Intercellular gas spaces are ubiquitous in the sporophytes of extant vascular plants. They function in the movement of gases in photosynthetic structures of land plants and in the movement of oxygen to below-ground structures, especially in plants of waterlogged soils (Raven, 1996). The view for the six decades up to 1980 was that these gas spaces were havens of diffusion, undisturbed by the mass flow of gases except when wind flapped leaves or flexed shoots, although the occurrence of convective flow of gases in plants was reported in the 19th century, beginning with Dutrochet in 1841 (Dacey, 1980, 1981; Groβe, 1996). The modern era of studies began with Dacey (1980, 1981) who showed that water lilies had internal winds apparently related to the O2 supply to roots in anoxic sediments. Internal winds have subsequently been found in a number of the wetland and emergent aquatic flowering plants, often in those with multiple photosynthetic shoots arising from a common rhizome (Brix et al., 1992; Groβe, 1996; Skelton & Allaway, 1996; Evans et al., 2009). William and Jean Armstrong have made major contributions through their work on Phragmites australis (e.g. Armstrong et al., 1992; Afreen et al., 2007). In this issue of New Phytologist, Armstrong & Armstrong (pp. 202–215) extend these studies to Equisetum telmateia where they found the fastest pressurized gas-flow rates measured so far, and model the possibility of internal winds in some Carboniferous relatives of Equisetum with the life form of rhizomatous trees (i.e. Calamites and similar genera).
‘Armstrong & Armstrong have given a plausible account of how the occurrence of high gas-flow rates relate to root and root-hair structure and the ecology of the plant.’
The key feature of humidity-induced convection that underlies gas flow in E. telmateia is the small (< 1 μm) aperture of stomata in younger, undamaged shoots. These small apertures impose such a high resistance to pressurized mass flow of gases that they prevent such flow; small apertures also restrict diffusive gas movement. The presence of photosynthetic pigments and of a water supply to the shoot means that the usual process of H2O evaporation into intercellular gas spaces takes place, with the latent heat provided by that fraction of the absorbed solar radiation which is, necessarily, not stored as photosynthetic products. With widely open stomata, H2O vapour loss by diffusion with little mass flow is the dominant gas flux between shoot and atmosphere, with minor fluxes of O2 out and CO2 in, and a very small pressure difference between shoot gas spaces and the atmosphere. With stomata nearly closed, all these fluxes (and especially the mass flow component) are decreased, and the pressure difference between intercellular gas spaces and the atmosphere increases owing to the addition of more water vapour to the internal atmosphere. If there is no means of relieving this pressure by mass flow of gas through a relatively high-conductance pathway to, and out of, other parts of the plant, then the shoot temperature increases until energy balance is achieved, with increased heat loss by convection (in the atmosphere), radiation and (limited) latent heat in transpiration through stomata. Relief of the pressure by convective gas flow requires a high-conductance pathway from the younger live shoots through the rhizome and out to the atmosphere through older shoots attached to the same rhizome. The presence of such a high-conductance pathway, and the associated mass flow of gas in E. telmateia, involves relatively uninterrupted gas spaces occupying a large fraction of the volume of the rhizomes and emergent shoots of this perennial plant (Armstrong & Armstrong). Gas is vented to the atmosphere mainly through damaged shoots or stubble (Armstrong & Armstrong).
Why is gas flow through E. telmateia so rapid relative to other plants?
The anatomy of E. telmateia has a particularly high computed conductance for pressurised flow of gas through the schizogenous aerenchyma channels (vallecular canals), of which there are four or five in branches, c. 10 in rhizomes and c. 30 in stems. These channels occupy less of the cross-sectional area of axes than does the central pith cavity, but, unlike the pith cavity, have no flow-impeding nodal diaphragms. Measurements show higher conductance on a per-axis basis for the aerenchyma channels than for the pith cavity for axis segments long enough to include nodes (Armstrong & Armstrong).
In other species of Equisetum, there was no detectable convective flow in two species, slow flow in a further two species and relatively rapid flow inEquisetum palustre, especially when its small size was taken into account (Armstrong & Armstrong). These differences in flow, and those between E. telmateia and other wetland species, are related mechanistically to anatomical differences in the axes and/or to differences in stomatal behaviour.
Functionally, gas flow through wetland plants probably relates in natural selection to the internal provision of O2 to below-ground parts, which are in an environment with a low O2 concentration and a low O2 diffusivity relative to plants in ‘normal’ soil. For E. telmateia, Armstrong & Armstrong have given a plausible account of how the occurrence of high gas-flow rates relate to root and root-hair structure and the ecology of the plant.
What other mechanisms generate pressure within gas spaces and produce gas flow through plants?
Thermal osmosis occurs in parallel with humidity-driven pressurization, involving thermal expansion of gases in gas spaces separated from the cooler atmosphere by stomata or other pores with an aperture of < 1 μm. Although favoured by some (Groβe, 1996) as the dominant driving force for some cases of convective flow involving living photosynthetic tissue, the consensus from modelling is that thermal osmosis is a minor contributor relative to humidity-driven flow in most cases (Armstrong et al., 1992; Bendix et al., 1994; Afreen et al., 2007; Armstrong & Armstrong, this issue).
The Venturi effect, characterized for ‘standing dead’ culms of P. australis by Armstrong et al. (1992), also contributes to flow into, through and out of rhizomes. Wind (external)-driven mass flow of gas to the atmosphere from erect culms generates reduced pressure inside with gas entry in snapped culms on the same rhizome opening to the atmosphere nearer the substratum where wind speeds are lower. This might operate in E. telmateia.
A further process that generates gas pressure is the accumulation of photosynthetically produced O2. Most relevant here is crassulacean acid metabolism (CAM), where the malic acid accumulated at night using CO2 entering (on land) through open stomata is decarboxylated; the resulting CO2 is refixed in C3 photosynthesis behind closed stomata and the resulting O2 accumulates in the daytime (Ekern, 1965; Lüttge, 2002). This pressurization, and the enrichment of the accumulated gas in O2, was discovered in 1800 by Alexander von Humbold, who significantly influenced Darwin (and died 1859) and it was subsequently, independently rediscovered (Ekern, 1965; Lüttge, 2002). This pressurization is unlikely to lead to a through-flow of gases because CAM plants lack gas efflux sites when pressurization occurs, and plants in which internal winds have been found do not show CAM (references in Groβe, 1996; Skelton & Allaway, 1996; Smith & Winter, 1996; Armstrong & Armstrong, this issue; Evans et al., 2009).
Humidity-driven flow in the past: the case of Calamites in the Carboniferous
A major innovation in Armstrong & Armstrong is the quantitative modelling of humidity-driven flow through members of the Calamitales; these Carboniferous wetland rhizomatous trees are closely related to Equisetum. From the similarity of calamitalean gas-channel anatomy to that of E. telmateia, Armstrong & Armstrong suggested that large-scale humidity-driven flow could have occurred, ventilating the large underground rhizome and the associated prolific roots. Armstrong & Armstrong also point out the possible problems of O2 supply to the wetland ‘stigmarian axes’ bearing ‘stigmarian rootlets’ involved in anchorage and uptake of water and nutrients in the lepidondendrid (lycopod) trees such as Lepidondendron and Sigillaria, which co-existed with the Calamitales. These trees have aerenchymatous ‘parichnos strands’ in the leaves running radially into the trunk and abundant aerenchyma in the stigmarian rootlet systems (Fig. 1): would gas diffusion have been adequate in these large (up to 30 m in height) plants (Hirmer, 1927; Stewart & Rothwell, 1993; Taylor et al., 2009), granted the lack of demonstrated continuity of aerenchyma between the trunk near leaves and leaf scars and the rootlets? Groβe (1996) discusses the occurrence of gas mass flow influx and efflux through single floating leaves of Nelumbo nucifera, with mass fluxes of gas to and from the rhizome through separate aerenchyma channels in a single petiole; such a mechanism may not seem likely in lepidodendrids on anatomical grounds.
Future research needs
From a personal viewpoint I highlight three topics for further development. One is the further development of modelling of possible convective flow within extinct plants, continuing the work of Armstrong & Armstrong using other anatomically well-preserved material of wetland plants from the Carboniferous coal measures (Stewart & Rothwell, 1993; Taylor et al., 2009).
A second is investigation of the possibility that convective ventilation of below-ground axes could divert respiratory CO2 from loss to the medium to movement through aerenchyma, with a corresponding increase in the possibility of photosynthetic refixation where the gas efflux sites are photosynthetic. This refixation from aerenchyma is best developed in the stomata-less submerged plants and in some of the amphibious plants of the isoetid life form (Raven et al., 1988), although here there is no flow atmosphere–plant–atmosphere and the CO2 flux is purely diffusive. Experimentally, it is important to distinguish refixation from the gas-phase flux with that of CO2 carried in solution in the xylem (Aubrey & Teskey, 2009).
The third topic is analysis of the resource cost of the convective gas flow system to set against the perceived benefits. One cost could be a smaller return as rate of photosynthate production from a given investment in photosynthetic machinery as a result of the low diffusive conductance of stomata or any other barrier to mass flow of gas between the atmosphere (Dacey, 1980, 1981, 1987; Armstrong et al., 1992; Groβe, 1996; Armstrong & Armstrong, this issue) and the cellular site of photosynthesis in the pressure-generating part of the system.