Abstract
• Background and Aims Claims that submerged roots of alder and other wetland trees are aerated by pressurized gas flow generated in the stem by a light-induced thermo-osmosis have seemed inconsistent with root anatomy. Our aim was to seek a verification using physical root–stem models, stem segments with or without artificial roots, and rooted saplings.
• Methods Radial O2 loss (ROL) from roots was monitored polarographically as the gas space system of the models, and stems were pressurized artificially. ROL and internal pressurization were also measured when stems were irradiated and the xylem stream was either CO2 enriched or not. Stem photosynthesis and respiration were measured polarographically. Stem and root anatomy were examined by light and fluorescence microscopy.
• Key Results Pressurizing the models and stems to ≤10 kPa, values much higher than those reportedly generated by thermo-osmosis, created only a negligible density-induced increase in ROL, but ROL increased rapidly when ambient O2 concentrations were raised. Internal pressures rose by several kPa when shoots were exposed to high light flux and ROL increased substantially, but both were due to O2 accumulation from stem photosynthesis using internally sourced CO2. Increased stem pressures had little effect on O2 transport, which remained largely diffusive. Oxygen flux from stems in high light periods indicated a net C gain by stem photosynthesis. Chloroplasts were abundant in the secondary cortex and secondary phloem, and occurred throughout the secondary xylem rays and medulla of 3-year-old stems. Diurnal patterns of ROL, most marked when light reached submerged portions of the stem, were modified by minor variations in light flux and water level. Low root temperatures also helped improve root aeration.
• Conclusions Pressurized gas flow to submerged roots does not occur to any significant degree in alder, but stem photosynthesis, using internally sourced CO2 from respiration and the transpiration stream, may play an important role in root aeration in young trees and measurably affect the overall carbon balance of this and other species.
Keywords: Alnus glutinosa, root aeration, diffusion, pressure flow, stem photosynthesis, radial oxygen loss, thermal transpiration, thermo-osmosis
INTRODUCTION
The aeration of the belowground parts of some wetland plant species is known to be much improved by various forms of pressurized gas flow from the aboveground parts. The flows known to be most effective involve a through-flow of gases, sometimes between emergent stems via the rhizome system (Armstrong and Armstrong, 1990; Armstrong et al., 1992; Brix et al., 1992; Sorrell and Boon, 1994), or between emergent leaves via the underground stem to other emergent leaves (Dacey, 1980, 1981; Grosse and Bauch, 1991; Bendix et al., 1994), or even from one part of an emergent leaf via the underground system and back through other channels in the same petiole and leaf (Mevi-Schultz and Grosse, 1988). The major pressurizing mechanisms driving these flows have been identified as thermal transpiration (=thermo-osmosis), humidity-induced diffusion and Venturi-induced pressure differences (Dacey, 1987; Armstrong et al., 1992, 1996a, b; Grosse, 1996). The roots are not considered to be a part of the through-flow system, but they can derive obvious benefit from it, since the gas stream will usually raise the O2 concentration at the root–rhizome junction, thus increasing O2 diffusion into the roots. The removal of CO2, C2H4, CH4 and other products will also be enhanced (Brix et al., 1996).
In 1985, Grosse and Schroder reported that ‘… experiments show evidence of [in addition to diffusion] an additional manifold higher O2 supply due to a gas transport in leaved as well as leafless trees of Alnus glutinosa (L.). This gas transport is directed from the stems to the roots and is driven by a thermo-osmotic pressurization within the air space system of the stems resulting from temperature gradients up to 3·6 °C between the stem and the ambient air following absorption of light energy by the brownish pigments of the bark’. In subsequent papers, it was claimed that the O2 flow to the roots of several other wetland tree species was improved by a thermo-osmotically driven pressure flow, i.e. Taxodium distichum, Betula pubescens and Populus tremulus (Grosse et al., 1992), Alnus japonica and Alnus hirsuta (Grosse et al., 1993) and Salix alba, S. cinerea and S. viminalis (Grosse et al., 1996). However, no evidence of improved root ventilation was found in Alnus incana, Fraxinus excelsior or Acer pseudoplatanus (Schröder, 1989a), or in Acer pseudopatanus and Ilex aquifolium (Grosse et al., 1992).
For several reasons, we doubted that there could be a significant pressure flow into submerged tree roots. For example, to create a through-flow, high surface tension pressures would have to be overcome to force gas from submerged lenticels, and the internal gas pressures measured by Grosse and Schröder (1985) in the alder stem did not exceed 17 Pa. In these circumstances, we would anticipate only transitory non-through-flow convection to the roots as they are pressurized. Similarly, it is difficult to envisage how a significant pressure flow could enter submerged non-woody roots since, in ventilation terms, they are effectively blind-ended tubes (Beckett et al., 1988).
If pressure flow was not responsible for the reported improvements in the O2 supply to the roots of alder and the other species, an obvious alternative explanation could be that photosynthetically generated O2 might have been the cause. Although the brown stems of alder do not immediately prompt such an explanation, abundant chloroplasts can be found in the secondary cortex, the secondary phloem, secondary rays and the medulla of stems up to 4 years old (see below). However, if the only CO2 source for photosynthesis is that diffusing through lenticels from the atmosphere, the photosynthetically generated O2 would simply counter-diffuse away; any increase in O2 concentration in the secondary cortex could be expected to be tiny and not sufficient to have a significant influence on O2 transport to the roots (Armstrong, 1979). A marked increase in O2 concentration in the stem would require either that there should be an internal CO2 source and/or that the resistance to O2 escape should be greater than the resistance to CO2 entry. The latter occurs where leaves and stems are submerged (Waters et al., 1989). However, Stringer and Kimmera (1993) have reported significant petiolar photosynthesis in Populus deltoides dependent upon CO2 carried in the xylem stream; concentrations up to 9 mm were recorded in the xylem stream in plants from submerged soils. Hibbard and Quick (2002) have reported C4-type photosynthesis in the petioles of tobacco, with the carbon source being supplied from the vascular system and not from the stomata. Preliminary studies on willow (S. viminalis) and alder in our laboratory (Moir, 1993; Dockerty, 1994; Hughes, 1996) found that ROL from roots increased significantly only if (a) partially submerged stems were illuminated; (b) gas pressures applied to a cut emergent end of the stem were sufficient to cause bubbling below the water line (pressures >2000 Pa); or (c) the emergent stems were exposed to above ambient O2 concentrations. We also found that alder twigs fresh from the field would produce a regular bubbling from submerged lenticels and the cut end when they were partially submerged and illuminated (W. Armstrong and J. Armstrong, unpubl. res.).
We have now extended these observations and report here on (a) the pressurization generated by stem photosynthesis in the black alder, Alnus glutinosa (L.) Gaertn.; (b) the effect that stem photosynthesis has on O2 diffusion from the stems to the roots; and (c) the insignificant effect pressurization per se has on transport to submerged roots. The latter was tested using (a) a purely physical artificial root model; (b) a combined model in which alder stems were equipped with artificial roots; and (c) whole plants with submerged root systems.
BACKGROUND
Thermal transpiration (thermo-osmosis)
Thermal transpiration (thermo-osmosis) and humidity-induced diffusion and pressurization in plants require the presence of a microporous partition between the aerial atmosphere and the internal atmosphere of the plant with pores sufficiently small so as to offer more resistance to a pressure flow than to gas phase diffusion. For thermal transpiration to generate pressures within these plants, it is necessary for the internal atmosphere under the micropores to be constantly warmer than the external atmosphere (Grosse, 1996). Solar warming will tend to cause gas expansion and pressurization of the internal atmosphere. Pressure is not readily dissipated through the micropores because of their resistance (although they may be leaky to some extent) and, if the gases can escape more readily along a more distal venting path with lower resistance, the internal gas density will become less than the external atmosphere, so inducing a diffusive inflow of atmospheric gases. Provided that the solar warming is continuous, the inwardly diffusing gases will be sufficient to rise and sustain above ambient internal pressures and continuously drive a pressurized gas flow along the venting path.
In alder, Buchel and Grosse (1990) identified the phellogen under the lenticels as a microporous partition that should support thermo-osmosis: pores through the vascular cambium were not smaller than 100 nm, but the mean diameters of pores through the sub-lenticellular phellogen were as low as 14 nm. Pores as large as 3 µm diameter can be sufficient to support some pressurization (Takaishi and Sensui, 1963) and flow (Armstrong et al., 1996b), but they could be expected to be very leaky.
Original evidence underlying the supposition of a significant pressure flow to the roots of alder
The suggestion that thermal transpiration causes a thermo-osmotically driven pressure flow into the roots of alder and other wetland tree species was based largely on the results of two types of data. Both were obtained from experimental assemblies similar to that shown in Fig. 1. Six-month- to 1-year-old seedlings were used. For plants in leaf, the lower half of the woody shoot above the root collar was enclosed in an upper chamber as shown; for leafless plants, the whole of the woody shoot above the root collar was enclosed (Grosse and Schröder, 1984, 1985; Schröder, 1989a). The shoot base and the whole of the root system were in a lower chamber, the stem base and several root bases being enclosed in a gas pocket above either saturated soil or a culture solution; the remainder of the root system was submerged. The top chamber was fitted with an injection port and the lower was fitted with a sample port. Pressure gradients between the chambers and the external atmosphere were minimized using long narrow ‘breather’ tubes. Ethane was injected into the upper chamber and sampled from the head space in the lower chamber, and the experiments were carried out in the dark and with a radiant light source (a 150 W incandescent lamp) as illustrated. Grosse and Schröder (1985) and Grosse et al. (1992) measured differences in temperature (ΔT °C) between the inside of the stems and the surrounding atmosphere. Grosse and Schröder (1984) also recorded differences in pressure (ΔP) between the inside of the stem and the atmosphere. Grosse et al. (1992) did not measure pressure differences but did record changes in O2 concentration in the soil solution around the roots using a Clark-type electrode sealed into the bottom of the root chamber.
Fig. 1.
Type of assembly used to collect the data on which claims of pressurized gas flow to alder roots and other species were based. Both in the dark and with the stem irradiated, 10 mL quantities of a tracer gas, ethane (100 %), were injected into the top chamber and the lower chamber was sampled periodically. Faster rates of ethane transfer to the lower vessel in the light were interpreted as being due to a thermo-osmotically induced pressure flow from the irradiated shoot. Modified from Grosse et al. (1992).
The results showed that ethane was transported from the upper to the lower chamber through the plant and that the rates of ethane release into the lower chamber increased rapidly when the stems were irradiated by incandescent light, reaching a new equilibrium after approx. 20 min. For those species that showed a marked response, the increased rate as a percentage of the dark condition was 189–261 for Alnus glutinosa, 177 for Taxodium distichum and 144 for Populus tremula. In A. glutinosa, ΔTs of up to 3·6 °C were accompanied by ΔPs of up to 17 Pa, and the latter was considered to be the driving force increasing the transport to the lower chamber. O2 uptake by roots was less when the stems were irradiated, and this also was taken as indicative of an increased O2 transport from stem to root due to a pressurized flow. Translated into an increased O2 supply through the stem, examples of the readings obtained under irradiance as a percentage of the dark condition were T. distichum, 453 and for Betula pubescens, 252.
MATERIALS AND METHODS
Plant material
Stem material
One metre leafless lengths of stems were harvested from mature partially coppiced black alder at North Cliffe Woods, East Yorkshire between February and May 2004, cut into 20–25 cm long segments and the lower ends placed in tap water for transference to the laboratory. The water level was maintained at a depth of 1–2 cm as necessary until a cutting was required for an experiment. The cutting would then be trimmed back at both ends by approx. 1 cm, and a 1 cm length of the ‘bark’ was stripped back from what had been originally the lower end. This end was now placed uppermost and a small water reservoir was attached to it by means of rubber tubing sleeving on to the stripped surface (e.g. see Fig. 2B). This reservoir was topped up with tap water or CO2-enriched de-ionized water as required. The water usually began to percolate immediately through the xylem. If percolation failed, the ends of the cutting were re-cut and the procedure was repeated. These cuttings were employed for various types of experiment as described later; sometimes they were fitted with artificial roots and with pressure take-off and application tubes.
Fig. 2.
(A) Physical model of the stem–root system to investigate the potential effects on root aeration of (a) gas pressurization per se; (b) partial through-flow convection; and (c) a higher source O2 concentration. Radial O2 losses measured in an anaerobic medium with the silicone rubber tip on the tip of the microcapillary slotted through the sleeving Pt electrode. (B) Views of the type of assembly used to study the effects of light on pressurization and O2 production by stem cuttings. Effects were recorded as ROL measurements on artificial silicone rubber-tipped microcap ‘roots’. Root bases were attached to the stem over cavities prepared by removing lenticels and discs of cork and secondary cortex beneath. The lower vessel was filled with anaerobic medium: 0·05 % agar : water + 7 mm KCl. Xylem of stem cuttings was pre-charged with CO2-enriched water (4 or 6·9 mm) from a plastic reservoir. Lenticel numbers per cm of stem, 16·3 ± 2·8 s.e. (n = 14); lenticel area (mm2 per cm2), 6·2 ± 0·6 s.e. (n = 14). (C) Sectional diagram of an artificial silicone rubber-tipped ‘root’ of the type attached to stem cuttings. The root tip is positioned within a sleeving Pt electrode, and values of diffusive resistances along the path from root base to electrode surface are indicated. (D) The relationship between radial O2 loss from the artificial roots and O2 concentration within the root–shoot junction. O2 electrolysis from which ROL is derived is also shown. (E) Top and side elevation diagrams to show the positioning of cuttings and whole plants in relation to the light, and the heat sink arrangement for measuring the effects of light on ROL and pressurization. (F) Sectional view of apparatus for determining the effect of light on the potential for pressurization in stem cuttings. Container dimensions: internal length 200 mm, internal diameter 55 mm. (G) Sectional view of apparatus for measuring the photosynthetic O2 production rate of stem cuttings.
Whole plants
Two- and three-year-old plants were purchased from a local nursery during December 2003, replanted in a sandy loam in 20 cm pots and placed in a cool greenhouse. In early spring before bud break, the pots were inserted into larger non-draining containers, the soil was flooded and maintained at approx. 4 cm above the surface, and the plants placed near the window of a south-facing laboratory until dormancy broke and beyond. Black plastic card was fitted around the stem base at the water level and over the rim of the pot to shade any adventitious roots that were produced. This also helped to prevent algal growth in the water. Whole plants were experimented upon in various ways as described later; before doing so they were removed from their pots and the soil was gently washed away from the roots.
Artificial root-stem model
To explore the effects on root radial O2 loss (ROL) of (a) pressurization per se; (b) a partial through-flow such as might arise if gas could be forced out through lenticels at some distant point, e.g. below the water table; and (c) an increased partial pressure at the O2 source, an artificial root and ‘stem’ system was constructed as shown in Fig. 2A. The gas space system of a stem/woody root was mimicked by two 1 µL microcap capillaries joined in series and connected at one end (the top) via a syringe tip and large bore connectors to an acrylic T-piece. At its bottom end, this capillary train was connected to one arm of a small glass T-piece. Two more 1 µL microcap capillaries tipped by a 0·5 cm length silicone rubber tubing (o.d. 1·0 mm; i.d. 0·5 mm) were connected to the other in-line arm of the glass T-piece to mimic the gas space system and O2-permeable apex of a short adventitious root. The silicone rubber tube was blocked at the tip by a piece of glass rod to prevent flooding. Provision for a return pressurized through-flow to the atmosphere from the junction between ‘stem’ and ‘adventitious root’ was made using a train of three 1 µL microcap capillaries connected to an elbow of the stem of the glass T-piece. This was to mimic the gas escape route through leaky unsubmerged lenticels or the situation where an internal pressure build-up might be sufficient to blow gas from submerged lenticels. To mimic the situation where gas pressure is insufficient to blow gas from submerged lenticels, the basal end of this capillary train was sealed with Terostat.
The ‘stem–root’ assembly was immersed in stagnant anaerobic 0·05 % agar : water (w/v) (+ supporting electrolyte: 7 mm KCl) with only the acrylic T-piece and the upper end of the by-pass capillary train remaining emergent. Compressed air, or air at atmospheric pressure, or raised partial pressures of O2 were supplied to one arm of the acrylic T-piece; any applied pressures were recorded using a pressure transducer (Digitron 2026P, Radiospares, UK) attached to the other arm. The 0·5 cm long O2-permeable tip of the artificial root was positioned exactly within a 0·5 cm long cylindrical Pt cathode (i.d. 0·225 cm) and ROL from the root tip was measured polarographically (Armstrong, 1979) using appropriate polarizing voltages chosen from current–voltage curves obtained using a polarograph with digital output (Barman Electronics, Skipsea, UK). A saturated Ag : AgCl half-cell was used as the anode.
Alder stems with artificial roots
In an attempt to simplify the logistics of this study, some alder stem cuttings were fitted with artificial roots (Figs. 2B, C). Alder cuttings do not readily form adventitious roots but the use of artificial roots meant that the complicating effects of root respiration could be avoided. To fit the root, a small disc of bark including a lenticel and subtending secondary cortex was first removed using a razor blade; the acrylic connector which formed the base of the artificial root was placed over the cavity and glued to the stem using quick-setting Araldite. To effect a good and long-lasting seal, it was necessary for the stem surface to be properly dry. On some occasions, the stems were also fitted with connectors and tubes for gas pressure sampling in, or application to, the secondary cortex; details of construction are provided later. Stem pressures were measured by pressure transducer (CPFM, FCO11 Mk. 2, Furness Controls, UK) with the analogue output directed to a suitably scaled chart recorder (BD 112, Kipp & Zonen, The Netherlands). Stems were pre-percolated with either 4 or 6·9 mm CO2-enriched water.
Measurements of ROL from the artificial roots could be used to calculate the O2 concentrations developed within the secondary cortex of the stem at the ‘root–shoot’ junction. At equilibrium, the ROL, Q/t (g s−1), from an artificial root can be expressed as:
 |
(Armstrong, 1979) and ΔC = Cstem – Celectrode, where Cstem is the O2 concentration in the stem at the root base, and Celectrode is the O2 concentration at the electrode surface. Since Celectrode is always zero, ΔC = Cstem.
ΣR, the total diffusive resistance between the gas space in the root–shoot junction and the electrode surface (4·475 × 105 s cm−3 in the example shown in Fig. 2C), is the sum of the longitudinal diffusive resistance through the microcaps (Rc), and the radial resistances (RSi) and (Rls) through the wall of the silicone rubber ‘root’ tip and the liquid shell between ‘root’ tip and the electrode, respectively. These are calculable as: Rc = Lc/Doa Axc, RSi = ri loge(ro/ri)/DSi ASi, and Rls = (Coa/Cow) × ro loge(re/ro)/Dow Asi, where Lc is the length of capillary, Doa is the diffusion coefficient of O2 in air (0·205 cm2 s−1 at 23 °C), Axc is the cross-sectional area of the capillary gas space, ri and ro are the inner and outer radii of the silicone rubber tip (0·025 and 0·05 cm), DSi is the effective O2 diffusion coefficient in the silicone rubber (taken as 4·2 × 10−6 cm2 s−1), Coa is the O2 concentration in air, Cow is the concentration of O2 in air-saturated water, re is the electrode radius (0·1125 cm) and Dow is the O2 diffusion coefficient in the liquid shell between the silicone rubber ‘root’ tip and the electrode (2·267 × 10−5 cm2 s−1 at 23 °C). Multiplying by the factor (Coa/Cow) adjusts the liquid shell resistance to account for the large concentration drop at the air–water interface. For this root, the relationship between ROL and O2 partial pressure in the root–shoot junction is shown in Fig. 2D.
To investigate the effects of photosynthesis on pressurization and ROL, cuttings were illuminated by a high-pressure sodium vapour lamp in the configuration illustrated diagrammatically in Fig. 2E. Heat loading was kept to a minimum by an intervening glass plate, cooling fan and a heat sink cooling tower fed continuously with tap water. The side opposite the light source was illuminated by the reflected light from a three-piece wrap-round mirror placed opposite the light source.
Exploring the potential for pressure development in photosynthesizing stem cuttings
Water charged with 6·9 mm CO2 was percolated through the xylem of a stem cutting for several hours from an attached reservoir such as that shown in Fig. 2B. The top and bottom ends of the cutting were then sealed using Terostat-filled aluminium caps, a weight was attached to the bottom cap and the cutting was then submerged upright in de-ionized water in a transparent acrylic tower as shown in Fig. 2F. A hole through the lid of the tower connected the head space to a compressed air source and a pressure transducer. A three-way tap was inserted in the gas line tubing that connected the compressed air source to the tower. The container was illuminated as in Fig. 2E using a 500 W high-pressure sodium vapour lamp. When the light was turned on and whenever photosynthetically induced gas bubbling from the lenticels had commenced, the air pressure of the head space was raised by a fixed amount, using the compressed air source. At first, the bubbling was only temporarily stopped by this procedure and when it recommenced the pressure was again raised. These steps were repeated until there was no more bubbling. Temperatures ranged between 21 and 23 °C.
Measuring photosynthetic rates of stem cuttings
The xylem of a stem cutting was pre-percolated with either CO2-enriched (6·9 mm) or air-saturated water before the cutting was sealed into a glass tube as shown in Fig. 2G. The bottom of the tube was fitted with a rubber bung. The tube had three side ports: the top port was fitted with a thermocouple screened from direct light, a Clark-type polarographic O2 electrode probe (YSI 5331, Yellow Springs Inc., OH), similarly screened, was sealed into the middle port, and the lower port was left open. However, the gas space around the cutting was isolated from the outside atmosphere by water that had percolated through the xylem filling the bottom of the tube and the lower port. The diameter of the lower port was such that the surface tension forces offered sufficient resistance to create a small head of water above it and the water then leaked away at the same rate as it percolated through the xylem. In this way, the gas pressure differences, and gas exchange, between the inside and outside of the tube were minimized. Again the cuttings were exposed to radiation from a high-pressure sodium vapour lamp located behind a glass screen and heat sink as shown in Fig. 2E. Reflected light from a three-piece mirror supplemented the light regime imposed on the cutting. Direct PAR was 550 µmol m−2 s−1 and the reflected PAR was approx. 350 µmol m−2 s−1.
Oxygen concentrations in the gas space around the cutting were measured in the light and in the dark with the output from the polarograph monitored continuously using a flat-bed chart recorder. Temperatures in the gas space were recorded to allow correction for any temperature-induced concentration (−0·37 % °C−1) and membrane permeability coefficient (+4 % °C−1) changes to the readings. In the event, the temperatures within the chamber which averaged 22–23 °C never changed by more than 1 °C during the observation period. Net photosynthesis in the light (μmol CO2 m−2 stem surface s−1) and stem respiration in the dark (μmol O2 m−2 stem surface s−1 or μmol O2 cm−3 stem s−1) were calculated from the slopes of the recorded traces and the assumption that 1 mol of O2 evolved indicated the consumption of 1 mol of CO2 reduced, and vice versa. Dark measurements were made with the glass tube assembly wrapped in black polythene.
Effects of light, pressure, root temperature and O2 concentrations around the shoot on ROL from the adventitious roots of whole plants
Whole plants were removed from their pots, the soil gently washed from the roots and then placed with their shoots in the air and their root systems in a stagnant O2-depleted 1/4-strength Hoaglands 0·05 % agar–water medium. A 2000 mL glass measuring cylinder was first used but, later, water-jacketed glass containers were employed: for smaller root systems, this was a cylindrical vessel (i.d. = 90 mm, o.d. = 112 mm, length = 440 mm), while for large root systems it was rectangular (internal dimensions 170 × 170 × 300 mm; external dimensions 220 × 220 × 330 mm). The agar medium had previously been bubbled with O2-free N2 for 24 h during which time the only escape route from the head space was through a 1 cm long, 1 mm diameter vent. With tap water continuously circulated through the outer jacket of the vessel, the rooting medium could be reduced from room temperature (21–23 °C) to approx. 11–12 °C. The rooting medium was separated from the atmosphere by a partition consisting of either a rubber bung or a split wooden plate (10 mm thick) secured in/on the top of the vessel. The shoot and cylindrical electrode cables were secured in appropriately sized holes in the partition, the shoot being packed round with wet cotton wool to complete the seal; holes were also made for the insertion of the fibre junctions of Ag/AgCl anodes. The shoot was additionally supported by a clamp half way up the stem. The rooting medium surface was raised to the top surface of the partition and in short-term experiments was maintained at this level whenever possible. However, in longer term experiments, transpirational water loss sometimes lowered the liquid level by a few centimetres before replenishment could be effected. To enable changes to be made in the O2 regime around the shoot, a short collar-like chamber with inlet and outlet ports was fabricated from an acrylic syringe body. This was fitted around the shoot base, and sealed top and bottom with Terostat. The ports were normally left open, but occasionally were used to change the atmosphere around the shoot base from air to 100 % O2. To enable gas pressure in the secondary cortex of the stem to be raised artificially, a small disc of bark including a lenticel and subtending secondary cortex was first removed using a razor blade, and the acrylic base of a syringe needle (obliquely cut) was sealed on to the stem around the cavity using quick-setting Araldite. Narrow non-permeable plastic tubing linked the syringe needle to the pressure transducer via a three-way tap. To enable a high level of illumination (PAR 550 µmol m−2 s−1) to be applied to the plant without creating an excessive heat load, the plants were positioned as for previous experiments (Fig. 2E). Heavy-duty black polythene sheeting was applied as required to keep the submerged parts of the shoot and root in darkness.
Sleeving Pt electrodes were used to monitor continuously the ROL from the adventitious roots at approx. 2–7 mm from the tips (a) under different light regimes; (b) under different temperatures in the rooting medium; (c) under different O2 regimes around the shoot base; and (d) as gas pressures in the secondary cortex of the stem were raised artificially using compressed air applied through the acrylic connectors attached to the stem.
Anatomy
Transverse sections of stems and roots were cut either by hand or by sledge-microtome using fresh material. They were examined by light microscopy (Olympus BX40) or by fluorescence microscopy (Olympus CH with fluorescence attachment). A digital camera (Olympus Camedia 4040Zoom) was used for the photography.
Light absorption by stem tissues
To investigate the degree to which photosynthetically active radiation was absorbed by the different tissues, discs of different thickness cut from the stem surface were used. A thin coating of a clear lubricating jelly was applied to the inner surface and the disc then placed over the sensing element of a LI-COR light meter (Model LI-189). A ring of Terostat was used to prevent light leaking in though the side of the sensing unit. The light flux was measured with the stem discs held normal to the various strength of light source, and any light flux passing right through the disc was recorded.
RESULTS AND DISCUSSION
Artificial root–stem model: effects on ROL of raised gas pressure per se, raised O2 partial pressures and partial pressurized through-flow
With the base of the model open to the atmosphere and the diffusion gradients between atmosphere and electrode surface at equilibrium, the ROL from the silicone rubber apex of the artificial root–stem model gave an electrolytic reduction current for O2 of 3·904 µA (radial O2 flux = 123·6 ng cm−2 root surface min−1). When compressed air was applied at the acrylic T-junction and the pressure raised in a stepwise manner, the ROL quickly increased and equilibrated at each stage (Fig. 3). However, each stepwise increase in flux was tiny and, even at the comparatively high pressure of 2530 Pa, ROL was only 126·5 ng cm−2 root surface min−1 (3·996 µA), just 2·35 % greater than for the original non-pressurized condition: a 0·003 % rise per 100 a. This can be accounted for almost entirely by the increased density of the gases at 2530 Pa according to the equation
 |
where atmospheric pressure is 101 300 Pa, ROL2530 is the ROL at the applied pressure of 2530 Pa and ROLatm is the original ROL when the basal end of the model was open to the atmosphere. The predicted rise based upon this equation was very slightly greater than the observed increase (Fig. 4), and any difference is probably attributable to experimental error. The sometimes steep decline in ROL value between applied pressures (Fig. 3) due to a momentary release of pressure before application of the next raised value is illustrative of how rapidly the density change is expressed and of how sensitive is the cylindrical Pt electrode polarographic technique.
Fig. 3.
Stem–root model. The effects on radial O2 loss of stepwise increases in gas pressure applied at the top of the model with the side arm sealed. ROL is expressed in terms of the O2 electrolysis current. ROL in ng cm−2 min−1 = 4·974 × [electrolysis current/surface area of root tip inside the sleeving Pt electrode].
Fig. 4.
Stem–root model. Observed and predicted relationships between ROL and applied pressure. ROL is expressed in terms of the O2 electrolysis current. The inset shows the observed data plotted with a full (not selectively expanded) y-axis. ROL in ng cm−2 min−1 = 4·974 × [electrolysis current/surface area of root tip inside the sleeving Pt electrode].
When the Terostat was removed from the basal end of the artificial stem creating a by-pass akin to leaking lenticels exposed to a gaseous head space at atmospheric pressure e.g. as in Fig. 5, the ROL increased rapidly and substantially (by 43·5 %) to reach a new equilibrium value of 177·3 cm−2 root surface min−1. This is the type of effect that we believe probably led to the increased ethane movement in the original experiments of Grosse et al. (1985, 1985) although the pressurization in that case might have been a combination of photosynthetic O2 production, thermal transpiration and humidity-induced diffusion. In this example, the pressure was not released between stages.
Fig. 5.
Stem–root model. Effects on ROL of stepwise increases in gas pressure with the side arm sealed followed by effects on ROL with the side arm open permitting convective gas flow through the stem capillaries only.
A further contrast to the minimal effect on ROL of pressurization per se is the result obtained by increasing the O2 partial pressure at the base of the model (Fig. 6): exposure to 35 % O2 increased the ROL by 46 %, and again, removing the Terostat to create a partial through-flow increased the ROL to approx. 106 % of the original diffusion-only value obtained with air around the base.
Fig. 6.
Stem–root model, showing (A) the effect on ROL of increasing the partial pressure of the O2 source to 30·5 % with the side arm sealed and no convection; (B) as for (A) but with the side arm open permitting convection of a 30·5 % O2 stream through the stem capillaries only; (C) as for (B) but with air (20·7 % O2) convecting through the stem capillaries. With the y-axis expanded, the inset shows the initial change in ROL with time after exposure to 30·5 % O2 at (A).
Alder stems with artificial roots—effects of photosynthesis
Examples from two separate experiments are shown. In the first, Fig. 7, the ROL from the artificial roots was measured initially with the submerged (lower) portion of the stem in darkness and the emergent (upper) part receiving low light (approx. 30 µmol m−2 s−1 PAR) from a north-facing window. The xylem of the stem had previously been perfused with water containing 4 mm CO2. Under these conditions, the ROL was low (15 ng cm−2 min−1 in the example shown) and this indicated an O2 partial pressure of only 1·35 kPa in the secondary cortex of the stem and base of the artificial root. When the sodium vapour light was switched on to give a PAR flux of 550 µmol m−2 s−1 over the emergent part of the stem, the ROL rose to a new equilibrium value of approx. 85 ng cm−2 min−1, indicating now an O2 partial pressure of 7·5 kPa in the stem–artificial root junction. When the covers were removed, and the whole stem illuminated at 550 µmol m−2 s−1, the ROL again increased, this time to 336 ng cm−2 min−1, showing that the O2 partial pressure at the stem–artificial root junction must have risen to approx. 30 kPa (30 % O2). This could only have come about due to O2 generation, presumably by stem photosynthesis. A second artificial root attached to this stem behaved similarly, but the pressure sensors had become leaky during these manipulations and no pressure data for the stem were obtained. Earlier, however, a sensor on the upper part of the stem registered 2000 Pa (Furness Controls CPFM pressure transducer outputting to a suitably scaled chart recorder) when the whole stem was submerged and illuminated, and some bubbling from the lenticels began at approx. 1920 Pa.
Fig. 7.
Partially submerged alder cutting with artificial roots: effects of light on O2 concentrations in an artificial root–shoot junction (i.e. in the secondary cortex of stem) and ROL from artificial roots with (A) the submerged part in the dark and emergent stem illuminated with a PAR of 30 µmol m−2 min−1, (B) the emergent part of the stem illuminated with a PAR of 550 µmol m−2 min−1 and (C) the emergent and submerged parts of the stem illuminated at 550 µmol m−2 min−1.
In the second experiment, the pressure sensors functioned well but the ROL data could only be collected digitally from one root because of electrical interference. However, ROL from the second root was monitored on a chart recorder and the data (not shown) followed the same pattern as for the digitally monitored root. The cutting was pre-percolated with de-ionized water containing CO2 at 6·9 mm. When the lower (submerged) portion of the twig was in darkness, and the upper (emergent) part of the shoot was receiving only north light, the ROL from the artificial roots was barely detectable (Fig. 8A). When the emergent part of the twig was illuminated (550 µmol m−2 s−1 PAR), there was a 17 min delay before there was any obvious effect on the ROL; it then rose and by 45 min had reached a quasi-equilibrium value of approx. 15 ng cm−2 min−1. Gas pressures in both the upper and lower sections of the twig were initially negative but began to rise as soon as the upper portion of the stem was illuminated. Their ultimate equilibration at between 2·5 and 3·0 kPa coincided approximately with the equilibration in ROL. Pressure in the emergent portion was greater than in the darkened submerged stem, and this could indicate some release of pressure between the two sampling points. When the covers were removed and the lower portion of the stem was also illuminated, the pressures began to rise once more and very quickly became coincident top and bottom. Unfortunately, however, the pressure recording for the upper section went off-scale at approx. 4·0 kPa and no data were available from this sensor until the pressures declined later in the experiment. The second pressure rise was approximately the same as the first (approx. 2·5 kPa) but the rise in ROL was very much greater in the second phase, equilibrating at approx. 390 ng cm−2 min−1 (Fig. 8A). This ROL value equates with a stem–artificial root junction O2 partial pressure of 32 kPa and indicates, as in experiment 1, that photosynthesis, not pressurization, was the major contributor to root aeration. We attribute the supplementary large rise in ROL to the additional photosynthetic generation and accumulation of O2 in the lower portion of the stem where its escape is hindered because it cannot readily diffuse into the water. In diffusion terms, the lower portion of the stem would therefore have become the effective O2 source and the atmosphere around the emergent shoot the principal sink. In the first stage of the experiment, the O2 generated in the upper part of the stem probably fed the lower stem section and the root by diffusion only and, judging by the almost undetectable initial ROL, the ‘effective resistance’, i.e. synergism between diffusive resistance and stem O2 consumption, was very substantial. In the second stage of this second experiment, the pressures generated did cause some obvious convective flow, because, at approx. 4·5 kPa gas bubbles began to escape from the submerged part of the stem. However, since resistance to escape will have been greater below than above the water line, it seems reasonable to assume that if there was any directional component to convective flow within the stem at this stage, it would have been directed upwards from the lower to the upper part.
Fig. 8.
A partially submerged alder cutting with artificial roots: effects of light on root–shoot junction (i.e. in the secondary cortex of stem) O2 concentrations, ROL and pressurization in submerged and emergent parts of a stem. (A) At 200 s, the emergent stem only is exposed to a PAR of 550 µmol m−2 min−1 (submerged part in the dark); at 3125 s, both emergent and submerged parts are illuminated. (B) Continued from (A) and showing (i) the gradual decline in ROL and stem pressures which began at approx. 6000 s after exposing the emergent and submerged parts to the light, and (ii) the final rapid fall in pressure and ROL when the light was switched off.
Neither pressurization nor the high values of ROL were sustained indefinitely in this experiment. After 90 min, the pressures began to fall. Shortly afterwards, this was followed by a steady decline in ROL (Fig. 8B) which continued until the lights were switched off. Bubbling was very much reduced when the internal pressures reached 4·0 kPa. When the lights were switched off, the pressure differential between the upper and lower stem sections became evident again and the pressures fell rapidly, eventually becoming negative again as at the start of the experiment. The decline in ROL also accelerated after the lights were switched off but took 20 min longer to reach its initial value and was still appreciable when the pressures had already become negative. This again indicates a lack of dependence between ROL and pressurization per se. The finally negative pressures might indicate internal stem temperatures lower than ambient resulting in an outwardly directed thermo-osmosis (Idso, 1981; Steinburg, 1996).
We are uncertain as to why pressurization and ROL were not sustained at their peak values in the light and did not investigate this further. Several reasons could be suggested: (a) CO2 availability may have begun to decline since the twigs were not re-charged with CO2-enriched water during the experiment; (b) O2 enrichment and declining CO2 may have increased photorespiration; (c) respiratory demand in the shoot might have been stimulated by substrate accumulation; or (d) substrate accumulation might have reduced photosynthesis by a negative feedback. With whole plants and adventitious roots, the ROL did not show this decline if the submerged parts were maintained at 12 °C (Figs 15 and 16).
Alder stems with artificial roots—effects of artificially applied pressure
Following on from the previous experiment, ROL was monitored as pressure from a compressed air source was applied artificially to the upper section of stem via the feed tube previously used to sense internal pressures. No pressures could be recorded from the lower stem section because of a leaky joint. The results are shown in Fig. 9. Again it is evident how little pressurization itself affects root aeration: an increase in ROL only became really obvious when the pressures had exceeded 7 kPa and a continuous bubbling from lenticels on the stem near the junction with the artificial root had begun (Fig. 9B). However, after the stem was illuminated above and below the water line to stimulate photosynthesis, ROL increased very rapidly (Fig. 9A).
Fig. 9.
Partially submerged alder cutting with artificial roots: effects on ROL of applying pressure to the emergent stem—the same stem cutting as for Fig. 8. (A) ROL vs. time and indicating pressure changes followed by a PAR of 550 µmol m−2 min−1 to emergent and submerged parts. (B) ROL vs. pressure.
Photosynthetic rates of stem cuttings
With CO2-enriched water (6·9 µm) percolating through the xylem, the net photosynthetic rates for the stem cuttings (measured as O2 output) averaged 1·28 µmol m−2 s−1 at a light flux of 550 µmol m−2 s−1 PAR, while dark respiration measured as O2 uptake averaged 2·46 µmol m−2 s−1 (Table 1). Without CO2 enrichment, i.e. when only de-ionized water in equilibrium with air was percolated through the xylem, net photosynthesis was only marginally positive (data not shown).
Table 1.
Net photosynthesis and dark respiration of cuttings of 3-year-old alder stems—xylem pre-charged with CO2-enriched water (6·9 mm)
| Net photosynthesis (μmol m−2 s−1) |
Dark respiration (μmol m−2 s−1) |
|---|---|
| 1·28 ± 0·49 (5) | 2·46 ± 1·02 (4) |
Based on measurements of O2 evolution and uptake, respectively. PAR: 550 µmol m−2 s−1.
Means ± s.d. (replicates). Twig dimensions length = 150–170 mm; diameter = 16·5–19 mm; xylem stream flow 0·4–0·5 mL min−1; temperature 22–24 °C.
Although the net photosynthesis value for the stem is relatively low when compared with those attainable by the leaves of some herbaceous species, it is not far below rates cited for leaf photosynthesis in some deciduous tree species (Larcher, 1969). It is interesting to note that the chlorophyll content of young aspen twigs can almost reach that of the leaves, can increase with age and develop lower chlorophyll a/b ratios indicative of becoming shade adapted (Aschan et al., 2001).
The potential for pressure development in photosynthesizing stem cuttings
When cuttings pre-charged with CO2-enriched water (6·9 mm) were submerged and illuminated, gas bubbles began to emerge and escape from the lenticels. We did not measure the O2 concentration of the bubbles, but it is clear from the artificial root ROL measurements (Figs 7 and 8) and the photosynthetic measurements that the gas bubbles would have been O2 enriched. Applied pressures were raised in 2–4 kPa steps. Such steps were sufficient to temporarily stop the bubbling and, if and when bubbling resumed, the pressure was raised again. The pressure required permanently to stop bubbling varied between cuttings but was always >15 kPa and <30 kPa and was reached within approx. 35 min, with the fastest rate of pressurization being 2 kPa min−1. We are uncertain as to why the cutting failed to pressurize beyond 25 kPa but it would seem to indicate that photosynthesis had either slowed substantially or stopped. It is quite likely that the O2 concentrations within the cutting reached levels high enough to have an inhibitory effect on photosynthesis.
The rapid development of high internal pressures can readily be explained in terms of the net rates of photosynthetic O2 output into what is a relatively small volume of gas space in the stem. The gas space is principally in the secondary cortex. The latter is approx. 200 µm thick and, from transverse sections, we estimate it to have a fractional porosity of 0·1. On this basis, the volume of internal gas space in the cutting would have been approx. 0·18 cm3.
Chloroplast distribution and light penetration of alder bark and wood
Despite the frequently brown appearance of the stems, transverse sections of stems viewed by light and fluorescence microscopy revealed the presence of chloroplasts in the secondary cortex, secondary phloem, the ray tissues of the stele and even the medulla (Fig. 10A–E). The distribution of starch grains and chloroplasts in the stele seemed to be coincident.
Fig. 10.
(A) Transverse section of sector of a 1-year-old alder stem: red autofluorescence under blue light is due to chloroplasts in secondary cortex, secondary phloem, secondary xylem and medulla. Scale bar = 200 µm. (B) As (A) but at higher magnification showing individual chloroplasts fluorescing red in secondary phloem and in secondary xylem ray cells. Scale bar = 25 µm. (C) Sector of a 3-year-old alder stem showing red autofluorescence from chloroplasts in medulla and the inner two annual rings. Scale bar = 200 µm. (D) Transverse section of a 3-year-old alder stem through a lenticel. Some algal cells are fluorescing red on the outer surface of the flaking lenticel. Some red fluorescence is seen from chloroplasts in secondary cortex and secondary phloem. The pink colour in banded tissues of lenticel is not fluorescence but is due to anthocyanins. Scale bar = 100 µm. (E) Transverse section of a 1-year-old alder stem with transmitted white light. Chloroplasts (green) in secondary xylem rays and medulla. Scale bar = 200 µm. (F) Three-year-old alder sapling (plant 1) prepared for ROL and pressure measurements and for applying above ambient O2 concentrations to the shoot base. Note the sleeving Pt electrodes around the tips of two of the adventitious roots. An acrylic chamber around the shoot base and the pressure sampling/delivery point on the emergent part of the stem are visible. (G) As (F) showing the submerged stem base and sleeving Pt electrodes on two adventitious roots. (H) Transverse section of 16 cm long adventitious alder root at 15 mm from the base showing aerenchyma (aer) in the primary cortex and well developed secondary xylem (sx) and secondary phloem (sp). Arrows point to non-aerenchymatous gas spaces (black) visible in the secondary cortex (sc) and running radially in the secondary rays. Scale bar = 200 µm. (I) As (H) but at 49 mm from the base. Aerenchyma (aer) visible in the primary cortex and early normal development of secondary xylem (sx), secondary phloem (sp) and secondary cortex (sc). Lignified tissues, including primary xylem (px) stained red using phloroglucinol and concentrated HCl. Scale bar = 100 µm. (J) Root system of a 3-year-old alder after 4 months in a waterlogged soil. Adventitious roots produced from the submerged stem base during this period have grown to a length of up to 36 cm. The white tips are visible, as are their thick woody bases protruding at right angles to the stem. (K) Transverse section 40 mm from the tip of longest adventitious root shown in (J) (length = 36 cm). Lignified primary xylem (px) stained red using phloroglucinol and concentrated HCl but arrows point to abnormal development of secondary xylem with large thin-walled non-lignified cells. Scale bar = 100 µm.
We deduced from this that light must indeed be capable of penetrating even to the centre of 3-year-old alder stems and, to try to obtain some quantitative estimate of this, bark peels of different thickness and halved stem segments were affixed to the PAR sensor and the light flux penetration measured. The results are shown in Table 2.
Table 2.
Light penetration through various tissue groups from 3-year-old stems of alder
| PAR incident on upper and lower surfaces (µmol m−2 s−1) |
||||
|---|---|---|---|---|
| Tissue thickness and type |
Rep 1 |
Rep 2 |
Surface |
|
| 270 µm, periderm (phellem, phellogen and secondary cortex) | 1089 | 460 | Upper | |
| 86 | 40 | Lower | ||
| 600 µm, periderm, secondary phloem and cambium) | 1089 | 460 | Upper | |
| 4 | 1·8 | Lower | ||
| 9 mm, radial penetration to medulla | 1089 | 1089 | Upper | |
| 0·14 | 0·06 | Lower | ||
Most of the light (90 %) was absorbed in the periderm (i.e. phellem, phellogen and secondary cortex), and much of the remainder in the secondary phloem. Although the methodology was crude, it showed that some light did reach even into the medulla since a fully darkened sensor gave a zero reading. It is conceivable also that the results will under-estimate the flux reaching the chloroplasts deep within the stem since the sensor averages the flux over the sensor disc whereas it is most likely that the cell walls will function as a fibre-optic system.
Effects of light, pressure, root temperature and O2 concentrations around the shoot on ROL from the adventitious roots of whole plants
Several whole plants were monitored in a variety of ways and experimental assemblies. Experiments on two plants will be described.
Plant 1
The plant shown in Fig. 10F was monitored with its roots in the vessel shown and later in a water-jacketed container to cool the root system. A pressure delivery connector was affixed to the stem at a point 5 cm above the supporting bung (and agar surface) and just above an acrylic sleeve for applying O2 to the stem base. Two roots (13·5 and 12·5 cm) were fitted with cylindrical Pt electrodes and the whole assembly was positioned relative to the sodium vapour light, mirror and north-facing window as in Fig. 2E.
The first observations were made with the roots and shoots at room temperature (i.e. 22 °C). With only north light (50 µmol m−2 s−1 PAR) incident on the shoot above the water line and the submerged parts in darkness, no ROL was detected (Fig. 11A). Increasing the light flux to the emergent shoot to 550 µmol m−2 s−1 had no detectable effect on the ROL from the roots but, when the covers were removed, exposing the submerged parts to light, the ROL soon increased and rose rapidly and levelled out at approx. 12 ng cm−2 min−1 (Fig. 11A). Although the plot indicates a very significant rise, it should be realized that 12 ng cm−2 min−1 is a relatively low value of ROL for a wetland plant root. For a rice root of this length, a value of 80–100 ng cm−2 min−1 would not be exceptional. With alder, however, although the roots were at least partially aerenchymatous (Fig. 10H, I), their apical parts were not and their point of origin was 6 cm below the water table and the secondary cortex of the submerged stem was not aerenchymatous.
Fig. 11.
Alder sapling (plant 1) showing the effects of light on ROL from a 13·5 cm long intact root. (A) ROL was unaffected by north light (50 µmol m−2 s−1) or supplementary light (550 µmol m−2 s−1) on an emergent shoot only but rose rapidly when emergent and submerged parts were exposed to a PAR of 550 µmol m−2 min−1 followed by a steady decline. ROL returned rapidly to background (approximately zero) when the light was switched off. (B) Continuation from (A) showing a similar pattern but lower ROL peak after exposure of emergent and submerged parts to 550 µmol m−2 s−1 PAR. (C) Continuation from (B) showing two further cycles of light application and a further diminished ROL peak despite some CO2 enrichment of the bathing medium.
From the earlier observations on the stem cuttings, we attribute the rise in ROL chiefly to photosynthetic O2 production in the submerged parts of the stem. However, ROL soon declined and, after approx. 45 min, had almost halved in value. When the light was switched off, and despite some north light reaching the submerged parts, the ROL became undetectable once more. A second exposure to the high light regime also produced a rapid rise in ROL (Fig. 11B) but the peak value was lower than before and again the ROL soon began to decline once more. On this occasion, the light remained on for longer and the ROL had fallen by 80 % before the light was switched off. As with the stem cuttings and their artificial roots, we are still uncertain as to what caused the decline in ROL. It could have been that there was insufficient internal CO2 to sustain photosynthesis, and so for a third and fourth light dark cycle we charged the agar around the basal submerged region of the stem with 20 mm CO2. This did not have any obvious enhancing effect on the ROL (Fig. 11C) but it may have been insufficient to raise internal CO2 levels. ROL could also have declined for other reasons, e.g. an increase in stem and/or root respiration. This could have been brought about by an increase in root/stem temperatures or by increased availability of respiratory substrate arising from stem photosynthesis. The development of a barrier to O2 loss from the root is also a possibility. However, the following manipulations subsequently performed on the same plant and monitoring the same roots do not seem to indicate the existence of much of a barrier.
When O2 was circulated through the acrylic collar on the emergent stem base, an increase in ROL was detectable within 3 min (Fig. 12) and after 48 min it had reached 42 ng cm−2 min−1 and was still rising. When the emergent shoot and the submerged basal portion of stem were illuminated, the ROL rose more rapidly and at approx. 80 ng cm−2 min−1 had still not fully equilibrated when the experiment was halted. The rather slow equilibration when the O2 was applied may have been a function of such a large and complex root system.
Fig. 12.
An alder sapling (plant 1) showing the effects of exposing the emergent stem base to pure O2 followed by exposure of emergent and submerged parts to a PAR of 550 µmol m−2 min−1.
Lowering the temperature of the submerged parts also had a very significant effect on ROL. The same plant was transferred to a container with a cooling water jacket and ROL was measured with the emergent stem and the basal 5 m of submerged stem receiving only north light. The data show (Fig. 13) that the ROL from the apex of an 18 cm long root (root–shoot junction at 5 cm below the water line) was initially undetectable at 22 °C. However, once the medium began to cool, ROL soon became detectable. By the time that the whole medium had reached approx. 12 °C (after 140 min), the ROL had begun to equilibrate at approx. 13–14 ng cm−2 min−1. Subsequently, the medium was allowed to return to room temperature (22 °C) and ROL became undetectable once more (Fig. 13 inset).
Fig. 13.
An alder sapling (plant 1) showing the effects on ROL from an 18 cm long root of (A) cooling submerged parts from 22 to 11 °C—emergent parts and top 5 cm of submerged parts illuminated but only by north light at approx. 50 µmol m−2 s−1, and (B) allowing the submerged parts to warm up again to room temperature (approx. 22 °C).
Since the cooling of the root system had produced detectable ROL without the need for high intensity illumination to emergent or submerged sections of stem, and since A. glutinosa normally inhabits rather cold wet soils, subsequent experiments were all performed with the rooting medium held at 11–12 °C. First, using the same plant and the roots at 12 °C, the effect on ROL of artificial pressurization of the stem was examined. The results are shown in Fig. 14. Unfortunately, the submerged sensor point had failed so that only the applied pressure could be recorded. However, it was clear from the bubbling that took place from the basal parts of the submerged stem that pressurization within the stem, and presumably also the root, was substantial. As with the artificial stem–root model, and the stem cuttings with artificial roots, pressurization per se had little effect on ROL. The increase in applied pressure from 6 kPa to approx. 23 kPa only raised the ROL from the living root by 4 % (i.e. from 17·2 to 17·9 ng cm−2 min−1). This is an even smaller rise than was predictable based on the perceived density change due to pressurization, but this was probably due to some pressure leakage between the application point and the root base; hypertrophied lenticels may have contributed to this. Nevertheless, compared with the increases in ROL that were recorded on exposure (a) to light, (b) to O2 around the emergent shoot base and (c) to cooling, the effect of pressurization per se is almost undetectable.
Fig. 14.
An alder sapling (plant 1). Observed and predicted effects of increasing pressurization applied to the connector on an emergent shoot 50 mm above the water line. The root system is held at 11 °C, the emergent parts and the top 5 cm of submerged parts receiving only north light at approx. 50 µmol m−2 min−1.
Subsequently, with no pressure applied to the stem, exposure to the sodium vapour light raised the ROL to 39 ng cm−2 min−1 and, unlike previous experiments where roots were at room temperature, ROL did not subsequently decline until the lights were switched off again (data not shown). This observation was repeated over two daily cycles (results not shown). Internal pressures recorded within the emergent stem did not exceed 26 Pa in the light. This is a much lower figure than obtained with the cuttings, but whole plants are more likely to have imperfections such as dead buds and small branch scars which are more leaky than the phellogen, and this would reduce the ability for pressurization in the secondary cortex. They also have hypertrophied lenticels above the water line that may be more leaky to gases than normal lenticels Furthermore, CO2 concentrations in the xylem stream may have been much lower than in the cuttings where the xylem had been deliberately pre-charged with CO2-enriched water. In waterlogged soils, however, the water absorbed by the roots is likely to be CO2 enriched.
Plant 2
This larger plant required the rectangular water-jacketed vessel to accommodate the root system which was maintained at 11–12 °C and the ROL was recorded from two roots (each approx. 12 cm in length) over several daily cycles. The emergent shoot and 5·5 cm of the submerged portion were exposed to natural north light. Initially, however, the ROL was recorded (a) with the submerged part wholly in the dark, and the emergent part receiving north light (37 µmol m−2 s−1); (b) with the submerged part in the dark and the lower emergent part illuminated at 550 µmol m−2 s−1 PAR; and (c) with the emergent shoot and 5·5 cm of the submerged stem exposed to 550 µmol m−2 s−1 PAR. This showed (Fig. 15) that for a 12 cm long root (root base 5 cm below the surface) there was for (a) a detectable ROL at the tip of approx. 5 ng cm−2 min−1, for condition (b) this rose to approx. 12 ng cm−2 min−1 and for condition (c) it exceeded 55 ng cm−2 min−1.
Fig. 15.
An alder sapling (plant 2). Effects of different light regimes on ROL; the root system is held at approx. 11 °C. Initial reading with submerged parts in the dark and the emergent parts receiving north light at approx. 37 µmol m−2 s−1 followed by (A) the emergent parts exposed to a PAR of 550 µmol m−2 s−1 and (B) emergent parts and the top 5 cm of submerged parts receiving 550 µmol m−2 s−1.
The daily record of ROL for this root (roots at 11–12 °C) with only north light up to 75 µmol m−2 s−1 PAR reaching the emergent stem and 5·5 cm of submerged stem is shown in Fig. 16. Diurnal cycling is obvious and, extrapolating from the results of the previous experiments, this must be attributable to photosynthesis rather than pressurization. Short-term fluctuations during daylight hours seemed to be a function of different amounts of cloud cover, and on day 4 poor lighting in the afternoon (5040–5400 min) was such as to foreshorten the period of elevated ROL. Another feature of the system that affected the ROL was the level of the rooting medium which was lowered by transpiration. The medium was not topped up on day 2 and, interestingly, resulted in much higher night and subsequent daytime readings. By the time it was topped up, the level had fallen by 7 cm, exposing many lenticels and the root base. A much shorter diffusion path length under these circumstances presumably accounted for the higher readings. However, the daytime reading was still supplemented by photosynthesis. In other cases where the fall in level had only been small, the situation is less clear; topping up on day 5 produced an initial decline in flux but there was still photosynthetic supplementation, while on day 6, topping up resulted only in a temporary slowing down of the daytime flux rise.
Fig. 16.
An alder sapling (plant 2). Diurnal patterning in root ROL modified by additional fluctuations caused by variations in light flux during the day and water level changes. Monitored roots were initially apprrox. 12 cm long. The basal 5·5 cm of submerged stem were unshaded and the plant was exposed to north light from the laboratory window (PAR ≤70 µmol m−2 s−1). Roots grew slowly for the duration of the experiment. Root chamber temperature was 11–12 °C.
Anatomy and morphology of root systems
The adventitious roots used for monitoring ROL were fleshy with schizogenously developed sub-apical aerenchyma similar to that in Fig. 10I. Plants maintained outdoors from April to July in pots of saturated loam with the water surface 3–4 cm above the soil developed adventitious roots above the soil surface that ultimately became thick and woody at the base and penetrated to the bottom of the pots under still flooded conditions. The example in Fig. 10J had achieved a length of approx. 36 cm. However, the anatomy of these and even 16 cm long roots did show some evidence of ‘O2 stress’ near to the tips. The basal regions showed a normal anatomy (Fig. 10H, I) but, near the tips, although the primary xylem elements in transverse sections had a relatively normal appearance and were lignified, the secondary xylem was clearly abnormal: in place of thick-walled elements, only inflated thin-walled cells were being produced. (Fig. 10K); in some, there was a large central cavity in the medulla (not shown). We have noted a similar phenomenon with metaxylem development in pea roots growing in agar (M. T. Healy and W. Armstrong, unpubl. res.): metaxylem was replaced by inflated thin-walled cells or a cavity when the apical O2 concentrations (as detected by ROL) approached zero. As these roots aged, the apical regions received more O2 again (Armstrong and Healy, 1984); the roots grew on and normal metaxylem development began again. The result was a root with a discontinuous metaxylem.
GENERAL DISCUSSION
The aeration of the submerged organs of non-woody herbaceous wetland plant species is fairly well understood. Where the stem base is at or near ground level, and provided that shoots are not greatly submerged, aerenchyma formation is usually adequate for the transport of sufficient quantities of O2 to support root and rhizosphere aeration (Armstrong, 1979; Colmer, 2003). Species having deeply submerged rhizomes usually supplement diffusive gas exchange with some through-flow convection which brings O2, at close to atmospheric levels, into the rhizome and to the root–shoot junctions from where it can diffuse into the roots (J. Armstrong et al., 1996; Brix et al., 1996). Some submerged aquatics such as Isoetids may rely on photosynthetic O2 production to aerate the root and rhizosphere, and in these species a major CO2 source for photosynthesis may be that derived from the sediment and transported to the leaves via aerenchyma or the stele (Pedersen and Sand-Jensen, 1992; Pedersen et al., 1995; Sand-Jensen et al., 2005).
Root aeration in woody wetland species is less well understood, particularly those species that do not have obvious adaptations to the wetland environment such as the pneumatophores and aerenchymatous stilt roots of mangroves. Hook and Brown (1972) and Hook and Schlotens (1978) have noted, however, that the tolerance of some American hardwoods correlates with the presence or absence of gas spaces traversing the vascular cambium and that, under vacuum, bubbling of gas from lenticels below the water line could be more readily induced for flood-tolerant than flood-intolerant species. Rhizosphere oxidation was also a characteristic of tolerant, but not of intolerant species. Furthermore, hypertrophied lenticels at or around the water table, and the production of aerenchymatous ‘water roots’ were also features of flood-tolerant species. From field observations, Hook and Schlotens (1978) note that for such species it can be assumed, but not proven, that O2 diffuses through the bark above the water line in sufficient quantities to oxygenate their roots—a distance of >100 cm. However, the secondary cortices of submerged stems and the woody part of root systems are not generally furnished with aerenchyma. The respiratory demand and relatively low porosity of a non-aerenchymatous secondary cortex will be restrictive to O2 transport over long distances (Armstrong, 1979). Not surprisingly, therefore, we have found in the present study that when the roots of alder were at room temperature (22–24 °C), there was insufficient O2 to oxygenate the rhizosphere of roots up to 15 cm in length and arising from the stem only 5 cm below the surface. There may of course still have been some O2 within the root cortex but this was insufficient to traverse or give a measurable flux across the epidermal–hypodermal cylinder.
How then might significant root aeration be sustained over longer distances? A suggestion which had seemed to promise such a solution was that in plants such as alder, there might be a convective gas flow driven by above ambient pressures in the stem arising in sunlight from stem warming and a resulting thermal osmosis (Grosse and Schröder 1984, 1985; Schröder, 1989b). Unfortunately, so far as the roots themselves are concerned, because they are effectively blind-ended tubes, pressurization should do no more than increase O2 concentrations in terms of the pressure-induced density increase, and we have confirmed that this effect is tiny (Figs 3–15). Some improvement in root aeration might come about if pressurization in the stem was sufficiently high to cause a though-flow into the submerged stem or woody part of the root system by forcing gas from the submerged lenticels. Such an effect was found in the present study only when pressures much in excess of those likely to be generated by thermal osmosis or humidity-induced diffusion were applied (Fig. 9). Bubbling from lenticels did not usually begin until the internal pressure had exceeded 2–4 kPa and, even if the phellogen was behaving as a perfect microporous membrane, and there was a temperature differential across it of 10 °C, the pressure generated by thermal transpiration would only be 1·7 kPa. Humidity-induced diffusion could theoretically generate a pressure of 4·6kPa if the internal temperature was 10 °C above ambient (Armstrong et al., 1996a, b). However, we took every precaution to reduce the heat load on the stem and it is doubtful whether the internal temperatures were much above ambient. Indeed, because of evaporation, temperatures within the secondary cortex might well have been below ambient and the potential for thermal transpiration negative (Idso, 1981; Steinburg, 1996). In Fig. 8A, negative pressures were recorded before the lights came on. Further to this, it would be most unusual if the phellogen were to behave as a perfect microporous partition: much of the stem material that we used had bud scars that proved to be more leaky than lenticels. The low pressures recorded by Grosse and Schröder (1985) indicate some considerable leakiness in their system. The pressurizations that we recorded under high illumination were accompanied by increased ROL but not in proportion to the pressure. Furthermore, the ROL from the artificial roots showed that the O2 concentrations in the stems were often much above ambient (48 % was the highest figure recorded; data not shown), a feature that could only be due to photosynthetic O2 production and accumulation within the stem, and not pressurization by thermal osmosis or humidity-induced diffusion.
The anatomical investigation revealed the presence of chloroplasts throughout the stem, in the secondary cortex, secondary phloem, secondary wood rays and the medulla (Fig. 10A–E), and it was possible to deduce that the pressure values and rates of pressurization that we recorded were more than matched by the pressurization that would have accompanied the rates of O2 output into the relatively small volume of gas space present in the stems. Furthermore, the stems achieved even greater pressures when fully submerged, and gas loss could only occur when the pressures were sufficient to overcome the surface tension in the lenticels and the pressures artificially applied above the submerging fluid. In these circumstances, there was no likelihood of either thermal osmosis or humidity-induced diffusion creating the pressures, only photosynthesis. With the stems partially emergent, we believe that the pressurizations were caused chiefly by photosynthetic O2 production and that the pressures developed were a function of (a) the degree to which the photosynthesis was dependent upon internally derived CO2 and (b) the balance between the rate of photosynthesis and the resistance to leakiness of the phellogen, lenticels, bud scars, etc., above the water line. Hypertrophied lenticels immediately above the water line, which characterize whole plants, may make the stems more leaky than the cuttings that we used where the lenticels were unmodified. It is also possible that sub-lenticellular phellogen will vary in its degree of leakiness during the year and the lenticels themselves can be capped by low porosity layers of ‘closing cells’ during dormant periods (Eames and MacDaniels, 1947).
Our measurements of photosynthesis on cuttings showed that a net output of O2 was increased by supplementing the CO2 concentrations in the xylem fluid. Although we did not attempt to measure levels of CO2 in the xylem of alder, the idea that in trees, respiratory CO2, derived from the roots and trunk, might be a source of carbon for photosynthesis is not new (MacDougal and Working, 1933; Zelawski et al., 1970; Stringer and Kimmerer, 1993; Hari et al., 1999; Aschen et al., 2001; Pfanz et al., 2002; Damesin, 2003). Stringer and Kimmerer (1993) reviewed earlier work on this subject and pointed out that ‘woody plants, especially tall trees, have a relatively long hydraulic path which provides ample time for CO2 to equilibrate between the pneumatic system of the soil plant continuum and the xylem path’. MacDougal (1927) found that, on a volume basis, CO2 concentrations at 12–20 cm into the interior of Quercus agrifolia trunks varied from 1·2 % in January to 15 % in August; in Salix lasiolepis, there was a minimum of 5 % in January rising to 10% in July; and for a Populus species he found 1·4 % in February which had risen to 8 % by the time the first seasonal increase in diameter had begun in early March. He also noted that the proportion of CO2 rose on days when the outside of the trunk was wetted by rain and that in other species with green trunks, such as Parkinsonia microphylla and the tree cactus Carnegiea gigantean, O2 concentrations sometimes exceeded atmospheric. However, he did not believe that photosynthesis could have been responsible for this. In their experiments, Stringer and Kimmerer (1993) found substantial amounts of CO2 (up to 2 mm) in xylem perfusate from 1-year-old stems of P. deltoides, a value that rose to 9·4 mm when the plants were in flooded soils. They also found that when excised leaves were allowed to transpire 1 mm [14C]NaHCO3, 99·6 % of the label was fixed in the light. More recently, for tobacco, Hibberd and Quick (2002) have reported a C4 type of photosynthesis in the cells of stems and petioles that surround the xylem and phloem, and that these cells are supplied with the carbon for photosynthesis from the vascular system and not from stomata; here the carbon source proved to be malate transported from the roots. It would be interesting to know whether this type of fixation might occur in alder and other flood-tolerant species.
Our study is different from those reported above in so far as it is the secondary tissues of the stem that are responsible for the photosynthesis and that chloroplasts are present in abundance even in what appear at first sight to be brown stems. We would also expect that the CO2 will reach the photosynthesizing cells via the xylem stream, perfusion or radial gas phase diffusion from there to the secondary phloem and cortex and by ‘long-distance’ gaseous diffusion via the secondary cortex of belowground parts. Other recent studies have concluded that substantial stem photosynthesis can take place in woody species (Aschan et al., 2001; Pfanz et al., 2002); the measurements, usually based on CO2 fluxes, rarely indicate a positive carbon gain in the light but rather a conservation of carbon through the recycling of a high proportion of the stem respiratory CO2. Undoubtedly we would have reached a similar conclusion for Black Alder had we relied on the measurement of CO2 fluxes rather than O2. By measuring O2 flux, however, we have been able to show a positive (net) gain in carbon by stem photosynthesis during the light period, with the stem respiratory CO2 source having been supplemented by the dissolved CO2 percolating through the xylem. In the field, it is very likely that xylem stream CO2 levels will be similarly enhanced (McGuire and Teskey, 2004) and particularly in wet soils, as found by Stringer and Kimmerer (1993) for P. deltoides.
The significance of the results in a wider ecological context is difficult to assess. Clearly, when the stems of young alder trees are illuminated, there may be the potential for a photosynthesis-derived increase in O2 concentration in the secondary cortex and this will increase the diffusive O2 transport to the roots. The extent to which this will occur will depend upon many factors, not least of which will be the CO2 availability. Our results suggest that if the basal parts of the stems were to be standing in water and illuminated, it could be that the O2 concentrations produced could be even higher. However, at night, when the major O2 source for the roots would be via the lenticels on the emergent stem, the O2 transport to the roots would be less than if the stems were fully emergent.
The data indicate that soil temperatures will play an important role in determining the extent to which the O2 produced in the stem will support root penetration. Clearly, a cooler regime around the roots much improved the O2 regime in the root and the rhizosphere (Fig. 13), but how this might translate into achievable rooting depths in the field we can only speculate. It was interesting that in plants maintained outdoors, adventitious roots developed from just below the water line ultimately became thick and woody at the base and penetrated to the bottom of the pots. However, in the early stages of secondary xylem, the appearance of inflated thin-walled cells (Fig. 10K) instead of thick-walled lignified elements (Fig. 10I) is probably indicative of some ‘O2 stress’ near to the tips.
In terms of improved root aeration, it may be only young trees in their establishment phase that will benefit from stem-derived photosynthesis. We have shown here that 2- to 3-year-old alder trees can benefit in this way, and previous studies in our laboratory revealed that 2- to 3-year-old Salix species can behave similarly (Moir, 1993; Dockerty, 1994; Hughes, 1996). Whether more mature trees with thicker bark at their base will behave similarly remains to be determined. However, with increasing girth, the proportion of gas-filled non-respiring elements in the xylem of trunk and woody roots could become considerable and might function as a major diffusive conduit for O2 transport to deeper roots. MacDougal and Working (1933) and Hook and Brown (1972) found that considerable gas movements (including radial ones) could be induced in the mature trunks of trees, and Gill (1975) noted that the mature root system of alder can survive a substantial period of flooding and remain alive. Further studies on this topic are needed.
Although the emphasis of this study has been on the possible role of stem photosynthesis in tree root aeration, we believe that our observations might point to something of much wider significance: the possibility that stem photosynthesis, supported by xylem-borne CO2, might also be of some importance for the general carbon balance of the plant.
Acknowledgments
We are very grateful to Mr Mike Bailey, Chemistry Department, University of Hull, for constructing various pieces of essential glassware for this project. We thank Dr D. J. Boatman for permission to take cuttings of alder from North Cliffe Wood, East Yorkshire, and Dr Tim Colmer and Dr Hans Brix for helpful comments on the manuscript.
LITERATURE CITED
- Armstrong J, Armstrong W. 1990. Light-enhanced convective throughflow increases oxygenation in rhizomes and rhizosphere of Phragmites australis (Cav.) Trin. ex Steud. New Phytologist 114: 121–128. [DOI] [PubMed] [Google Scholar]
- Armstrong J, Armstrong W, Beckett PM. 1992.Phragmites australis: Venturi- and humidity-induced convections enhance rhizome aeration and rhizosphere oxidation. New Phytologist 120: 197–207. [Google Scholar]
- Armstrong J, Armstrong W, Beckett PM, Halder JE, Lythe S, Holt R, Sinclair A. 1996. Pathways of aeration and the mechanisms and beneficial effects of humidity- and Venturi-induced convections in Phragmites australis (Cav.) Trin. ex Steud. Aquatic Botany 54: 177–198. [Google Scholar]
- Armstrong W. 1979. Aeration in higher plants. In: Woolhouse HWW, ed. Advances in botanical research Vol. 7. New York: Academic Press, 225–332. [Google Scholar]
- Armstrong W, Healy MT. 1984. Oxygen diffusion in pea. III. Changes in the oxygen status of the primary pea root attributable to an ageing of the tissues. New Phytologist 96: 179–185. [Google Scholar]
- Armstrong W, Armstrong J, Beckett PM. 1996. Pressurised ventilation in emergent macrophytes: the mechanism and mathematical modelling of humidity-induced convection. Aquatic Botany 54: 121–136. [Google Scholar]
- Armstrong W, Armstrong J, Beckett PM. 1996. Pressurised aeration in wetland macrophytes: some theoretical aspects of humidity-induced convection and thermal transpiration. Folia Geobotanica et Phytotaxonomica 31: 25–36. [Google Scholar]
- Aschan G, Wittmann C, Pfanz H. 2001. Age-dependent bark photosynthesis of aspen twigs. Trees 15: 431–437. [Google Scholar]
- Beckett PM, Armstrong W, Justin SHFW, Armstrong J. 1988. On the relative importance of convective and diffusive gas-flows in plant aeration. New Phytologist: 110: 463–468. [Google Scholar]
- Bendix M, Tornberg T, Brix H. 1994. Internal gas transport in Typha latifolia L., and Typha angustifolia L. I. Humidity-induced pressurisation and convective throughflow. Aquatic Botany 49: 75–90. [Google Scholar]
- Brix H, Sorrel BK, Orr PT. 1992. Internal pressurisation and convective gas flow in some emergent freshwater macrophytes. Limnology and Oceanography 37: 1420–1433. [Google Scholar]
- Brix H, Sorrell BK, Schierup HH. 1996. Gas fluxes achieved by in situ convective flow in Phragmites australis Aquatic Botany 54: 151–163. [Google Scholar]
- Büchel HB, Grosse W. 1990. Localisation of the porous partition responsible for pressurized gas transport in Alnus glutinosa (L.) Gaertn. Tree Physiology 6: 247–256. [DOI] [PubMed] [Google Scholar]
- Colmer TD. 2003. Long distance transport of gases in plants: a perspective on internal aeration and radial oxygen loss from roots. Plant Cell and Environment 26: 17–36. [Google Scholar]
- Dacey JWH. 1980. Internal winds in water-lilies: an adaptation for life in anaerobic sediments. Science 210: 1017–1019. [DOI] [PubMed] [Google Scholar]
- Dacey JWH. 1981. Pressurised ventilation in the yellow water-lily. Ecology 62: 1137–1147. [Google Scholar]
- Dacey JWH. 1987. Knudsen-transitional flow and gas pressurisation in leaves of Nelumbo Plant Physiology 8: 199–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Damesin C. 2003. Respiration and photosynthesis characteristics of current-year stems of Fagus sylvatica: from the seasonal pattern to an annual balance. New Phytologist 158: 465–475. [DOI] [PubMed] [Google Scholar]
- Dockerty SR. 1994.An investigation into claims of pressurised gas flows and their effects on aeration in willow and alder cuttings. Project Report. Hull: Department of Applied Biology, University of Hull. [Google Scholar]
- Gill CJ. 1975. The ecological significance of adventitious rooting as a response to flooding in woody species with special reference to Alnus glutinosa (L.) Gaertn. Flora 164: 85–97. [Google Scholar]
- Grosse W. 1996. The mechanism of thermal transpiration (= thermo-osmosis). Aquatic Botany 54: 101–110. [Google Scholar]
- Grosse W, Bauch C. 1991. Gas transfer in floating-leaved plants. Vegetatio 97: 185–192. [Google Scholar]
- Grosse W, Meyer, D. 1992. Effects of pressurized gas transport on nutrient uptake during hypoxia of Alder roots. Botanica Acta 105: 233–236. [Google Scholar]
- Grosse W, Schröder, P. 1984. Oxygen supply of roots by gas transport in Alder trees. Zeitschshrift für Naturforschung 39c: 1186–1188. [Google Scholar]
- Grosse W, Schröder P. 1985. Aeration of roots and chloroplast-free tissues of trees. Berichte der Deutschen Botanischen Gesellschaft 98: 311–318. [Google Scholar]
- Grosse W, Frye J, Lattermann S. 1992. Root aeration in wetland trees by pressurized gas-transport. Tree Physiology 10: 2285–2295. [DOI] [PubMed] [Google Scholar]
- Grosse W, Schulte A, Fujita H. 1993. Pressurized gas transport in two Japanese alder species in relation to their natural habitats. Ecological Research 8: 151–158. [Google Scholar]
- Grosse W, Jovy K, Tiebel H. 1996. Influence of plants on redox potentials and methane production in water-saturated soil. Hydrobiologia 340: 93–99. [Google Scholar]
- Hari P, Nygren P, Korpilahti E. 1999. Internal circulation of carbon within a tree. Canadian Journal of Forest Research 21: 514–515. [Google Scholar]
- Hibberd JM, Quick WP. 2002. Characteristics of C4 photosynthesis in stems and petioles of C3 flowering plants. Nature 415: 451–454. [DOI] [PubMed] [Google Scholar]
- Hook DD, Brown CL. 1972. Permeability of the cambium to air in trees adapted to wet habitats. Botanical Gazette 133: 304–310. [Google Scholar]
- Hook DD, Schlotens JR. 1978. Adaptations and flood tolerance of tree species. In: Hook DD, Crawford RMM, eds. Plant life in anaerobic environments. Ann Arbor, MI, Ann Arbor Science Publishers Inc., 299–331. [Google Scholar]
- Hughes JP. 1996.An investigation of the relative importance of throughflow and non-throughflow convection in root aeration of Salix cinerea sub sp. cinerea. Project Report. Hull, Department of Applied Biology, University of Hull. [Google Scholar]
- Idso SB. 1981. A technique for evaluating the potential for mass flow of gases in plants. Plant Science Letters 23, 47–53. [Google Scholar]
- Larcher W. 1969. The effect of environmental and physiological variables on the carbon dioxide gas exchange of trees. Photosynthetica 3: 167–198. [Google Scholar]
- MacDougal DT. 1927. Composition of gases in trunks of trees. Carnegie Institution of Washington Year Book 26: 162–163. [Google Scholar]
- MacDougal DT, Working EB. 1933.The pneumatic system of plants, especially trees. Publication No. 441, Washington, DC: Carnegie Institution of Washington. [Google Scholar]
- 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: 571–578. [DOI] [PubMed] [Google Scholar]
- Mevi-Schultz J, Grosse W. 1988. A two-way gas transport system in Nelumbo nucifera Plant Cell and Environment 11: 27–34. [Google Scholar]
- Moir SE. 1993.Investigations of gas transport induced by pressurisation in Salix viminalis and Populus nigra var. italica. Project Report. Hull: Departemt of Applied Biology, University of Hull. [Google Scholar]
- Pedersen O, Sand-Jensen K. 1992. Adaptations of submerged Lobelia to aerial life form: morphology, carbon sources and oxygen dynamics. Oikos 65: 89–96. [Google Scholar]
- Pedersen O, Sand-Jensen K, Revsbech N-P. 1995. Diel pulses of O2 and CO2 in sandy lake sediments inhabited by Lobelia dortmana Ecology 76: 1536–1545. [Google Scholar]
- Pfanz H, Aschan G, Langenfeld-Heyser R, Wittmann C, Loose M. 2002. Ecology and ecophysiology of tree stems: corticular and wood photosynthesis. Naturwissenschaften 89: 147–162. [DOI] [PubMed] [Google Scholar]
- Sand-Jensen K, Pedersen O, Binzer T, Borum J. 2005. Contrasting oxygen dynamics in the freshwater isoetid Lobelia dortmanna and the marine seagrass Zostera marina Annals of Botany 96: 613–623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schröder P. 1989. Aeration of the root system in Alnus glutinosa L. Gaertn. Annals of Forestry Science 46 (Suppl): 310s–314s. [Google Scholar]
- Schröder P. 1989. Characterization of a thermo-osmotic gas transport mechanism in Alnus glutinosa (L.) Gaertn. Trees 3: 38–44. [Google Scholar]
- Sorrell BK, Boon PI. 1994. Convective gas flow in Eleocharis sphacelata R.Br.: methane transport and release from wetlands. Aquatic Botany 47: 197–212. [Google Scholar]
- Steinburg SL. 1996. Mass and energy exchange between the atmosphere and leaf influence on gas pressurization in aquatic plants. New Phytologist 134: 587–600. [DOI] [PubMed] [Google Scholar]
- Stringer JW, Kimmera TW. 1993. Refixation of xylem sap CO2 in Populus deltoides Physiologia Plantarum 89: 243–251. [Google Scholar]
- Takaishi T, Sensui Y. 1963. Thermal transpiration effect of hydrogen, rare gases and methane. Transactions of the Faraday Society 59: 2503–2514. [Google Scholar]
- Waters I, Armstrong W, Thomson CJ, Setter TL, Adkins S, Gibbs J, Greenway H. 1989. Diurnal changes in radial oxygen loss and ethanol metabolism in roots of submerged and non-submerged rice seedlings. New Phytologist 113: 439–451. [Google Scholar]
- Zelawski W, Riech FP, Stanley RG. 1970. Assimilation and release of internal carbon dioxide by woody plant shoots. Canadian Journal of Botany 48: 1351–1357. [Google Scholar]