What's the flux? Unraveling how CO2 fluxes from trees reflect underlying physiological processes


(Author for correspondence: tel +49 3641 576164; email trumbore@bgc-jena.mpg.de)

Tree stems and branches emit carbon dioxide (CO2) at rates that per unit area can rival emissions from leaves or the soil surface and summed over a forest stand can comprise 14–30% of the total CO2 efflux (Chambers et al., 2004; Ryan et al., 2009). Stem CO2 fluxes have predictable patterns of variation with growth rate, stand age, and elevation (Chambers et al., 2004; Ryan et al., 2009; Robertson et al., 2010). Over the past decade observations of diel covariation of CO2 efflux with sapflux rates measured in tree stems have led to the conclusion that internal transport of CO2 within the stem strongly influences the measured CO2 efflux at the surface (Teskey et al., 2008). In this issue of New Phytologist, Bloemen et al. (pp. 555–565) report on a tracer experiment that demonstrates not only upward transport of 13CO2 added to the transpiration stream, and emission of this label along the stem, but also fixation of a significant fraction of the added CO2 in canopy branches, petioles and, to a minor extent, leaves. The study of Bloemen et al. adds to the growing literature that demonstrates the utility of isotope labeling studies to understand allocation and carbon (C) cycling in trees (Powers & Marshall, 2011; Epron et al., 2012).

Dynamic approaches for measuring continuous diurnal CO2 fluxes and transport in the transpiration stream need to be more widely applied.’

Processes influencing stem CO2 efflux

A number of factors can influence the efflux of CO2 measured by a flux chamber covering a segment of tree stem (Fig. 1). The cambium is the site of formation of new tissue, that is, of growth, while maintenance respiration produces CO2 in all living tissues. The C being respired may derive from recent photosynthetic products transported in the phloem (e.g. Powers & Marshall, 2011) and from storage reserves. The pathways for respiration may vary with time or tree species: recently 18O/16O measurements in oxygen (O2) provided the first evidence for the alternative oxidase pathway contributing to respiration in some tree stems (Angert et al., 2012a). CO2 may also be locally fixed by photosynthetic tissues found under the bark before it is lost to the atmosphere.

Figure 1.

Sources, sinks and transport processes in tree stems result in the net CO2 flux observed when a chamber encloses part of the stem surface (modified after Teskey et al., 2008). The degree to which the stem CO2 efflux records the underlying process of respiration (minus any fixation) depends on the saturation state and rate of transport of the xylem water. This balance can change from day to night and with altered temperature as sun warms the outer stem. Inset: Relationship between CO2 efflux (inline image) and O2 influx (inline image) in the cases of net soluble C transport into or out of the volume covered by a stem respiration chamber. At times of low transport (e.g. at night), the two fluxes should be approximately equal. Note that, if respiration is the main process, a local imbalance in inline image: inline image means that there must be a compensating imbalance elsewhere – e.g. if CO2 is transported to the canopy and emitted without consumption of O2.

Low rates of diffusion, especially across the cambium, can cause high CO2 concentrations in stems, and internal O2 concentrations can drop to very low levels (Spicer & Holbrook, 2005; Teskey et al., 2008). CO2 is highly soluble, and will dissolve in (or exsolve from) stem water, depending on local saturation conditions, which in turn are controlled by factors such as temperature and pH. Uptake of CO2 directly from the soil atmosphere, once thought potentially important, has largely been shown to be minor (see summary in Bloemen et al.). Hence the source of CO2 emitted to the atmosphere from the bark surface can reflect a combination of local growth and maintenance respiration, other local processes producing CO2 (including potentially decomposition in heartwood) or CO2 from respiration in other tissues (e.g. roots) that has been transported into the volume beneath a chamber in solution. However, there can also be net export in the xylem water stream, as indicated by the fate of the tracer added by Bloemen et al. The measured chamber flux at any given time is thus the complex result of transport in, transport out and respiration minus photosynthesis in local tissues. Use of a dark chamber will exclude local photosynthesis.

Observations of a relationship between sapflux and CO2 efflux provide a clue as to whether CO2 is net imported or exported from the volume of stem under a chamber attached to the stem surface (see Fig. 1, modified from Teskey et al., 2008). Other evidence for net CO2 transport away from the region of efflux measurement comes from lower-than-expected efflux rates compared with what is expected given the construction costs of wood (Ryan et al., 2009), and potentially from higher efflux rates in canopy branches (Teskey et al., 2008). Changes in local temperature and/or pH can change respiration rates and also cause changes in CO2 solubility (Kunert & Mercado Cárdenas, 2012).

Stem anatomy, including bark thickness and tree hydraulics, likely influences the importance of the mechanisms and can help explain observations such as changes in CO2 efflux with stand age or tree size, or differences between similar trees growing in different environments (Ryan et al., 2009). Bloemen et al. report results from labeling Populus deltoides, the eastern cottonwood tree, which has very high transpiration rates and generally is found in riparian zones. As noted by Ubierna et al. (2009) most studies that have reported relationships between sapflux and CO2 efflux have been made in tree species with high sapflux rates and small conducting area. By contrast, the large conifer trees investigated by Ubierna et al. (2009), with lower overall sapflux, did not demonstrate such relationships, and even crown removal did not change the rates of CO2 efflux from stems they studied.

What do these results mean for interpretation of other ecosystem CO2 efflux measurements?

A major conclusion of Bloemen et al. is that the transport of the tracer from the tree base to the canopy indicates that root respiration can be a source of at least some of the CO2 emitted in the canopy. While the high CO2 concentrations at the base of trees do argue for a belowground source, Bloemen et al. did not successfully introduce enough label via roots to demonstrate definitively the transfer of root CO2 up the stem. Aubrey & Teskey (2009) have argued that up to 50% of root respired CO2 may be transported upward and diffuse out higher in the tree stem or in branches. Grossiord et al. (2012), using isotopic differences to distinguish plant and decomposition derived soil respiration, detected a day-time reduction in autotrophic respiration from soil, albeit the ‘missing’ root respiration they infer is transported up the tree stem amounted to only a 17% underestimation of the autotrophic CO2 efflux on a daily basis.

Tracer studies by Powers & Marshall (2011) as well as Bloemen et al. show that 13C-labeled CO2 added to the xylem stream indeed is transported upward, emitted and a fraction refixed in the canopy. In the Bloemen et al. study, an estimated 6–17% of the added tracer was fixed in photosynthetic tissues in branches and petioles. Hence recycling of CO2 within the plant is potentially quite important – perhaps especially so when CO2 concentrations in the atmosphere were lower than those of today (Teskey et al., 2008).

Hibberd & Quick (2002) provide an additional mechanism for internal C transport, based on the capture of CO2 by PEP-carboxylase, after which it can be removed from the site of respiration as malate. Their labeling experiments indicate that malate transported in the xylem enters the bundle-sheath cells, and can be used for photosynthesis, in this ‘C4 like’ mechanism. Bloemen et al. found that most of their labeled CO2 was fixed in branches and in leaf petioles, which agrees well with transported C being fixed in bundle-sheath cells.

How can we derive an estimate of the ‘real’ stem respiration flux?

The various effects of temperature and transpiration velocity can affect CO2 efflux rates over a day–night cycle. One way to estimate fluxes might be to choose to sample at night, when transpiration flux is near zero (Teskey & McGuire, 2002). However, this is also the coolest time of day, so this might underestimate daytime respiration in tissues (such as the cambium) that may warm significantly over the daytime period (Kunert & Mercado Cárdenas, 2012). A second method is to measure CO2 evolution or O2 uptake (Teskey & McGuire, 2002; Spicer & Holbrook, 2005) on excised wood. Apart from damaging the tree (or the tissues with heat generated on sampling), such methods must be used with care as the degassing of high CO2 in wood pores can initially yield too-high CO2 fluxes (Teskey & McGuire, 2002).

Another possibility is to use in situ O2 uptake as a measure of respiration (Angert & Sherer, 2011). Because O2 is much less soluble in water than CO2, the molar flux of O2 into stems should roughly equal that of CO2 out if transport is minimal, given the stoichiometry of the respiration substrate in most woody tissues. In cases where CO2 respired elsewhere is transported and emitted in the stem and canopy, we would expect CO2 release to exceed O2 uptake (i.e. inline image/inline image > 1). In cases where locally derived CO2 is exported, the ratio of CO2 release to O2 uptake will be ≤ 1 (Fig. 1 inset).

In Amazonian tropical forest trees, Angert et al. (2012b) found the CO2 efflux from the stems was on average only 0.66 (± 0.18) of the O2 influx, in other words about one third of the CO2 respired in the stem section beneath the chamber is not locally emitted, but transported away. Some of this ‘removed CO2’ can be carried by the xylem stream, as shown by Bloemen et al. However, the flux of CO2 that can be removed in this way is constrained by the chemistry of the carbonate system, and is governed by the xylem pH which is seldom > 7 (Teskey et al., 2008). Angert et al. (2012b) concluded, based on estimates of stem [CO2] and xylem pH, that the rate of dissolved inorganic C export might not remove CO2 at the rate required, and suggest that the C might be exported in organic (e.g. malate), rather than inorganic form.

Progress in understanding the sources and magnitudes of CO2 fluxes in tree stems is being made rapidly, and linking actively respiring tissues local to, and remote from, the point of measurement. Future studies taking advantage of pulse labeling in trees with different water-use strategies would be useful for resolving conditions where stem xylem water transport in stems significantly impact soil CO2 efflux. Radiocarbon measurements of stem CO2 can help resolve questions about whether the C being respired (and potentially translocated) derives from storage reserves vs fresh photosynthetic products. Dynamic approaches for measuring continuous diurnal CO2 fluxes and transport in the transpiration stream need to be more widely applied. New sensor methods for O2 measurement can add information that allows separation of transport from local physiological processes. Future studies should also focus on quantitative measurements of the photosynthetic fraction supported by both inorganic, and organic C, transported internally in the xylem.

Meanwhile, we need to be careful about invoking the process ‘respiration’ when really we are measuring CO2 flux. Ultimately the CO2 emitted from a stem is produced by physiological processes, but the challenge remains identifying what portion is produced by local tissues, which will facilitate much-needed mechanistic understanding of factors controlling autotrophic respiration.