 and δ13C. Science 287: 24672470.
  • Benecke U, Nordmeyer AH. 1982. Carbon uptake and allocation by Nothofagus solandri var. cliffortioides (Hook. f.) Poole and Pinus contorta Douglas ex. Loudon ssp. contorta at montane and subalpine altitudes. In: WaringRH, ed. Carbon Uptake and Allocation in Subalpine Ecosystems as a Key to Management. Corvallis, OR, USA: Forest Research Laboratory, Oregon State University, 921.
  • Bowling DR, McDowell NG, Bond BJ, Law BE, Ehleringer JR. 2002. 13C content of ecosystem respiration is linked to precipitation and vapor pressure deficit. Oecologia 131: 113124.
  • Bowling DR, Pataki DE, Ehleringer JR. 2003a. Critical evaluation of micrometeorological methods for measuring ecosystem–atmosphere isotopic exchange of CO2. Agricultural and Forest Meteorology 116: 159179.
  • Bowling DR, Sargent SD, Tanner BD, Ehleringer JR. 2003b. Tunable diode laser absorption spectroscopy for ecosystem–atmosphere CO2 isotopic exchange studies. Agricultural and Forest Meteorology 118: 119.
  • Brugnoli E, Farquhar GD. 2000. Photosynthetic fractionation of carbon isotopes. In: LeegoodRC, SharkeyTC, Von CaemmererS, eds. Advances in Photosynthesis – Photosynthesis: Physiology and Metabolism, Vol. 9. Dordrecht, the Netherlands: Kluwer, 399434.
  • Brugnoli E, Hubick KT, Von Caemmerer S, Wong SC, Farquhar GD. 1988. Correlation between the carbon isotope discrimination in leaf starch and sugars of C3 plants and the ratio of intercellular and atmospheric partial pressures of carbon dioxide. Plant Physiology 88: 14181424.
  • Brugnoli E, Scartazza A, Lauteri M, Monteverdi MC, Máguas C. 1998. Carbon isotope discrimination in structural and non-structural carbohydrates in relation to productivity and adaptation to unfavourable conditions. In: GriffithH, ed. Stable Isotopes, Integration of Biology, Ecological and Geochemical Processes. Oxford, UK: BIOS Scientific, 133146.
  • Buchmann N, Brooks JR, Flanagan LB, Ehleringer JR. 1998. Carbon isotope discrimination by terrestrial ecosystems. In: GriffithH, ed. Stable Isotopes, Integration of Biology, Ecological and Geochemical Processes. Oxford, UK: BIOS Scientific, 203222.
  • Canadell JG, Mooney HA, Baldocchi DD, Berry JA, Ehleringer JR, Field CB, Gower ST, Hollinger DY, Hunt JE, Jackson RB, Running SW, Shaver GR, Steffen W, Trumbore SE, Valentini R, Bond BY. 2000. Carbon metabolism of the terrestrial biosphere: a multitechnique approach for improved understanding. Ecosystems 3: 115130.
  • Cernusak LA, Farquhar GD, Pate JS. 2005. Environmental and physiological controls over the oxygen and carbon isotope composition of the Tasmanian blue gum, Eucalyptus globulus. Tree Physiology. (In press.)
  • Chew Y. 2003. Infestation of sooty beech scale insect (Ultracoelostoma spp.): determinants and relationship with host tree growth. MSc thesis, University of Canterbury, Christchurch, New Zealand.
  • Cias P, Tans PP, Trolier M, White JWC, Francey RJ. 1995. A large Northern Hemisphere terrestrial CO2 sink indicated by the 13C/12C ratio of atmospheric CO2. Science 269: 10891102.
  • Damesin C, Lelarge C. 2003. Carbon isotope composition of current-year shoots from Fagus sylvatica in relation to growth, respiration and use of reserves. Plant, Cell & Environment 26: 207219.
  • Dungan RJ, Kelly D. 2003. Effect of host-tree and environmental variables on honeydew production by scale insects (Ultracoelostoma sp.) in a high elevation Nothofagus solandri forest. New Zealand Journal of Ecology 27: 169177.
  • Duranceau M, Ghashghaie J, Badeck F, Deleens E, Cornic G. 1999. δ13C of CO2 respired in the dark in relation to δ13C of leaf carbohydrates in Phaseolus vulgaris L. under progressive drought. Plant, Cell & Environment 22: 515523.
  • Ekblad A, Högberg P. 2001. Natural abundance of 13C in CO2 respired from forest soils reveals speed of link between tree photosynthesis and root respiration. Oecologia 127: 305308.
  • Farquhar GD, Von Caemmerer S, Berry JA. 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149: 7890.
  • Farquhar GD, Ehleringer JR, Hubick KT. 1989. Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 40: 503537.
  • Fernandez I, Mahieu N, Cadisch G. 2003. Carbon isotopic fractionation during decomposition of plant materials of different quality. Global Biogeochemical Cycles 17: 1075.
  • Ferretti DF, Lowe DC, Martin RJ, Brailsford GW. 2000. A new gas chromatograph–isotope ratio mass spectrometry technique for high-precision, N2O-free analysis of δ13C and δ18O in atmospheric CO2 from small air samples. Journal of Geophysical Research – Atmospheres 105: 67096718.
  • Fessenden JE, Ehleringer JR. 2002. Age-related variations in δ13C of ecosystem respiration across a coniferous forest chronosequence in the Pacific Northwest. Tree Physiology 22: 159167.
  • Flanagan LB, Brooks RJ, Varney GT, Berry SC, Ehleringer JR. 1996. Carbon isotope discrimination during photosynthesis and the isotope ratio of respired CO2 in boreal forest ecosystems. Global Biogeochemical Cycles 10: 629640.
  • Formanek P, Ambus P. 2004. Assessing the use of δ13C natural abundance in separation of root and microbial respiration in a Danish beech (Fagus sylvatica L.) forest. Rapid Communications in Mass Spectrometry 18: 897902.
  • Gessler A, Rennenberg H, Keitel C. 2004. Stable isotope composition of organic compounds transported in the phloem of European beech – evaluation of different methods of phloem sap collection and assessment of gradients in carbon isotope composition during leaf-to-stem transport. Plant Biology 6: 110.
  • Ghashghaie J, Duranceau M, Badeck FW, Cornic G, Adeline MT, Deléens E. 2001. δ13C of respired CO2 in the dark in relation to δ13C of leaf metabolites: comparison between Nicotiana sylvestris and Helianthus annuus under drought. Plant, Cell & Environment 24: 505515.
  • Grant WD, Beggs JR. 1989. Carbohydrate analysis of beech honeydew. New Zealand Journal of Zoology 16: 283288.
  • Griffis TJ, Baker JM, Sargent SD, Tanner BD, Zhang J. 2004. Measuring field-scale isotopic CO2 fluxes with tunable diode laser absorption spectroscopy and micrometeorological techniques. Agricultural and Forest Meteorology 124: 1529.
  • Grodzinski B, Jiao JR, Leonardos ED. 1998. Estimating photosynthesis and concurrent export rates in C3 and C4 species at ambient and elevated CO2. Plant Physiology. 117: 207215.
  • Hanson PJ, Amthor JS, Wullschleger SD, Wilson KB, Grant RF, Hartley A, Hui D, Hunt ER Jr, Johnson WH, Kimball JS, King AW, Luo Y, McNaulty SG, Sun G, Thornton PE, Wang S, Williams M, Baldocchi DD, Cushman RM. 2004. Oak forest carbon and water simulations: model intercomparisons and evaluations against independent data. Ecological Monographs 73: 443489.
  • Henn MR, Chapela IH. 2001. Ecophysiology of 13C and 15N isotopic fractionation in forest fungi and the roots of the saprotrophic–mycorrhizal divide. Oecologia 128: 480487.
  • Hobbie EA, Werner RA. 2004. Intramolecular, compound-specific, and bulk carbon isotope patterns in C3 and C4 plants: a review and synthesis. New Phytologist 161: 371385.
  • Högberg P, Nordren A, Buchmann N, Taylor AFS, Ekblad A, Högberg MN, Nyberg G, Ottosson-Löfvenius M, Read DJ. 2001. Large-scale forest girdling shows that current photosynthesis drives soil respiration. Nature 411: 789792.
  • Hollinger DY. 1989. Canopy organisation and foliage photosynthetic capacity in a broad-leaved evergreen montane forest. Functional Ecology 3: 5362.
  • Horwath WR, Pretziger KS, Paul EA. 1994. 14C allocation in tree–soil systems. Tree Physiology 14: 11631176.
  • Jäggi M, Saurer M, Fuhrer J, Siegwolf R. 2002. The relationship between the stable carbon isotope composition of needle bulk material, starch, and tree rings in Picea abies. Oecologia 131: 325332.
  • Keeling CD. 1958. The concentration and isotopic abundances of atmospheric carbon dioxide in rural areas. Geochimica et Cosmochimica Acta 13: 322334.
  • Keeling CD. 1961. The concentration and isotopic abundances of carbon dioxide in rural and marine air. Geochimica et Cosmochimica Acta 24: 277298.
  • Keitel C, Adams MA, Holst T, Matzarakis A, Mayer H, Rennenberg H, Geßler A. 2003. Carbon and oxygen isotope composition of organic compounds in the phloem sap provides a short-term measure for stomatal conductance of European beech (Fagus sylvatica L.). Plant, Cell & Environment 26: 11571168.
  • Kelliher FM, Ross DJ, Law BE, Baldocchi DD, Rodda NJ. 2004. Limitations to carbon mineralization in litter and mineral soil of young and old ponderosa pine forests. Forest Ecology and Management 191: 201213.
  • Lee X, Wu H-J, Sigler J, Oishi C, Siccama T. 2004. Rapid and transient response of soil respiration to rain. Global Change Biology 10: 10171026.
  • Leuning R. 1995. A critical appraisal of a combined stomatal–photosynthesis model for C3 plants. Plant, Cell & Environment 18: 339357.
  • Lin G, Ehleringer JR. 1997. Carbon isotopic fractionation does not occur during dark respiration in C3 and C4 plants. Plant Physiology 114: 391394.
  • McDowell NG, Bowling DR, Bond BJ, Irvine J, Law BE, Anthoni P, Ehleringer JR. 2004a. Response of the carbon isotopic content of ecosystem, leaf, and soil respiration to meteorological and physiological driving factors in a Pinus ponderosa ecosystem. Global Biogeochemical Cycles 18: 1013.
  • McDowell NG, Bowling DR, Schauer A, Irvine J, Bond BJ, Law BE, Ehleringer JR. 2004b. Associations between carbon isotope ratios of ecosystem respiration, water availability and canopy conductance. Global Change Biology 10: 17671784.
  • Mikan CJ, Zak DR, Kubiske ME, Pregitzer KS. 2000. Combined effects of atmospheric CO2 and N availability on the below-ground carbon and nitrogen dynamics of aspen mesocosms. Oecologia 124: 432445.
  • Moing A, Escobar-Gutiérrez A, Gaudillère JP. 1994. Modeling carbon export out of mature peach leaves. Plant Physiology 106: 591600.
  • Morales CF. 1991. Margarodidae (Insecta: Hemiptera). Fauna of New Zealand 21. Auckland, New Zealand: DSIR.
  • Pataki DE, Ehleringer JR, Flanagan LB, Yakir D, Bowling DR, Still CJ, Buchmann N, Kaplan JO, Berry JA. 2003. The application and interpretation of Keeling plots in terrestrial carbon cycle research. Global Biogeochemical Cycles 17: 1022.
  • Pate JS, Arthur D. 1998. δ13C analysis of phloem sap carbon: novel means of evaluating seasonal water stress and interpreting carbon isotope signatures of foliage and trunk wood of Eucalyptus globulus. Oecologia 117: 301311.
  • Randerson JT, Still CJ, Fung IY, Doney SC, Tans PP, Conway TJ, White JWC, Vaughn B, Suits N, Denning AS. 2002. Carbon isotope discrimination of arctic and boreal biomes inferred from remote atmospheric measurements and a biosphere–atmosphere model. Global Biogeochemical Cycles 16: 1028.
  • Richardson SJ, Allen RB, Whitehead D, Carswell FE, Ruscoe WA, Platt KH. 2005. Temperature and net carbon availability both determine seed production in a temperate tree species. Ecology. (In press.)
  • Scartazza A, Mata C, Matteaucci G, Yakir D, Moscatello S, Brugnoli E. 2004. Comparisons of δ13C of photosynthetic products and ecosystem respiratory CO2 and their responses to seasonal climate variability. Oecologia 140: 340351.
  • Schmidt H-L, Gleixner G. 1998. Carbon isotope effects on key reactions in plant metabolism and 13C-patterns in natural compounds. In: GriffithH, ed. Stable Isotopes, Integration of Biology, Ecological and Geochemical Processes. Oxford, UK: BIOS Scientific, 1325.
  • Tcherkez G, Nogués S, Bleton J, Cornic G, Badeck F, Ghashghaie J. 2003. Metabolic origin of carbon isotope composition of leaf dark-respired CO2 in French bean. Plant Physiology 131: 237244.
  • Tcherkez G, Farquhar GD, Badeck F, Ghashghaie J. 2004. Theoretical considerations about carbon isotope distribution in glucose of C3 plants. Functional Plant Biology 31: 857877.
  • Thompson MV, Holbrook MN. 2004. Scaling phloem transport: information transmission. Plant, Cell & Environment 27: 509519.
  • Walcroft AS, Silvester WB, Whitehead D, Kelliher FM. 1997. Seasonal changes in stable carbon isotope ratios within annual rings of Pinus radiata reflect environmental regulation of growth processes. Australian Journal of Plant Physiology 24: 5768.
  • Wardle JA. 1984. The New Zealand Beeches: Ecology, Utilization and Management. Christchurch, New Zealand: New Zealand Forest Service.
  • Whitehead D, Hall GMJ, Walcroft AS, Brown KJ, Landsberg JJ, Tissue DT, Turnbull MH, Griffin KL, Schuster WSF, Carswell FE, Trotter CM, James IL, Norton DA. 2002. Analysis of the growth of rimu (Dacrydium cupressinum) in south Westland, New Zealand, using process-based simulation models. International Journal of Biometeorology 46: 6675.
  • Xu L, Baldocchi DD. 2004. Seasonal variation in carbon dioxide exchange over a Mediterranean annual grassland in California. Agricultural and Forest Meteorology 123: 7996.
  • Xu C, Lin G, Griffin KL, Sambrotto RN. 2004. Leaf respiratory CO2 is enriched relative to leaf organic components in five species of C3 plants. New Phytologist 163: 499505.
  • Yakir D, Sternberg LDL. 2000. The use of stable isotopes to study ecosystem gas exchange. Oecologia 123: 297311.
  • Zimmermann MH, Braun CL. 1971. Trees, Structure and Function. Berlin, Germany: Springer.
  • Zimmermann MH, Ziegler H. 1975. List of sugars and sugar alcohols in sieve-tube exudates. In: ZimmermannMH, MilburnJA, eds. Encyclopedia of Plant Physiology, Vol. 1: Transport in Plants, 1: Phloem Transport. New York, USA: Springer, 480503.