Long-distance transport in non-vascular plants


  • J. A. RAVEN

    Corresponding author
    1. Division of Environmental and Applied Biology, School of Biological Sciences, University of Dundee, Biological Sciences Institute, Dundee DD1 4HN, UK
      Correspondence: John A. Raven. E-mail: j.a.raven@dundee.ac.uk
    Search for more papers by this author

Correspondence: John A. Raven. E-mail: j.a.raven@dundee.ac.uk


Many macroalgae have significant spatial differentiation involving higher rate resource use at a site than of acquisition of that resource from the environment at that site. Long-distance symplasmic transport of solutes occurs in some large green algae where the solutes are moved in streaming cytoplasm. In some large brown algae there is evidence of long-distance symplasmic transport of organic C and other solutes. Structural and physiological data suggest that while the transport in ‘sieve tubes’ of Macrocystis might be by a Munch pressure flow mechanism the transport in many other brown algae is less likely to be by this mechanism. Less is known of long-distance symplasmic transport in red algae. In terrestrial bryophytes transpiration occurs and in some liverworts and many mosses (but not in hornworts) there are files of dead cells in their tissues which may, and in some cases certainly, function in long-distance apoplasmic water transport. The hydraulic conductivity of these conduits is poorly characterized. Long-distance symplasmic transport in some mosses have been characterized both structurally and physiologically, but in other mosses and in liverworts the evidence is only structural. Most of these symplasmic transport pathways seem to have a high resistance to flow.


The term ‘vascular plants’ (tracheophytes) gives a central role to long-distance transport in defining the photosynthetic organisms which dominate the land surface today. Xylem is involved in transporting water, soil-derived nutrients and organic acid anions from roots to shoots, whereas the phloem distributes photosynthate and stored organic C from photosynthetic structures and storage regions. The requirement for long-distance transport of water and, by mass flow, solutes, can be related to several evolutionary selective pressures. These can be considered in the context of the interfacial habitat occupied by vascular plants. These organisms obtain photons and CO2 from their aerial environment, leading to a large demand for transpired water (hundreds of grams per gram dry matter gain), and water and nutrients other than carbon, from the soil. The requirement for an endohydric water-conducting system becomes apparent when cuticularized plants exceed a height of a few tens of centimetres (Niklas 1997; Raven 1977, 1984b, 1999a, b), and a phloem (or phloem-like) apparatus which moves organic solutes from sites of photosynthesis or storage to sites of chemo-organotrophic growth or storage has a similar relationship to size as does the xylem (Raven 1977). The selection for taller plants (with necessarily deeper root systems: Raven & Edwards 2001) involves competition for light and shading, rather than being shaded by competitors, and the improved potential for dispersal of wind-borne spores by their release by structures at the top of the canopy in turbulently mixed air. The taller the plant, the greater the quantitative need for xylem in movement of transpiratory water (Raven & Handley 1987).

The same quantitative arguments apply to non-vascular plants in a similar polarized environment, for example, tall (= 10 cm) erect gametophytes of bryophytes which are endohydric; we shall see that these organisms have xylem analogues (hydrome and other water-containing cells) and phloem analogues (leptome and other photosynthetic-conducting cells) in their gametophyte phase and their sporophyte phase. Ectohydric, including aquatic, bryophytes lack hydrome. Bryophytes contain rhizophytic (those with rhizoids in fine-grained sediments such as mud or sand) and haptophytic (those attached to large stones or rocks) members.

The other group of non-vascular photosynthetic organisms are the algae. All of the large algae are aquatic and, if intertidal, are poikilohydric since they have no water source on rocks at low tide. These algae have no endohydric conducting system. Many large algae, including both haptophytes and rhizophytes have significant differentiation among parts which are specialized in nutrient uptake, photosynthesis, reproduction and growth. The requirement for transport of solutes among these differentiated parts involve a range of symplasmic mass flow mechanisms in these polyphyletic organisms.

The ‘non-vascular’ organisms will be dealt with in the inverse order from that used above to reflect the level of complexity of organisms and the evolution of the embryophytes (bryophytes and tracheophytes) from charophycean green algae (Karol et al. 2001).


The diversity of algal higher taxa with representatives having specialized long-distance symplasmic transport structures

The prerequisites for long-distance symplastic transport in the algae are differentiation (Bell & Mooers 1997) among regions with an imbalance between resource requirements and resource acquisition from their environment, and a sufficient distance between the differentiated regions to make diffusion ineffective in moving resources at the rates required to account for the observed rates of growth (Raven 1981, 1984a, 1997).

Taxa which have representatives which are large enough, and sufficiently differentiated, to have symplasmic long-distance transport by mass flow over significant portions of the pathlength, are the classes Charophyceae sensu lato and the classes Bryopsidophyceae and Dasycladophyceae (Ulvophyceae sensu lato) of the Chlorophyta, the class Phaeophyceae of the Heterokontophyta, and the class Florideophyceae of the Rhodophyta (Raven 1981, 1984a, 1997; Van den Hoek, Mann & Jahns 1995; Graham & Wilcox 2000).

As a background to considering translocation in differentiated macroalgae we distinguish between organisms which consist largely of (a) giant cell(s), in which large cells are those which are 1 mm3 in volume, and those which do not. The former category include the order Charales in the Charophyceae (sensu lato), some members of the Bryopsidophyceae and Dasycladophyceae of the Chlorophyta, as well as Griffithsia in the Florideophyceae (Raven 1981, 1984a, 1997; Van den Hoek et al. 1995). In the Charales and Griffithsia differentiation is intercellular, with little intracellular differentiation at the morphological, anatomical and physiological levels, noting that Griffithsia is less differentiated than are the Charales. Individuals of the green algal classes Bryopsidophyceae and Dasycladophyceae consist of a single giant cell; despite this there is very considerable differentiation within the thallus, reaching its extreme in species of Caulerpa (Bryopsidophyceae).

In the Charales and Griffithsia any long-distance transport must mainly occur through the large cells which make up most of the thallus; these cells also perform other functions (e.g. photosynthesis and/or nutrient absorption; mechanical support; storage). Transfer between cells (including, in Charales, small internodal cells) involves plasmodesmata or, in Griffithsia, their presumed analogues the pit plugs (= pit connections) (Raven 1981, 1984a, 1997). In the Bryopsidophyceae and Dasycladophyceae long-distance transport in the (single) cell which performs all other organismal functions occurs by actomyosin-based cytoplasmic streaming (as does movement within cells of the Charales).

The larger and more differentiated Phaeophyceae and Floridiophyceae (other than Griffithsia) are made up of relatively small cells (more obviously in the Phaeophyceae than the Rhodophyceae). In some cases the cells which are known to, or may be, involved in long-distance transport consist of long cells, such as the solenocysts of the phaeophycean Saccorhiza which can be as long as 75 mm (see Raven 1984a). Central cells in the thalli of many floridiophycean red algae are also more than 1 mm long, for example, 8 mm long in the freshwater Lemanea (Raven 1984a). More commonly the cylindrical cells which are certainly or probably involved in long-distance transport in the Phaeophyceae are relatively short, 1 mm or less, with a diameter of 10–62 µm, and often largest at the circular end walls (Schmitz 1981, 1990; Raven 1984a). The pores in the end walls with diameters ranging from the 40 nm characteristic of plasmodesmata up to 4·5 µm which is almost as large as the largest pores in the end walls of flowering plant sieve plate (Schmitz 1981, 1990; Raven 1984a). The situation for the red algae is much less clear, with the elongate cells believed to function in long-distance transport separated by cross walls containing pit plugs (= pit connections) which have been called ‘transfer connections’ when found in walls separating the alleged conducting cells (Schmitz 1981, 1990; Raven 1984a).

Multiple evolutionary origins of long-distance symplasmic transport structures in algae

Before considering the mechanism of these various transport processes in more detail, it would be appropriate to consider the extent to which the various structures have been subject to multiple evolutionary origins in response to increased demands for long-distance transport with increased size and differentiation. In the Chlorophyta the long-distance transport within cells in the Charales, Bryopsidophyceae and Dasycladophyceae is via the actomyosin system. Actin and myosin are ubiquitous in eukaryotes. The involvement of this mechanochemical in long-distance transport apparently arose independently in the three higher taxa of green algae (Raven 1984a; Van den Hoek et al. 1995). The plasmodesmata connecting the large cells of the Charales via small nodal cells are phylogenetically close to the plasmodesmata of embryophytes but lack desmotubules (Cook et al. 1997; Raven 1997; Cook & Graham 1999).

Little can be said about the phylogeny of the poorly characterized long-distance transport system of red algae (Hartman & Eschrich 1969; Schmitz 1981, 1990; Raven 1984a, 1997).

In brown algae, the ancestral (plesiomorphic) state is to have plasmodesmata, and the parenchymatous condition has evolved at least four times (Graham & Wilcox 2000; Draisma et al. 2001). Whether the anatomically defined long-distance transport system could have evolved in pseudoparenchymatous ancestors prior to the evolution of the parenchymatous condition cannot be decided on the basis of present evidence; certainly long-distance transport systems are found in the (mainly) pseudoparenchymatous Desmarestiales (Clayton 1990). Raven (1984a) named the Fucales, Durvilleales, Laminariales and Desmarestiales as having anatomically distinguishable long-distance transport systems. Schmitz (1990) added the Dictyotales (Katsaros & Galatis 1988). The Durvillaeales have now been (re)subsumed into the Fucales (Draisma et al. 2001; Rousseau et al. 2001). Molecular phylogenetic analysis has also shown that the Laminariales sensu lato are paraphyletic with the Laminariales sensu stricto (Alariaceae, Laminariaceae and Lessoniaceae), the Phyllariaceae and Halosiphonaceae, and the Chordariaceae as distinct clades no more closely related than are one or more of them to the Ectocarpales sensu lato, the Ascoseirales and the Desmarestiales (Draisma et al. 2001; Rousseau et al. 2001; Sasaki et al. 2001; Yoon et al. 2001). The monotypic Antarctic order Ascoseirales (Clayton 1987, 1990) also have cells whose structure, at least when immature, suggests a role in long-distance transport (Clayton & Ashburner 1990).

The current tally of extant brown algal orders (or families incertae sedis) with cells whose anatomy suggests a role in long-distance transport is shown in Table 1. It is not possible to say with certainty that the seven higher taxa listed showed independent evolution of anatomically defined conducting cells (see Clayton & Ashburner 1990), even with more (molecular) phylogenetic evidence than was available to Clayton & Ashburner (1990). Indeed, the discovery of cells which could be conducting cells in a species of Dictyopteris from the Dictyotales (Katsaros & Galatis 1988) and the subsequent finding that the Dictyotales is the earliest-branching order of the Phaeophyceae (Draisma et al. 2001; Rousseau et al. 2001) open the possibility that conducting cells evolved in early, but relatively complex, brown algae and that their current distribution is the result of numerous evolutionary losses. Of course, molecular phylogenies do not inform us directly about the evolution of large, complex, thalli, so that multiple independent origins of complex pseudoparenchymatous (most Desmarestiales) or parenchymatous (Dictyotales, Fucales, Ascoseirales, Ectocarpales s.l., Laminariales s.s., Phyllariaceae, some Desmarestiales) (Graham & Wilcox 2000) thalli with attendant evolution of long-distance transport structures, is a distinct possibility. Multiple independent origins of complex thalli certainly fit with the different development patterns among the large complex brown algae, with parenchymatous, apical blade growth in the Fucales and Dictyotales, intercalary parenchymatous blade growth in the Ascoseirales, Ectocarpales s.l., Laminariales s.s and Phyllariaceae, and intercalary pseudoparenchymatous (or, more rarely, parenchymatous) blade growth in the Desmarestiales (Clayton & Ashburner 1990; Graham & Wilcox 2000). Such independent evolution of complex thalli is consistent with several independent origins of long-distance transport structures. This would seem plausible in view of the coenocytic, large-celled conducting apparatus of the Ascoseirales and of the Phyllariaceae, and the smaller cells in a longitudinal file in the other taxa in Table 1.

Table 1.  Occurrence of cells with structural (or structural and functional) indications of a role in long-distance symplasmic transport in brown algae
OrderSieve pore diameter‘Sieve cell’
‘Sieve cell’
evidence of
of long-distance
Dictyotales60–100 nm30 µm300 µmSchmitz (1990); Katsaros &
Galatis (1988
Fucales (including
50 (− 100?) nm  (+) Raven (1984a); Moss (1983);
Roberts (1979); Fielding, Carter &
Smith (1987
); Schmitz (1990);
Penot, Dumay & Pellegrini (1985)
AscoseiralesNo cross walls
40 µm> 100 mm?(+)Clayton (1987, 1990); Clayton &
Ashburner (1990
); Gomez,
Wiencke & Weyknam (1995
Gomez & Wiencke (1998)
Desmarestiales400 nm10–30 µm300 µm(+)Moe & Silva (1981); Peters et al.
); Raven (1984a); Schmitz
Laminariales400–6000 nm £ 1·2 mm+Lüning, Schmitz & Willenbrinck
); Schmitz (1981); Davison
& Stewart (1983
); Raven (1984a);
Lüning (1990); Schmitz (1990);
Mizuta et al. (1996)
Phyllariaceae (not
sensu stricto)
50 nm (in walls
solenocysts and
 75 mm+Emerson, Buggeln & Bal (1982);
Raven (1984a); Henry & South
); Schmitz (1990)
Ectocarpales sensu lato
 14 µm
(10–20 µm)
14 mm
(50–150 µm)
 Guimarães et al. (1986); Schmitz
); Draisma et al. (2001)

Quantitative analysis on long-distance symplasmic transport in algae Phaeophyceae

Other entries in Table 1 concern the extent to which functional data are available for the operation of the putative conducting cells in long-distance transport. The widest range of evidence is available for the Laminariales s.s and especially for Macrocystis, Laminaria and Alaria of growth at the expense of stored organic C (and N), i.e. in excess of net carbon accumulation by photosynthesis in the growing zone (see Raven 1984a). Radioactive labelling gives evidence (via autoradiography) that the putative conducting cells are indeed the conduits of transport, as well as of the speed of transport (Schmitz 1981, 1990; Raven 1984a). It is also possible to measure the composition of sap bleeding from cut conducting elements for Macrocystis (Schmitz 1981, 1990; Raven 1984a). Finally, quantitative evidence on localized rate of growth of the alga in excess of the rate of local net (24 h) acquisition of external resources can be used, with the transverse sectional area of the conduits to compute the flux (specific mass transfer) of nutrients (C, N, Fe) through the pathway (Raven 1984a; Buggeln 1985; Dunton 1985; Dunton & Schell 1986; Lobban & Harrison 1997).

For Macrocystis, where there are estimates for C of all three of these values (concentration, speed and flux). Transport by mass flow should follow Eqn 1.


Flux has the units mol C m−2 (conduit transverse area) s−1, speed has the units m s−1 and concentration has the units mol m−3 (conduit volume). Entering concentration and flux into Eqn 1 yields an estimate of the flux which has been compared with measurements (Raven 1984a); the predicted flux is just over twice the measured value. Considering the different conditions for the collection of the data for the various inputs this agreement is satisfactory, and suggests that the transport of organic C around Macrocystis involves mass flow of the soluble contents of the conducting cells. For other brown alga there are fewer data on long-distance transport than is available for Macrocystis. Thus, although there are flux and speed data for Alaria, there seem to be no data on the concentration of solutes in the conducting cells (Raven 1984a). The C flux in Alaria is less than in Macrocystis (less than 10%), and the speed is also rather lower (Raven 1984a) so that the translocation could be accommodated by much lower solute concentrations than in Macrocystis or with a smaller fraction of the cross-sectional area of the conduits involved in the flux, consistent with the vacuolation in the conducting cells of Alaria (Raven 1984a).

For Saccorhiza there is evidence for long-distance translocation in the solenocyst/adelocyst apparatus in the form of 14C measurements of the speed of 14C organic C movement, but without estimates of the flux or the concentration of organic C in the conduits. The sporophyte of Saccorhiza is annual, so the role of translocation in many perennial members of the Laminariales s.s. in mobilizing stored organic C from mature regions to meristems for early spring (winter) growth at high latitudes presumably does not apply here. However, some members of the Laminariales s.s. (e.g. Nereocystis luetkeana) also have annual sporophytes.

The functional evidence for long-distance transport in the Desmarestiales and Ascoseirales comes from work showing that intercalary growth occurs more rapidly than can be explained in terms of local net photosynthesis over a 24 h period (Drew & Hastings 1992; Gomez, Thomas & Wiencke 1995;Gomez & Wiencke 1998).

For the Fucales (including Durvillaea) there is evidence for long-distance transport based on the transport speed of organic 14C, but little evidence is available as to apical growth of thalli in excess of local supply of photosynthate. Although the speed of transport in the Fucales is relatively low, calculations by Raven (1984a) show that it is adequate to account for the apical growth rate even if no photosynthate is supplied from the apical 20 mm of the thallus.

As to the mechanism of long-distance transport, i.e. the means by which the energy required to cause mass flow is applied, little progress seems to have been made in terms of the energy requirements (i.e. defining the resistance to flow) and the means by which metabolic energy is applied to overcoming this resistance since the publications by Schmitz (1981, 1990) and Raven (1984a). The quantitative analysis by Raven (1984a) suggests that long-distance transport in Macrocystis could occur over at least 1 m by a Münch pressure-flow mechanism (Münch 1930), granted the enucleate state of the mature functional conduits of Macrocystis (as in the sieve tube elements of angiosperms and gnetaleans) and the likelihood that the sieve pores up to 6 µm in radius (and which occupy up to half of the sieve plate area) are open during translocation. This latter point about the structural status of sieve pores could be resolved by the use of confocal microscopy (Van Bel & Knoblauch 2000).

For other brown algae (including many Laminariales s.s) the contents of the putative conducting cells seem less conducive to mass flow within the cell than is the case in Macrocystis (Clayton & Ashburner 1990; Schmitz 1990) even though there is evidence of nuclear degeneration in mature conducting cells of Desmarestia (Clayton & Ashburner 1990; Schmitz 1990). Furthermore, the sieve pores in these other brown algae are invariably smaller than in Macrocystis although this is partially offset by the larger number per unit area of sieve plate; the flux through a pore is, other things being equal, proportional to the fourth power of the radius of the pore, although the Hagen–Poiseuille equation only applies when pores are much longer than they are wide. Another factor is the frequency of sieve plates along the transport pathway; this varies by at least three orders of magnitude (Table 1). Although pressure flow seems a less plausible likely mechanism, at least for organisms with very narrow pores in frequent sieve plates, no plausible alternative has been proposed.


Evidence as to the extent and mechanism of long-distance transport in red algae is restricted to the work of Hartman & Eschrich (1969) and Turner & Evans (1986) on the structure of the putative conduits and the speed of tracer (14C) movement and the seasonal variation in such 14C movement in Delesseria sanguinea. Other evidence of long-distance transport is the seasonality of growth of perennial algae where the gain in biomass in growing organs exceeds the local net carbon gain from photosynthesis minus respiration. The extreme case is that of growth of new blades in darkness (Powell 1986; Bowen 1971; Lüning 1984; Lüning & Schmitz 1988) in Constantinea subulifera, Maripelta rotata and Delesseria sanguinea. There seems to be no information on the transverse sectional area of the conduits in organs of red algae in relation to the flux of material on a per organ basis, so there are no data on the flux of material in long-distance transport in red algae. Also unknown is the concentration of solutes in the conduits, so that the mass flow equation used for Macrocystis (Raven 1984a), i.e. flux=concentration × speed has not been applied to red algae (Raven, Johnston & MacFarlane 1990).

Although it is difficult to account for the published speed of tracer movement (175 µm s−1) in Delesseria sanguinea other than by mass flow, the occurrence of pit plugs between cells poses a major problem. The polysaccharide plug in the cell wall would impose a very significant resistance on intercellular symplasmic mass flow (remembering that the plug and the wall are separated by plasmalemma). This is exacerbated if, as is usually the case, one or more membranes separate the cytosol of each cell from the plug (Pueschel 1990), even if the membrane(s) have proteinaceous pores which permit passage of water and solutes of Mr < 800–1000 (Raven 1984a; Bauman & Jones 1986). A lower resistance for symplasmic intercellular would result from the partial or complete dissolution of pit plugs; while this is known for the putative conduits of Delesseria (Hartman & Eschrich 1969; Turner & Evans 1986), the extent to which it is found throughout the long-distance transport pathway is unclear. The putative conduits for long-distance transport in Delesseria also have less complex intracellular structure than do the cells involved in growth, photosynthesis, nutrient uptake and storage (Hartman & Eschrich 1969; Turner & Evans 1986).

Over short distances (less than, or equal to a few millimetres) there are well authenticated examples of symplasmic transport in red algae related to reproduction in the development of the carposporophyte and in red algal parasitic or other red algae (Callow, Callow & Evans 1979; Turner & Evans 1979; Goff & Zuccarello 1994; Van den Hoek et al. 1995; Graham & Wilcox 2000).


Long-distance symplasmic transport in green algae is, as noted above, a function of cytoplasmic streaming in acellular organisms (Ulvophyceae s.l) and in multicellular organisms composed largely of very large cells symplasmically connected via smaller cells (Charales). Direct measurements (light microscopy; Laser-Doppler estimates) of the rate of streaming are 2·5–6 µm s−1 for the coenocytic Dasycladophyceae (Acetabularia calyculus) and Bryopsidophyceae (Caulerpa prolifera) and up to 100 µm s−1 for the largest cells of the Charales (Table 9·11 of Raven 1984a).

Although most estimates of the movement of tracer solutes in these organisms give values similar to those of the speed of streaming (e.g. Box, Andrews & Raven 1984), some estimates for the Charales yield higher values (Table 9·11 of Raven 1984a). Although some artefacts are possible in tracer measurements, this discrepancy is currently unresolved.

There are very few data sets bearing on the quantitative significance of this stream of cytoplasm in long-distance transport of solutes. Large data sets on the inorganic ion concentration in the cytoplasm are not paralleled by measurements of low Mr organic components, and especially of sugars. Raven (1984a) cites sucrose concentrations of 100 mol m−3 in the cytoplasm of the euryhaline characean Lamprothamnion (see also Okazaki, Sakano & Tazawa 1987) and Ding et al. (1992) measured sucrose concentrations in the streaming endoplasm of Chara corallina of 14–25 mol m−3. Before the data of Ding et al. (1992) were published Raven (1981) had used data of Raven & Smith (1978) on the carbon used in the growth of the apex of Chara corallina in excess of what is supplied by local (24 h) photosynthesis minus respiration, the volume of streaming cytoplasm and its speed of flow to compute the minimum sucrose concentration that is needed to account for the required carbon supply to the growing apex. These calculations took into account the cycling of the streaming cytoplasm around the cells, with osmotic and volume control implications for the removal of sucrose (and other solutes) at the apical end and re-supply at the basal end, and concluded that a sucrose concentration of not less than 10 mol m−3 was needed. This accords well with the later measurements of Ding et al. (1992) of 14–25 mol sucrose m−3.

The computations by Raven (1981) were based on earlier calculations of Hope & Walker (1975) for the major osmoticum Cl; the Cl requirement of the growing apex cannot be completely supplied by plasmalemma influx at the apex, but could be supplied by cytoplasmic streaming. Similar arguments apply to P transported as phosphate (Mimura, Reid & Smith 1998) and N transported as amino-acids (Mimura, Sakano & Tazawa 1990) as well as the other solutes in Table 2. For calcium the total concentration in the endoplasm is unlikely to all be symplasmically mobile, since some will be in endoplasmic reticulum (Tazawa, Kikuyama & Okazaki 2001). The published data on characean plasmodesmata indicates a much higher Mr cut-off (= 45 000) than for the ‘normal’ (lower limit) value of 800 for higher plant plasmodesmata (Kikuyama et al. 1992; Cook et al. 1997; Cook & Graham 1999; Graham & Wilcox 2000), so that many Ca-chelating compounds, including calmodulin, could move symplasmically in these algae.

Table 2.  Comparison of the composition of whole plants of Nitella spp. With the composition of the streaming cytoplasm of Nitella spp. (Raven 1981; 1984a; Sakano & Tazawa 1984; Mimura et al. 1990; Ding et al. 1991; Mimura et al. 1998; Tazawa, Kikuyama & Okazaki 2001) (Data in parentheses for Chara spp.)
ElementWhole Nitella/mol
element (m3 plant)−1
Streaming cytoplasm of
Nitella/mol element m−3
  • a

    From concentration of sucrose in streaming cytoplasm (Ding et al. 1991).

  • b

    From concentration of sucrose and amino-acids plus amide in streaming cytoplasm, assuming an average of 4C per amino-acid plus amide (Ding et al. 1991; Mimura et al. 1990).

  • c

    From concentration of amino acids plus amides in streaming cytoplasm, assuming a mean of 1·4 N per amino-acid plus amide (Mimura et al. 1990).

  • d

    From concentration of inorganic phosphate in streaming cytoplasm (Mimura et al. 1998).

C3760168–300 a (480–640) b
N169–226112 c
P9–1310–20 d
Ca (total)15(2–7) 8–17
Ca (free)0·0001
Mg4·5(3–7) 7–8
Cl113(10) 21–27

This discussion of the symplasmic transport in characeans shows that the concentration of symplasmically mobile elements in the streaming cytoplasm and the speed at which the cytoplasm streams is adequate to supply shoot apices with the materials needed for their growth in addition to the photosynthesis and nutrient uptake occurring in the apices (see Raven 1981, 1984a). The same apparently applies to movement of sediment-derived nutrients along rhizoids to the shoot and the reverse for photosynthate to rhizoids (Raven 1981, 1984a).


Raven (1984a) considers the evidence related to long-distance apoplasmic transport in pseudoparenchymatous and parenchymatous marine algal thalli. The most obvious structural feature which could permit apoplasmic mass flow is the dead central region of the stout stipe of an Australian species of the red alga Cryptonemia (Wetherbee & Kraft 1981; Raven 1984a). However, there is no evidence as to the occurrence of longitudinal mass flow, or any mechanism that would drive such a flow other than flexing of the thallus under the action of waves and currents in a submersed alga (Raven 1984a).

The driving force of evaporation is available for longitudinal transport of solutions in thalli of intertidal macroalgae when they are exposed to air, provided there is differential evaporation over different parts of the thallus (e.g. by part of the thallus being overlaid by another alga). However, there is no external source of water to make good evaporative water losses in such situations, and net water loss occurs in these desiccation-tolerant organisms during tidal emersion (Jones & Norton 1979).

For the brown alga Ascophyllum nodosum the measured longitudinal apoplasmic fluxes of a number of radioactively labelled solutes are no faster than can be accounted for by diffusion (see Raven 1984a). In Laminaria, in which (unlike Ascophyllum) there is good evidence for long-distance symplasmic transport, apoplasmic solute fluxes are lower than in Ascophyllum (Raven 1984a). The basis for this difference in apoplasmic fluxes is not clear; it is also unclear whether longitudinal diffusivity is greater than radial diffusivity (Raven 1984a). In any case, diffusive apoplasmic fluxes are not quantitatively significant in long-distance longitudinal transport.


Gas-phase transport within macroalgal thalli has no well-authenticated role in the fluxes of metabolic gases between the medium and sites of metabolism, or among sites of metabolism in the thallus (Raven 1984a, 1996).


The bryophyte grade of organization in extant plants is found in embryophytes which lack the characteristic feature of vascular (tracheophyte) plants, i.e. xylem defined by the presence of lignin and of (ultra) structural features (Edwards 1993, 2000, 2002; Friedman & Cook 2000). Based on extant organisms the bryophytes and tracheophytes can also be distinguished on the basis of unbranched sporophytes (one terminal sporangium per sporophyte) in bryophytes and branched sporophytes with many sporangia per sporophyte in the tracheophytes. However, in the Silurian and Lower Devonian some embryophytes had branched sporophytes but lacked true xylem; these organisms are termed polysporangiophytes (Kenrick & Crane 1997; Edwards 2000; Friedman & Cook 2000) and will not be considered further here. Bryophytes have more cell types and have more tissue differentiation than do algae, suggesting that long-distance transport is needed to accommodate excesses of requirement for a resource at a site over the acquisition of that resource from the medium at that site (Bell & Mooers 1997; Raven 1999a, b; Introduction).

There are three clades of extant bryophytes, i.e. the hornworts, liverworts and mosses. There is still uncertainty as to the branching order of these clades in the evolution of basal embryophytes (Kenrick & Crane 1997; Qiu et al. 1998; Duff & Nickrent 1999; Qiu & Lee 2000; Renzaglia et al. 2000; Karol et al. 2001). The hornworts and many liverworts have thalloid gametophytes, whereas other liverworts and all mosses have leafy gametophytes; both kinds of gametophyte have rhizoids, and some have fungal associates that may act like mycorrhizas (Kenrick & Crane 1997; Read et al. 2000). The sporophyte phase is always dependent on the gametophyte for the supply of water and nutrients, including C and energy in the case of liverworts and some mosses; hornworts and many mosses have a very substantial capacity for sporophyte photosynthesis.


The long-distance transport of water and soil-derived nutrients from soil (or rock surface) to transpiration surfaces in bryophytes can involve either ectohydric or endohydric routes. The ectohydric pathway involves the surface of thalli, and is associated with the lack of a waxy cuticle on the surface (Richardson 1981; Raven 1999a, b; Ligrone, Duckett & Renzaglia 2000; Proctor 2000; Raven 2002). Raven (1999a, b), Proctor (2000) and Raven (2002) point out that ectohydric water transport is often correlated with desiccation tolerance. All bryophytes, like all algae, are poikilohydric, so only have a limited capacity to control their water content when the rate of tanspiratory water loss exceeds the rate of water uptake. Endohydric water transport over long distances in bryophytes involves dead cells with various degrees of (premortem!) modification which alter (increase) hydraulic conductivity.

The bryophyte gametophytes which lack a specialized endohydric conducting system comprise all of the hornworts, many liverworts (especially the thalloid examples) and the Andraeidae, Sphagnidae and some Bryidae among the mosses (Ligrone et al. 2000). The gametophytes of the Marchantiales with their intercellular gas spaces, gas-conducting pores and waxy cuticle, are endohydric, but the vertical pathlength for water movement is only a few hundred micrometres and the water flux can be sustained with modest driving forces through walls of unspecialized parenchyma cells (Raven 2002). Although there is usually little obvious differentiation of structures involved in ectohydric water movement in bryophyte gametophytes, Sphagnum has specialized water-conducting cells termed hyalocytes (Mozingo et al. 1969; Ligrone et al. 2000).

The sporophyte phase of bryophytes often lack an endohydric conducting system. There is generally held to be no endohydric conducting system in the erect sporophytes of hornworts. However, the elongate cells in the centre of the sporophyte (the columella) are dead at maturity, although they seem to lack a role in water transport (Ligrone et al. 2000). The non-photosynthetic sporophytes of liverworts (Proctor 1982) also lack an endohydric conducting system (Ligrone et al. 2000). Some mosses in the Bryidae (e.g. Grimmia pulvinata) lack endohydric conducting systems in their sporophytes, but appear to use ectohydric conduction via the upper leaves of the gametophyte in the supply of water to the developing sporophyte (Ligrone et al. 2000). How such a water supply can function in a cuticularized sporophyte is not clear. The absence of an endohydric conducting system in the Bryidae seems to be a derived (apomorphic) character (Ligrone et al. 2000; Newton et al. 2000).

The bryophytes with endohydric conducting systems include the gametophyte phase of liverworts in the Calobryales, and Metzgeriales (but not the Marchantiales or Jungermaniales) and in the gametophyte phase of mosses such as Takakia and of many members of the Bryidae (but not the Andraeidae or Sphagnidae) and in the sporophytes of many of the Bryidae (Hébant 1977; Ligrone & Duckett 1996; Ligrone et al. 2000). These hydroid cells have a major role in water transport in the mosses in which they occur (Grubb 1961; Hébant 1977; Ligrone et al. 2000) although the evidence for functionality of the putative water-conducting cells in liverworts is less extensive. Table 2 gives quantitative data on the structure of hydroids of mosses and liverworts.

The occurrence of hydrome in mosses is usually correlated with a water-repellent cuticle and, in sporophytes, with the occurrence of stomata (which may be less responsive to environmental signals than those of most tracheophytes) and intercellular gas spaces (Raven 2002). Water-repellency of the surface of the gametophytes of the Bryidae limits ectohydric water movement. Such waxy cuticles need not significantly restrict photosynthesis by decreasing permeability to CO2 (and O2) if the water-repellent wax layer is discontinuous. The water-repellency decreases the restriction on diffusive CO2 entry for photosynthesis by limiting the occurrence of surface water layers on the plant surface during rain or dew events. In sporophytes the water repellency is correlated with water resistance (low permeability of the cuticle to CO2 and H2O) which limits H2O loss (and CO2 entry) when the stomata are closed. This permits the homoiohydric condition for the sporophyte, although the sporophyte is parasitic on the poikilohydric gametophyte. The sporophytes of hornworts are also potentially homoiohydric, but are limited by their parasitism or the poikilohydric gametophyte and the lack of a well-authenticated endohydric conducting system in the sporophyte.

There are very few data on which to make estimates of the conductance of the endohydric conducting system of bryophytes. The diameter of the hydroids (Table 3) can be used, with the Poiseuille–Hagen equation, to compute a conductance and, with modifications, of varying spacing, thickness, and frequency and size of pores (Table 2), in the longitudinal transport pathway (see Schultze & Castle 1993; Comstock & Sperry 2000). Such anatomical measurements of hydroid diameter suggest that the conductivity of at least the larger elements could be at or above the lower end of the range for fern tracheids (10−9 m2 s−1 Pa−1) (Woodhouse & Nobel 1982; Gibson, Calkin & Nobel 1984; Gibson et al. 1985; Calkin, Gibson & Nobel 1985, 1986; Niklas 1985, 1992; Schultze, Gibson & Nobel 1987; Veres 1990). This conclusion has the proviso that the Hagen–Poiseuille equation is modified by the occurrence of transverse walls with phylogenetically related differences in wall thickness and in the occurrence of pores in these walls, with differences in both the number and size of pores as well as in the distance between transverse walls (conduit length): Ligrone et al. (2000); Table 3.

Table 3.  Occurrence of water-conducting cells and photosynthate-conducting cells in bryophytes (Ligrone et al. 2000)
TaxonInternal water-conducting cellsPhotosynthate-conducting cells
HornwortsGametophyte, sporophyte: absentGametophyte, sporophyte: absent
Gametophyte: slightly elongate cells (50–60 × 10 µm)
with thin walls (0·25–0·50 µm) and large pores
(300–600 nm) produced by lysis of primary wall
associated with modified plasmodesmata
Gametophyte: (Haplomitrium) shoots and ‘roots’:
elongate cells with organelles associated with
endoplasmic reticulum; disappearance of large vacuoles
at maturity; high frequency of plasmodesmata in end
Sporophyte: absentSporophyte: absent
(Pallaviciniaceae) Gametophyte: elongate cells
(300 µm × 8 µm) with tapering ends and thickened
walls (1·0–1·7 µm) perforated by pores
(250–600 nm) produced by dissolution of secondary
wall material, associated with modified
plasmodesmata; polyphenols in cell walls.
Gametophyte: (Pellia) elongate cells with organelles
associated with endoplasmic reticulum; disappearance of
large vacuoles at maturity; high frequency of
plasmodesmata in end walls.
Sporophyte: absentSporophyte: absent
Marchantiales s.l
Gametophyte, sporophyte: absent.Gametophyte: (Asterella) elongate cells with polarized cell
contents, organelles associated with endoplasmic
Sporophyte: absent
Gametophyte, sporophyte: absentGametophyte; sporophyte: absent
Gametophytes: slightly elongate cells (80 × 12 µm)
with thin walls (0·3 µm) and small pores (120 nm)
derived from plasmodesmata.
Gametophytes: conducting parenchyma cells as in Bryidae
(other than leptoids) as described below. Found in leafy
shoot, mature caulonema, rhizoids.
Sporophytes: absentSporophytes: conducting parenchyma in seta
Gametophytes, sporophytes: absent 
Gametophytes, sporophytes: absentGametophytes: conducting parenchyma cells as in Bryidae
(other than leptoids) as described below, but derived
from a subapical secondary meristem that is unique to
the Sphagnales.
Sporophytes: absent
Gametophytes: highly elongate cells (200–1500 × 10–
25 µm) with thin walls (0·5 µm) and remnants of
plasmodesmata (50 nm) but not true pores. Walls
loosely textured; polyphenols in cell walls. Most
strongly developed in Polytrichales; reduced in
some ‘advanced’ members.
Polytrichales:Gametophytes: leptoids are elongate cells (500 µm × 10–
20 µm) with thin walls (0·5 µm) and plasmodesmata in
the end walls with inflated centres (100 nm) but ends
constricted around the desmotubule. Polarization of
cells with most organelles at sink end; nacreous walls,
partial degradation of nucleus, microtubules associated
with organelles, large vacuoles absent.
Sporophytes: as for gametophytes.Sporophytes: leptoids in ‘conducting parenchyma’ (see
below) in seta. Other orders: Gametophytes and sporophytes: conducting
parenchyma cells are elongate (150–250 × 10–20 µm)
with walls 0·5 µm thick and up to 30 plasmodesmata
per µm2 in end walls. Polarized cell contents,
microtubules associated with organelles, vacuoles absent
at maturity.

The driving force which moves the solution through the water-conducting cells in bryophytes is a pressure gradient. The obvious driving force is the transpirational loss of water which has been shown in tracheophytes to decrease the pressure potential in the water-conducting cells nearest to the sites of transpiration, causing a pressure gradient which moves water from the uptake sites to the transpirational termini (Steudle 2001).

The other potential driving force, especially under conditions of low transpiration, is that of ‘root’ pressure. Of course, there are no true roots in bryophytes, although Grubb (1970) makes an excellent case for the root-like nature of below-ground organs of the liverwort Haplomitrium (Calobryales) and of Takakia which is now known to be a moss (Kenrick & Crane 1997; Ligrone et al. 2000). Whether ‘root’ pressure occurs in bryophytes with water-conducting cells, regardless of whether they have ‘roots’, is still uncertain (Grubb 1961, 1970; Raven & Edwards 2001; Raven 2002).

The phylogeny of the water-conducting cells of bryophytes is considered by Ligrone et al. (2000). Ligrone et al. consider that three independent origins of the perforate (disrupted plasmodesmata as a feature of terminal differentiation) water-conducting cells in calobryalean and metzegerialian liverworts and in the moss Takakia is likely, and that it is unlikely that any of these elements are related to S-type tracheids in tracheophytes (see Edwards 2002). The hydroids of bryoid mosses are imperforate (plasmodesmata occluded during differentiation) and are not homologous with tracheids (or vessels) of tracheophytes or ‘hydroids’ of the non-vascular polysporangiophyte Aglaophyton.


Symplasmic long-distance transport of photosynthate in bryophytes has been most investigated in the gametophyte phase of the Polytrichales (moss; Bryidae) (Hébant 1977; Richardson 1981; Scheirer 1990; Ligrone et al. 2000). For these mosses there is evidence as to the speed of 14C photosynthate transport, the conduits (leptoids) involved and the quantitative ultra structure of the leptoids (Reinhart & Thomas 1981; Thomas, Schiele & Scheirer 1988, Thomas, Schiele & Damberg 1990a, Thomas et al. 1990b; Thomas & Lombard 1991; Table 3). However, there is no evidence as to the concentration of solutes in the leptoids or the flux (specific mass transfer) along them. Accordingly, even a crude test of moss flow cannot be performed. Less work has been performed on the sporophyte phase of polytrichaceous mosses which also have leptoids.

For other bryophytes the physiological evidence for long-distance symplasmic transport is even less complete than it is for the Bryidae. However, for Sphagnum (Sphagnidae) Rydin & Clymo (1989) have shown that organic C carbon and phosphate (in an unknown form) are symplasmically transported from older to younger parts of the plant. Translocation is presumably an essential feature of the growth of Sphagnum in oligotrophic waterlogged environments, as is symplasmic transport of C, N and P from older to younger parts of unattached, dense populations of the characean Chara hispida (Andrews et al. 1984a, b; Box et al. 1984).

The cells in Sphagnum which are presumably involved in this transport are less differentiated than the leptoids of the Polytrichales and are termed ‘conducting parenchyma cells’ (Hébant 1977; Ligrone & Duckett 1998a, b; Ligrone et al. 2000; Table 3).

Putative long-distance symplasmic transport elements (conducting parenchyma cells) which are less differentiated than leptoids occur in mosses in the Bryidae (other than the Polytrichales), and in Takakia as well as in the Sphagnidae, and in some thalloid metzgerialean and marchantialean liverworts as well as the leafy liverwort Haplomitrium (Ligrone & Duckett 1994a, b, 1998a; Ligrone et al. 2000; Table 3). There seems to be no evidence on long-distance symplasmic transport in such organisms at speeds faster than can be accounted for by diffusion other than the work with 14C-labelled photosynthate in the marchantiaceous liverwort Conocephalum conicum (Trebacz & Fenson 1989), which showed that the speed of movement was consistent with the cytoplasmic streaming speeds found in this organism.

Ligrone et al. (2000) consider the published data, and provide previously unpublished results, on the cytology of putative ‘food-conducting cells’ of bryophytes (liverworts and mosses) (Table 2 of Ligrone et al. 2000; Table 3). Ligrone et al. (2000) also (their Table 3) indicate the additional cytological features found in leptoids of the Polytrichales, some of which parallel (their Table 3) the fine-structural features of tracheophyte ‘food-conducting cells’, with especial reference to sieve elements. In terms of a possible mass flow of solution the presence of desmotubules in the plasmodesmata of the transverse walls in the leptome and the absence of any major enlargement of the diameter of plasmodesmata at their constricted ends (as opposed to the median enlargement) is a major distinguishing feature of leptoids as opposed to the sieve pores of sieve cells and sieve elements (Ligrone et al. 2000).

These fine structural data, while not giving many clues as to the mechanism of long-distance transport, are helpful in considering the phylogeny of symplasmic long-distance transport in embryophytes. Ligrone et al. (2000) rightly consider that the leptoids and conducting parenchyma cells of mosses (Table 3) have an evolutionary origin independent of the sieve cells and sieve elements of tracheophytes, and that the striking similarities between leptoids and tracheophyte food-conducting cells is an example of homeoplasy (parallel evolution). Ligrone et al. (2000) consider it possible that the similarities between putative food-conducting cells of liverworts and of mosses indicates a common origin (i.e. that these conducting cells are a plesiomorphy of liverworts and mosses). Such a view is most consistent with topologies with hornworts as the earliest-branching clade of tracheophytes and with mosses not being the sister group of tracheophytes. Neither of these requirements are in accord with all of the available data (Kenrick & Crane 1997; Qiu et al. 1998; Duff & Nickrent 1999; Ligrone et al. 2000; Qiu & Lee 2000; Renzaglia et al. 2000).


The gametophyte phase of bryophytes lacks intercellular gas spaces except in the case of the marchantiaceous liverworts (see Raven 1996). However, the thallus of these liverworts is analogous to a heterobaric leaf of a vascular plant, i.e. with vertical cellular partitions preventing lateral transport in the gas phase over distances of more than a few millimetres. In the case of the marchantiaceous liverworts there is a single pore in the upper epidermis for each gas-phase compartment in the thallus. The movements of gas within these liverworts cannot be classified as long distance. The absence of long-distance transport in the gametophytes of robust mosses (e.g. Polytrichum, Dawsonia) could reduce the O2 supply to below-ground organs in waterlogged soils (Raven 1996).

In the sporophyte phase of hornworts and bryid mosses there are intercellular gas spaces and stomata. The gas spaces generally occur in photosynthetic tissue with stomata in the epidermis (capsule of mosses; most of the sporophyte of hornworts) so that there is a minimal distance for metabolic gases to move radially. However, gas spaces also occur in the seta of moss sporophytes where stomata are less common; the function of these gas spaces is not clear. There are even gas spaces among the conducting parenchyma cells (Fig. 7e of Ligrone et al. 2000); gas spaces among xylem or phloem cells (or hydroids or leptoids) are almost unknown (Raven 1991, 1994; Edwards 1993).


Apoplasmic and symplasmic long-distance transport in lower plants involves ‘xylem-like’ and ‘phloem-like’ transport in some bryophytes, and ‘phloem-like’ or cytoplasmic streaming-based symplasmic transport in some macroalgae. In only relatively few cases have detailed physiological studies been performed of speed of transport, concentration of the transported solutes and specific mass flow. Recent work has emphasized fine structural and phylogenetic aspects of the transport pathways. Recent technical advances (e.g. confocal microscopy) have not been adequately employed in studies of long-distance transport in lower plants.


The support of colleagues past and present is gratefully acknowledged.

Received 15 February 2002;received inrevised form 28 May 2002;accepted for publication 29 May 2002