SEARCH

SEARCH BY CITATION

Keywords:

  • growth;
  • Norway spruce (Picea abies);
  • root respiration;
  • seasonality;
  • stem efflux;
  • stem respiration;
  • temperature;
  • xylem sap flow

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Stem CO2 concentrations (stem [CO2]) undergo large temporal variations that need to be understood to better link tree physiological processes to biosphere–atmosphere CO2 exchange.
  • During 19 months, stem [CO2] was continuously measured in mature subalpine Norway spruce trees (Picea abies) and jointly analysed with stem, soil and air temperatures, sap flow rates, stem radius changes and CO2 efflux rates from stem and soil on different time scales.
  • Stem [CO2] exhibited a strong seasonality, of which over 80% could be explained with stem and soil temperatures. Both physical equilibrium processes of CO2 between water and air according to Henry's law as well as physiological effects, including sap flow and local respiration, concurrently contributed to these temporal variations.
  • Moreover, the explanatory power of potential biological drivers (stem radius changes, sap flow and soil respiration) varied strongly with season and temporal resolution. We conclude that seasonal and daily courses of stem [CO2] in spruce trees are a combined effect of physical equilibrium and tree physiological processes. Furthermore, we emphasize the relevance of axial diffusion of CO2 along air-filled spaces in the wood, and potential wound response processes owing to sensor installation.

Introduction

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

The net carbon balance of an ecosystem is the balance of carbon gain by assimilation and carbon loss by respiration. While assimilation can be clearly assigned to green plant tissues aboveground, respiration originates from both aboveground and belowground plant tissues, as well as soil organisms, and their dependences on physical and biochemical factors or their interactions are not well yet understood (Trumbore, 2006). Recent studies indicate that CO2 release measured in a specific tissue does not always reflect local CO2 respiration and that stem CO2 efflux (EFStem), measured usually by chambers attached to the stem, does not necessarily reflect true stem respiration resulting from translocation of CO2 dissolved in xylem water (Teskey et al., 2008). Furthermore, Aubrey & Teskey (2009) stated that much of root-respired CO2 enters the xylem stream, challenging the paradigm of root-respired CO2 diffusing completely into the soil atmosphere and thus questioning the ability to derive root respiration rates by measuring soil CO2 efflux (EFSoil) only. Consequently, if part of root-respired CO2 remains inside the tree, it can be refixed by chloroplast-containing tissues (Gansert & Burgdorf, 2005; Dima et al., 2006; McGuire et al., 2009) and might contribute to tree assimilation. Thus, common approaches to partition net ecosystem CO2 fluxes into gross primary productivity and total respiration by response functions probably have to be revised.

Trees are an important link between the soil and the atmosphere, both in terms of water and carbon translocation. Stem internal CO2 (stem [CO2]) consists of CO2 respired locally by different stem tissues, such as bark, phloem, cambium and living cells within the sapwood, and of CO2 respired by roots or lower sections of the stem, dissolved in xylem water and being dislocated via sap flow (Moore et al., 2008; Aubrey & Teskey, 2009). Hence, understanding tree internal CO2 fluxes is essential when aiming at a better understanding of stem respiration and measured EFStem, but also of processes involved in soil respiration and links between forest ecosystem assimilation and respiration.

The contribution of respired CO2 to either the CO2 remaining inside the tree or to the CO2 diffusing to the atmosphere is determined by respiration rates of individual tissues/organisms and by diffusion barriers (Sevanto et al., 2011; Steppe et al., 2012), all of which are a function of the actual phenological phase. For example, respired CO2 in the bark travels through air-filled spaces when released to the atmosphere, whereas CO2 respired by sapwood also has to cross water-saturated tissue and the nearly gas-impermeable cambium (Steppe et al., 2007). Thus, translocation of CO2 via sap flow is supposed to affect EFStem especially during day-time, while at night, when transpiration but also corticular and woody photosynthesis are close to/at zero, measured EFStem is assumed to better represent the actual stem respiration rate (McGuire & Teskey, 2004).

Tree internal CO2 fluxes, such as import, export and recycling, could serve as one explanation for high spatial and temporal variability of previously reported EFStem rates (Ceschia et al., 2002; Damesin et al., 2002; Teskey & McGuire, 2005; Steppe et al., 2007). Moreover, they might explain the documented temperature hysteresis or even decoupling of EFStem from temperature (Saveyn et al., 2008a), which is in conflict with the commonly accepted theory that respiration should be related exponentially to temperature, as can be expected for purely biologically driven processes. Further factors have been identified that may also alter internal stem [CO2] and EFStem, such as substrate limitations (Stockfors & Linder, 1998; Pruyn et al., 2005), diurnal growth patterns (Daudet et al., 2005) and changes in stem water status (Saveyn et al., 2007b, 2008b). Together, these aspects make interpretation of respiration even more difficult.

Recently, Teskey et al. (2008) proposed that a correct estimate of stem respiration needs to account for internal CO2 fluxes, which can be assessed by measuring CO2 concentrations in situ in stems or branches of trees. Inserting a microelectrode (McGuire & Teskey, 2002) or a nondispersive infrared CO2 sensor (Teskey & McGuire, 2007) into the stem allows for continuous and high-resolution measurements of stem [CO2]. However, there are few studies that report continuously measured stem [CO2] in trees and most of these were conducted in the glasshouse and/or were restricted to short periods of several days or weeks (Teskey & McGuire, 2002; Saveyn et al., 2007b; Cerasoli et al., 2009). The results of these studies vary widely in terms of flux magnitude and their interpretation (Bowman et al., 2005; Teskey & McGuire, 2007; Saveyn et al., 2008a). Furthermore, different relationships for internal CO2 fluxes and EFStem have been reported. While Teskey & McGuire (2005, 2007) found EFStem to be directly proportional to stem [CO2] in sweet-gum (Liquidambar styraciflua) and sycamore (Platanus occidentalis), EFStem of poplar trees was either uncoupled from stem [CO2] or the two were inversely related to each other (Saveyn et al., 2008b); no significant correlation could be found for rimu (Dacrydium cupressinum; Bowman et al., 2005). Manipulations of xylem sap CO2 concentrations in poplar and oak trees were reflected in EFStem (Teskey & McGuire, 2002) but crown removal and hence breakdown of the transpiration stream had no effect on EFStem of loblolly pine trees, leading to the conclusion that diffusion of CO2 from the xylem to the stem surface is restricted in pine trees (Maier & Clinton, 2006). In agreement with this conclusion, Ubierna et al. (2009) could not detect any changes in the isotopic signal of EFStem of large conifer trees (Abies grandis, Thuja plicata and Larix occidentalis), neither after labelling their water source nor after crown removal. Hence, the abiotic and biotic drivers that affect the dynamics of stem [CO2] and EFStem are not yet identified.

Thus, in this study, we combine ecophysiological measurements with continuous stem [CO2] measurements that were conducted in a mature subalpine Norway spruce forest over 19 months (May 2009 to December 2010). We address the following objectives: to quantify the degree of daily and seasonal variations of stem [CO2] explicable by pure physical equilibrium processes between CO2 in gas and CO2 dissolved in water, as described by Henry's law; to test interdependences of stem [CO2] and stem, soil and air temperatures, sap flow rates and stem radius changes, as well as EFStem and EFSoil; and to address the temporal consistency of such interdependences at seasonal and daily scales over the 19 months of study. To our knowledge, this is the longest time-series of continuous in situ measurements of stem [CO2] and provides a deeper insight into the dynamic interdependences of the plant–soil system.

Materials and Methods

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

Study site

The Norway spruce forest Seehornwald Davos is located at 1639 m above sea level at 46°48′55.2″N, 9°51′21.3″E in the eastern part of the Swiss Alps. The site belongs to the long-term forest ecosystem monitoring network (LWF, run by the Swiss Federal Research Institute WSL, Switzerland), and several national and international research and monitoring networks. The mean age of the dominant Norway spruce (Picea abies (L.) Karst.) trees is 240 yr, reaching up to 450 yr. The patchy understory is dominated by dwarf shrubs of Vaccinium myrtillus L. and Vaccinium gaultherioides L., and moss mats (Hylocomium splendens (Hedw.) Schimp. and Dicranum scoparium Hedw.). Soils are acidic (pH 3.5–4.5, Jörg, 2008) and classified as rustic podzols and chromic cambisols (IUSS Working Group WRB, 2007). Mean annual precipitation is 1020 mm (1901–2008) with a distinct seasonality (c. 40% of annual rainfall occurs during the summer months June to August). Mean annual air temperature is 4.0°C. With a mean canopy height of 25 m and a leaf area index of 3.9 m2 m−2, the forest can be considered a moderately productive forest stand within this region of the Alps (Etzold et al., 2011).

CO2 concentrations in the stem

Continuous stem [CO2] measurements were conducted during the years 2009 and 2010 on two mature spruce trees (tree 148 and tree 296) by inserting manufacturer calibrated nondispersive infrared (NDIR) CO2 sensors (GMP221 Carbon dioxide probe 0…20 vol%; Vaisala Oyj, Vantaa, Finland) into the xylem, following instructions by McGuire & Teskey (2002) and Saveyn et al. (2008b). Briefly, we drilled 60 mm deep holes with a diameter of 20 mm into the stem, inserted the NDIR sensor and sealed the space between sensor housing and stem with rubber foam (Terostat-IX; Henkel AG & Co.KGaA, Düsseldorf, Germany). Seals were visually checked several times per year. The NDIR sensors were placed 25 cm above the respective stem efflux collars. Measurement precision of the sensor CO2 reading is given as ± 0.15 vol% CO2 (accuracy is ± 2% of the absolute CO2 reading) by the manufacturer. Temperature dependency of the method was corrected according to the manufacturer (−0.1 vol%/°C). Stem [CO2] is given in volume percentage (vol%, 1 vol% = 10 000 ppm).

In addition, CO2 dissolved in the xylem sap (sap [CO2], in mM) was calculated by applying Henry's law, extended by dissociation factors, according to Teskey et al. (2008) and McGuire et al. (2007):

  • display math(Eqn 1)

where sap [CO2] is total dissolved inorganic carbon in solution, K1 and K2 are first and second acidity constants, KH is Henry's constant, and pCO2 is the partial pressure of CO2 over the solution. Henry's law is based on the dependency of CO2 dissolution in water on temperature, pH and partial pressure of CO2: the dissolution of CO2 in water increases with decreasing temperature and increasing pH. As sap pH can only be assessed by destructive measures, we assumed a constant pH of 5.6 as was given for loblolly pine (Carter & Larsen, 1965). Henry's law is only valid for temperatures above 0°C. Therefore, sap [CO2] is displayed for periods with stem temperatures above 0°C only. Stem temperature (TStem) was continuously measured 2 cm below the stem surface with copper-constantan thermocouples (type T). Stem [CO2] measurements were available from 1 April 2009, TStem measurements from 28 April 2009.

We calculated model stem [CO2] and sap [CO2] according to Henry's law under the assumption of a constant CO2 concentration either in the liquid or in the gas phase, a constant pH of 5.6 and the measured TStem. Temporal variability of model stem and sap [CO2] values was thus solely attributable to the physical effect of temperature changes, as changes in the CO2 pool were excluded. Using different sap pH input values (e.g. 5.6, 6.0 and 6.5) changed only the value of the respective model CO2 concentrations, but not their temporal amplitudes. Differences between model and measured stem [CO2] were interpreted as net import or export of CO2 into or out of the stem segment investigated. For the model stem [CO2] we assumed a constant sap [CO2] value of 3.5 mM (stem [CO2]sap=3.5), and for the sap [CO2] a constant stem [CO2] value of 4 vol% (sap [CO2]stem=4), according to the annual mean values of our measurements.

Stem efflux measurements, EFStem

Stem CO2 efflux (EFStem) was measured on four spruce trees (the two trees 296 and 148 with stem [CO2] measurements and two additional trees 147 and 411) growing in direct vicinity of each other. EFStem was measured campaign-wise with a custom-made chamber connected to a portable infrared gas analyser (Li-8100; Li-cor Inc., Lincoln, NE, USA), operating in closed mode. The chamber was made of a half cylinder of acryl glass (PMMA GS; Evonik Industries AG, Darmstadt, Germany), with the edges covered by neoprene. The chamber covered a stem surface area of 144 cm2 and enclosed a volume of 770 cm3. During the measurements it was placed and fixed on collars that were permanently installed at breast height on the south- and north-facing sides of each tree stem. Collars were sealed to the tree bark with rubber foam and Terostat-IX. In addition, the chamber was covered with aluminium foil to prevent stem photosynthesis and heating by direct radiation, and was equipped with a small fan, a temperature sensor and a pressure vent. The EFStem rates of all trees were measured consecutively with the same installation device (chamber connected to the IRGA). For one measurement cycle, each stem side (North, South) was measured twice within 10 min. Thus, EFStem per tree and at any point in time is given as an average of four measurements. All four trees were measured within 1 h. During the day (06:00–20:00 h), each tree was measured every hour, during the night (21:00–05:00 h) every 2 h (i.e. n = 20 for each tree per diurnal cycle). EFStem was measured during two campaigns in 2008 (6/7 August, 3/4 September), and three campaigns in 2009 (1/2 July, 22/23 July, 12/13 September). The EFStem measurements were fitted to temperature variations of each tree, using the following equation:

  • display math(Eqn 2)

where EFStem.10 is EFStem estimated at 10°C, Q10 is the relative increase in EFStem for a temperature increase of 10°C and TStem is stem temperature (°C).

As sap flow is hypothesized to affect EFStem by translocation of CO2, EFStem is assumed to represent stem respiration under low sap flow conditions only (McGuire & Teskey, 2004). In order to analyse the potential effect of sap flow on EFStem, we established tree-specific temperature response functions of EFStem for conditions when sap flow was almost zero, that is, during the night (18:00–6:00 h), using the following equation:

  • display math(Eqn 3)

Coefficients a and b are estimated by the model. Adjusted r2 values of tree-specific models were 0.43, 0.73, 0.73 and 0.85 (< 0.001) for trees 296, 148, 147, and 411, respectively. Assuming a comparable temperature dependence of respiration during the day, a theoretical EFStem was estimated for daylight conditions and compared with measured EFStem (dEFStem = measured EFStem – predicted EFStem).

Continuous stem radius changes

Stem radius changes (DR, difference in radius over time) were measured with automated point dendrometers (ZB06; natkon.ch, Hombrechtikon, Switzerland) mounted c. 1.5 m aboveground on the north-west side of each stem. The DR was measured every 10 s, averaged and stored every 10 min. The measured temperature sensitivity of the ZB06 of 0.27 μm °C−1 was corrected using a linear temperature response function. The electronic resolution of the dendrometers (in combination with the logger used) was c. 1 μm. Variations in DR reflect tree water status on the short term (minutes to weeks) and wood growth in the long term (months to years; Steppe et al., 2006; Zweifel et al., 2006).

Sap flow

Sap flow of the sampled trees was continuously assessed with two Granier-type sap flow sensors (x600M; UP GmbH, Cottbus, Germany) per tree (Granier, 1985). One sensor was mounted at the north and one at the south exposed side of the stem, c. 1.5 m above the ground. For our analyses, we used the mean of both readings per tree. Sensors were insulated against temperature and solar radiation fluctuations with Styrofoam and bubble wrap with a reflective surface.

Soil CO2 efflux, EFSoil

Continuous soil CO2 efflux measurements (EFSoil) were conducted during the snow-free period in 2009 and 2010, starting mid-May and ending mid-November. An automatically operating soil chamber (LI-8100; Li-cor Inc.) was installed c. 200 m away from the sampling trees, measuring soil CO2 efflux every 30 min (Etzold et al., 2011). Soil temperature (TSoil) was measured in 5 cm soil depth. The continuous EFSoil data were cross-checked with additional manual EFSoil measurements on nine nearby plots during the growing period in 2009, using a portable closed chamber system (LI-8100 with LI-8100-103 chamber; Li-Cor Inc.). A tight and highly significant linear relationship between the continuous and the manual EFSoil measurements confirmed the representativeness of our continuous EFSoil measurements (continuous EFSoil = 0.60 × manual EFSoil + 1.8, adj. r2 = 0.95, P < 0.001, n = 6; campaign averages as a function of daily-averaged continuous EFSoil measurements). Manually measured data were not used for further analyses.

Definition of tree physiological phases

Tree physiological processes within 1 yr were assigned to different consecutive phases as follows (Table 1, Fig. 1): dormancy (period with hardly any transpiration of trees, mainly during winter time; mean daily sap flow rate < 10% of the summer maximum rate); spring (period with recovery from winter stem contraction by rehydration of depleted cell water reserves (Zweifel & Häsler, 2000) leading to radial stem expansion, but not yet to radial stem growth (as in Zweifel et al., 2010)); wood growth (period of radial stem growth according to Zweifel et al. 2010); late summer (period after 97.5% radial stem growth is reached and mean daily sap flow > 10% of summer maximum rates). For further analyses, we separated dormancy into consecutive periods with air temperatures below and above 0°C.

Table 1. Correlation coefficients of mean daily values of stem CO2 concentration and stem temperature (TStem), soil temperature (TSoil), air temperature (TAir), stem radius changes (DR), detrended stem radius changes (DRdetrended, correlated to stem [CO2]detrended) and sap flow for different tree physiological phases (see text for definitions), determined for Norway spruce (Picea abies) trees in the years 2009 and 2010Thumbnail image of
image

Figure 1. Annual cycles of tree and soil physiological processes at a subalpine Norway spruce (Picea abies) forest in 2009 and 2010. (a) Stem CO2 concentrations (stem [CO2]) of two Norway spruce trees (trees 296 and 148) and a model stem [CO2] curve calculated according to Henry's law, with the measured stem temperature and assuming a constant CO2 concentration in the sap of 3.5 mM and a constant pH of 5.6 (stem [CO2]sap=3.5). (b) Stem radius changes (DR) for two Norway spruce trees (trees 296 and 148). (c) Sap flow rates for two Norway spruce trees (trees 296 and 148). (d) Efflux of CO2 from the soil (EFSoil). Tree physiological phases within a year are labelled: dorm, dormancy; sprg, spring; grow, wood growth; sum, late summer (as defined in the Materials and Methods section).

Download figure to PowerPoint

Statistical analyses

All calculations and analyses were conducted with the R statistical software package 2.11.1 (R Development Core Team, 2011). For all tree physiological phases of the two study years we calculated correlations (Spearman rank) of stem [CO2] with TStem, TSoil, TAir, DR and sap flow, as well as cross-correlations with a maximum time lag of ± 3 d. We also smoothed the time-series of stem [CO2] and DR (smooth.spline function in the R statistical software) and calculated correlation matrices with the remaining residuals in order to eliminate the growth trend (stem [CO2]detrended and DRdetrended). For cross-correlations, a positive time lag in days means that stem [CO2] lags behind the cross-correlation variable, a negative lag means that the cross-correlation variable lags behind stem [CO2]. For each tree physiological phase, stem [CO2] was predicted by general linear models (GLM) with a minimum subset of required predictor variables (TStem, TSoil, DR and sap flow). The best model was identified by a stepwise selection of variables in both directions with the stepAIC function from the MASS package in the R statistical software. For multiple regression models, we calculated the standardized beta coefficients in order to determine the relative explanatory power of the variables within each model.

For analyses on a diurnal time-scale, we excluded the long-term seasonality of the variables by normalizing hourly values by the respective daily maxima (in %). For all measures of absolute values (DR, stem and sap [CO2], TStem), we also calculated the changes over each 30-min time-step (∆).

Results

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

Seasonality of stem [CO2]

Stem [CO2] of mature Norway spruce trees increased as soon as cambial activity started in spring, was highest during July and August, and decreased slowly with the cessation of growth during late summer (Fig. 1a,b). Rather outstanding and nonrecurring in its magnitude was the stem [CO2] peak during the early growth period in 2009 (Fig. 1a). The temporal course of modelled stem [CO2], based on the assumption of a constant sap [CO2] of 3.5 mM (stem [CO2]sap = 3.5), was closely related to measured stem [CO2] during both vegetation periods (spring until late summer, 2009: adj. r2 = 0.31, 2010: adj. r2 = 0.68), but did not reflect the general increase of stem [CO2] in the two wood growth phases. During winter 2009/10, especially from January to March 2010, stem [CO2] showed large fluctuations and substantial differences between the two trees (Fig. 1a).

Stem [CO2] and temperature conditions of air, soil and stem

Temperature as a main driver for respiration processes was measured in air (TAir), soil (TSoil) and stems (TStem; Fig. 2a). Stem [CO2] was most closely related to TStem and TSoil, whereas TAir always explained less variation in stem [CO2] than the other two temperatures (rmax in Table 1). TStem was more closely related to stem [CO2] during the early active phases of the year (spring and wood growth), whereas TSoil was more closely related in late summer and during dormancy (> 0°C). Interestingly, the correlation of stem [CO2] to TStem and TAir was best with a time lag of 1 d, whereas the correlation to TSoil was best without a time lag (Table 1). Overall, > 80% of the seasonal variation in stem [CO2] could be explained by TStem when excluding winter days with a mean daily TStem below 0°C and the growth period in 2009 (Fig. 3). Under freezing conditions, additional drivers seemed to affect the relationship between stem [CO2] and TStem, resulting in less variance being explained (Table 1, Fig. 3). The exceptional increase and peak of stem [CO2] in the wood growth period 2009 showed a weaker relationship with temperature than during the remaining vegetation period (Table 1, Fig. 3a), pointing towards a process affecting stem [CO2] which did not occur in 2010.

image

Figure 2. Climate conditions at the Norway spruce forest during the two study years 2009 and 2010. (a) Air, soil and stem temperatures (TAir, TSoil, TStem). (b) Photosynthetic photon flux density (PPFD). (c) Soil moisture and precipitation. (d) Snow depth. Tree physiological phases within a year are labelled: dorm, dormancy; sprg, spring; grow, wood growth; sum, late summer; as defined in the Materials and Methods section.

Download figure to PowerPoint

image

Figure 3. Relationships between daily means of stem CO2 concentrations (stem [CO2], average of both trees) and stem temperatures (TStem) of Norway spruce (Picea abies) trees for the years (a) 2009 and (b) 2010. We distinguished the following time-periods: In 2009: April to June (stem [CO2] = 0.15TStem + 0.28; adj. r2 = 0.37, < 0.001), June to December with TStem > 0°C (stem [CO2] = 0.25TStem + 3.0; adj. r2 = 0.85, < 0.001) and June to December with TStem < 0°C (stem [CO2] = 0.06TStem + 3.5; adj. r2 = 0.40, < 0.001). In 2010: January to December with TStem > 0°C (stem [CO2] = 0.18TStem + 3.0; adj. r2 = 0.84, < 0.001), and January to December with TStem < 0°C (stem [CO2] = 0.09TStem + 3.7; adj. r2 = 0.33, < 0.001).

Download figure to PowerPoint

With a linear regression model, we tested which combinations of two factors (out of our continuous long-term measurements: TStem, TSoil, DR and sap flow) explained the variation in daily mean stem [CO2] best (Table 2). For almost all phases (except wood growth 2009, dormancy >°C 2010), TStem or TSoil were the only or the main explanatory factors.

Table 2. Standardized beta coefficients and performance of general linear models (GLM) to explain mean daily stem CO2 concentrations by mean daily TStem, TSoil, DR and sap flow rates for different tree physiological phases (see text for definitions) for Norway spruce (Picea abies) trees in the years 2009 and 2010
Period yearDays T Stem T Soil DRSap flow r 2 inline image TStem
  1. r2, coefficient of determination of the best GLM; r2maxTStem, coefficient of determination derived from rmax of TStem given in Table 1 (i.e. including time lags); ‘–’, not included in the GLM.

  2. *, Significant at < 0.001.

Spring 200990–1471.46–0.510.96*0.94*
Wood growth 2009147–2220.400.700.86*0.46*
Late summer 2009222–2820.920.420.96*0.79*
Dormancy > 0°C 2009282–3460.98−0.200.85*0.86*
Dormancy < 0°C 2009/10346–770.40−0.670.61*0.18*
Dormancy > 0°C 201077–1100.310.860.75*0.03
Spring 2010110–1651.24−0.600.77*0.72*
Wood growth 2010165–2281.14−0.380.92*0.94*
Late summer 2010228–288−0.291.260.96*0.90*
Dormancy > 0°C 2010288–3350.78−0.190.86*0.86*
Dormancy < 0°C 2010335–3650.73−0.340.68*0.56*

Stem [CO2] in relation to stem radius changes and sap flow

Stem growth rates, and therefore DR dynamics, were different between the two observation years (Fig. 1b). The year 2009 was characterized by an early start of the vegetation period because of extraordinarily high temperatures in April and May (Fig. 2a) with plenty of melt water available to the trees (Fig. 2c,d). Thus, in 2009, most of the cambial activity took place during June and July (wood growth period; Fig. 1b). By contrast, the winter of 2009–2010 was extraordinarily dry (Fig. 2c), while the wood growth period of 2010 was characterized by large amounts of rainfall, low PPFD, and relatively low temperatures (Fig. 2). These climatic conditions led to low sap flow rates (Fig. 1c), frequently reduced cambial activity and thus reduced growth (Fig. 1b). Thus, the wood growth period started nearly 1 month later in 2010 than that in 2009, which is indicated by DR crossing the zero line in Fig. 1b. However, DR showed no consistent relationship with stem [CO2] (Table 1). For example, during the most active period for stem physiological processes, the wood growth period, a very close positive relationship between DR and stem [CO2] was found in 2009, but a negative relationship was found in 2010 (Table 1).

Eliminating the overall seasonal growth trend from the time series of DR and stem [CO2] and thus strengthening the short-term, water-related information content of DR, led to a consistent negative correlation between stem [CO2]detrended and DRdetrended with a time lag of 1–2 d (except late summer 2009 and dormancy < 0°C 2009–2010; see Table 1), suggesting increasing stem [CO2] with shrinking stems. In addition, daily means of stem [CO2] were positively related to daily means of sap flow with a time lag of 1–3 d (Table 1), with highest correlations for the spring rehydration periods before cambial activity started.

When testing multiple variables for explaining the seasonal course of stem [CO2], DR improved the multiple regression model (Table 2) in combination with temperature during wood growth 2009, late summer 2009 and 2010, and dormancy > 0°C late 2010. Sap flow appeared as a second (negative) factor beside TStem in only two cases, but with marginal improvements of the respective models only (r2max TStem in Table 2).

Stem [CO2] and soil efflux EFSoil

Daily mean stem [CO2] and daily mean EFSoil showed consistent seasonal patterns and synchronous day-to-day fluctuations (Fig. 1a,d). TSoil (> 1°C) as a potential driver for EFSoil was closely linearly related to TStem (TStem = 1.3*TSoil + 0.4; adj. r2 = 0.92, < 0.001). The linear temperature dependency of stem [CO2] on TStem (Fig. 3) and the exponential temperature dependency of EFSoil on TSoil (data not shown; see Fig. 4 in Etzold et al., 2011) resulted in close exponential relationships of the form ‘stem [CO2] = a − b * exp(−c × EFSoil)’ for both years (Fig. 4, adj. r2 > 0.8).

image

Figure 4. Relationships between daily mean stem CO2 concentrations (stem [CO2], average of both Norway spruce (Picea abies) trees) and daily mean soil efflux rates (EFSoil) for four different stem temperature classes of the years (a) 2009 and (b) 2010 (2009: stem [CO2] = 12.36–10.37e−0.12EFSoil; adj. r2 = 0.84, < 0.001. 2010: stem [CO2] = 7.53–7.53e−0.32EFSoil; adj. r2 = 0.83, < 0.001). The nonlinear models include all days of the respective year for which EFSoil was measured. The period from April to June 2009 (stem [CO2] = 1.33EFSoil −2.49; adj. r2 = 0.72, < 0.001) is displayed separately and is not included in the model.

Download figure to PowerPoint

Confounding effects of temperature on stem [CO2] and sap [CO2]

In order to amplify diurnal patterns, we selected sunny and warm summer days with a distinct diurnal cycle of tree physiological activity, which were defined by a mean day-time TAir above 10°C and PPFD above 800 μmol m−2 s−1 (6:00–18:00 h; = 210). At this time-scale, stem [CO2] dynamics were affected by physical equilibrium processes between water and gas-filled spaces in the wood owing to temperature changes according to Henry's law, as shown by the relationship between stem [CO2] and model stem [CO2]sap=3.5 (adj. r2 = 0.49; < 0.001). In general, the diurnal cycle of stem [CO2] lagged 3 h behind TStem and stem [CO2]sap=3.5 (= 0.96, < 0.001, Fig. 5a,b). The comparison of sap [CO2] calculated from measured stem [CO2] to that calculated from constant stem [CO2] of 4 vol% showed only small deviations (adj. r2 = 0.85, < 0.001, Fig. 5b).

image

Figure 5. Normalized diurnal courses of stem CO2 concentrations (stem [CO2]) and related measures of Norway spruce (Picea abies). Displayed are absolute values (a–c, f), rates of change () over 30-min time-intervals for stem [CO2]), sap [CO2], TStem and DR as well as fluxes of sap flow and soil efflux (EFSoil) (d–f). Values are normalized by the respective daily maxima. Only sunny days with a mean daytime (06:00–18:00 h) PPFD > 800 μmol m−2 s−1 (n = 210 d) are included. Means ± 1 SE are given. (a) Measured stem [CO2] and the corresponding model calculation with a constant [CO2] in the sap of 3.5 mM (stem [CO2]sap=3.5) according to Henry's law (with pH = 5.6). (b) Stem temperature (TStem), [CO2] in the sap calculated with measured stem [CO2] (sap [CO2]) as well as with a constant stem [CO2] of 4 vol% (sap [CO2]stem=4%; see the Materials and Methods section for details). (c) Stem radius changes (DR). (d) Rate of change of stem [CO2] ( stem [CO2]). (e) Rate of change of sap [CO2] ( sap [CO2]) and TStem (∆TStem). (f) Rate of changes in DR (DR), sap flow and soil respiration (EFSoil).

Download figure to PowerPoint

However, the physical temperature effect could only explain about half of the diurnal courses of stem and sap [CO2] (Fig. 5a,b), leaving the remaining explanatory power to other tree physiological processes, for example sap flow and stem radius changes. Deviations between the measured and the modelled stem and sap [CO2] appeared most prominent during the onset (8:00–10:00h) and cessation (17:00–20:00h) of sap flow (compare Fig. 5a,b with Fig. 5f). During the onset of sap flow in the morning, all processes measured distinctly changed their dynamics, in agreement with the fast-changing stem and sap [CO2] (Fig. 5d,e): sap flow showed the sharpest increase (Fig. 5f), TStem reached its largest positive rate of change (Δ TStem, Fig. 5e), EFSoil showed a small local efflux peak (Fig. 5f) and the stem radius experienced its fastest contraction per hour of the entire day (ΔDR, Fig. 5f). During the day, as long as sap flow was present until c. 20:00 h (Fig. 5f), stem [CO2] increased slower than stem [CO2]sap=3.5 (Fig. 5a). During the night, stem [CO2] was constant, while stem [CO2]sap=3.5 decreased (Fig. 5a,d) and sap [CO2] increased faster compared with sap[CO2]stem=3.5 (Fig. 5b,e), indicating an accumulation of CO2.

Stem [CO2] and stem efflux (EFStem)

EFStem normalized to 10°C (EFStem.10) followed a similar course to that of stem [CO2], both lagging behind TStem for c. 2–4 h (Fig. 6). However, stem [CO2] could only explain 47% of the daily variations in EFStem (= 0.005), raising the question of whether EFStem is a representative measure of local stem respiration or whether EFStem is confounded by the translocation of CO2 via sap flow. In order to quantify the potential effect of sap flow on EFStem, a theoretical EFStem rate was estimated from temperature response functions established during night under low sap flow rates that were then extrapolated to daylight conditions. Comparing the predicted vs the measured EFStem (dEFStem = measured EFStem – predicted EFStem) suggested that measured EFStem was smaller than expected from sole temperature response functions in most of the cases (dEFStem < 0; Fig. 7), indicating a dilution effect/net CO2 export by sap flow of −1.1 to −1.7 μmol m−2 s−1 (i.e. slope of the regressions during the wood growth periods). However, this was only true for the wood growth period, as during late summer sap flow showed no effect on EFStem.

image

Figure 6. Mean diurnal courses of (a) stem temperature (TStem), stem CO2 efflux (EFStem) and stem CO2 efflux corrected for a reference temperature of 10°C (EFStem.10) of four Norway spruce (Picea abies) trees. Data were collected during five measurement campaigns in summer 2008 and 2009. (b) Mean diurnal courses of stem [CO2] and sap [CO2], measured in two of the four sampling trees during the measurement campaigns in 2009 (trees 296 and 148). Values are normalized by the respective daily maximum. Two-hourly means ± 1 SE are given.

Download figure to PowerPoint

image

Figure 7. Differences between predicted and measured stem efflux rates dEFStem (measured EFStem – predicted EFStem) in relation to relative sap flow rates. Data were collected on four Norway spruce (Picea abies) trees during five measurement campaigns in the wood growth and late summer periods (as defined in the Materials and Methods section) in the years 2008 and 2009. Predicted EFStem rates were derived from temperature response functions, established during night (18:00–08:00 h) under low sap flow rates and extrapolated to daylight conditions (see the Materials and Methods section for details). Relative sap flow values refer to the maximum sap flow rate of the respective tree. Regression lines for data collected in the wood growth period (solid line) and in late summer (tinted dashed line) are given. Adj. r2 and slope of the regression line for the wood growth period are shown (for all four trees: < 0.001). None of the regressions for the late summer period were significant.

Download figure to PowerPoint

Discussion

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

Continuously measured stem CO2 concentrations (stem [CO2]) in Norway spruce at a subalpine site varied between 1 vol% and 10 vol% within the study period of 19 months (Fig. 1), which is consistent with previous reports on other tree species (Teskey et al., 2008). These concentrations are amazingly high compared with the CO2 concentration in air (0.039 vol%), indicating a significant accumulation of CO2 inside tree stems owing to high diffusion barriers (Sorz & Hietz, 2006; Sevanto et al., 2011; Steppe et al., 2012). Stem [CO2] exhibited a strong seasonality (Fig. 1, Table 1), similar to the only other long-term study we are aware of (Eklund, 1990; Eklund & Klintborg, 2000). Based on our long-term data set, we had the unique opportunity not only to quantify variations in stem [CO2] at different time scales (see later), but also to show the high consistency of relationships between stem [CO2] and its potential drivers over almost 2 yrs (Tables 1, 2). In particular, TStem and TSoil were strong driving factors for stem [CO2] in both years. Furthermore, we found tight relationships between stem [CO2] and EFSoil during both vegetation periods, while other tree physiological measures (e.g. DR, sap flow) affected stem [CO2] differently for different seasons in both years, helping us to identify possible cause–effect mechanisms.

Stem [CO2] and its dependency on Henry's law

Henry's law describes the equilibrium process between all forms of CO2 dissolved in water and CO2 in air in relation to temperature, pH and partial pressure (Teskey & McGuire, 2002; Saveyn et al., 2007a; Teskey et al., 2008). The equilibrium dynamics driven by temperature changes have been found to play an important role when interpreting stem [CO2] on seasonal (Fig. 1a) and diurnal scales (Fig. 5a, b). We found that 31% (2009) and 68% (2010) of the seasonal variability of stem [CO2] (Fig. 1a), and 49% of the diurnal variability during summer (Fig. 5a), were attributable to physical effects of temperature changes (assuming constant CO2 concentration and sap pH). Thus, equilibrium processes according to Henry's law need to be accounted for before CO2 sink and source processes within a stem segment can be identified and interpreted in terms of tree physiology.

Although sap pH is known to have a substantial effect on the dissolution of CO2 in water (Teskey et al., 2008), we did not consider potential temporal variations of pH when calculating a model stem [CO2]. However, seasonal and diurnal changes of sap pH in poplar trees were reported to be rather small, probably indicating a minor impact on changes of solubility of CO2 in water over time (Aubrey et al., 2011).

Stem [CO2] on a seasonal time-scale

In spring and during both wood growth phases, measured changes in stem [CO2] were distinctly larger than expected from Henry's law, that is, 80% larger during the 2009 wood growth period and 35% larger than the modelled stem [CO2]sap=3.5 in 2010. This is a clear indication that during these periods considerable amounts of CO2 were produced, either because of local cambial respiratory activity or it was imported from belowground (Teskey et al., 2008). Throughout most of the physiological phases during the 19 months of study, we found stem [CO2] to be most closely related to TStem (Tables 1, 2). However, during the 2009 wood growth phase and late summer 2009 and 2010, TSoil had a higher explanatory power than TStem (Table 1), indicating a potentially large influence of root-respired CO2 transported up the tree. This coincides with root growth of Norway spruce that usually peaks in late spring and again during autumn when stem growth has already ceased (Puhe, 2003; Davidson et al., 2006).

An additional indirect indication for a coupling of belowground and aboveground CO2 is the close relationship between stem [CO2] and the efflux of CO2 from the soil (EFSoil; Figs 1, 4). This relationship indicates either a coincidence of independent processes driven by closely related courses of stem and soil temperatures or a possible causality between them. EFSoil was found to be exponentially related to TSoil (Etzold et al., 2011), whereas stem [CO2] was linearly related to TStem (Fig. 3). Assuming stem [CO2] is purely a function of local stem respiration, we would expect an exponential relationship to TStem and not a linear one. Furthermore, we would not expect a close relationship to EFSoil. However, assuming a 50% or higher fraction of root-respired CO2 transported via the xylem, we would expect a high correlation between EFSoil and stem [CO2], which was actually the case (adj. r2 = 0.83). Thus, it seems to be most likely that a large amount of CO2 produced in the roots is transported up the tree. Low sap flow rates seemed to increase stem [CO2] (Fig. 5) and EFStem (Fig. 7), whereas high sap flow rates seemed to induce a dilution effect and therefore a decrease in stem [CO2] (Fig. 7). Belowground growth phenology might not only affect changing explanatory powers of TStem and TSoil on stem [CO2] during the year but also pinpoint to belowground sources of stem CO2 during wood growth and late summer phases.

By contrast, during dormancy (> 0°C), sap flow was low (Fig. 1) and little CO2 transport from belowground can be expected. Thus, local stem respiration should be the dominant source for stem [CO2] during this period. However, during these dormant periods, TSoil still played an important role in explaining stem [CO2], either as single factor (Table 1) or in combination with TStem (only 2010; Table 2). At these times, TStem and TSoil were less coupled than during the rest of the year as soils were insulated by a layer of snow, keeping TSoil almost constant at c. 0°C, whereas TStem followed TAir (Fig. 2a). Thus, in addition to local respiration, we hypothesize that CO2 respired from the roots under the snow might diffuse up the stem along gas-filled spaces. With a diffusion coefficient of CO2 in air of 1.6 × 10−5 m2 s−1 (Nobel, 1991), changes in [CO2] in the roots could theoretically affect the stem section measured via air-filled spaces in the wood within 2–3 d. According to Sachs (1887), gas makes up to 17% of the xylem volume of fir trees (and we assume a similar value for spruce) and could build the necessary pathway for CO2 diffusion. Thus, the origin of stem [CO2] during dormancy phases with temperatures > 0°C could be either root- or local stem-respired CO2.

Furthermore, during dormancy with freezing conditions, large deviations between stem [CO2] of the two trees were observed (Fig. 1a). During these phases, temperature dropped below the threshold for frost-induced shrinkage of bark. Such conditions led to a (reversible) decrease in bark thickness of 1 mm or more (Fig. 1b), whereas xylem cells are reported to remain largely unaffected in size (Zweifel & Häsler, 2000). Such rapid and large frost-/freeze-induced variations in DR are known to induce pressure changes inside the tissues (Robson & Petty, 1993; Sevanto et al., 2012), which might be responsible for changes in stem [CO2]. These freeze/thaw processes may also explain the observed deviations in stem [CO2] between the two trees (Fig. 1a) and the differences in stem [CO2] levels, as it is well known that such processes do not occur simultaneously in all stems (Zweifel & Häsler, 2000). Overall, during winter stem [CO2] seems to be dominated by local respiration, but also affected by physically induced CO2 concentration changes, such as pressure changes and/or diffusion processes.

Nonrecurring stem [CO2] peak during wood growth phase 2009

In the year 2009, TStem and DR could explain most of the variability of stem [CO2], during the wood growth phase, while in the year 2010, TStem and sap flow were the most important predictors (Table 2). Although this is partly explicable with different growth dynamics and thus different respiration rates in the 2 yrs, it does not satisfactorily explain the exceptional peak of stem [CO2] from April to June 2009. This period appears as an anomaly in all our analyses (Tables 1, 2, Figs 1, 3, 4) and could be induced by a physiological response to sensor installation, as the hole drilled into the stem to place the sensor was 2 cm diameter and 6 cm deep. Increased stem CO2 efflux as wound response is a well-described phenomenon and usually interpreted as increased respiration owing to healing processes, such as cell repair, callus formation, lignification and suberization (Bloch, 1941; Uritani & Asahi, 1980; Schmitt & Liese, 1993; Levy et al., 1999). The fact that the stem [CO2] peak occurred 3 months after installation might indicate a delayed wound response after tree dormancy ceased and radial stem growth started at the end of May (Dunn et al., 1990; Schmitt & Liese, 1993; Dujesiefken et al., 2005).

Furthermore, annual amplitudes of stem [CO2] signals seem to have decreased over the two study years (mean amplitude ± standard deviation of stem [CO2] within a 7-d window during vegetation period: 2009, 0.51 ± 0.50; 2010, 0.45 ± 0.26). However, data sets even longer than in this study (19 months) would be needed to reliably test potential sensor signal degradation, which might occur because of wound closure and thus increased diffusion resistances between sensor and measurement tissue of interest.

Stem CO2 on a diurnal time-scale

On a diurnal scale, stem [CO2] in the morning (8:00–10:00 h) was found to increase more slowly than expected from Henry's law (Fig. 5a, only sunny summer days were analysed). One explanation could be that sap flow transports water with lower [CO2] into the measured stem section (Teskey & McGuire, 2007), thus acting as a dilution factor in the first hours after dawn (Fig. 5). Furthermore, water was withdrawn from the bark (as shown by shrinking stems, Fig. 5c,f), in parallel with the onset of sap flow. As bark water is reported to contain much lower [CO2] than xylem water (Cernusak & Marshall, 2000; Wittmann et al., 2006), this could be a second dilution factor.

During the second half of the day, the increase in stem [CO2] was also slightly lower than expected according to Henry's law (Fig. 5a), meaning that stem CO2 could have been diluted by sap flow, as also indicated by the analyses of sap flow in combination with EFStem (Fig. 7) and the findings of Teskey & McGuire (2007). With decreasing sap flow in the early evening, stem [CO2] remained constant, while modelled stem [CO2]sap=3.5 decreased, suggesting an accumulation of locally respired CO2 under low sap flow rates during night. The different effects of sap flow rates on EFStem in growth- and non-growth periods (Fig. 7) might have to do with the generally higher stem [CO2] during the growth period (Fig. 1). During this time, the stem (growth) respiration is assumed to be highest (Ryan, 1990) and the difference between locally produced CO2 and imported CO2 from belowground might be biggest.

Conclusions

Stem [CO2] was strongly related to stem and soil temperatures, which in turn affect physical equilibrium processes within the stem according to Henry's law but also affect different respiration processes, making disentangling of single factors difficult. As the relationships of stem [CO2] to tree physiological measures changed over time, long-term data sets are critical for the interpretation of stem [CO2] dynamics. Clear indications were found for CO2 translocation within the tree. However, whether these observations resulted from sap flow or axial diffusion, or a combination of both, remains unclear. Nevertheless, although the fate of stem [CO2] is not yet known, it has huge implications on the partitioning of net ecosystem CO2 exchange fluxes. If a large fraction of sap CO2 is recycled internally, then gross primary productivity estimated based on light response curves as well as ecosystem respiration estimated based on temperatures will both be underestimated. Furthermore, as the potential for refixation of root-respired and translocated CO2 changes strongly with phenophases, the underestimates might change during the year.

Acknowledgements

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

We are grateful to Rudolf Häsler, Johannes P. Böhm, Arnold Streule and Peter Plüss for long-term technical support and for maintenance at this site. We thank Kathy Steppe, John Marshall and anonymous reviewers for their valuable inputs, the Long-term Forest Ecosystem Research (LWF), the Swiss Air Quality Monitoring Network (NABEL) of FOEN, SLF and Empa for providing data. This research project was partly funded by the ETH Zurich, Switzerland (ETH research grant TH-10 06-2).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Aubrey DP, Boyles JG, Krysinsky LS, Teskey RO. 2011. Spatial and temporal patterns of xylem sap pH derived from stems and twigs of Populus deltoides L. Environmental and Experimental Botany 71: 376381.
  • Aubrey DP, Teskey RO. 2009. Root-derived CO2 efflux via xylem stream rivals soil CO2 efflux. New Phytologist 184: 3540.
  • Bloch R. 1941. Wound healing in higher plants. Botanical Review 7: 110146.
  • Bowman WP, Barbour MM, Turnbull MH, Tissue DT, Whitehead D, Griffin KL. 2005. Sap flow rates and sapwood density are critical factors in within- and between-tree variation in CO2 efflux from stems of mature Dacrydium cupressinum trees. New Phytologist 167: 815828.
  • Carter MC, Larsen HS. 1965. Soil nutrients and loblolly pine xylem sap compositions. Forest Science 11: 216220.
  • Cerasoli S, McGuire MA, Faria J, Mourato M, Schmidt M, Pereira JS, Chaves MM, Teskey RO. 2009. CO2 efflux, CO2 concentration and photosynthetic refixation in stems of Eucalyptus globulus (Labill.). Journal of Experimental Botany 60: 99105.
  • Cernusak LA, Marshall JD. 2000. Photosynthetic refixation in branches of Western White Pine. Functional Ecology 14: 300311.
  • Ceschia E, Damesin C, Lebaube S, Pontailler JY, Dufrene E. 2002. Spatial and seasonal variations in stem respiration of beech trees (Fagus sylvatica). Annals of Forest Science 59: 801812.
  • Damesin C, Ceschia E, Le Goff N, Ottorini JM, Dufrêne E. 2002. Stem and branch respiration of beech: from tree measurements to estimations at the stand level. New Phytologist 153: 159172.
  • Daudet FA, Ameglio T, Cochard H, Archilla O, Lacointe A. 2005. Experimental analysis of the role of water and carbon in tree stem diameter variations. Journal of Experimental Botany 56: 135144.
  • Davidson EA, Richardson AD, Savage KE, Hollinger DY. 2006. A distinct seasonal pattern of the ratio of soil respiration to total ecosystem respiration in a spruce-dominated forest. Global Change Biology 12: 230239.
  • Dima E, Manetas Y, Psaras GK. 2006. Chlorophyll distribution pattern in inner stem tissues: evidence from epifluorescence microscopy and reflectance measurements in 20 woody species. Trees – Structure and Function 20: 515521.
  • Dujesiefken D, Liese W, Shortle W, Minocha R. 2005. Response of beech and oaks to wounds made at different times of the year. European Journal of Forest Research 124: 113117.
  • Dunn JP, Potter DA, Kimmerer TW. 1990. Carbohydrate reserves, radial growth, and mechanisms of resistance of oak trees to phloem-boring insects. Oecologia 83: 458468.
  • Eklund L. 1990. Endogenous levels of oxygen, carbon dioxide and ethylene in stems of Norway spruce trees during one growing season. Trees – Structure and Function 4: 150154.
  • Eklund L, Klintborg A. 2000. Ethylene, oxygen and carbon dioxide in woody stems during growth and dormancy. In: Savidge R, Barnett JR, Napier R, eds. Cell and molecular biology of wood formation. Edinburgh, UK: Heriot–Watt University, 4356.
  • Etzold S, Ruehr NK, Zweifel R, Dobbertin M, Zingg A, Pluess P, Häsler R, Eugster W, Buchmann N. 2011. The carbon balance of two contrasting mountain forest ecosystems in Switzerland: similar annual trends, but seasonal differences. Ecosystems 14: 12891309.
  • Gansert D, Burgdorf M. 2005. Effects of xylem sap flow on carbon dioxide efflux from stems of birch (Betula pendula Roth). Flora 200: 444455.
  • Granier A. 1985. Une nouvelle méthode pour la mesure du flux de sève brute dans le tronc des arbres. Annales des Sciences Forestieres 42: 193200.
  • IUSS Working Group WRB. 2007. World reference base for soil resources 2006. In: World Soil Resources Reports. Rome, Italy: FAO.
  • Jörg S. 2008. Böden im Seehornwald bei Davos und deren Vorrat an Kohlenstoff und Stickstoff. Zürcher Hochschule für Angewandte Wissenschaften, Wädenswil, and WSL, Birmensdorf.
  • Levy PE, Meir P, Allen SJ, Jarvis PG. 1999. The effect of aqueous transport of CO2 in xylem sap on gas exchange in woody plants. Tree Physiology 19: 5358.
  • Maier CA, Clinton BD. 2006. Relationship between stem CO2 efflux, stem sap velocity and xylem CO2 concentration in young lobolly pine trees. Plant, Cell & Environment 29: 14711483.
  • McGuire MA, Cerasoli S, Teskey RO. 2007. CO2 fluxes and respiration of branch segments (Platanus occidentalis L.) examined at different velocities, branch diameters, and temperatures. Journal of Experimental Botany 58: 21592168.
  • McGuire MA, Marshall JD, Teskey RO. 2009. Assimilation of xylem-transported 13C-labelled CO2 in leaves and branches of sycamore (Platanus occidentalis L.). Journal of Experimental Botany 60: 38093817.
  • McGuire MA, Teskey RO. 2002. Microelectrode technique for in situ measurement of carbon dioxide concentrations in xylem sap of trees. Tree Physiology 22: 807811.
  • McGuire MA, Teskey RO. 2004. Estimating stem respiration in trees by a mass balance approach that accounts for internal and external fluxes of CO2. Tree Physiology 24: 571578.
  • Moore DJP, Gonzalez-Meler MA, Taneva L, Pippen JS, Kim H-S, DeLucia EH. 2008. The effect of carbon dioxide enrichment on apparent stem respiration from Pinus taeda L. is confounded by high levels of soil carbon dioxide. Oecologia 158: 110.
  • Nobel PS. 1991. Physicochemical and environmental plant physiology. London, UK: Academic Press.
  • Pruyn ML, Gartner BL, Harmon ME. 2005. Storage versus substrate limitation to bole respiratory potential in two coniferous tree species of contrasting sapwood width. Journal of Experimental Botany 56: 26372649.
  • Puhe J. 2003. Growth and development of the root system of Norway spruce (Picea abies) in forest stands – a review. Forest Ecology and Management 175: 253273.
  • R Development Core Team. 2012. R: a language and environmental for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.
  • Robson DJ, Petty JA. 1993. A proposed mechanism of freezing and thawing in conifer xylem. In: Borghetti M, Grace J, Raschi A, eds. Water transport in plants under climatic stress. Cambridge, UK: Cambridge University Press, 7585.
  • Ryan MG. 1990. Growth and maintenance respiration in stems of Pinus-Contorta and Picea-Engelmannii. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 20: 4857.
  • Sachs J. 1887. Lectures on the physiology of plants. Oxford, UK: Clarendon.
  • Saveyn A, Steppe K, Lemeur R. 2007a. Daytime depression in tree stem CO2 efflux rates: is it caused by low stem turgor pressure? Annals of Botany 99: 477485.
  • Saveyn A, Steppe K, Lemeur R. 2007b. Drought and the diurnal patterns of stem CO2 efflux and xylem CO2 concentration in young oak (Quercus robur). Tree Physiology 27: 365374.
  • Saveyn A, Steppe K, Lemeur R. 2008a. Report on non-temperature related variations in CO2 efflux rates from young tree stems in the dormant season. Trees – Structure and Function 22: 165174.
  • Saveyn A, Steppe K, McGuire MA, Lemeur R, Teskey RO. 2008b. Stem respiration and carbon dioxide efflux of young Populus deltoides trees in relation to temperature and xylem carbon dioxide concentration. Oecologia 154: 637649.
  • Schmitt U, Liese W. 1993. Response of xylem parenchyma by suberization in some hardwoods after mechanical injury. Trees 8: 2330.
  • Sevanto S, Holbrook NM, Ball MC. 2012. Freeze/thaw-induced embolism: probability of critical bubble formation depends on speed of ice formation. Frontiers in Plant Science 3: 107.
  • Sevanto S, Holtta T, Holbrook NM. 2011. Effects of the hydraulic coupling between xylem and phloem on diurnal phloem diameter variation. Plant, Cell & Environment 34: 690703.
  • Sorz J, Hietz P. 2006. Gas diffusion through wood: implications for oxygen supply. Trees – Structure and Function 20: 3441.
  • Steppe K, Cochard H, Lacointe A, Ameglio T. 2012. Could rapid diameter changes be facilitated by a variable hydraulic conductance? Plant, Cell & Environment 35: 150157.
  • Steppe K, De Pauw DJW, Lemeur R, Vanrolleghem PA. 2006. A mathematical model linking tree sap flow dynamics to daily stem diameter fluctuations and radial stem growth. Tree Physiology 26: 257273.
  • Steppe K, Saveyn A, McGuire MA, Lemeur R, Teskey RO. 2007. Resistance to radial CO2 diffusion contributes to between-tree variation in CO2 efflux of Populus deltoides stems. Functional Plant Biology 34: 785792.
  • Stockfors J, Linder S. 1998. Effect of nitrogen on the seasonal course of growth and maintenance respiration in stems of Norway spruce trees. Tree Physiology 18: 155166.
  • Teskey RO, McGuire MA. 2002. Carbon dioxide transport in xylem causes errors in estimation of rates of respiration in stems and branches of trees. Plant, Cell & Environment 25: 15711577.
  • Teskey RO, McGuire MA. 2005. CO2 transported in xylem sap affects CO2 efflux from Liquidambar styraciflua and Platanus occidentalis stems, and contributes to observed wound respiration phenomena. Trees – Structure and Function 19: 357362.
  • Teskey RO, McGuire MA. 2007. Measurement of stem respiration of sycamore (Platanus occidentalis L.) trees involves internal and external fluxes of CO2 and possible transport of CO2 from roots. Plant, Cell & Environment 30: 570579.
  • Teskey RO, Saveyn A, Steppe K, McGuire MA. 2008. Origin, fate and significance of CO2 in tree stems. New Phytologist 177: 1732.
  • Trumbore S. 2006. Carbon respired by terrestrial ecosystems – recent progress and challenges. Global Change Biology 12: 141153.
  • Ubierna N, Kumar AS, Cernusak LA, Pangle RE, Gag PJ, Marshall JD. 2009. Storage and transpiration have negligible effects on 13C of stem CO2 efflux in large conifer trees. Tree Physiology 29: 15631574.
  • Uritani I, Asahi T. 1980. Respiration and related metabolic activity in wounded and infected tissues. New York, USA & London, UK: Academic Press.
  • Wittmann C, Pfanz H, Loreto F, Centritto M, Pietrini F, Alessio G. 2006. Stem CO2 release under illumination: corticular photosynthesis, photorespiration or inhibition of mitochondrial respiration? Plant, Cell & Environment 29: 11491158.
  • Zweifel R, Eugster W, Etzold S, Dobbertin M, Buchmann N, Häsler R. 2010. Link between continuous stem radius changes and net ecosystem productivity of a subalpine Norway spruce forest in the Swiss Alps. New Phytologist 187: 819830.
  • Zweifel R, Häsler R. 2000. Frost-induced reversible shrinkage of bark of mature, subalpine conifers. Agricultural and Forest Meteorology 102: 213222.
  • Zweifel R, Zeugin F, Zimmermann L, Newbery DM. 2006. Intra-annual radial growth and water relations of trees – implications towards a growth mechanism. Journal of Experimental Botany 57: 14451459.