Production of adventitious roots is regarded as an important morphological trait contributing to waterlogging tolerance (Jackson & Drew, 1984). In many wetland species, adventitious roots dominate root biomass (e.g. the glycophytic Rumex spp.; Laan et al., 1989) and such roots also form in wetland halophytes (e.g. S. virginicus, Naidoo & Mundree, 1993; Halosarcia pergranulata, Pedersen et al., 2006). The influence of salinity on development of adventitious roots during flooding has, however, rarely been studied. Production of ‘aboveground’ (i.e. aquatic) roots was not influenced by salinity (highest treatment was 400 mM NaCl) in the perennial grass S. virginicus (Naidoo & Mundree, 1993), whereas saline floodwater reduced adventitious root formation in the salt-tolerant semi-xerophyte tree Eucalyptus camaldulenis (42 dS m−1; van der Moezel et al., 1988) and the halophytic tree Melaleuca ericifolia (49–60 dS m−1; Salter et al., 2007). In Casurina obesa, adventitious roots were produced in saline (42 dS m−1) floodwater, but these roots deteriorated in health compared with those of plants flooded with nonsaline water (van der Moezel et al., 1988). Further information is needed on whether salinity has specific effects on adventitious rooting, beyond those that would be expected as a result of overall reductions in growth at high external NaCl concentrations.
Aerenchyma and internal O2 movement
Aerenchyma consists of large interconnected gas channels that greatly enhance O2 movement into submerged portions of plants (Armstrong, 1979). The most comprehensive study of the capacity to develop root aerenchyma (quantified as gas volume per unit root volume, i.e. porosity) was a survey of 91 species (nonwetland, ‘intermediate’, and wetland plants) grown in nonsaline waterlogged conditions in pots (Justin & Armstrong, 1987); 19 of the species studied are listed as halophytes by Aronson (1989). Before we consider these data on root porosity for halophytes, the difficulty in distinguishing the ‘arbitrary boundary’ between halophytes and nonhalophytes (discussed in Flowers & Colmer, 2008) should be noted, and considered for Phragmites australis. Aronson (1989) included P. australis as a halophyte; however, P. australis occurs predominately in fresh waters, but also in brackish marshes. Phragmites australis‘types’ with relatively high salt tolerance have been described (e.g. Lissner & Schierup, 1997; Vasquez et al., 2006; Takahashi et al., 2007) and although P. australis occurs in some saline marshes, in some cases the plants have been shown to be accessing interstitial water of salinity considerably lower than that of the surface water (e.g. Lissner & Schierup, 1997; Adams & Bate, 1999), so that some authors have referred to P. australis as a ‘pseudo-halophyte’ (Adams & Bate, 1999). The majority of studies of internal aeration in P. australis have, with two exceptions (discussed later in this section), been conducted in fresh water.
Root porosity in the two ‘nonwetland’ halophytes (Festuca rubra and Cochlearia danica) was ~2 to 4%, in the four ‘intermediate’ halophytes (Halimione portulacoides, Suaeda maritima, Juncus bufonius and Lepidium latifolium) it was ~2 to 23%, in 13 of the ‘wetland’ halophytes (e.g. Cochlearia anglica, Salicornia europaea, Puccinellia maritima and Spartina anglica) it was ~3 to 33%, and in P. australis it was 52% (Justin & Armstrong, 1987). Although roots of some wetland halophytes developed high porosity (e.g. S. anglica, 33%; P. australis, 52%) the values (except for that of P. australis) were generally less than those in roots of several glycophytic wetland species in the study (e.g. Juncus inflexus, 53%) (Justin & Armstrong, 1987). Some halophytes can develop high root porosity (e.g. 48% in Spartina patens; Burdick & Mendelssohn, 1990) almost as high as in roots of glycophytic wetland species (cf. 53% in J. inflexus; Justin & Armstrong, 1987). Thus, we conclude that the range of root porosity in halophytes is similar to that in glycophytes and the limited data available indicate that, as in glycophytes, the capacity to develop high root porosity tends to be greatest in species originating from flood-prone habitats.
The study by Justin & Armstrong (1987) was conducted under nonsaline conditions. The question of whether salinity influences aerenchyma formation has been addressed on few occasions. Sporobolus virginicus grown in N2-bubbled solution contained 32% aerenchyma in rhizomes when in fresh water and 28% when in coastal water (~410 mM Na+) (Donovan & Gallagher, 1985). Naidoo & Mundree (1993) reported that salinity reduced the air spaces in the culm of S. virginicus; the area of aerenchyma decreased from 30–32% to ~22% when salinity was increased from 0 or 100 mM to 200 or 400 mM NaCl. By way of contrast, in Spartina alterniflora and S. patens, increasing salinity from 25 to 325 mM NaCl did not affect root porosity measured as root specific gravity (Naidoo et al., 1992). The lack of influence of salinity on root porosity in S. patens is consistent with the high porosity (48%) of roots collected from plants growing in a salt marsh (Burdick & Mendelssohn, 1987).
In addition to high porosity, other features also enhance the capacity for longitudinal O2 diffusion in roots (e.g. root diameter, numbers of laterals, relative stele volume, and a barrier to radial O2 loss (ROL)) (Armstrong, 1979; Colmer, 2003). A barrier to ROL in cell layers exterior to the aerenchyma is present in the basal zones of roots of numerous wetland species; the barrier restricts O2 losses to the rhizosphere, and thus enhances longitudinal diffusion towards the root tip (Armstrong, 1979; Colmer, 2003). For example, roots of waterlogging-tolerant species in the genus Hordeum, including the halophyte Hordeum marinum, form a barrier to ROL when grown in an O2-deficient stagnant medium (Garthwaite et al., 2003). In roots of H. marinum the barrier is formed by putative suberin deposits in cell walls in the hypodermal layer (Garthwaite et al., 2008); suberin deposits have also been implicated in barrier formation in roots of P. australis (Soukup et al., 2007). Whether salinity influences the formation of a barrier to ROL in roots of wetland halophytes that possess this feature, and whether having a barrier to ROL influences the regulation of ion ‘exclusion’ by roots in saline substrates, are both unknown.
Diffusion is the mechanism by which O2 moves along aerenchyma in roots (Beckett et al., 1988). In rhizomes of some wetland species, through-flows (i.e. pressure-driven mass flows of gas) occur, but even in these species O2 still diffuses into, and along, roots subtending rhizomes (Colmer, 2003). Through-flows along rhizomes (and petioles or culms) enable plants to grow deeper into waters and/or anoxic sediments (Vretare-Strand, 2002) than would be possible if O2 moved only via diffusion through the entire plant body. Such through-flows increase the O2 available to rhizomes at depth, and higher O2 at root–rhizome junctions enhances internal O2 supply to roots (through-flow mechanisms, as studied in water lilies and common reed, are reviewed in Colmer, 2003). Through-flows have been documented in several wetland species, including the halophyte S. alterniflora (Hwang & Morris, 1991).
The influence, if any, of salinity on through-flows has not been studied for halophytes; but through-flows have been evaluated for P. australis at brackish and low-salinity sites in Europe (Rolletschek et al., 1999; Rolletschek & Hartzendorf, 2000). At brackish sites, growth of P. australis was inhibited so that plants had smaller sheath areas and thinner culms than those in low-salinity sites; these changes reduced the ventilation capacity of the plants at the brackish sites. Resistances to flows in rhizomes (and/or exit culms) were also higher, further reducing ventilation in plants at the brackish sites compared with the low-salinity sites. Through-flows in halophytes might therefore also be influenced by salinity at sites where the salt concentration impedes shoot growth.
Shallow rooting in aerobic upper soil layers, in addition to some internal O2 movement, presumably contributes to the means by which some marsh halophytes (e.g. the annual species of Salicornia) with low root porosity (namely 3–6%) can tolerate flooding in salt marshes (Mahall & Park, 1976; Pearson & Havill, 1988a). In addition, tolerance of high sulphide concentrations might also contribute to the persistence of Salicornia spp. in marsh soils (Ingold & Havill, 1984). Superficial rooting presumably also contributes to survival of other salt marsh species with low root porosity (e.g. ~3% in the annual leaf-succulent Suaeda maritima, Justin & Armstrong, 1987), whereas for others (e.g. perennial Sarcocornia perennis) growth is restricted to nonreducing, well-drained sandy sediments in tidal marshes (Davy et al., 2006). Small, locally raised areas of sediments can also support ‘patches’ of S. perennis within some marshes (Davy et al., 2006). We have not been able to locate data on anoxia tolerance per se in salt marsh species with low root porosity, although ADH activity has been measured (e.g. Salicornia spp.; Pearson & Havill, 1988b). Anoxia tolerance might be important during short periods of tidal submergence, so it remains to be resolved whether this aspect of waterlogging tolerance differs between these marsh species with low root porosity and those with higher root porosity. Aspects of anoxia tolerance in roots of halophytes are discussed in the next section.
Even in species with large volumes of aerenchyma in rhizomes and roots, O2 deficiency can still occur in some parts of the roots in flooded soils (e.g. S. patens; Burdick & Mendelssohn, 1990). Deeper parts of root systems, and even some tissues within individual roots (e.g. apices and stele), may suffer suboptimal O2, while other parts remain aerobic (Armstrong, 1979; Armstrong et al., 2000). Tissue anoxia results in the cessation of oxidative phosphorylation, and thus a severe energy deficit. Anaerobic carbohydrate catabolism, together with other essential traits (e.g. regulation of ATP-consuming processes), enables survival during anoxia; the period of survival depending on the supply of carbohydrates, the species, and the tissue or organ (Gibbs & Greenway, 2003; Greenway & Gibbs, 2003). Fermentative metabolism provides at least some ATP during anoxia, albeit only 3 to 35% of the rate of energy production in aerobic cells (Gibbs & Greenway, 2003). The anoxia-tolerant coleoptiles of rice (Oryza sativa) (Huang et al., 2005; Lasanthi-Kudahettige et al., 2007) and anoxia-sensitive Arabidopsis (Klok et al., 2002; Gonzali et al., 2005), both glycophytes with sequenced genomes, have been used in comprehensive gene expression and proteomic studies. It is beyond the scope of the present short review to consider the findings of these studies in glycophytes; the point made here is that such studies are lacking for wetland halophytes as suitable genetic resources are not presently available.
Biochemical studies related to anoxia tolerance in wetland halophytes are largely restricted to assessments of alcohol dehydrogenase (ADH) activities in roots; even ethanol production data per se appear to be lacking. Several studies demonstrate large increases in ADH activity in roots of S. patens when in waterlogged (or O2-deficient) conditions (e.g. 6-fold in the first weeks after flooding, Burdick & Mendelssohn, 1990; 17-fold, Naidoo et al., 1992). Interestingly, although the increase in ADH activity was 17-fold at 25 mM NaCl, it was only 4-fold at 325 mM NaCl, a reduction attributed by Naidoo et al. (1992) to ‘unhealthy looking roots’ in the high-salinity treatment. By contrast, ADH activities in roots of S. alterniflora were much lower than in those of S. foliosa, and were not affected by increasing salinity from 25 to 325 mM NaCl (Naidoo et al., 1992). Nevertheless, induction of ADH in roots of S. patens by waterlogging in controlled pot experiments is consistent with 3- to 15.8-fold higher activities of this enzyme in roots of this species sampled from a marsh, compared with a dune, environment (Burdick & Mendelssohn, 1987). Thus, at least ADH in roots of halophytes appears to respond similarly to that in roots of some freshwater marsh species (e.g. Smith et al., 1986). We expect that other aspects of anoxic metabolism in halophytes will probably reflect those in glycophytes, although further studies, particularly to measure rates of ethanol production in anoxia and of pyruvate decarboxylase activity (the rate-limiting enzyme in conversion of pyruvate to ethanol, Gibbs & Greenway, 2003), should also be assessed for a range of halophytic species. In addition, the hypothesis of Rivoal & Hanson (1993) that alternative end-products to ethanol, especially lactate, might benefit whole-plant C balance for some halophytes with anoxic roots should also be further explored. We refer the reader to Gibbs & Greenway (2003) for further consideration of lactate production, as well as other end-products (e.g. alanine), during fermentative metabolism in plants. Anoxia tolerance and metabolism in halophytes should be assessed at a range of external NaCl concentrations.