|Home | About | Journals | Submit | Contact Us | Français|
Two main strategies that allow plants to cope with soil waterlogging or deeper submergence are: (1) escaping by means of upward shoot elongation or (2) remaining quiescent underwater. This study investigates these strategies in Lotus tenuis, a forage legume of increasing importance in areas prone to soil waterlogging, shallow submergence or complete submergence.
Plants of L. tenuis were subjected for 30 d to well-drained (control), waterlogged (water-saturated soil), partially submerged (6 cm water depth) and completely submerged conditions. Plant responses assessed were tissue porosity, shoot number and length, biomass and utilization of water-soluble carbohydrates (WSCs) and starch in the crown.
Lotus tenuis adjusted its strategy depending on the depth of submergence. Root growth of partially submerged plants ceased and carbon allocation prioritized shoot lengthening (32 cm vs. 24·5 cm under other treatments), without depleting carbohydrate reserves to sustain the faster growth. These plants also developed more shoot and root porosity. In contrast, completely submerged plants became quiescent, with no associated biomass accumulation, new shoot production or shoot elongation. In addition, tissue porosity was not enhanced. The survival of completely submerged plants is attributed to consumption of WSCs and starch reserves from crowns (concentrations 50–75 % less than in other treatments).
The forage legume L. tenuis has the flexibility either to escape from partial submergence by elongating its shoot more vigorously to avoid becoming totally submerged or to adopt a non-elongating quiescent strategy when completely immersed that is based on utilizing stored reserves. The possession of these alternative survival strategies helps to explain the success of L. tenuis in environments subjected to unpredictable flooding depths.
Flooding of the soil drastically reduces the availability of soil oxygen, and this can damage the growth and survival of the roots on which plants depend (Armstrong, 1979; Colmer, 2003; Voesenek et al., 2006). When water is sufficiently deep to cover part or all of the shoot, the stress is intensified as shoots become directly affected by shortages of carbon dioxide and oxygen imposed by slow gas diffusion through water. The severity of the stress thus depends on water depth. Accordingly, two contrasting strategies may be required to deal with partial or complete submergence. In rice (Metraux and Kende, 1983), maize (Jackson, 1989) and Paspalum dilatatum (Insausti et al., 2001) submergence of the shoot base can promote elongation (escape strategy) while complete submergence can also strongly induce this in rice seedlings (Setter and Laureles, 1996), Rumex palustris (Voesenek et al., 2006) and other species. In response to partial submergence, tolerant plants increasingly allocate carbon and nutrients towards shoot growth, permitting faster elongation in order to ‘escape’ from the water (Striker et al., 2008). In contrast, some plants are able to tolerate long periods of complete submergence; they maintain their basic metabolism by respiring reserve carbohydrates until water drawdown (Blom et al., 1994; Parolin, 2009). This second strategy is called quiescence (Bailey-Serres and Voesenek, 2008). To date, the ability to adopt the quiescence strategy has been described for certain specific different cultivars of rice (Oryza sativa; Setter and Laureles, 1996), for one overwintering ecotype of Ranunculus repens (Lynn and Waldren, 2003) and for Rumex crispus (Laan and Blom, 1990; Voesenek et al., 1990). However, the possibility that individual plants of one species or ecotype can adopt either strategy depending on water depth has not been examined before. We now address this question in the legume Lotus tenuis.
In addition to elongation and quiescence effects, submergence of some or all the plant can trigger changes in root porosity (Armstrong, 1979; Colmer, 2003) and in the carbohydrate metabolism of reserve organs (e.g. crowns and tap roots) to maintain root respiration and facilitate shoot elongation in the absence of adequate rates of photosynthesis (Colmer and Greenway, 2005). In legume species, previous reports on this topic mostly focused on flood-sensitive Medicago sativa (alfalfa). The usual observation was severely damaged plants and an accumulation of water-soluble carbohydrates (WSCs) in crowns [from 2–6 % to 11–16 % dry weight (d. wt.)] due to the lack of growth and sugar demand under waterlogged conditions and with little change in starch levels (Barta, 1988; Castonguay et al., 1993). However, there have been no comparable studies for a flood-tolerant legume such as L. tenuis, which is able to maintain carbon fixation and plant growth during long-term flooding (Striker et al., 2005).
If plants become completely submerged this can lead to a dramatic decline of the carbon balance and energy status of the plants and to plant death (Vartapetian and Jackson, 1997; Bailey-Serres and Voesenek, 2008). However, some plants are able to sustain their basic metabolism by consuming reserve carbohydrates (Laan and Blom, 1990; Schlüter and Crawford, 2001; Dixon et al., 2006). In some wetland species, this is supplemented by underwater photosynthesis (Mommer and Visser, 2005). Starch reserves can be mobilized during submergence and can readily provide soluble sugars for plant respiration in anaerobic tissues (Perata et al., 1992; Jackson and Ram, 2003; Colmer and Greenway, 2005). However, this capacity – which reflects the plant's ability to survive with little or no oxygen or photosynthesis – differs greatly among plant species (Crawford, 1992; Blom et al., 1994). For instance, the consumption of starch reserves in tubers of Scirpus maritimus allows the plant to survive complete submergence for nearly 3 months, while rhizomes of Juncus effusus survive only 4 d under anaerobic conditions (see Crawford, 1992). In accordance, our study also evaluated the potential use of carbohydrates in crowns of L. tenuis plants subjected to submergence. Consequently, we included in the present study the role of crown carbohydrates (WSCs and starch) in responses of L. tenuis to partial or complete submergence over 30 d. These were combined with analyses of shoot elongation, tissue porosity, morphological traits, shoot number, plant biomass and the utilization of crown carbohydrates. We show that increasing flooding depth induces alternative strategies by otherwise identical plants that are centred on faster shoot elongation under partial shoot submergence conditions and on quiescence under complete submergence linked to utilization of crown reserve carbohydrates.
The genus Lotus is native to the Mediterranean basin and widely distributed throughout the world, with several species adapted to stressful soil conditions poorly tolerated by other forage legumes such as M. sativa (alfalfa), Trifolium pratense (red clover) or Trifolium repens (white clover) (Blumenthal and McGraw, 1999; Striker et al., 2005; Teakle et al., 2006). In this regard, L. tenuis Waldst. & Kit. ex Willd. (syn. Lotus glaber Mill.) is a notable perennial with a long growing season. It possesses a well-developed crown (2 cm of the transition zone between shoots and roots that functions as a storage reserve tissue able to support shoot growth) and a taproot with extensive lateral root branching. Leaves are pentafoliate and the species is an important forage crop for flood-prone soils and is naturalized in several plant communities of the Flooding Pampa grasslands of Argentina (Striker et al., 2005, 2006, 2008).
Seeds of L. tenuis (‘Pampa INTA’) were germinated in an incubator (25 °C) in polystyrene boxes containing cotton and absorbent white paper saturated with water. After 2 d, germinated seeds were transplanted to 4 L plastic pots (three seedlings per pot) filled with sand and topsoil (1 : 1) from a lowland grassland of the Flooding Pampa (3 × 33 % organic carbon; for further details see Lavado and Taboada, 1987) and transferred to a glasshouse at the Faculty of Agronomy at the University of Buenos Aires. Seedlings were subsequently thinned to one plant per pot and grown for 6 months at 22 ± 6 °C and a photosynthetic photon flux density (PPFD) of 1400 ± 40 µmol m−2 s−1.
Four treatments were applied for 30 d to 6-month-old plants following a fully randomized design with eight replicates: (1) control – watered daily to field capacity; (2) waterlogging – saturated soil flooded to 2–5 mm above the soil; (3) partial submergence – water level maintained 6 cm above the soil surface; and (4) complete submergence – plants submerged in water 8–10 cm deep to ensure no leaves emerged above water level during the experiment. Soil oxygen availability was determined by measuring oxygen diffusion rates (ODRs) at a depth of 5 cm with platinum microelectrodes (Letey and Stolzy, 1964). Flood water was replaced weekly and root zone temperature checks showed these were similar for all treatments (P = 0·88).
Shoot and root porosity was measured at the end of the flooding period by pycnometry (Jensen et al., 1969). The porosity in aerial (shoot) and submerged tissues (roots) allowed comparison of the capacity for internal aeration from shoot to root among L. tenuis plants under the different treatments.
Shoots per plant were counted and measured at the beginning and end of the experiment. Shoot length was expressed as the average of the five longest shoots per plant. Both variables provide information on the strategy developed by the plants in response to submergence.
Plants were harvested for biomass and chlorophyll at the beginning and end of the experiment using randomly chosen individuals (n = 8). Plants was separated into leaves, stems, shoots, crown and roots for dry mass determinations after drying for 72 h at 80 °C. In addition, the ratio of leaf to total shoot biomass and of crown to total biomass was calculated for each plant. Leaf chlorophyll was measured in fully expanded young leaves using a portable chlorophyll meter (Model SPAD-502, Minolta, Ramsey, NJ, USA) at the end of the experiment.
For WSCs and starch, samples of crown tissue weighing approx. 0·1 g (d. wt.) were milled and heated to 80 °C in 25 mL of 80 % ethanol for 1 h. The supernatant was filtered twice, diluted with 25 mL of distilled water and heated to 60 °C for 1 h. All supernatants were made up to 100 mL with 80 % ethanol and aliquots mixed with anthrone reagent [200 mg of anthrone in 75 % (v/v) H2SO4], left to stand for 30 min, shaken and heated for 10 min in an 80 °C water bath (Yemm and Willis, 1954). Absorbance was recorded at 620 nm, and total soluble carbohydrate concentration was calculated based on a glucose calibration curve. To determine starch, the remaining residues were oven-dried at 37 °C for 24 h, suspended in 10 mL of distilled water and heated in a boiling water bath for 4 h. The resulting homogenate was incubated with 0·7 mL of a solution containing 18 U mL−1 of amyloglucosidase in an acetate buffer at pH 4·5 and 37 °C overnight. The amounts of glucose released were analysed by the anthrone method as described above (Yemm and Willis, 1954). WSCs and starch content (mg per plant crown) were calculated as the product of WSCs or starch concentration (mg g−1) and crown biomass (g) at the beginning and end of the experiment. The difference between these two values represented the change in WSCs or starch content during the course of the 30 d experiment.
All variables were analysed using one-way analysis of variance (ANOVA) with Tukey's honestly significant difference tests. These were done a posteriori to identify differences among treatments (P < 0·05). Normality and the homogeneity of variances were previously verified. The variables involving proportions were arcsine √x transformed before analyses. In addition, orthogonal contrasts were performed to compare data between day 0 and day 30. Statistical analyses were performed using the STATISTICA package for Windows (StatSoft, Tulsa, OK, USA). All results are presented as non-transformed means ± s.e.
The ODR under control conditions remained between 74·3 and 78·3 × 10−8 g cm−2 min−1, indicating well-oxygenated soil conditions. In the waterlogged soil, the ODR decreased slowly over 6 d to values reported as non-limiting for root growth (25·5 ± 3·2 × 10−8 g cm−2 min−1; see Stolzy and Letey, 1964). By day 10, ODRs in waterlogged soil decreased further to 1·5 ± 0·7 × 10−8 g cm−2 min−1, and similar values were recorded up to the end of the 30 d long experiment. The ODR under partial and complete submergence conditions decreased rapidly from 77·3 ± 3·7 to 6·9 ± 1·1 × 10−8 g cm−2 min−1 during the first 3 d, and remained near zero until the end of the experiment, indicating anaerobic soil.
Root porosity was lower under control conditions (12·8 %) and higher in treatments involving excess water (P < 0·05, Table 1). Values ranged from 21·6 to 32·5 % for plants growing under waterlogged or in partial or completely submerged conditions (Table 1). In contrast, constitutive shoot porosity in well-drained plants was always higher than root porosity (e.g. 24·6 % vs. 12·8 %) and did not increase under soil waterlogging or under complete submergence. An increase in shoot porosity was seen only in partially submerged plants (P < 0·01).
The number of shoots of L. tenuis increased during the course of the experiment except in completely submerged plants (Fig. 1). The magnitude of the increase was lower in partial submergence conditions than under control and waterlogging conditions. Thus, the shoot number of partially submerged plants was 22·1 % less than in control and waterlogged plants (P < 0·05; Fig. 1). Instead, partially submerged plants prioritized elongation of the emerging shoots since the five longest shoots of the plant were 30 % longer than under control or waterlogging conditions (P < 0·01; Fig. 1). In contrast, completely submerged plants did not increase the number or length of their shoots (P > 0·7) but, importantly, all plants survived the 30 d complete submergence treatment. The lack of change in these variables is evidence of the quiescence syndrome of L. tenuis plants when wholly submerged in water.
Increasing flooding depth had a negative effect on growth in biomass (Fig. 2). At the end of the experiment, total plant biomass of partially submerged plants was 30·3 % below that of control plants and 64·4 % less than in waterlogged plants (P < 0·05). Under complete submergence, biomass accumulation during the experiment was negative (P = 0·072; cf. day 0 vs. day 30 in Fig. 2). Final shoot biomass was progressively lowered by increased flooding depth: waterlogged, partially submerged and completely submerged plants finishing with a shoot biomass of 21·7, 33·5 and 64·3 %, respectively, below that of well-drained plants (P < 0·05 for all cases; Fig. 2). Different parts of the shoot were affected differently. Leaf biomass was depressed more than stem biomass as flooding depth increased. Accordingly, the leaf-to-shoot biomass ratio was smaller in partially submerged and completely submerged plants than in control and waterlogged plants (P < 0·05; Table 2). Completely submerged plants had a lower leaf-to-shoot biomass ratio and much lower chlorophyll leaf content (Table 2).
Crown biomass at final harvest was less under partial submergence and complete submergence than in other treatments (33·3 and 81·3 % lower mass, respectively; P < 0·05 in all cases; Fig. 2). However, important differences were noted: in partially submerged plants, crown biomass was little different from the initial mass 30 d previously, while it was much lower than the starting mass in completely submerged plants (cf. day 0 vs. day 30 in Fig. 2). The reduction in the crown biomass of completely submerged plants was disproportionately larger than the reduction in total biomass, as indicated in the lower crown-to-total-biomass ratio (Table 2). Root growth was also differentially affected by the treatments: it ceased in both partially submerged and completely submerged plants (P > 0·25; cf. day 0 vs. day 30 in Fig. 2), but continued similarly under control and waterlogged conditions (Fig. 2). Thus, partially submerged and completely submerged plants registered a root biomass that was 42·8 and 63·0 % lower, respectively, than that of control and waterlogged plants by the end of the experiment (P < 0·05; Fig. 2).
WSCs and starch accumulated in plant crowns during the course of the experiment under control, waterlogging and partial submergence conditions (Fig. 3). The magnitude of the accumulation differed among treatments for WSCs but not for starch content (Fig. 3). Crown WSCs were similar in concentration in well-drained controls, waterlogged and partially submerged plants but substantially depressed by complete submergence (P < 0·01; Fig. 3). Total accumulation per crown over 30 d (ΔWSCs) was, however, reduced by both waterlogging and partial submergence, while a large loss in total WSCs per crown was seen in completely submerged plants over time. Similar but less marked effects were also seen for starch. Importantly, complete submergence was the only treatment that provoked reductions in WSC and starch concentration (Fig. 3).
The research reveals that L. tenuis plants select either an escape strategy based on promoted underwater shoot elongation or a quiescence strategy based on suppressed elongation and greater consumption of stored reserves. The choice depends on whether the shoot is partially or completely covered by the flood water. This may be the first report showing that individual plants of a given species or ecotype can respond in either way depending on the circumstances. The species is therefore able to change its growth strategy depending on the degree of plant submergence.
When faced with partial submergence, root growth ceases in L. tenuis plants, and carbon allocation prioritizes an acceleration of shoot extension without incurring a substantial depletion of carbohydrate reserves (WSCs and starch). As expected for a flood-tolerant species, partially submerged plants increased shoot tissue porosity (Jackson, 1989; Jackson and Armstrong, 1999). In contrast, completely submerged plants suspended their growth, with no biomass accumulation, no production of new shoots or shoot elongation, while maintaining low and unchanged tissue porosity. These responses (i.e. lack of positive growth responses and no increase in porosity) comprise a quiescence strategy that enhances the prospects of survival. Such survival is related to the conspicuous consumption of WSCs and starch reserves stored in crowns to sustain their basic metabolic requirements in the absence of significant amounts of net underwater photosynthesis. The switch between escaping from water and remaining quiescent thus depends on water depth. This helps to explain the success of L. tenuis in environments subjected to varying and unpredictable flooding depths.
Internal plant aeration is critical for growth and survival during waterlogging and partial submergence (Armstrong, 1979; Vartapetian and Jackson, 1997; Colmer, 2003). In waterlogged L. tenuis plants, this process does not seem to constrain biomass accumulation as shoot and root biomass was similar to that of well-drained control plants (Fig. 2). In this sense, waterlogged plants presented all leaves above the water level, maximizing photosynthesis, oxygen capture (Laan et al., 1990; Grimoldi et al., 1999) and oxygen transport to roots through increased root and shoot porosity (presumed aerenchyma tissue; Table 1). In deeper water (i.e. partial submergence) only some leaves were submerged and plants showed increased shoot porosity, providing improved potential for longitudinal gas transport and aeration to benefit inundated parts (Colmer, 2003). Under these conditions, carbon allocation was directed exclusively towards shoots at the expense of root growth (Fig. 2). This latter response could be related to increasing the elongation rate of shoots (Fig. 1; Grimoldi et al., 1999), possibly associated with increased accumulation of ethylene gas within submerged parts that promotes elongation in co-operation with gibberellins and a decline in the inhibitor abscisic acid (Voesenek et al., 1990, 2006; Fukao and Bailey-Serres, 2008; Jackson, 2008). A higher rate of shoot elongation maintains an increased proportion of the leaf area above water. This, in turn, facilitates atmospheric oxygen entry and the maintenance of photosynthesis (Laan et al., 1990; Insausti et al., 2001; Striker et al., 2005) and avoids underwater leaf senescence (yellowing) that is also ascribed to accumulated ethylene in submerged leaves (Jackson et al., 1987). In the present work, photosynthesis by partially submerged plants appeared to contribute markedly to shoot growth in length and mass and avoided the need for consumption of carbohydrate reserves from the crown (Fig. 3). This notion is supported by the relatively higher leaf-to-shoot ratio and the greenness of the leaves of partially submerged plants (Table 2), as well as maintenance of stomatal conductances and photosynthesis as shown in earlier work on L. tenuis plants that were partially submerged for 25 d (Striker et al., 2005).
Plants subjected to complete submergence are intensely stressed. All the leaves are underwater and thus cannot capture the carbon dioxide and oxygen needed to sustain vigorous photosynthesis and aerobic metabolism. In this situation, plant responses were directed towards survival instead of growth (Fig. 2) as seen previously in R. crispus (Laan and Blom, 1990), and submergence-tolerant forms of O. sativa (Setter and Laureles, 1996). In L. tenuis there was also an absence of shoot elongation or biomass accumulation under complete submergence (Figs 1 and and2).2). Both strategies have been demonstrated as beneficial for plant survival of submergence. Setter and Laureles (1996) evaluated six cultivars of O. sativa and reported that survival of 14 d complete submergence was strongly correlated with the lack of shoot elongation of rice plants. Moreover, these authors proposed that shoot elongation competes with maintenance processes for energy. Therefore, energy management appears to be crucial for submergence tolerance (Fukao and Bailey-Serres, 2004). It is possible that, as in rice, L. tenuis also possesses a gene such as the Sub1A gene in rice (Fukao and Bailey-Serres, 2008) that responds to submergence ethylene by repressing the elongation-promoting action of other ethylene-responsive genes. Our results also suggest that the energy used by plants to survive 30 d of complete submergence is drawn from WSC consumption and starch breakdown in the crowns (Fig. 3), as also reported for tap roots in submerged R. crispus plants (Laan and Blom, 1990) and culms of submerged rice plants (Ram et al., 2002). Furthermore, in our study, underwater photosynthesis seemed unable to provide significant energy under submerged conditions (see review by Mommer and Visser, 2005): plants had a reduced leaf-to-shoot biomass ratio (Fig. 2; for rice, see Ella and Ismail, 2006) and leaves evidenced a lower chlorophyll content with respect to other treatments (Table 2; for rice, see Ella et al., 2003). In addition, recent work on six wetland species revealed that the values for underwater photosynthesis of Acorus calamus, a species known to survive lengthy submergence by consuming stored carbohydrates in rhizomes (Schlüter and Crawford, 2001), were three times slower than among other species (Colmer and Pedersen, 2008). In this regard, it is possible to speculate that the capacity for underwater photosynthesis decreases among species which rely on carbohydrate consumption to survive complete submergence. This hypothesis and its implications merit further investigation.
Crown carbohydrates accumulated in all growing plants (control, waterlogged and partially submerged), and were only consumed under complete submergence, the most stressful condition imposed in our experiment. These results concur with previous works on the use of stored carbohydrates in L. tenuis faced with extreme carbon starvation: plants that were simultaneously subjected to intense defoliation and partial submergence (Striker et al., 2008). The use of WSCs and starch from storage organs as an energy supply for maintenance processes associated with plant survival appeared to be critical when submergence time is prolonged (Crawford, 1992; Bailey-Serres and Voesenek, 2008). The transformation of starch into respirable sugars under anoxia occurred in the crowns of L. tenuis (this study, Fig. 3) and is also known to take place in rhizomes of A. calamus (Schlüter and Crawford, 2001) and tubers of Potamogeton pectinatus (Dixon et al., 2006). Although crown reserves allow L. tenuis to survive prolonged submergence, a longer recovery time for crown reserves and to develop new functional leaves after submergence needs to be considered for grazing management purposes. Otherwise, plants biomass production, vigour and persistence could drop if plants are defoliated immediately after the water subsides.
The results obtained in this experiment are significant for the ecophysiology and management of this forage crop species. Future work exploring the responses of other genotypes of L. tenuis (e.g. natural populations) would be interesting to assess the degree of intraspecific variation in the reported ability of plants to adjust their strategy depending on flooding depth. The optimal survival strategy will depend on the water regime of the specific environment in which L. tenuis is incorporated as part of the forage resources. In areas suffering from shallow flooding events, such as plant communities located along intermediate topographical positions, plant responses aimed at maintaining leaves above the water level to allow for oxygen capture and carbon fixation are considered sufficient for plant growth and survival. In contrast, in lowland areas in which plants periodically deal with waterlogging to complete submergence, L. tenuis demonstrated the ability to adjust its responses, repressing shoot elongation and changing to a quiescence strategy. This flexibility has not been reported before for individual plants. Previous studies only revealed how different species or ecotypes possess contrasting quiescence of escape strategies for dealing with deep or more shallow submergence [e.g. O. sativa cultivars (Setter and Laureles, 1996) R. crispus vs. R. palustris (Laan and Blom, 1990; Voesenek et al., 1990) and contrasting forms of Ranunculus repens (Lynn and Waldren, 2003)]. The developmental plasticity of L. tenuis thus contributes to its persistence in flood-prone environments where other popular legumes do not thrive.
Valuable comments from Professor Timothy D. Colmer (University of Western, Australia) helped in different stages of this work. We thank Rolando J. C. León (University of Buenos Aires, Argentina) for his invaluable support throughout the study and the Bordeau family, owners of Estancia Las Chilcas, who facilitated our work on their land for soil extraction. This study was funded by grants from the University of Buenos Aires (UBA G-421) and Agencia Nacional de Promoción Científica y Tecnológica ANPCyT Foncyt-PICT 20-32083. M.M. was supported by a fellowship from University of Buenos Aires.