Woody tissue photosynthesis and its contribution to trunk growth and bud development in young plants



    1. Laboratory of Plant Ecology, Faculty of Bioscience Engineering, Ghent University, Coupure links 653, B-9000 Gent, Belgium
    2. Department of Integrative Biology and Centre for Stable Isotope Biogeochemistry, University of California, Berkeley, CA 94720, USA
    Search for more papers by this author

    Corresponding author
    1. Laboratory of Plant Ecology, Faculty of Bioscience Engineering, Ghent University, Coupure links 653, B-9000 Gent, Belgium
      K. Steppe. Fax: +32 9 224 44 10; e-mail: kathy.steppe@UGent.be
    Search for more papers by this author

    1. Department of Forest Resources, University of Idaho, PO Box 441133, Moscow, ID 83844-1133, USA
    Search for more papers by this author

    1. Department of Integrative Biology and Centre for Stable Isotope Biogeochemistry, University of California, Berkeley, CA 94720, USA
    Search for more papers by this author

K. Steppe. Fax: +32 9 224 44 10; e-mail: kathy.steppe@UGent.be


Stem photosynthesis can contribute significantly to woody plant carbon balance, particularly in times when leaves are absent or in ‘open’ crowns with sufficient light penetration. We explored the significance of woody tissue (stem) photosynthesis for the carbon income in three California native plant species via measurements of chlorophyll concentrations, radial stem growth, bud biomass and stable carbon isotope composition of sugars in different plant organs. Young plants of Prunus ilicifolia, Umbellularia californica and Arctostaphylos manzanita were measured and subjected to manipulations at two levels: trunk light exclusion (100 and 50%) and complete defoliation. We found that long-term light exclusion resulted in a reduction in chlorophyll concentration and radial growth, demonstrating that trunk assimilates contributed to trunk carbon income. In addition, bud biomass was lower in covered plants compared to uncovered plants. Excluding 100% of the ambient light from trunks on defoliated plants led to an enrichment in 13C of trunk phloem sugars. We attributed this effect to a reduction in photosynthetic carbon isotope discrimination against 13C that in turn resulted in an enrichment in 13C of bud sugars. Taken together our results reveal that stem photosynthesis contributes to the total carbon income of all species including the buds in defoliated plants.


It is known that a considerable amount of CO2 is produced inside the stems and branches of woody plants because of mitochondrial respiration by living cells. A part of this respired CO2 leaves the woody organs via radial diffusion through all tissues layers including the cambium and bark. However, the relatively gas-impermeable cambium (Kramer & Kozlowski 1979) and bark form a strong barrier to radial diffusion (Lendzian 2006; Steppe et al. 2007). Consequently, a part of the respired CO2 does not escape to the atmosphere, causing stem internal CO2 concentrations to be generally very high. Macdougal & Working (1933) reported values up to 260,000 ppm in mature Quercus agrifolia trunks, which was over 900 times higher than the atmospheric CO2 concentration at the time they did their study. Trees and other woody plant species take advantage of this high internal CO2 concentration by re-fixing a portion of it via photosynthesis in stems. Stems of many plant species contain active chloroplasts, which efficiently assimilate CO2. Chlorophyll in stems is mainly located in the cortex and, therefore, the term corticular photosynthesis is commonly used when referring to stem CO2 assimilation (Pfanz & Aschan 2001). However, chlorophyll can also be found in the ray cells of wood (Wiebe 1975; Rentzou & Psaras 2008) and even in the pith (Vancleve et al. 1993; Berveiller, Kierzkowski & Damesin 2007; Rentzou & Psaras 2008). We therefore use the more general term ‘woody tissue photosynthesis’ in the work presented here. Woody tissue photosynthesis has been reported in a wide variety of plant species and is widespread in trees (see Teskey et al. 2008, table 3 for an overview).

During photosynthesis in the leaves of C3 plants, discrimination against the heavy isotope of carbon (13C) occurs mainly due to two processes: differential diffusion of 13CO2 and 12CO2 from the ambient air into the leaf and carboxylation inside the chloroplasts by principle carboxylating Rubisco (Brugnoli & Farquhar 2000). The result of these discrimination processes at leaf level is a depletion in 13C relative to the atmosphere and therefore a lower ratio of 13C/12C for all fixed plant carbon. Stems rarely, if ever, take up atmospheric CO2 directly, but instead re-fix stem internal CO2 (Cernusak & Marshall 2000; Pfanz et al. 2002), which is already depleted in 13C because it originates from respiration of plant organic material (Cernusak et al. 2001). Depleted source CO2 and further discrimination against 13C if fixed by Rubisco in woody organs result in photosynthates that are very depleted in 13C. Evidence of this has been given by Cernusak et al. (2001).

Studies that have examined the functional importance of woody tissue photosynthesis have usually determined rates of photosynthesis and inferred from these how much stems might contribute to whole-plant carbon gain (e.g. Schaedle & Foote 1971; Nilsen & Sharifi 1994, 1997; Nilsen, Rundel & Sharifi 1996; Levy & Jarvis 1998; Damesin 2003; Cernusak et al. 2006). It has rarely been examined via direct growth measures to which extent woody tissue assimilation contributes to the plant's carbon budget. One study has evaluated the significance of stem photosynthesis for carbon gain of the stem itself via direct growth analysis (Bossard & Rejmanek 1992). No significant difference was found in relative growth rate of branches, relative growth rate of basal diameter and stem biomass between control plants and plants with stems wrapped in aluminium foil of the shrub Cytisus scoparius, suggesting that in this species stem photosynthesis did not significantly contribute to stem carbon gain.

To our knowledge, direct measurements of the contribution of stem photosynthesis to carbon gain of plant organs other than the stems themselves have never been performed. Cernusak et al. (2001) found evidence that stem assimilates rapidly enter substrate pools for respiration and tissue synthesis of the stem itself. In some extreme situations, woody organs may contribute to the carbon gain of other organs. There is some evidence that in periods of large sink demands, such as during spring growth, branches use carbohydrates stored in the stem (Sprugel, Hinckley & Schaap 1991; Lacointe et al. 2004). Pfanz & Aschan (2001) pointed out that woody tissue photosynthesis may be of particular importance in the case of defoliation events, caused by insect attacks, leaf fungal pathogens, grazing, fire, etc. Two studies investigated the role of stem photosynthesis in survival and recovery after defoliation. Bossard & Rejmanek (1992) found that the shrub C. scoparius continued to grow after complete defoliation, suggesting that stem photosynthesis made a significant contribution to whole-plant biomass production after defoliation in this species. C. scoparius has functional stomata in its bark, which allows the uptake of airborne CO2 (Bossard & Rejmanek 1992). Nevertheless, in most woody plant species, the source CO2 for corticular and stem photosynthesis is CO2 released during stem respiration (internal re-fixation of CO2). Recently, Eyles et al. (2009) found that stems of Eucalyptus globulus trees had an increased capacity to re-fix internal CO2 following partial defoliation and suggested that stem re-fixation represents a previously unexplored mechanism to minimize the impact of foliar loss.

In the present study, we aimed at investigating the contribution of stem photosynthesis to the woody plant carbon balance at two levels: at the level of the stem itself and at the level of development of new leaves after defoliation. We therefore conducted a combined study of light exclusion of stems (coverage with aluminium foil and with shade cloth) and complete defoliation in three woody California native evergreen species: Prunus ilicifolia (Nutt. ex Hook. & Arn.) Walp. ssp. Lyonii (Eastw.) P.H. Raven, Umbellularia californica (Hook. & Arn.) Nutt. and Arctostaphylos manzanita Parry ‘Dr Hurd’. Two main questions were addressed: (1) do stem assimilates contribute to the carbon gain in the stem itself and (2) does stem photosynthesis enhance recovery of defoliated plants by providing carbon for the formation of new leaves?

Our specific working hypotheses were: (1) if stem assimilates contribute to stem carbon gain, stem growth will be slower in plants with covered trunks versus trees with uncovered trunks; (2) if stem assimilates contribute to the formation of new leaves after a defoliation event, bud growth will be slower in covered plants compared to uncovered plants; and (3) if stem assimilates contribute to the formation of new leaves after a defoliation event, bud sugars of covered plants (before they become autotrophic) will be less depleted in 13C compared to bud sugars of uncovered plants.


Plant material and experimental conditions

Experiments were performed on three long-lived woody evergreen species: P. ilicifolia ssp. Lyonii, U. californica and A. manzanita‘Dr Hurd’, all California natives. The species were selected based on their visible trunk pigment content. Umbellularia has deep green bark, but has no visible green pigments in the wood, Arctostaphylos has a so-called peeling bark with green pigments in both the bark and the wood and Prunus has fewer pigments in the bark and no visible pigments in the wood. Nine young woody plants of each species, approximately 1 m in height and with a trunk diameter of 0.9–1.1 cm at 0.10 m from the soil surface, were grown in 19 L pots with commercial potting mixture (Conrad Fafard, Agawam, MA, USA). Just before the onset of the experiment the plants were fertilized with a commercial fertilizer (all purpose planting and growing food; Lilly Miller, Walnut Creek, CA, USA). The pots were irrigated with tap water three times a week.

The plants were grown in a greenhouse held at 25 ± 5 °C and between 45 and 70% relative humidity at the Jane Gray Facility housed at the University of California Botanical Garden, Berkeley. Plants were grouped together, but with space between them (about 30 cm between pots). In addition to natural daylight, artificial light was provided by halide lamps (Agrosun Gold Halide 400 W; Hydrofarm, Petaluma, CA, USA) for a 12 h period (0700–1900 h). In this period, minimum photosynthetic active radiation (PAR) at crown level was 300 ± 70 µmol m−2 s−1. An important limiting factor for woody tissue photosynthesis is irradiance. For photosynthesis to occur, adequate PAR must reach the light-harvesting complexes of the chloroplasts (Pfanz et al. 2002). Two PAR sensors (LI-190 Quantum Sensor; Li-Cor Inc., Lincoln, NE, USA) were therefore mounted on the trunk of three plants about 0.15 m above soil surface: one in radial (measuring PAR intercepted horizontally by the trunk) and one in axial (measuring PAR intercepted vertically by the trunk) direction. A third PAR sensor was installed at the crown top level, measuring vertically intercepted PAR. This sensor was placed in an open spot between the plants, approximately 1 m above soil surface.

Experimental design

On 21 January 2009, stem diameters were measured at 0.10 m and 0.15 m from the soil surface. Trunks of three plants of each species were wrapped with aluminium foil (100% light exclusion), trunks of three plants were wrapped with aluminium shade cloth (50% light exclusion) and the remaining three plants served as controls. Aluminium foil and shade cloth were wrapped very loosely so that gaseous diffusion from trunks was not restricted. On 21 April 2009, trunk diameters were measured at the same positions. Next, all plants were entirely defoliated and the few buds present were removed. The bud-free branches of the plants with covered trunks were then covered with the same material (aluminium foil or shade cloth). Ten mature leaves were immediately frozen in liquid nitrogen and afterwards stored at −20 °C until chlorophyll and stable isotope analysis. The remaining leaves were oven dried for 4 d at 65 °C for determining leaf dry weight (DW). In the next 3 weeks, buds (which developed under the cover if present) were harvested three times (1, 7 and 12 May) when they were not autotrophic yet, immediately put in liquid nitrogen and stored at −20 °C until stable isotope analysis. Bud sampling at these three time instances was necessary to prevent them developing into autotrophic plant organs (which happens at the time of budburst). On 13 May 2009, entire plants were harvested. Part of the functional roots (diameter < 0.2 cm) was immediately immersed in liquid nitrogen and then stored at −20 °C until stable isotope analysis. For each plant, phloem tissues from two 0.3 cm diameter branches and two trunk segments were separated from the xylem tissues. These tissues were immediately frozen in liquid nitrogen and stored at −20 °C until chlorophyll and stable isotope analysis could be performed. The remaining roots, branches and trunk were oven dried for 4 d at 65 °C to determine DW.

Chlorophyll concentration

The amounts of chlorophyll a and b were calculated as described by Wellburn (1994). To this end, 220 mg fresh weight (FW) from leaves, 100 mg FW from trunk and branch phloem, 200 mg FW from branch xylem and 800 mg FW from trunk xylem was put in tubes containing 5 mL N,N-dimethylformamide. The tubes were incubated for 3 h in a water bath at 65 °C and then left at room temperature for 1 h. Tubes were kept in darkness at all times. The absorbance of the solution was measured at 647 nm and 664 nm with a spectrophotometer (model UV 160, Shimadzu, Kyoto, Japan). After measurements, the plant material was collected and oven dried for 4 d at 65 °C to determine the corresponding DW.

Sugar extractions

Soluble sugar extraction was performed according to Brugnoli et al. (1988) with minor modifications. Frozen plant material (leaves, branch phloem, trunk phloem, roots) was first lyophilized and then ground in a ball mill. The weight of the dried materials was measured. Next, 150 mg finely ground tissue was weighed into tubes, 150 mg polyvinylpolypyrrolidone (PVPP; Sigma-Aldrich, St. Louis, MO, USA) was added to remove phenolics and the mixture was suspended in 5 mL deionized water. The samples were vortexed and then put on a shaker at room temperature for 45 min before being centrifuged for 20 min at 12 000 g. The supernatants were passed through columns of ion exchange resin DOWEX-50W X 8, 50–100 mesh (H+ form) (Sigma-Aldrich) to remove amino acids and ion exchange resin DOWEX 1X2, Cl-, 50–100 mesh (Cl- form) (Sigma-Aldrich) to remove organic acids. The columns were washed with 5 mL deionized water and the sugar containing eluate collected. The samples were freeze-dried and stored in a desiccator before carbon isotope analysis. We acknowledge, as recently demonstrated by Richter et al. (2009), that this method for soluble sugar extraction is not entirely accurate because the eluate may not only contain bulk soluble sugars, but also sugar-like substances, such as polyols (e.g. myo-inositol), which may be isotopically different from sugars.

Carbon isotope analysis

Sugar samples were analyzed for their carbon isotope composition using a PDZ Europa Scientific 20-20 continuous flow isotope ratio mass spectrometer (IRMS) coupled to an ANCA/SL elemental analyzer. Analyses were performed at the Centre for Stable Isotope Biogeochemistry, University of California at Berkeley, CA, USA. The isotope composition was calculated from the sample isotopic ratio 13C/12C (Rs) measured by the IRMS and was expressed using the conventional delta notation (in‰ notation) according to the relationship:


where RPDB is the international standard (Pee Dee Belemnite) = 0.0112372. Long-term precision for C isotope measurements was ±0.2‰ based on repeated measurements of three laboratory working standards (bovine liver, peach leaf and sucrose).

Statistical analysis

Effects of species, treatment and treatment × species on total chlorophyll concentration (a + b) were evaluated by two-way analysis of variance (anova). The effects of treatment and species on stem diameter increment, sugar δ13C values and the ratio bud DW : plant DW were examined by one-way anova with Tukey post hoc tests. SPSS, 15th edition (SPSS Inc., Chicago, IL, USA) was used for all analyses.


Trunk PAR interception

Radially and axially intercepted PAR at trunk level was high (Fig. 1). On a sunny day in April (before defoliation), axial values were 43, 67 and 79% and radial values 77, 73 and 89% of maximum PAR (1760 µmol m−2 s−1) at crown level for Prunus, Umbellularia and Arctostaphylos, respectively. Total axially intercepted PAR was 24, 34 and 50% and total radially intercepted PAR was 56, 52 and 50% of the total daily PAR sum measured at crown level for Prunus, Umbellularia and Arctostaphylos, respectively. Hence, shading by crowns did not dramatically reduce PAR at trunk level. Axial PAR interception was highest in Arctostaphylos because the openness of the crown is quite large in this species.

Figure 1.

Daily pattern of photosynthetically active radiation (PAR) in the greenhouse at crown level, at trunk level in axial direction and at trunk level in radial direction for (a) Prunus (b) Umbellularia and (c) Arctostaphylos. Measurements were taken on 15 April 2009, before defoliation.

Chlorophyll concentrations

A highly significant species effect was found for chlorophyll a + b concentration in all tissues (Table 1). For trunk phloem and branch phloem, highest chlorophyll a + b concentrations were found in Umbellularia. Trunk xylem exhibited most striking differences: chlorophyll a + b concentration was up to 10 times higher in Arctostaphylos compared to Prunus and Umbellularia. In Prunus and Umbellularia, chlorophyll was almost absent (0.01–0.02 mg g−1) in trunk xylem. Leaf chlorophyll a + b concentrations were slightly lower in Arctostaphylos compared to the other species.

Table 1.  Chlorophyll a + b concentration (mg per g of DW) of different tissues of the three studied species and for the three different treatments
SpeciesTreatmentChlorophyll a + b (mg g−1)
Trunk phloemBranch phloemTrunk xylemBranch xylemLeaf
  • *

    P < 0.05,

  • **

    P < 0.001.

  • Shade cloth and aluminium foil was applied during 3 months and 3 weeks on the trunks, and during 3 weeks (after defoliation) on the branches. Leaves were not covered. Data are means ± SE (n = 3). F values are given for the effect of species, treatment and the interaction between them.

  • na, not applicable; ns, not significant.

PrunusControl1.26 ± 0.052.11 ± 0.590.02 ± 0.000.30 ± 0.038.66 ± 0.49
Shade cloth1.19 ± 0.172.07 ± 0.110.02 ± 0.010.26 ± 0.058.76 ± 1.02
Aluminium0.71 ± 0.131.84 ± 0.140.01 ± 0.010.22 ± 0.068.10 ± 0.64
UmbellulariaControl2.16 ± 0.342.43 ± 0.960.02 ± 0.000.17 ± 0.028.26 ± 0.52
Shade cloth2.75 ± 0.502.64 ± 0.310.02 ± 0.010.20 ± 0.088.91 ± 1.05
Aluminium1.81 ± 0.572.37 ± 0.390.02 ± 0.010.15 ± 0.048.38 ± 1.56
ArctostaphylosControl0.95 ± 0.231.46 ± 0.130.16 ± 0.020.33 ± 0.076.90 ± 0.87
Shade cloth0.97 ± 0.201.32 ± 0.110.14 ± 0.030.31 ± 0.036.66 ± 0.97
Aluminium0.35 ± 0.221.08 ± 0.430.10 ± 0.040.27 ± 0.067.16 ± 0.94
Treatment11.35**0.89 ns4.28*2.52 nsna
Species × treatment1.01 ns0.15 ns3.04*0.44 nsna

Treatments had a significant effect on chlorophyll a + b concentration in phloem as well as in xylem of trunks. Long-term coverage of the trunk with aluminium foil caused a decrease in phloem chlorophyll a + b concentration. The largest relative reduction in trunk phloem chlorophyll concentration happened in Arctostaphylos (63%) (Table 2), which is the species that had the lowest initial chlorophyll a + b concentration (Table 1), whilst the smallest relative reduction happened in Umbellularia (16%) (Table 2) which had the largest initial chlorophyll a + b concentration (Table 1). Shade cloth coverage had no consistent effect (Table 1). In Arctostaphylos, aluminium coverage also caused a clear decrease in trunk xylem chlorophyll concentration. This effect was not visible in Prunus and Umbellularia because of the very low initial concentrations. Again, shade cloth had no apparent effect. In branches, the effect of treatments was not significant. Branches were only covered for three weeks (after defoliation), which was apparently not sufficient to cause a decrease in chlorophyll concentration.

Table 2.  Relative reduction in chlorophyll a + b concentration of trunk phloem and in trunk diameter growth of the three studied species
SpeciesReduction in chlorophyll a + b concentrationReduction in trunk diameter growth
  1. Relative reductions were calculated as the difference between the control value and the value measured after aluminium coverage, normalized for the control value.


Trunk growth

For all treatments, stem growth was greater in Prunus compared to Umbellularia and Arctostaphylos. Aluminium coverage had a significantly negative effect on trunk diameter increment in all species (Fig. 2). The largest relative reduction in trunk diameter increment happened in Umbellularia (56%) despite having the lowest chlorophyll a + b reduction following aluminium coverage (Table 2). Stem growth of shade cloth covered trunks was not significantly different from controls (Fig. 2).

Figure 2.

Relative growth rate of trunk diameter over a 3 month period (start of coverage until defoliation). Relative trunk diameter increment was calculated as the increment in diameter (in mm) divided by time (in days) and by the initial diameter of the woody plant (in mm). Data are mean ± SE with n = 3. Statistical differences (P < 0.05) between i) treatments for a given species (lower case letters) and ii) species for a given treatment (capital letters) were determined.

δ13C of soluble sugars in leaves at defoliation, and of trunk and branch phloem, and root and bud biomass after defoliation

Leaf sugar δ13C was significantly different among species (Fig. 3). The most negative values were found in Arctostaphylos and the least negative values in Prunus. Coverage of trunks had no significant influence on sugar δ13C at leaf level (Fig. 3).

Figure 3.

δ13C of soluble sugars in leaves. Data are mean ± SE with n = 3. Statistical differences (P < 0.05) between i) treatments for a given species (lower case letters) and ii) species for a given treatment (capital letters) were determined.

Sugar δ13C of trunk phloem was significantly lower in Arctostaphylos compared to the other species (Fig. 4a). Trunk phloem sugar δ13C was significantly affected by aluminium coverage in defoliated Prunus and Arctostaphylos. The same trend was observed in Umbellularia although the effect was not significant. Shade cloth coverage had no effect on trunk phloem sugar δ13C. The same trends were observed for branch phloem (Fig. 4b).

Figure 4.

δ13C of soluble sugars in (a) trunk phloem (b) branch phloem (c) roots and (d) (non-autotrophic) buds after defoliation. Trunk phloem, branch phloem and roots were collected at the end of the experiment (13 May), buds were collected three times during the experiment (1, 7 and 12 May). Data are mean ± SE with n = 3. Statistical differences (P < 0.05) between i) treatments for a given species (lower case letters) and ii) species for a given treatment (capital letters) were determined. No value could be obtained for buds of aluminium foil covered Arctostaphylos plants because of extremely low bud biomass.

The δ13C signal of root sugars was only significantly different between species in control plants, where Arctostaphylos had lower values than Prunus (Fig. 4c). Root sugar δ13C was not affected by the treatments. However, although not significant, there was also a tendency in aluminium and shade cloth covered Arctostaphylos plants to have lower δ13C in root sugars than Prunus and Umbellularia.

Bud sugar δ13C in Prunus was as high as −20.6‰ and significantly higher compared to the other species (Fig. 4d). Sugar δ13C was also higher in buds compared to all other tissues in Prunus (2–3.6‰) (Fig. 4d versus 4a,b,c) and higher in buds compared with trunk and branch phloem in Arctostaphylos (1.8–2.4‰) (Fig. 4d versus 4a,b). In Umbellularia, differences between sugar δ13C of buds and other tissues were smaller (0.1–1.0‰) but also present. In bud sugars, aluminium coverage resulted in significant higher δ13C than in controls in Prunus and Umbellularia (Fig. 4d). Bud biomass of aluminium foil covered Arctostaphylos plants was too small to be analysed for stable isotope composition. Bud sugar δ13C of shade cloth covered plants was not different from controls.

Bud biomass

Recovery from defoliation was clearly species dependent: Arctostaphylos exhibited the lowest relative bud biomass, in covered as well as control plants (Fig. 5). The ratio between bud DW and total plant DW was significantly reduced under aluminium coverage for Prunus and Arctostaphylos. A similar but not significant trend was observed for Umbellularia. In none of the species, shade cloth coverage affected bud biomass.

Figure 5.

Ratio of bud dry weight (DW) over total plant DW. Data are mean ± SE with n = 3. Statistical differences (P < 0.05) between i) treatments for a given species (lower case letters) and ii) species for a given treatment (capital letters) were determined.


Trunk PAR interception

A primary requirement for stem photosynthesis to occur is light. Measurements of PAR at trunk level revealed that shading by the crown was limited and did not cause a serious reduction in PAR at trunk level. However, the light that reaches the stem surface does not necessarily reach the chloroplasts. Light must pass through the epidermal, peridermal and/or rhytidomal layers of the bark. The amount of light that passes through these layers varies between <10 and 50% of incident light, depending on species and age of the stem (Aschan & Pfanz 2003; Filippou, Fasseas & Karabourniotis 2007).

The light environment inside a stem is not only determined by incident light at the stem surface, however. Additional light can be delivered axially through the stem vascular tissue that can leak out to the surrounding living tissues (Sun et al. 2003). Nevertheless, this axially transmitted light likely has only a minor role in woody tissue photosynthesis. Firstly, the conducting elements (vessels, fibres and tracheids) are dead in mature tissues and thus not able to perform any metabolic activity. Secondly, the type of light that is most efficiently conducted axially is in the infrared and near-infrared regions (720–910 nm) (Sun et al. 2003), while photosynthetic machinery preferentially uses wavelengths of 400–700 nm.

Chlorophyll concentrations

The amount of chlorophyll a + b in trunk phloem of Prunus and Arctostaphylos was within the range (0.38–1.89 mg per g of DW) reported by Pfanz et al. (2002) and Berveiller & Damesin (2008) for several deciduous and coniferous trees, while the values in Umbellularia were slightly higher. Interestingly, chlorophyll was not only found in phloem tissues, but the amount of chlorophyll in the trunk xylem of Arctostaphylos was exceptionally high compared to the other species. Arctostaphylos is a species with a so-called peeling or exfoliating bark. When stems grow, old dead bark layers peel off, so that the barriers for light transmission decrease and light can penetrate deep into the stems. In the other species, bark layers are thicker, restricting light transmission.

Long-term coverage with aluminium foil resulted in a decrease of the amount of chlorophyll a + b in trunks. The same result was obtained by Bossard & Rejmanek (1992), who wrapped stems of C. scoparius plants in aluminium foil for 66 d. In our study, 50% light exclusion with shade cloth had no effect at all. Wittmann, Aschan & Pfanz (2001) also observed minor changes in pigment content of Fagus sylvatica L. twigs after shading (80% light exclusion) during an annual cycle. Probably, the amount of light that reached the trunk surface was still sufficient to allow maximum photosynthesis. We did not measure light saturation levels of trunk photosynthesis in this study, but it has been shown in other studies and for other species that these can be very low. For example, Wittmann et al. (2001) showed that in F. sylvatica L. twigs grown under high light, light saturation of twig photosynthesis was as low as 207 µmol PAR m−2 s−1. The amount of PAR that reached uncovered trunks in our study was up to 1400 µmol PAR m−2 s−1 under sunny conditions (Fig. 1), so that the amount of PAR under 50% light exclusion could still be adequate for light-saturated trunk photosynthesis.

Trunk growth and bud biomass

Long-term coverage with aluminium foil not only clearly reduced trunk chlorophyll concentration, but also radial trunk growth in all species. These results confirm our first working hypothesis that stem photosynthesis makes a significant contribution to stem carbon gain for the focal species. The largest relative reduction in trunk diameter increment in Umbellularia, despite the lowest relative chlorophyll a + b reduction following long-term coverage with aluminium foil, suggests that Umbellularia relied more on stem photosynthesis for stem growth compared to the other two species. It is known that carbon assimilation in stems partially compensates for the carbon loss by stem respiration (e.g. Levy & Jarvis 1998; Damesin 2003). When stem assimilates are used as a carbon source for respiration, other assimilates (e.g. leaf assimilates) become available for other purposes, such as tissue synthesis. Our measurements indicate that stem assimilation supports the construction of new stem tissue.

Bud development was severely reduced in aluminium foil covered plants, confirming our second working hypothesis that stem assimilates contribute to the formation of new leaves. This was further supported by our isotopic data. Bud sugar δ13C values mirrored trunk and branch phloem sugar δ13C (discussed below).

δ13C of soluble sugars in leaves at defoliation, and of trunk and branch phloem, and root and bud biomass after defoliation

The δ13C values of leaf sugars at defoliation were not affected by the light exclusion treatments applied to the stem tissue, but they differed among species. Carbohydrate reserves in leaves indeed originate from leaf photosynthates, which δ13C value can be predicted from the model for photosynthetic discrimination (Farquhar, O'Leary & Berry 1982) as a function of the ratio Ci/Ca (Ci and Ca the leaf internal and the atmospheric CO2 concentration, respectively). This suggests that leaf photosynthate δ13C values are independent from stem performance, but related to species-specific leaf photosynthetic activity. Among the focal species, Arctostaphylos had a light-saturated leaf conductance three and four times higher than Prunus and Umbellularia, respectively (data not shown). Consequently, this species had the lowest δ13C values for leaf sugars when compared with the other species. The carbon fixed by the leaves constitutes the primary carbon source for the different plant organs. Accordingly, Arctostaphylos had more depleted sugars in the trunk and branch phloem. The δ13C of sugars in roots showed the same trend, although species differences were not statistically significant in these organs.

After defoliation, new buds have to be produced with stored carbohydrates. Because δ13C of bud sugars did not reflect δ13C of root sugars (Fig. 4d versus 4c), but did mirror δ13C of trunk and branch phloem sugars (Fig. 4d versus 4a,b), it seems more likely that the latter organs contribute more to carbon gain in buds than roots. Several studies have concluded that defoliation caused a decrease in starch concentration in stems and branches of young trees, whereas it did not affect starch concentration in roots (e.g. VanderKlein & Reich 1999; Cerasoli et al. 2004). In our study, Arctostaphylos, the species with the most depleted trunk and branch phloem sugars, also had the most depleted bud sugars.

Interestingly, for all species, bud sugars were more enriched in 13C than sugars in the trunk or branch phloem. Damesin & Lelarge (2003) found that at budburst δ13C of sucrose in buds was lower than δ13C of sucrose in twigs of F. sylvatica trees, but δ13C of glucose and fructose was higher in buds than in twigs. We determined bulk sugar δ13C, which is an integrated measure for δ13C of sucrose, glucose and fructose. Considering the more general concept of carbon sources versus carbon sinks, another interesting study can be added for comparison. Maunoury-Danger et al. (2009) found that when sprouting occurs in Solanum tuberosum tubers in the dark, sucrose δ13C in the heterotrophic sprouts (i.e. carbon sinks) was nearly 2‰ higher than in the tubers themselves (i.e. carbon sources). A plausible explanation for the observations in our and other studies would be that 13C favouring fractionation processes occur at the level of sugar export. However, Bathellier et al. (2008) and Maunoury-Danger et al. (2009) demonstrated that during ontogeny of Phaseolus vulgaris L. seeds and during sprouting of Solanum tuberosum tubers sucrose export was not associated with fractionation. Maunoury-Danger et al. (2009) alternatively explained this enrichment by arguing that 13C depleted carbon is consumed in respiratory metabolism and in tissue synthesis of the sprout, resulting in a 13C enrichment of sprout sucrose. Respired CO2 has already been found to be 13C depleted compared with sucrose in other heterotrophic organs such as roots (e.g. Badeck et al. 2005; Klumpp et al. 2005). It is probable that also in our study respiration and tissue synthesis in buds caused a 13C enrichment of bud sugars.

An interesting outcome of our experiment is that we observed an enrichment in the δ13C of bud sugars in aluminium covered trees relative to control trees and that the magnitude of this enrichment was similar to the observed enrichment in trunk and branch phloem (1–2‰). Cernusak et al. (2001) also found that aluminium foil covered trees had 1‰ and 0.8‰ more enriched wood and bark, respectively, compared to control trees. Because buds were not autotrophic yet when harvested, the sugars in these organs originated from either: (1) mobilization of stored reserves from other organs (branches, trunk); or (2) direct transport of newly assimilated sugars. Stored reserves (starch) are known to be more enriched than sugars (Bowling, Pataki & Randerson 2008). New assimilates were more enriched in plants covered with aluminium foil than in control plants. Therefore both factors could explain the observed 13C enrichment in bud sugars for aluminium foil covered plants. During this experiment, we did not measure starch content or δ13C value of starch. Therefore, we cannot speculate about the relative contribution of sugars versus starch towards bud development. Regardless of the C source, however, the observed enrichment in the δ13C of bud sugars, together with a lower bud biomass development for the aluminium foil covered trees, clearly alludes to a direct (via transport of stem assimilates) and/or indirect (via mobilization of reserves that were built with stem assimilates) contribution by stem photosynthesis to bud growth in defoliated woody plants.


Studies on the contribution of woody tissue photosynthesis to the carbon budget of trees and other woody plants are still underexplored (Pfanz 2008). Our results have now provided an insight in the role of woody tissue synthesis in the carbon gain of stems itself and in the carbon gain of developing leaves. Long-term light exclusion of assimilative trunks revealed that stem assimilates play a significant role in the trunk's carbon gain. Furthermore, defoliation treatments in combination with light exclusion from woody tissues demonstrated that stem assimilates also play a role in the carbon budget of newly developing leaves. Our findings point out the need, especially in young plants, to comprehensively measure both leaf and woody tissue carbon assimilation for assessment of whole-plant carbon budgets.


We wish to thank the Fund for Scientific Research–Flanders (F.W.O.–Vlaanderen) for the postdoctoral funding and the mobility allowance awarded to the first author and the Centre for Stable Isotope Biogeochemistry at UC Berkeley for funding the stable isotope analyses. We are also indebted to Paul Brooks and Stefania Mambelli of the Centre for Stable Isotope Biogeochemistry at UC Berkeley, for assistance with stable isotope analyses and the methods for extracting and preparing sugars for isotope analyses. We also thank the members of the Dawson Lab Group for their assistance and good humour during all the phases of the research.