Oxygen dynamics during submergence in the halophytic stem succulent Halosarcia pergranulata



    1. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia,
    2. Freshwater Biological Laboratory, Biological Institute, University of Copenhagen, Helsingørsgade 51, DK-3400 Hillerød, Denmark and
    Search for more papers by this author
  • H. VOS,

    1. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia,
    2. Plant Ecophysiology, Institute of Environmental Biology, Faculty of Science, Utrecht University, Sorbonnelaan 16, 3584 CA, Utrecht, the Netherlands
    Search for more papers by this author
  • T. D. COLMER

    Corresponding author
    1. School of Plant Biology, Faculty of Natural and Agricultural Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia,
    Search for more papers by this author

T. D. Colmer. Fax: +61 86488 1108; e-mail: tdcolmer@cyllene.uwa.edu.au


This study elucidated O2 dynamics in shoots and roots of submerged Halosarcia pergranulata (Salicornioideae), a perennial halophytic stem succulent that grows on flood-prone mudflats of salt lakes. Oxygen within shoots and roots was measured using microelectrodes, for plants when waterlogged or completely submerged, with shoots in light or in darkness, in a controlled environment. Net photosynthesis (PN) when underwater, at a range of dissolved CO2 concentrations, was measured by monitoring O2 production rates by excised stems. The bulky nature and apparently low volume of gas-filled spaces of the succulent stems resulted in relatively high radial resistance to gas diffusion. At ambient CO2, quasi-steady state rates of PN by excised succulent stems were estimated to be close to zero; nevertheless, in intact plants, underwater photosynthesis provided O2 to tissues and led to radial O2 loss (ROL) from the roots, at least during the first several hours (the time period measured) after submergence or when light periods followed darkness. The influence of light on tissue O2 dynamics was confirmed in an experiment on a submerged plant in a salt lake in south-western Australia. In the late afternoon, partial pressure of O2 (pO2) in the succulent stem was 23.2 kPa (i.e. ∼10% above that in the air), while in the roots, it was 6.2–9.8 kPa. Upon sunset, the pO2 in the succulent stems declined within 1 h to below detection, but then showed some fluctuations with the pO2 increasing to at most 2.5 kPa during the night. At night, pO2 in the roots remained higher than in the succulent stems, especially for a root with the basal portion in the floodwater. At sunrise, the pO2 increased in the succulent stems within minutes. In the roots, changes in the pO2 lagged behind those in the succulent stems. In summary, photosynthesis in stems of submerged plants increased the pO2 in the shoots and roots so that tissues experience diurnal changes in the pO2, but O2 from the H2O column also entered submerged plants.


Flooding can result in complete submergence of terrestrial plants in many situations, the occurrence depending on location and climatic events. The frequency, durations and depths of floods can influence species distributions in some landscapes; species present at some locations can be determined by, among other factors, ecophysiological responses to flooding (Voesenek et al. 2004).

When plants become submerged, the supply of O2 diminishes because diffusion of dissolved gases in H2O is 10 000-fold slower than that in air (Armstrong & Drew 2002). Oxygen deficiency can be most severe in the roots, particularly at night-time, so that respiration is inhibited (e.g. submerged rice, Waters et al. 1989) causing an ‘energy crisis’ in anoxic cells/tissues (Gibbs & Greenway 2003). In addition, the slow gas diffusion in H2O can limit CO2 uptake by leaves (Smith & Walker 1980), and furthermore, light availability decreases because of its attenuation in the H2O column (Holmes & Klein 1987; Kirk 1994), so that photosynthesis can be impeded. In flooded soils, O2 is consumed and then microorganisms use a range of alternative electron acceptors, so that compounds such as Mn2+, Fe2+, S2–, H2S and carboxylic acids can accumulate to levels that are toxic to plants (Ponnamperuma 1984; McKee & McKevlin 1993).

The sources of O2 potentially available to plants when completely submerged are (1) from the H2O column or (2) from photosynthesis (Pedersen, Binzer & Borum 2004). Submerged plants therefore experience marked changes in tissue O2 levels, as dependent on incident light, partial pressure of O2 (pO2) being lowest during the night and highest during daytime (Sorrell & Dromgoole 1987; Waters et al. 1989; Sand-Jensen et al. 2005). Oxygen in the shoot, be it produced in photosynthesis or via entry from the H2O column, diffuses throughout the plant via the aerenchyma and gas-filled intercellular spaces. Oxygen supply to roots in flooded soil is determined by the pO2 within the shoots, as well as the total physical resistance and the consumption of O2 in respiration, along the diffusion path (Armstrong 1979). Dynamics in root O2 supply in some submerged plants have been described [e.g. freshwater Lobelia dortmanna (Pedersen, Sand-Jensen & Revsbech 1995; Sand-Jensen et al. 2005), rice (Waters et al. 1989), Eriophorum angustifolium (Gaynard & Armstrong 1987) and seagrasses (Pedersen et al. 1998; Connell, Colmer & Walker 1999; Greve, Borum & Pedersen 2003)]. For the emergent halophyte Spartina alterniflora growing in a salt marsh, tidal inundation during darkness leads to a severe depletion of internal O2 (Gleason & Zieman 1981). This study is the first to evaluate these processes in a succulent terrestrial halophyte, Halosarcia pergranulata (P.G. Wilson) ssp. pergranulata (Salicornioideae), which can be seasonally submerged in inland salt lakes.

H. pergranulata is a perennial halophytic stem succulent that inhabits the margins and upper reaches of the flood-prone playas (i.e. mudflats) of salt lakes throughout southern Australia (Wilson 1980). Similarly, in North America, inland saline playas are typically colonized by succulent halophytes, namely, annual species of Salicornia (stem succulent) or Suaeda (leaf succulent) (Ungar 1974). However, 2 weeks of submergence resulted in death of Salicornia europaea at a salt lake in Ohio (Egan & Ungar 2000), whereas in a salt lake near Kalgoorlie in Australia, many individuals of H. pergranulata survived submergence lasting up to 10 months (English 2004). This study elucidated O2 dynamics in shoots and roots of H. pergranulata when waterlogged or completely submerged, in conditions of light and darkness, as well as determined the capacity of stems for photosynthesis when underwater.


Study site and plant materials

Yenyening Lakes was selected as our study site, as it was known that populations of H. pergranulata at this lake are submerged each winter. The lakes are ∼150 km east of Perth, Australia, and the site was on the northern shore (32°14′′48′ S, 117°10′′58′ E). Yenyening Lakes, like all salt lakes in the region, are ephemeral (Bowler 1981) and Na+ and Cl are the dominant ions (Geddes et al. 1981), having been deposited in the catchments over millennia as salt aerosols blown inland from the ocean (McArthur et al. 1989).

Submerged or emergent H. pergranulata plants (0.2–0.4 m in height) were collected by digging up individuals in largely intact blocks of soil (diameter, ∼0.15 m; depth, ∼0.10 m). Possibly, the submerged plants had been inundated for several weeks (0.25–0.5 m H2O depth), whereas the emergent plants that were used would not have experienced flooding by surface H2O exceeding 0.05 m.

Key environmental parameters at the site were measured (Table 1). H2O samples were collected from the middle depth of the H2O column, at sunset and sunrise, to measure pH, total alkalinity using Gran titration (Stumm & Morgan 1996), temperature and salinity. Free CO2 was calculated from total alkalinity, salinity, temperature and pH, using the equations of Stumm & Morgan (1996). Total salts in 25 mL samples of lake H2O were determined by evaporating the H2O and then weighing the salts. Soil redox potential (EH) was measured at 0.05–0.10 m depth, using platinum (Pt)-wire electrodes inserted into the soil, with the readings taken 24 h after insertion, and using an Ag/Ag-Cl reference together with the redox probes connected to an mV meter (Patrick, Gambrell & Faulkner 1996). Soil solution pH was measured for cores, at ∼0.05 m depth, using a portable pH electrode (pHScan WP2; Eutech Instruments Pte Ltd, Singapore). All parameters were measured in duplicate, except total salts (n = 5) and soil EH and pH (n = 4).

Table 1.  Environmental parameters from Yenyening Lakes on August 2005 during in situ measurements of intraplant O2 dynamics (see Fig. 3)
ParameterSunset (17:55 h)Sunrise (06:41 h)
  1. H2O column, soil O2 and incident quantum flux density [photosynthetically active radiation (PAR)] are shown in Fig. 3. CO2 concentration was calculated from total alkalinity, salinity, temperature and pH, using the equations of Stumm & Morgan (1996). Values given are means [n = 2 for all parameters except salinity (n = 5) and soil pH and EH (n = 4)]. H2O samples were taken in the middle of the H2O column. Soil EH was measured at 0.05–0.10 m, and pH was measured at 0.05 m depth.

  2. EH, soil redox.

 pH 8.11 ± 0.005 7.84 ± 0.005
 Total alkalinity (meq L−1)1.943 ± 0.0121.955 ± 0.048
 CO2 (µM) 11.8 ± 0.9 25.1 ± 3.2
 Temperature (°C) 15 13
 Salinity (kg m−3)   31 ± 0.35  
 EH (mV)−18 ± 49 
 pH 6.22 ± 0.08 

Pots containing the plants with soil from the lake were placed in a 20/15 °C day/night phytotron; submerged individuals were kept in aquaria containing H2O from the lake, whereas emergent individuals were waterlogged with lake H2O to 0.02 m above the soil surface. Experiments were conducted on intact individuals, or on excised branches. Photosynthesis experiments (see subsequent section) were conducted within 5 d of collection of the plants; branches were excised from plants in the phytotron and were then used immediately in these experiments. The branches are cylindrical and have been described as succulent articulated stems, but after some time, the outer succulent tissues on the older parts senesce, leaving a ‘woody’ stem base which can increase in girth via secondary growth. Experiments on O2 dynamics in intact plants (only previously waterlogged individuals) were conducted 8–20 d after collection. Following collection from the field, the plants continued to produce new adventitious roots and new succulent stems when placed in pots in the phytotron.

Intraplant O2 dynamics


Two days before the experiments, all adventitious roots (or basal portions of roots) above the soil were covered with a mixture of 90% washed river sand and 10% soil from Yenyening Lakes, and the plants were kept waterlogged. A pot containing H. pergranulata was placed in an aquarium allowing initial waterlogging followed by complete submergence in artificial lake H2O, with the following composition: K2SO4 (5 mol m−3), CaCl2·2H2O (10 mol m−3), Na2SO4 (4.5 mol m−3), NaHCO3 (1 mol m−3), MgSO4·7H2O (1.0 mol m−3), Fe-ethylenediaminetetraacetic acid (EDTA) (0.05 mol m−3), H3BO3 (6.25 × 10−3 mol m−3), MnSO4·H2O (5.0 × 10−4 mol m−3), ZnSO4·7H2O (5.0 × 10−4 mol m−3), CuSO4·5H2O (1.25 × 10−4 mol m−3), Na2MoO4·2H2O (1.25 × 10−4 mol m−3), also with NaCl at 399 mol m−3 (to give a total electrical conductivity (EC) equal to that measured for H2O at Yenyening Lakes at the time of plant collection). During submergence, the H2O was aerated to maintain atmospheric saturation of gases and to provide H2O circulation in the aquarium. Light was provided by a slide projector to a midcanopy quantum flux density [photosynthetically active radiation (PAR)] of approximately 350 µmol m−2 s−1 (LI-1400 data logger with LI-190S 2π sensor; Li-Cor, Lincoln, NE, USA).

Soil O2 concentration (and H2O column O2 concentration during submergence) was continuously monitored with a Clark-type O2 mini electrode (OX-500; Unisense, Aarhus, Denmark). To access roots, soil was gently removed using small pulses of artifical lake H2O to expose the basal portion of an adventitious root. A Clark-type O2 microelectrode with guard cathode with a tip diameter of 25 µm (Revsbech 1989; OX-25, Unisense) was inserted 150 µm into an adventitious root by using a micromanipulator (MM5; Märzhäuser, Wetzlar, Germany). After insertion, the exposed root was again carefully covered with 0.02–0.03 m of soil, and the setup was left for 1 h to allow re-establishment of the geochemical profiles around the root (Pedersen et al. 2004). In addition, an O2 microelectrode with a tip diameter of 50 µm (OX-50, Unisense) was inserted 250 µm into a succulent stem of the same branch from which the adventitious root originated. The microelectrodes were connected to a pA meter (PA8000, Unisense), and the outputs were logged every 10 s on a computer using an analogue-to-digital converter (ADC-16; Pico Technology, Cambridgeshire, England). H2O temperature was logged using type-K thermocouples connected to a resistance converter (TC-08, Pico Technology).

Radial O2 profiles of succulent stems were measured in steps of 10 or 50 µm, using an O2 microelectrode with a tip diameter of 50 µm. Branches of submerged plants, with succulent articulations on the upper half and woody stem on the lower half, were excised, and the cut basal end was sealed with petroleum jelly (to prevent gas entry via the cut end) before each branch was mounted in a Petri dish, and kept completely submerged in artificial lake H2O (as previously mentioned). Stirring was maintained by a 17-gauge syringe connected to an air pump, and light was provided by a fibre optic lamp (model LGPS; Olympus, Tokyo, Japan) at a quantum flux density (PAR) of 1500 µmol m−2 s−1. After each profile, hand sections were prepared to confirm the electrode trace through the succulent stem.


Intraplant O2 dynamics in roots and in a succulent stem of H. pergranulata were followed in situ for a plant at Yenyening Lakes. A three-channel underwater pA meter with built-in data logger (PA3000UP-OP, Unisense) and 50 µm tip diameter O2 microelectrodes (OX50-UW, Unisense) were used. Micromanipulators with microelectrodes were mounted on Al-stands fixed in the lake bottom, and the microelectrodes were positioned using changes in signal to detect the surface of roots or succulent stems (Borum et al. 2005). Oxygen microelectrodes were inserted 150 µm into two different adventitious roots and 250 µm into the succulent stem from which the two roots originated. One root was already growing completely into the soil, but the root–shoot junction was covered with an additional 10 mm of fine soil to prevent any radial exchange of O2 between the H2O column and tissue. By contrast, 25 mm of the root base, including the root–shoot junction, of the other root that was studied was exposed to the H2O column and left in this condition. The same electrode, data logger and light-sensor equipment that were used in the laboratory experiments were also used at the lake for measurements of temperature, H2O column and soil O2, and PAR. Data were recorded from the late afternoon throughout the dark period until late in the morning the following day.

Succulent stem tissues from submerged or emergent individuals were sampled in the field at sunrise to evaluate possible effects of submergence on tissue sugar status. Branches (terminal 5–7 segments) were excised from five replicate plants for each growth condition, rinsed with deionized (DI) H2O, blotted to remove surface H2O, fresh mass measured, wrapped in Al-foil, frozen in dry ice, transported to the laboratory and freeze dried; dry mass was recorded, and then the tissues were pulverized in a ball mill. Sugars were extracted twice from 20 mg dry mass in 1 mL of 80% ethanol boiled with reflux for 20 min. Total sugars in the extracts were measured using anthrone (Yemm & Willis 1954). Reliability of the method was verified by determining the recovery of spikes of glucose into the tissue, added immediately before extraction.

Net photosynthesis (PN)

Underwater PN was measured as net O2 production by excised stems. Three types of stems were evaluated: (1) green non-succulent (i.e. woody) stems from submerged plants; (2) succulent stems of submerged plants; and (3) succulent stems of emergent plants. Approximately 2 g of fresh tissue was mounted on mesh within a custom-built 77 mL cylindrical transparent perspex chamber, which was illuminated with PAR at 1500 µmol m−2 s−1. The chamber was filled with filtered (0.2 µm) natural lake H2O that was air saturated and adjusted to a total dissolved inorganic carbon (DIC) level of 10 mol m−3 using NaHCO3, and stirred with a magnetic stirrer. The lid was fitted with a Clark-type O2 mini electrode (OX500, Unisense) connected to a two-channel pA meter (PA2000, Unisense). The O2 signal was logged every 10 s using the same data logger equipment as previously described. The pH in the chamber was recorded using a pH electrode. A concentration of 0.1 m HCl or 0.1 m NaOH was injected to manipulate the pH and thus, the amount of available CO2. The tissue in the chamber was provided with light for 2 h before each experiment to deplete internal CO2; NaOH was then injected to obtain a pH of 8.4 and photosynthesis was recorded over a period of 15 min. In time intervals of 15 min, CO2 availability was increased in logarithmic steps by lowering the pH via injection of HCl until a final pH of 6.9. Free CO2 was calculated from total alkalinity, salinity, temperature and pH, using the equations of Stumm & Morgan (1996).

Diameters and lengths of the stems used were measured, and surface area was calculated. The tissues were then frozen in liquid N2, freeze dried and pulverized in a ball mill. Chlorophyll was extracted from subsamples (20 mg dry mass) in 100% methanol (1.25 mL) for 30 min, centrifuged [Micromax model 3591, International Equipment Company (IEC), Needham Heights, MA, USA] at a relative centrifugal force (rcf) of 100 g, and supernatants were collected, all at 4 °C in the dark. Chl a was determined by absorbance measurements at wavelengths of 665.2 and 652.4 nm, using a glass cuvette in a UV-visible spectrophotometer (model 1601; Shimadzu, Tokyo, Japan) and the equations in Wellburn (1994).

The data sets for the responses of photosynthesis to dissolved CO2 in the external medium were fitted to a Hill & Whittingham (1955) model to estimate apparent half saturation contant (Ks) and maximum velocity (Vmax). The Hill–Whittingham model was used to account for possible boundary layer effects on the CO2 flux from the medium to the plant tissue (Smith & Walker 1980).

PN in the air at a quantum flux density of 1500 µmol m−2 s−1 and ambient CO2 (350 µL L−1) was also measured on excised succulent stems of emergent H. pergranulata, using an infrared gas analyser (LI-6400, Li-Cor).

Radial O2 loss (ROL) from adventitious roots

Rates of ROL from intact adventitious roots when in an O2-free medium were measured using cylindrical root-sleeving O2 electrodes [inside diameter (i.d.), 2.25 mm; height, 5.0 mm] fitted with guides (Armstrong & Wright 1975; Armstrong 1994). Branches (n = 4) with several adventitious roots were excised, and the base of the branch and the roots were immersed in transparent perspex chambers containing deoxygenated solution of the following composition: 0.1% agar (w/v) and (in mol m−3) NaCl, 400; KCl, 5.0; CaSO4, 0.5. The chambers were covered with Al-foil so that the roots were in darkness. The branches were held with wet cotton wool in a rubber lid sealed onto the top of each chamber so that the shoots were initially in the air. The chambers were placed within empty aquaria (25 L), in preparation for subsequent submergence. Two hours later, the ROL was measured at 60, 40, 20, 10 and 5 mm behind the root tip. The electrode was then again positioned at 60 mm behind the root tip, and after the ROL returned to a new quasi-steady state, the shoots were completely submerged in artificial lake H2O (composition as given under the heading Intraplant O2 dynamics). The rates of ROL from the roots were measured with the shoots in light and then also in darkness. The experiment was conducted at 20 °C with a quantum flux density of 250 µmol m−2 s−1 at shoot height. Root diameters were measured using a microscope (Axiovert 100; Zeiss, Jena, Germany) with a calibrated eyepiece reticle. The ROL was calculated according to Armstrong & Wright (1975).

Stem and root anatomy

Transverse sections of succulent stems and roots were taken using a hand-held razor, viewed using a microscope (model SZH ILLD, Olympus) and photographed using a digital camera. The proportion of each root cross section occupied by the aerenchyma and intercellular gas-filled spaces was determined using the public domain software ImageJ 1.32j (NIH 2005).


A Student’s t-test in Microsoft (MS) Excel was used to test for differences in sugar concentrations between submerged and waterlogged plants. GraphPad Prism 4.0 was used to run the Hill & Whittingham (1955) models of the CO2 response curves, to estimate Ks and Vmax, and associated statistics.


Intraplant O2 dynamics

A series of replicated experiments was conducted to evaluate the pO2 in succulent stems and adventitious roots (in soil) of plants under a range of conditions, namely, waterlogged with shoots in the air, completely submerged and when in light or darkness. In all the experiments, the pO2 in the H2O column and in the waterlogged soil were monitored; in the H2O column, the pO2 remained at atmospheric saturation as a result of bubbling with air, whereas at 15–20 mm below the soil surface (the depth of the position measured within the roots studied), O2 was always below detection (i.e. below 0.05 kPa). Responses of the pO2 to the treatments will firstly be presented in traces with time (Figs 1–3), and the replicated data at each new quasi-steady state are summarized in Table 2.

Figure 1.

Partial pressures of O2 (pO2) in a succulent stem, root, soil and H2O column during transition from waterlogged to completely submerged Halosarcia pergranulata, at 20 °C. Measured for intact plants in pots containing soil and flooded in air-saturated artificial saline lake H2O, in a well-stirred aquarium. The plant used was collected from Yenyening Lakes, from an area with ∼0.01 m H2O. The estimated 90% response times from one quasi-steady state to another were 40 and 42 min in the stem and root, respectively. The proportion of cross-sectional area of gas-filled spaces (aerenchyma + intercellular spaces) at the point of electrode insertion into the root was 0.19. The horizontal arrow indicates H2O column O2 concentration (atmospheric saturation) during submergence.

Figure 2.

Partial pressures of O2 (pO2) in a succulent stem and root of submerged Halosarcia pergranulata, in the soil and the H2O column, during light–dark switches, at 20 °C. Measured for intact plants in pots containing soil and submerged in air-saturated artificial saline lake H2O, in a well-stirred aquarium. The plant used was collected from Yenyening Lakes, from an area with ∼0.01 m H2O. The estimated 90% response times from quasi-steady state in the dark to steady state in light were 3.6 and 29.4 min in the stem and root, respectively. Ninety per cent response times from quasi-steady state in light to quasi-steady state in the dark were 3.0 and 23.4 min in the stem and root, respectively. The proportion of cross-sectional area of gas-filled spaces (aerenchyma + intercellular spaces) at the point of electrode insertion into the root was 0.24.

Figure 3.

Partial pressures of O2 (pO2) in a succulent stem and two roots of submerged Halosarcia pergranulata, and in the H2O column and soil, from late afternoon to the following morning in Yenyening Lakes (lower panel). The quantum flux density [photosynthetically active radiation (PAR)] with time is shown in the upper panel. The plant was completely submerged, and the H2O depth was ∼0.25 m. The proportion of cross-sectional area of gas-filled spaces (aerenchyma + intercellular spaces) at the point of electrode insertion was 0.10 in the completely buried root and 0.15 in the root with its base exposed to the H2O column.

Table 2.  Quasi-steady state of intraplant partial pressures of O2 (pO2) (kPa) in succulent stems and adventitious roots during waterlogging or complete submergence in Halosarcia pergranulata, in light or darkness
TreatmentLight conditionsOrganpO2 (kPa)
  1. Measured for intact plants in pots containing lake soil and flooded in artificial saline lake H2O at 20 °C, in a well-stirred aquarium. Mean ± SE, n = 2–5.

WaterloggedLightSucculent stem15.2 ± 3.3 (n = 2)
Root 6.0 ± 2.6 (n = 2)
DarknessSucculent stem15.7 ± 1.9 (n = 3)
Root 5.1 ± 1.2 (n = 3)
SubmergedLightSucculent stem17.3 ± 2.0 (n = 4)
Root 2.2 ± 0.8 (n = 4)
DarknessSucculent stem 3.6 ± 1.6 (n = 5)
Root 0.7 ± 0.2 (n = 5)

The effect of submergence on tissue pO2 was evaluated for plants in darkness, to avoid the confounding effects of photosynthetically produced O2. Upon complete submergence, direct access to atmospheric O2 is abolished, causing a fast decline in the pO2 in both the stems and roots (Fig. 1). For the three replicate plants, the pO2 in succulent stems was reduced during submergence to 23% of the value when shoots were in the air, and to only 14% in roots (calculated from Table 2). The 90% response times to reach the new quasi-steady states in the succulent stem and root of the plant used in Fig. 1 were 40 and 42 min, respectively. Although the decreases in tissue pO2 were substantial upon submergence in darkness, some O2 remained present, even in the roots surrounded by anoxic soil (Fig. 1, Table 2), indicating inward diffusion of O2 from the H2O column into the shoots and down to the roots.

Succulent stems in the air did not differ in pO2 when in light or in darkness (Table 2); however, the values (15–16 kPa) measured at a depth of 250 µm in the stems were significantly below the pO2 in the air (20.6 kPa). When submerged, light conditions impacted on tissue O2 status; the quasi-steady state pO2 in succulent stems in light was fivefold higher than that in darkness (Table 2). Moreover, photosynthetically derived O2 also moved to the roots, resulting in a threefold higher pO2 in roots of submerged plants when in light, compared with that in darkness (Table 2).

The dynamics of pO2 in submerged plants caused by light–dark cycles are shown in Fig. 2. A conspicuous peak of increased pO2 always developed in the succulent stems soon after switching from darkness to light, and this pO2 peak was also apparent in the roots. In the succulent stems, the peaks reached up to 39 kPa, but had almost disappeared after 45–60 min, whereas the peak persisted longer in the roots. Upon switching from darkness to light, the 90% response times to reach the new quasi-steady states in the shoot and root were 3.6 and 29.4 min, respectively, while for switches from light to darkness, the 90% response times were 3.0 and 23.4 min, respectively (derived from Fig. 2). Overall, these data demonstrate the importance of underwater photosynthesis to submerged plants in determining tissue O2 status in shoots and roots.

The influence of light on tissue O2 dynamics in submerged plants was verified by diurnal measurements of pO2 in a stem and roots of a plant in Yenyening Lakes (Fig. 3). In the late afternoon, the pO2 in the succulent stem was 23.2 kPa (i.e. ∼10% above that in the air), while in the roots, it was 6.2–9.8 kPa. Upon sunset, the pO2 in the succulent stem declined within 1 h to below detection, whereas in the H2O column, it had only declined by 10% during the same period. The much lower pO2 in darkness in plants in the field, when compared with those in the laboratory (Fig. 2), presumably resulted from a much lower turbulent mixing in the lake H2O (estimated at less than 0.01 m s−1, at least at dusk when wind speed was low) than in the vigorously bubbled H2O in the laboratory experiments (estimated at ∼0.06 m s−1) which would have resulted in larger boundary layers adjacent to stems in the field situation. In the roots, the pO2 also declined markedly after sunset; however, the response was slower, and the pO2 did not decrease below 4.0 (root with base in the H2O column) or 1.2 kPa (root completely in soil). During the night, the pO2 in the H2O column continued to decline slowly, reaching a minimum of 17 kPa near sunrise. In the plant, the pO2 remained low during night-time, but was not constant, and the changes in tissue pO2 did not merely follow the trend of O2 concentration in the H2O column. We hypothesize that the periods with increasing tissue pO2 were due to higher turbulence in the H2O column (as surface wind was variable through the night) which would have decreased boundary layers adjacent to the plant, and thereby enhanced diffusion of O2 from the H2O column into the plant. At sunrise, tissue pO2 increased rapidly; in the succulent stem, the initial peak at 40.4 kPa was much higher than atmospheric levels (20.6 kPa), and this supersaturated O2 level then declined within 2.5 h to approximately the pO2 recorded in this tissue in the previous afternoon. The roots also showed increases in pO2 after sunrise, but these lagged behind those in the succulent stems. Although the pO2 in the submerged plant in the field was more dynamic than in plants under laboratory conditions, the data confirmed the importance of light in determining O2 status in submerged plants.

We also determined the concentrations of total soluble sugars in the succulent stems of emergent and submerged plants in the field. Samples were taken at sunrise, the time when tissue sugar concentrations are likely to be lowest (e.g. for Atriplex, Aslam et al. 1986). The total sugars were 56 ± 7.9 mg hexose equivalents g−1 dry mass in stems of emergent plants, and 30 ± 3.4 in those of submerged plants (means significantly different at P < 0.05). The ethanol-insoluble dry mass of emergent stems was 294 ± 10 mg g−1 total dry mass, and for submerged stems, it was 339 ± 13. So, when expressed on an ethanol-insoluble dry mass basis (hexose equivalents g−1 ethanol-insoluble dry mass), the sugar concentrations in the stems were 190 ± 27 mg g−1 (emergent plants) and 75 ± 9 mg g−1 (submerged plants). The fresh:dry mass ratios of the stems were also determined as emergent succulent stems, 13.8 ± 0.27; submerged succulent stems, 13.9 ± 0.66; and submerged non-succulent (i.e. woody) stems, 3.4 ± 0.11.

ROL from adventitious roots

The rates of ROL along adventitious roots when in a deoxygenated medium were measured to further evaluate internal O2 transport from shoots to roots in emergent and submerged conditions. Firstly, profiles of the ROL along adventitious roots were measured (shoots in the air and in light, roots in O2-free medium in darkness), and this showed highly variable rates and patterns along individual roots (Fig. 4a). In three of four cases, the ROL decreased towards the tip of roots, indicating that these roots do not contain a barrier to ROL. However, the fourth root displayed a marked increase in the ROL towards the tip, a pattern typical when a barrier to ROL is present (Armstrong 1979; Colmer 2003). This variability is likely due to the roots growing in a variable environment; even on the same branch, different proportions of the roots were exposed to the H2O column or soil. Such differences in growth conditions might be expected to result in a different expression of a barrier to ROL, at least as demonstrated in some other species (Colmer 2003). Unfortunately, we could not confidently distinguish these root types when conducting the experiments. Secondly, for the two longest roots used in the experiments, we then submerged these plants and evaluated the effect of light on the ROL at 60 mm behind the apex of the root. When light was switched off, the ROL decreased with a 90% response time of 12 and 14.5 min for the two roots (derived from Fig. 4b). The marked dependence of the ROL from the roots on light supply to the shoots of submerged plants supports the intraroot measurements of O2 dynamics previously described and extends those internal measurements by demonstrating effects also on rhizosphere O2 supply.

Figure 4.

Radial O2 loss (ROL) from intact roots of Halosarcia pergranulata, at 20 °C. (a) ROL measured along four different roots when shoots were in the air and in light and (b) after submergence, ROL time trace showing the change in O2 flux from roots at quasi-steady state with shoots in light to quasi-steady state in darkness. When submerged, the aquarium contained well-stirred, air-saturated artificial saline lake H2O, with the base of each branch with adventitious roots sealed into an O2-free medium in a perspex chamber within the aquarium. Measured for intact roots of branches collected from Yenyening Lakes from an area with ∼0.15 m H2O (in the field, approximately half of each branch was above H2O). The estimated 90% response times for ROL from roots of submerged branches, from quasi-steady state in light to quasi-steady state in darkness, ranged from 12 to 14.5 min. The proportion of cross-sectional area occupied by gas-filled spaces (aerenchyma + intercellular spaces) at 60 mm behind the apex of the roots used in the measurements (n = 4) was 0.13 ± 0.008.


The importance of underwater photosynthesis for intraplant O2 status (see Intraplant O2 dynamics) prompted us to directly evaluate the PN capacity of stem tissues. The maximum capacity (i.e. apparent Vmax) for underwater PN was higher in previously emergent versus submerged succulent stems, by 4.4-fold on a surface area basis (Fig. 5a) and by 4.2-fold on a Chl a basis. The apparent Ks on a surface area basis for the underwater PN CO2-response curve for submerged succulent stems was only 44% of the value for previously emergent succulent stems, but when expressed per unit Chl a, the previously emergent and submerged succulent stems did not differ in Ks (Table 3). A likely explanation for these differences in Ks is that the Hill & Whittingham (1955) model fitted to the data had only moderate predictive power for the previously submerged stems (r2 = 0.55 for surface area basis and 0.63 for Chl a basis), so that these Ks values have considerable uncertainty. We also measured the PN by non-succulent stems, as these increase as a proportion of the shoot during prolonged submergence (own observations); the ‘woody, green stems’ do photosynthesize, albeit at rates substantially slower than those of the succulent stems (Table 3). Finally, we measured the PN by emergent succulent stems when at ambient conditions in the air (350 µL L−1 CO2); on a surface area basis, the rate was 2.8 µmol CO2 m−2 s−1.

Figure 5.

Underwater net photosynthesis (PN) of excised succulent stems (apical five segments) from previously emergent or submerged Halosarcia pergranulata, on (a) surface area and (b) Chl a bases. Measurements were conducted in a well-stirred cuvette containing air-saturated, filtered H2O from Yenyening Lakes. The quantum flux density [photosynthetically active radiation (PAR)] was 1500 µmol m−2 s−1, at 20 °C. The Hill & Whittingham (1955) model was fitted to each data set to estimate the apparent half saturation constant (Ks) and maximum velocity (Vmax) (values shown in Table 3). The plants that were used were from Yenyening Lakes; waterlogged plants from an area with ∼0.01 m H2O and completely submerged plants from 0.25 to 0.35 m H2O depth. Excised tissues were mounted in the cuvette and provided with light for 2 h, to deplete internal CO2, before the commencement of the measurements.

Table 3.  Underwater CO2 kinetics for uptake by excised stems from previously emergent or submerged Halosarcia pergranulata
Parameters describing submerged PNOrgan
Previously emergent succulent stems
Mean ± SE
Submerged succulent stems
Mean ± SE
Submerged non-succulent stems
Mean ± SE
  1. Succulent stems were measured for both types of plants, whereas the non-succulent (i.e. ‘woody’) stems were only measured for previously submerged plants. The apparent half saturation constant (Ks) and maximum velocity (Vmax) of the CO2 response curves were estimated using the Hill & Whittingham (1955) model (r2 = 0.91 or 0.92, previously emergent succulent stems; 0.55 or 0.63, previously submerged succulent stems; 0.62 or 0.73, previously submerged non-succulent stems, for area-based and Chl a-based models, respectively). Means ± SE (n = 3–6) are shown with superscript letters indicating significant differences at P < 0.05, for values across a row. The net photosynthesis (PN) of the emergent succulent stems when in the air at 350 µL L−1 CO2 and photosynthetically active radiation (PAR) of 1500 µmol m−2 s−1 was 2.8 ± 0.2 µmol CO2 m−2 s−1 or 34.5 ± 2.5 nmol CO2 mg−1 Chl a s−1 (mean ± SE, n = 7).

Ks (µM) – area basis a266 ± 50ab118 ± 32b48 ± 14
Ks (µM) – Chl a basis a328 ± 33a396 ± 109a380 ± 93
Vmax (µmol O2 m−2 s−1)a0.97 ± 0.09b0.22 ± 0.03c0.084 ± 0.015
Vmax (nmol O2 mg−1 Chl a s−1)a16.5 ± 0.5b3.9 ± 0.5b2.4 ± 0.2

The Hill & Whittingham (1955) model fitted to our data was used to estimate the PN by the various stem tissues when submerged with CO2 in the medium as measured at Yenyening Lakes (average of CO2 values for sunrise and sunset was 18.5 µm, Table 1). On a surface area basis, underwater PN (µmol O2 m−2 s−1) at 18.5 µm CO2 was estimated at −0.09 to −0.11 for all three types of stem tissues (succulent stems previously submerged or emergent, and non-succulent woody stems of submerged plants). However, these estimates of PN at ambient CO2 concentrations in the floodwater have considerable uncertainty because, as previously stated, the model fitted to the data had only moderate predictive power, especially for the previously submerged succulent stems (r2 = 0.55 for surface area basis). For the actual measurements taken near ambient CO2 in the experiments shown in Fig. 5, the PN was just above zero (viz. 0.011 µmol O2 m−2 s−1) in three replicates, and it was just below zero (viz. −0.023 µmol O2 m−2 s−1) in two replicates. That the PN must have been above zero in stems of intact plants at ambient CO2 in the floodwater is indicated by the pO2 in the succulent stem of the plant measured in situ being 23.2 kPa in the late afternoon (i.e. ∼10% above that in the air; Fig. 3), so that even by the late afternoon, this submerged plant must have still produced more O2 in photosynthesis than it consumed in respiration.

Radial O2 profiles in submerged succulent stems

The bulky nature of the succulent stems raises an intriguing question regarding possible radial gradients in tissue pO2. These stems consist of an epidermis, then a layer of mesophyll tissue comprising cells with numerous chloroplasts, layers of enlarged ‘succulent cells’ without chloroplasts and a central cylinder containing the vascular tissues, but also numerous cells with chloroplasts (Fig. 6a). While gas-filled spaces are evident in the mesophyll tissue, none were seen in the layers of enlarged succulent cells. Radial profiles were measured through succulent stems when submerged in artificial lake H2O bubbled with air, when in darkness or in light. In darkness, pO2 declined markedly across the epidermis, was fairly constant across the mesophyll tissue, declined with distance into the succulent cells to a minimum of ∼1.5 kPa and then increased towards the central cylinder; within the central cylinder, O2 was similar to that in the outer mesophyll layer (Fig. 6a). In light, pO2 was relatively constant (14–16 kPa) across the outer mesophyll layer and throughout the succulent cells, but was more variable (8–15 kPa) within the central cylinder (Fig. 6b). The effect of switching from light to darkness was also assessed, with the electrode positioned within the layer of succulent cells. Oxygen showed a steep decline from 13.5 kPa to below detection (< 0.05 kPa) in just 27 min after the light was switched off (Fig. 6b, insert). After about 1 h, O2 had increased above this minimum to a new quasi-steady state level of 2 kPa (Fig. 6b, insert). Such patterns of pO2 initially reaching levels lower than in the quasi-steady state were also evident in the mesophyll cells of intact branches (Fig. 2). In summary, the radial profiles showed considerable radial O2 gradients across the succulent stems when in darkness.

Figure 6.

Radial O2 profiles through submerged succulent stems of Halosarcia pergranulata in darkness (a) and light (b) during quasi-steady state, at 20 °C. The coefficient of variance of intratissue pO2 in darkness was 0.69 versus 0.11 in light. Measured in well-stirred, air-saturated, artificial saline lake H2O, for branches excised from plants collected from Yenyening Lakes (completely submerged in 0.25–0.35 m H2O). The insert shows an example of a time trace during the switch from light to darkness, with an estimated 90% response time from one quasi-steady state to another of 12 min. Ø, outer diameter.


The succulent terrestrial halophyte H. pergranulata inhabits flood-prone playas (i.e. mudflats) of salt lakes (Wilson 1980) and can survive prolonged submergence (English 2004). However, submergence tolerance in halophytes has rarely been studied (for two Spartina spp., see Gleason & Zieman 1981), so we evaluated the key issue of intraplant O2 dynamics in shoots and roots of H. pergranulata when submerged. The present experiments demonstrate the importance of underwater photosynthesis in determining the O2 status of shoots and roots, and highlight the dynamic nature of O2 levels in the tissues of H. pergranulata when submerged. The diurnal patterns of pO2 in submerged organs of this succulent terrestrial halophyte are reminiscent of the patterns described for submersed aquatic species (Sand-Jensen et al. 2005), although the bulky nature of the succulent stems is a very different morphology to that of the specialized leaves (e.g. thin and dissected) of many aquatic species (Arber 1920; Schulthorpe 1967). Succulent stems would be expected to have substantial boundary layers (Vogel 1996), longer radial path length for intraorgan diffusion and also a lower gas-filled volume (i.e. lower porosity), all contributing to slower gas exchange with the H2O column, than in common aquatic leaves. Furthermore, absence of lacunae in the succulent tissues of the stems will also greatly impede longitudinal diffusion of O2 from the shoot to roots (Armstrong 1979). As such, when shoots of submerged plants were in darkness, pO2 in roots of H. pergranulata were much lower (average 0.7 kPa) than those of 6–8 kPa in roots of the aquatic species L. dortmanna (Sand-Jensen et al. 2005). The consequences of the unique characteristics of the succulent stems for intraplant O2 dynamics will be the focus of this discussion.

Succulent stems of another member of the Salicornioideae, the annual Salicornia fruticosa, have relatively high mesophyll resistance to CO2 diffusion to chloroplasts (stems in the air, Abdulrahman & Williams 1981). This study of H. pergranulata evaluated underwater photosynthesis and gas transport in the succulent stems of submerged plants. The consequence of diffusive boundary layers was evident from the O2 concentration at the surface of the stems, being only 71–80% of that in the stirred bulk medium (e.g. Fig. 6a). Moreover, within the stems, a relatively poor capacity for radial O2 diffusion was evident from the steep decline in pO2 within the first 0.1 mm, and also within the deeper layers of enlarged succulent cells when in darkness. Such large declines in pO2 across the stems indicate high resistance (and/or large respiratory demand for O2) across the radial diffusion path (cf. Armstrong 1979; Colmer & Greenway 2005). The high radial resistance to intratissue O2 diffusion in the succulent stems is in marked contrast to that in non-succulent shoot tissues; for example, in petioles (diameter = 2 mm) of the semiaquatic species Rumex palustris that contain aerenchyma, the radial profiles of O2 were rather flat (Mommer, Pedersen & Visser 2004).

The high resistance to intraplant gas movement might also contribute to development of the conspicuous peak in pO2 in tissues of submerged plants when first exposed to light after a period of darkness (e.g. Fig. 2). These peaks in pO2 might have resulted from high initial photosynthetic rates fuelled by respiratory CO2, which had built up during darkness in the succulent tissues. After some time, however, this internal CO2 would be depleted, and photosynthesis in the succulent stems presumably would become limited by slow diffusion of CO2 from the H2O column surrounding the shoot. A significant early morning peak of O2 also developed in succulent stems of the plant measured in situ in the field, even though the natural light level gradually increased (Fig. 3), rather than the sudden switches from complete darkness to high levels of light in the laboratory experiments. Such morning peaks in pO2 have not been observed for non-succulent aquatic plants submerged in the field (Greve et al. 2003; Borum et al. 2005; Sand-Jensen et al. 2005); perhaps build-up of respiratory CO2 does not occur in aquatic species, because their much thinner leaves and higher porosity throughout the plant should enable better equilibrium of internal gas composition with that of the surrounding H2O. A morning peak of ROL from the roots has, however, been reported for submerged rice under laboratory conditions (Waters et al. 1989). Waters et al. (1989) attributed the morning ‘surge’ in ROL to the combined effects of (1) CO2 accumulated by the end of the dark period fuelling the initial rates of photosynthesis as well as (2) a decreased consumption of O2 within the roots during the early morning, as a result of low substrate availability as tissue sugars had largely been consumed through the night.

The importance of longitudinal O2 diffusion from shoots to roots when in anoxic soil is well documented in waterlogged and fully submerged plants (Armstrong 1979; Colmer 2003). Similarly, the intraplant O2 measurements for H. pergranulata showed that root O2 supply was dependent on pO2 in shoots (e.g. as manipulated via light–dark switches; Figs 2 & 3). Here, we discuss the probable sources and pathway for longitudinal O2 diffusion in stems of H. pergranulata, when in darkness or in light. A key finding was that although the pO2 decreased markedly with distance into the succulent tissues when in darkness, pO2 increased again towards the central cylinder of the stems (Fig. 6a), indicating that these central tissues must receive O2 longitudinally from a distal source. As the cut end was sealed with petroleum jelly, and because the succulent stems showed poor radial entry of O2, we speculate that the woody portion of the stem might be the major area of O2 entry from the H2O column. In light, photosynthesis led to increased shoot pO2. The basal woody stems, the central cylinder within succulent cells and the outer mesophyll cell layers in the succulent tissues all contain chloroplasts; that both woody and succulent stems are photosynthetically competent was demonstrated (Table 3). The relative importance of these tissues when in light as sources of intraplant O2 to roots remains to be resolved. Although the succulent tissues are capable of 2.5-fold higher rates of photosynthesis than the basal non-succulent (i.e. woody) portion of the stems (Table 3), the woody parts are closer to the root/stem junctions, so that the total resistance to longitudinal O2 diffusion from these woody stems into the roots would be less than for the more distant succulent tissues. The recent finding that stem photosynthesis provides O2 to roots when the basal portion of the stem of Alnus glutinosa is submerged supports our hypothesis that photosynthesis in the basal woody portion of stems might be a major source of O2 to roots (Armstrong & Armstrong 2005).

When submerged at the CO2 concentrations measured in Yenyening Lakes (average of 18.5 µm), the PN by excised stems of H. pergranulata was close to zero (see Results), but for intact plants rooted in soil, underwater PN must have been positive, as quasi-steady state pO2 in the shoot tissues was maintained for several hours just below that in the air (i.e. 20.6 kPa) in our laboratory experiments (Fig. 2), and was ∼10% above that in the air in a submerged plant at Yenyening Lakes, even in the late afternoon (Fig. 3). Thus, in addition to the slow entry of CO2 from the floodwaters, photosynthesis in the stems of H. pergranulata must also rely on internal cycling of CO2 produced in respiration, or possibly even CO2 from the soil that might enter via the roots and move to the shoots (cf. Raven, Osborne & Johnston 1985). Nevertheless, consistent with the low rates of underwater PN at ambient CO2 concentrations is that the concentration of total sugars in the succulent stems of submerged H. pergranulata at sunrise was only 54 (total dry mass basis) or 39% (ethanol-insoluble dry mass basis) of that in the emergent plants. Similarly, sugars decline with time in submerged rice (Jackson & Ram 2003). The soluble sugar concentration of 30 mg g−1 dry mass in submerged H. pergranulata at sunrise compares favourably with levels in a range of halophytes sampled from a desert salt basin in India (Mohammed and Sen 1994); for example, the succulent Suaeda fruticosa contained 18–39 mg g−1 dry mass (time of sampling not specified).

Upon submergence, some plants produce new leaves with an improved capacity for exchange of gases with the H2O column, when compared with the existing ‘aerial leaves’ (Mommer & Visser 2005). Submergence-acclimated leaves of the semiaquatic species R. palustris show an increased capacity for entry of CO2, and these leaves exhibit a positive PN when underwater with CO2 at ambient levels, whereas non-acclimated leaves were unable to sustain a positive PN when underwater (Mommer & Visser 2005; Mommer, Pons & Visser 2006). For direct comparison with H. pergranulata, PN at 18.5 µm CO2 by leaves of R. palustris is estimated at 0.24 µmol O2 m−2 s−1 for acclimated leaves and at −0.20 µmol O2 m−2 s−1 for non-acclimated leaves (Mommer & Pedersen, personal communication). By contrast, comparison of the underwater CO2 uptake kinetics for the succulent stems of H. pergranulata from emergent and submerged plants (Table 3) shows no evidence for submergence acclimation to improve CO2 uptake when underwater in this species.

In summary, populations of H. pergranulata growing on playas (i.e. mudflats) of salt lakes in southern Australia can experience months of waterlogging and in some cases, complete submergence. The importance of photosynthetically derived O2 for internal aeration of H. pergranulata was demonstrated for this halophytic stem succulent, both in the laboratory experiments and in the field. Oxygen produced during photosynthesis increased pO2 in both the shoots and roots. Although the succulent stems are capable of higher rates of PN than the basal non-succulent (i.e. woody) stems, there appears to be substantial resistance to radial diffusion of gases across the succulent tissues. We speculate therefore that basal woody stems might be an important source of O2 for the roots; in light periods, O2 is produced in photosynthesis, and during darkness, O2 enters from the H2O column.


We greatly appreciate the assistance of I. Malik with photosynthesis measurements in the air. We thank E. Barrett-Lennard for suggesting the field site, and H. Greenway and H. Lambers for their comments on a draft of this manuscript. We thank the Western Australian Department of Conservation and Land Management for permission to take samples from Yenyening Lakes. O. Pedersen was supported by the Danish Research Agency (grant no. 21-04-0321).