Plasticity as a plastic response: how submergence-induced leaf elongation in Rumex palustris depends on light and nutrient availability in its early life stage

Authors

  • Heidrun Huber,

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    Search for more papers by this author
  • Xin Chen,

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    2. Plant Ecophysiology, Institute of Environmental Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, the Netherlands
    Search for more papers by this author
  • Marloes Hendriks,

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    Search for more papers by this author
  • Danny Keijsers,

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    Search for more papers by this author
  • Laurentius A. C. J. Voesenek,

    1. Plant Ecophysiology, Institute of Environmental Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, the Netherlands
    Search for more papers by this author
  • Ronald Pierik,

    1. Plant Ecophysiology, Institute of Environmental Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, the Netherlands
    Search for more papers by this author
  • Hendrik Poorter,

    1. Plant Ecophysiology, Institute of Environmental Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, the Netherlands
    2. Plant Sciences (IBG-2), Forschungszentrum Jülich, 52425 Jülich, Germany
    Search for more papers by this author
  • Hans de Kroon,

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    Search for more papers by this author
  • Eric J. W. Visser

    1. Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Experimental Plant Ecology, Heyendaalseweg 135, 6525 AJ Nijmegen, the Netherlands
    Search for more papers by this author

Author for correspondence:
Heidrun Huber
Tel: +31 24 3652401
Email: h.huber@science.ru.nl

Summary

  • Plants may experience different environmental cues throughout their development which interact in determining their phenotype. This paper tests the hypothesis that environmental conditions experienced early during ontogeny affect the phenotypic response to subsequent environmental cues.
  • This hypothesis was tested by exposing different accessions of Rumex palustris to different light and nutrient conditions, followed by subsequent complete submergence.
  • Final leaf length and submergence-induced plasticity were affected by the environmental conditions experienced at early developmental stages. In developmentally older leaves, submergence-induced elongation was lower in plants previously subjected to high-light conditions. Submergence-induced elongation of developmentally younger leaves, however, was larger when pregrown in high light. High-light and low-nutrient conditions led to an increase of nonstructural carbohydrates in the plants. There was a positive correlation between submergence-induced leaf elongation and carbohydrate concentration and content in roots and shoots, but not with root and shoot biomass before submergence.
  • These results show that conditions experienced by young plants modulate the responses to subsequent environmental conditions, in both magnitude and direction. Internal resource status interacts with cues perceived at different developmental stages in determining plastic responses to the environment.

Introduction

In their natural habitats, plants are exposed to a multitude of different environmental triggers, which vary in both space and time (Valladares et al., 2007). Plants have evolved the potential to respond to such variation by altering their traits, a process commonly referred to as phenotypic plasticity (Bradshaw, 1965; Sultan, 1995, 2010; Schmitt et al., 2003), resulting in the display of different phenotypes in response to different environmental conditions (Poorter et al., 2010). Phenotypic plasticity is adaptive if the phenotypic changes increase survival and performance in the environment compared with plants that do not express phenotypic changes. Owing to their effects on performance, plastic changes are expected to be under selection, and phenotypic plasticity has been proposed to be an adaptive strategy which evolves under spatially or temporally heterogeneous conditions (Gabriel & Lynch, 1992; Kingsolver & Huey, 1998; Herben & Novoplansky, 2010; Baythavong, 2011; Karban, 2011). While the responses to single or concurrent environmental cues have been tested in many experiments and are relatively well understood, we are only beginning to understand how cues experienced at different developmental stages may interact in determining phenotypic change (Weinig & Delph, 2001; Novoplansky, 2009; Halpern et al., 2010; Sultan, 2010).

Plastic changes may be morphologically constrained or selected against as a result of costs associated with plasticity (van Kleunen & Fischer, 2005; Weijschede et al., 2006; Bell & Galloway, 2008; Auld et al., 2010; Chen et al., 2011). Phenotypic plasticity can be costly if it requires investment into the organ responding plastically to its environment (Huber et al., 1998; Weijschede et al., 2006, 2008b; Dechaine et al., 2007). As a consequence, the magnitude of the plastic response may depend on carbohydrates accumulated previously, and may thus be limited when plants are growing in suboptimal conditions that constrain carbohydrate accumulation (Groeneveld & Voesenek, 2003). The degree of plasticity will also depend on the developmental status of the plant or organ (Watson et al., 1995), and it has been argued that conditions experienced at early developmental stages may constrain the response to subsequent environmental cues (Sultan, 2000; Weinig & Delph, 2001; Bruce et al., 2007). Furthermore, different environmental stimuli can interact in determining the expression of plasticity (Cipollini, 2004; McGuire & Agrawal, 2005; Dechaine et al., 2007; Bell & Galloway, 2008; Anten et al., 2009). Summarizing, the available data so far suggest that the pattern and magnitude of plastic responses are plastic themselves. However, as yet we know very little about the extent to which a specific phenotypic response to a given environment is affected by conditions experienced previously and about the underlying mechanisms.

In this paper we will show how elongation responses induced by complete submergence in Rumex palustris are affected by the quantities of light and nutrients experienced by plants before submergence. Flooding is a common stress for terrestrial plants inhabiting river floodplains. It can have severe effects on plant performance, especially as gas diffusion in water is extremely slow compared with air, resulting in oxygen depletion in the submerged plant (Bailey-Serres & Voesenek, 2008, 2010). Some plant species such as R. palustris have evolved flooding-induced shoot elongation as a plastic response to periodical flooding (Kende et al., 1998; Voesenek et al., 2004, 2006; Bailey-Serres & Voesenek, 2008). Such elongation is beneficial because it may bring the shoot tip above the water surface and thus restore contact with the air (Pierik et al., 2009).

Rumex palustris seedlings were grown in a factorial combination of two intensities of light and two nutrient conditions and subsequently subjected to complete submergence. We hypothesize that plants that were growing in more favorable conditions, and that were therefore able to accumulate more resources, will show higher degrees of plasticity in response to subsequent submergence. Previous research has shown that the environment can affect the amount of resources accumulated in the taproot of R. palustris, which in turn affects plasticity to subsequent environmental cues. In this species, submergence-induced plastic elongation responses have been shown to depend on the presence of sufficient starch in the tap root (Groeneveld & Voesenek, 2003). Therefore, early shading may hamper subsequent elongation in response to flooding. Moreover, low nutrient availability will require investment into roots and thus reduce investment into shoots and concomitant shoot elongation. The combination of light and nutrient conditions may determine the accumulation of storage compounds, which depends on the balance among photosynthesis, respiration, storage, and growth. Plant biomass and carbohydrate concentrations were determined immediately before submergence as a measure of accumulated resources. Total leaf length was measured both before and after submergence to quantify the responses to the preconditions and subsequent flooding. Accessions from three populations were used for the experiment to investigate the consistency of the responses.

Materials and Methods

The experiment was performed with the biennial plant species Rumex palustris Sm. (Polygonaceae), which is often found in temporally wet habitats such as riverine floodplains or in habitats with more stagnant water tables (Chen et al., 2009). R. palustris responds to flooding by increasing leaf elongation, which may shift the leaves above the water surface (Supporting Information, Fig. S1). Seeds typically germinate in nonflooded conditions in autumn and survive winter as vegetative rosettes. They restart growing in spring and usually flower in the summer, but long-lasting and/or deep floods during the growing period may delay flowering until the next year. In these habitats the species is subjected to a large temporal and spatial variation of habitat conditions primarily caused by submergence events of unpredictable timing, depth, and duration. Depending on the substrate and the vegetation cover, plants are subjected to a small-scale microsite variation in nutrient and light availability as well. Particularly if intervals between floods are relatively long, annual and clonal perennial plant species may produce substantial biomass in this habitat, leading to competition for light. Flooding leads to an increased heterogeneity of the immediate environment, as it results in deposition of substrate with variable combinations of clay and sand and physical removal of the vegetation cover. As a result, plants of the same population may be subjected to different resource availabilities.

In 2004, seeds were collected in 12 populations covering a range of habitats (Chen et al., 2009) and stored at room temperature after collection. A subset of three populations, all collected along the Waal distributary of the river Rhine system, was used in this experiment. The numbering of the populations conforms with the population numbering used by Chen et al. (2009). Seeds of Population 1 were collected near Doornenburg at the Klompenwaard floodplain (51°53′08.36″00N, 06°00′57.33″00E). This population occurred on a clay bank of a recently dug side channel of the main river bed, which is frequently flooded. The seeds from Population 4 were collected at a floodplain near Ewijk (51°52′45.20″00N, 05°44′51.77″00E) with a relatively sandy substrate and subjected to frequent flooding. The seeds for Population 5 were collected near Deest (51°53′38.81″00N, 05°39′33.07″00E), at an old riverbank with a clayish substrate, which is further away from the running water and flooded less frequently. We used the seed families of six mother plants for each population, subsequently referred to as accessions.

The seeds were germinated for 10 d in separate Petri dishes on filter paper moistened with tap water and kept at a day : night cycle of 12 h of light (photosynthetic photon flux density (PPFD), 20 μmol m−2 s−1, 25°C) : 12 h of dark (15°C). Thirty-two seedlings per accession were individually transplanted in 240 ml plastic pots filled with a 1 : 1 mixture of sieved potting compost and sand in a climate chamber at 19°C. The length of the light period was 16 h (PPFD, 180–200 μmol m−2 s−1; sodium lamps SON-T plus 600 W fluorescent light, and TLD Reflex 36W/840R, both from Philips, Eindhoven, the Netherlands) and the night period was 8 h. The pots were distributed randomly over the tables in the climate chamber. The first 4 d after transplanting, the pots were covered with transparent plastic sheets to prevent dehydration.

Light and nutrient pretreatments

One week after transplantation, plants were subjected to a factorial combination of two light and two nutrient treatments. These conditions will be referred to as pretreatments. Half of the plants were moved to a shade cage covered by a white cloth, which decreased the irradiance by c. 60%, to c. 80 μmol m−2 s−1. For logistic reasons shade treatments were performed on separate benches in the climate chamber. The benches did not differ in overall light (above shade cages), temperature and humidity conditions. Shaded as well as nonshaded plants were grown under two nutrient treatments. High-nutrient plants received 20 ml of a 0.25 strength modified Hoagland solution (Visser et al., 1996) at 7 and 14 d after planting. The plants subjected to the low-nutrient treatment received only tap water without additional fertilizer. The plants of the different nutrient treatments were grown in separate trays, so that nutrients leaking after watering could not be taken up by the nonfertilized plants. Nutrient treatments and accessions were randomized within light treatments. These pretreatments were continued until the sixth leaf was emerging which, depending on the developmental speed, lasted for 20–29 d. A total of 432 plants were subjected to the light and nutrient pretreatments, with six seedlings per accession being subjected to each pretreatment combination.

Initial harvest and carbohydrate determination

After the light and nutrient pretreatment phase, two plants were harvested for each accession × pretreatment combination in order to determine the plant traits and soluble carbon concentration before the onset of the submergence treatments. A total of 144 plants were used for this initial harvest. At this initial harvest, petiole and leaf lamina lengths of the fourth and all younger leaves were measured. Soluble carbohydrates were measured according to Chen et al. (2010). After careful and complete removal of the substrate, shoots and roots were immediately frozen in liquid nitrogen, and kept at –80ºC before freeze-drying. After freeze-drying, plants were divided into shoots and roots, weighed and ground before the extraction of soluble carbohydrates. Accurately weighed (c. 10 mg) homogeneous powder of shoot or roots was used for total nonstructural carbohydrate measurements. The powder was first boiled with 3 ml 2.7 M HCl for 1 h, and then neutralized with 2.7 ml 3 M NaOH. The concentration of the extracted carbohydrates was determined using the anthrone color reaction (Yemm and Willis, 1954). Anthrone reagent (2.5 ml; containing 0.04 mM anthrone, 6% (v/v) ethanol, and 75% (v/v) H2SO4) was added to a 50 μl sample, and incubated in a boiling water bath for 7.5 min, after which it was cooled down immediately in ice water. The absorbance was measured using a spectrophotometer (Shimadzu, UV-1205, Kyoto, Japan) at 625 nm. The total nonstructural carbohydrate concentration was calculated using a calibration curve based on solutions containing known amounts of glucose covering the range of the samples’ absorbance.

Submergence treatments

As previous research has shown that the potential of plants to elongate during submergence depends on the developmental stage of the petiole (Groeneveld & Voesenek, 2003), we decided to submerge the plants when a given plant had reached a specific developmental stage and not to submerge all plants at a given point in time. Otherwise differences in developmental speed among seed families would have obscured our ability to find and interpret differences in plastic elongation responses among accessions. Therefore, submergence treatments started as soon as the sixth leaf started to emerge, which was 20–29 d after transplantation. Note that the shading pretreatment was stopped at this time. A total of 288 plants were subjected to the submergence treatments. Two replicates of each pretreatment and accession combination were subjected to complete submergence and two replicates to drained conditions, where plants were surface-watered three times per wk. These drained conditions will also be referred to as control conditions. Ten 300 l basins (opaque polyethylene; 0.8 × 0.6 × 0.7 m) were filled with tap water 3 d before the first plants were submerged, so that the water temperature could adjust to the room temperature. Light and nutrient pretreatment as well as accessions were fully randomized within submergence treatments. In the basins, the water circulated with a flow rate of 1.2 l min−1. The temperature of the water at submergence was 19ºC. Before placing the plants into the basins, the pots were first submerged in a separate water-filled container, after which loose soil as well as algae growing on the top of the soil was removed. This was done to keep the water in the basins as clear as possible and to minimize algae growth. The submergence treatments were performed in a different climate chamber than the pretreatments with 16 h of daytime at PPFD = 100 μmol m−2 s−1 (similar lamps as described earlier) and 8 h of darkness.

After 14 d the plants were harvested and the roots were carefully washed free of soil substrate. Previous research has shown that leaves finish their elongation within 2 wk after submergence (Chen et al., 2009). Petiole and leaf lamina lengths of the fourth and all younger leaves were measured at the final harvest Plants were divided into roots, petioles, and leaf laminas. The DW of all plant compartments was determined after drying at 70°C for 48 h.

Statistical analyses

The aim of the paper was to assess the effect of presubmergence conditions on the response of plants to submergence and whether this effect was consistent among populations. Specifically, we wanted to know whether leaf length plasticity differed depending on light and nutrient availability during the pretreatment. Before the univariate analyses on the individual traits, we performed MANOVAs to investigate whether there is a significant treatment effect over a set of traits (SAS PROC GLM). We allocated the response variables into five logical groups, the first group containing leaf length at the initial harvest, the second group biomass data at the initial harvest, the third group the concentration and absolute amount of soluble carbohydrates in root and shoot tissue, the fourth group total leaf length at final harvest and the fifth group biomass of the different plant organs at the final harvest (Tables S1, S2).

Data of plants harvested before the onset of submergence treatments, including carbohydrate concentration and absolute amount of stored carbohydrates, were, after testing for the homogeneity of variances, analyzed by means of a three-way mixed-model ANOVA with nutrient and light conditions as the fixed effects and population and accession as random effects and accession nested within population (SAS PROC GLM). Data obtained at the final harvest were analyzed by means of a four-way mixed-model nested ANOVA (SAS PROC GLM). In this analysis the pretreatment conditions light and nutrient availability as well as submergence treatment were treated as fixed effects, population and accession were treated as random effects, with accession nested within populations. In this analysis, a significant interaction with submergence treatments indicates that pretreatment or population induced differences in plasticity, that is, plasticity of plasticity.

We performed regression analyses to test the hypotheses that carbohydrates accumulated before submergence determine plastic responses to subsequent submergence. For these analyses, plasticity was calculated as the percentage difference of mean petiole length of a given accession and pretreatment combination in the submerged relative to control conditions. Biomass, carbohydrate concentration and total carbohydrate content of roots and shoots were used to represent the amount of carbohydrates accumulated before the submergence treatments. The analyses were done on the accession mean values and all data were standardized to the mean before the regression analyses. All statistical analyses were done with the program package SAS. For the mixed-model ANOVAs the correct error terms were determined using the program PC EMS (Dallal, 1985).

Results

Biomass accumulation

Light and, to a lesser extent, nutrient availability significantly affected biomass accumulation at early developmental stages (Fig. 1a, Table 1). Several of these pretreatment effects were maintained even after the plants were subjected to subsequent submergence treatments (Fig. 1, Table 2). Plants subjected to high light conditions accumulated significantly more root and shoot biomass before the start of the submergence treatment (Fig. 1a, Table 1), and were still larger after being subjected to submergence (Fig. 1c, Table 2). While high nutrient availability did not affect initial biomass accumulation, it did result in increased investment into leaf laminas at the final harvest (Fig. 1, Tables 1, 2).

Figure 1.

Mean (± 1 SE) biomass of the different plant parts of Rumex palustris before the onset of the submergence treatments (a); and biomass of drained (b) and submerged (c) plants at the final harvest. Dark gray bars, roots; light gray bars, petioles; white bars, leaf laminas. Please note that, for the initial harvest, leaves had not been separated into petioles and leaf lamina, because for most leaves the short petioles and laminas could not be clearly distinguished. Pretreatments: HL, high light; LL, low light; HN, high nutrients; LN, low nutrients. For significance of the differences, see Tables 1, 2.

Table 1.   Initial values measured before the onset of the submergence treatments
 dfLength of fourth leafLength of fifth leafRoot biomassShoot BiomassCarbohydrate concentration (%): rootCarbohydrate content: rootCarbohydrate concentration (%): shootCarbohydrate content: shoot
  1. Result of a three-way mixed-model nested ANOVA testing for the effects of light and nutrient pretreatments (F-values and their significance) on growth and biomass accumulation of Rumex palustris as well as the relative concentration and absolute content of nonstructural carbohydrates at the time when submergence started. Leaves were counted in the sequence of appearance, and thus lower numbers refer to developmentally older leaves and higher numbers to developmentally younger leaves. The first value in the df column indicates the nominator, and the second value indicates the denumerator df. The levels of significance are as follows: ***,  0.001; **,  0.01; *,  0.05; $,  0.1. For the effects of accessions (populations) see Table S3.

Light1, 15151.9 ***53.6 ***14.7 **30.0 ***0.47.0 *37.8 ***50.3 ***
Nutrients1, 1517.8 ***1.40.21.812.4**6.4 *0.40.7
Light × nutrients1, 151.41.90.60.91.00.80.10.1
Population2, 150.70.51.50.14.7 *2.9$13.7 ***3.4 $
Light × population2, 151.50.21.40.01.72.7$0.20.2
Nutrients × population2, 150.81.90.81.53.8*1.14.7 *3.4 $
Light × nutrients × population2, 150.61.81.20.92.10.61.01.1
Table 2.   Values measured at the final harvest of Rumex palustris plants
 dfLength of fourth leafLength of fifth leafLength of sixth leafShoot BiomassPetiole biomassLamina biomassRoot biomassTotal Biomass
  1. Results of a four-way mixed-model nested ANOVA testing for the effects of light and nutrient pretreatments and submergence (F-values and their significance) on leaf length and biomass accumulation. Leaves were counted in the sequence of appearance, and thus lower numbers refer to developmentally older leaves and higher numbers to developmentally younger leaves. The first value in the df column indicates the nominator, and the second value indicates the denumerator df. The levels of significance are as follows: ***,  0.001; **,  0.01; *,  0.05; $,  0.1. For the effects of accessions (populations) see Table S4.

Light1, 1536.4 ***0.245.0 ***1.42.29.3 ***25.2 ***16.8 **
Nutrients1, 153.2$6.7 *0.11.60.414.5 **2.915.6 **
Light × nutrients1, 150.60.11.400.10.43.7$2.42.3
Submergence1, 15429.6 ***471.0 ***377.2 ***53.0 ***26.6 ***485.3 ***779.2 ***655.0 ***
Light × submergence1, 150.01.325.2 ***0.221.0 ***0.41.90.0
Nutrients × submergence1, 153.8$0.12.66.4 *0.34.9 *16.1 **13.3 **
Light × nutrients × submergence1, 151.40.01.50.10.10.02.20.2
Population2, 160.10.11.66.5 ***1.88.6 **2.04.3 *
Light × population2, 153.5$2.9$0.20.50.40.80.20.2
Nutrients × population2, 150.90.10.00.13.3$2.20.92.6
Light × nutrients × population2, 153.8 *1.52.50.10.01.51.01.2
Submergence × population2, 151.40.24.3 *4.5 *5.0 *7.3 **10.1 **10.3 **
Light × submergence × population2, 152.03.3$1.80.41.80.42.71.6
Nutrients × submergence × population2, 150.90.40.51.11.40.50.20.3
Light × nutrients × submergence × population2, 151.81.90.70.20.40.00.20.1

Submergence decreased total biomass by, on average, 47%, while biomass of drained plants increased by c. 50% (Fig. 1). Drained plants grew more when plants had high amounts of nutrients available before the onset of submergence, but this effect did not occur in submerged plants, resulting in a significant submergence × nutrient availability interaction (Fig. 1b,c, Table 2). In submerged but not in drained plants, high light availability before the onset of the submergence treatments increased total petiole weight (Table 2).

Populations differed in lamina biomass, as well as in the response of biomass accumulation in response to submergence (Fig. 1b,c, Tables 1, 2). Plants originating from Population 5 produced, on average, the highest biomass when grown under drained conditions, but populations did not differ in biomass accumulation under submergence. Accessions within populations differed significantly with respect to early and late biomass accumulation (Tables S3, S4). The biomass accumulation in response to early light availability differed significantly among accessions, as did biomass accumulation in response to submergence (Tables S3, S4).

Leaf length plasticity

Plants subjected to either high light or low nutrient availability produced significantly shorter leaves (Fig. 2a,d, Table 1). The effects of early light and nutrient availability on leaf length remained significant throughout the experiment in the developmentally older leaves (Fig. 2b,c, Table 2), indicating that the length that leaves achieved early in their development significantly affects their final length, even if environmental conditions change throughout development. For example, the developmentally older, fourth leaf of plants subjected to high-light conditions in early developmental stages remained significantly shorter after flooding, compared with similar leaves subjected to low light. This effect of pretreatment, however, decreased (Fig. 2e,f) or was even reversed in developmentally younger leaves, which were still in their earliest developmental stages at the onset of the submergence treatment (Fig. 2h,i, Table 2). The effect of high-light preconditions on the length of the fifth leaf did not persist after flooding, and the developmentally youngest sixth leaf showed the opposite preconditioning effect than the developmentally older leaves, as this leaf became significantly longer when pregrown under high light (Fig. 2d–i, Table 2).

Figure 2.

Rumex palustris mean (± 1SE) leaf length before the onset of submergence (initial length) and at final harvest (drained and submerged plants). The lengths of the total leaves as well as of its components are given (dark gray bars, petioles; light gray bars, laminas). Pretreatments: HL, high light; LL, low light; HN, high nutrients; LN, low nutrients. Leaves were counted in the sequence of appearance, and thus lower numbers refer to developmentally older leaves and higher numbers to developmentally younger leaves. Note that at the initial harvest petioles were not present in some of the developmentally younger (= fifth) leaves, therefore only total leaf length is given. Also at this time the developmentally youngest sixth leaves were not yet present. For significance of the differences, see Tables 1, 2. Pop, population.

Plants subjected to high nutrient availability at early developmental stages had, on average, slightly (but significantly) longer leaves at the end of the experiment. This positive effect of nutrient availability on leaf length remained significant up to the fifth, but not for the sixth leaf (Fig. 2, Table 2).

The three populations differed significantly in the degree of plasticity of the sixth leaf expressed upon submergence (Fig. 2h,i, Table 2). Depending on the light conditions experienced during early developmental stages, the ranking of the populations with respect to their submergence-induced plasticity changed to some extent, as indicated by the marginally significant population ×light interaction (Table 2). For example, plants from Population 4 which had been exposed to high-light conditions had, on average, the most plastic leaves, whereas leaf plasticity of plants from Population 4 that had been pretreated with low-light conditions was intermediary to the plasticity of Populations 5 and 1 (Fig. 2). In addition to population differentiation, leaf length differed significantly among accessions (Tables S3, S4). Accessions within populations also responded significantly differently to early light availability and submergence, indicating that there is much genetic variation in submergence-induced responses as well as in the effects caused by conditions experienced before submergence (Table S4). It also indicates that the ranking of the magnitude of plasticity among accessions differs depending on the environmental conditions experienced previously.

Nonstructural carbohydrates (TNC)

Light and nutrient availability significantly affected TNC concentration and content (Fig. 3, Table 1). While light had a stronger effect on carbohydrate status of the shoots (Fig. 4a, Table 1), nutrient availability more strongly affected carbohydrate status of the roots (Fig. 3b, Table 1). Plants subjected to high light availability had significantly higher shoot carbohydrate concentrations before the onset of submergence than plants pregrown under low light (Fig. 3, Table 1). The positive effect of carbohydrate concentration was reinforced by the positive effect of light availability on biomass accumulation, resulting in a 50% increase in total carbohydrate content of plants pregrown in high vs low light. Being subjected to high nutrient availability, on the other hand, resulted in a significantly lower total nonstructural carbohydrate (TNC) content. However, this effect depended on the population and was mainly visible in Populations 4 and 5, but not in Population 1 (Fig. 3b). Accessions differed significantly in the TNC concentration in roots and shoots, as well as in the total carbohydrate content in these two organs (Table S3). Carbohydrate content and concentration in roots and shoots also responded differently to light availability among accessions (Table S3).

Figure 3.

Mean (± 1 SE) concentration of soluble carbohydrates in Rumex palustris shoots (a) and roots (b), and total amount of soluble carbohydrates in roots (dark gray bars) and shoots (light gray bars) (c) at the onset of submergence. Pretreatments: HL, high light; LL, low light; HN, high nutrients; LN, low nutrients. For significance of the differences, see Table 1.

Figure 4.

Effect of biomass, carbohydrate concentration, and absolute carbohydrate content at the onset of submergence treatments on plastic increase in Rumex palustris of the sixth leaf in submerged vs drained conditions after 14 d. The values in the graphs show the standardized regression coefficients as well as the level of significance of the correlation between biomass or carbohydrates and leaf length plasticity (***, P ≤ 0.001; ns, not significant). Each point indicates the mean of an accession under a given set of pretreatment conditions. Different symbols indicate different light and nutrient pretreatments; closed triangle, low light, low nutrients; squares, low light, high nutrients; circles, high light, low nutrients; open triangles, high light, high nutrients.

Effects of carbohydrates on leaf length plasticity

One of the aims of the experiment was to test whether differences in size or carbohydrate storage before submergence explain variation in submergence-induced leaf elongation. In order to test this, we looked at three different measures of growth and storage, that is, biomass increment, carbohydrate concentration, and total carbohydrate content. We found that biomass accumulation before submergence did not affect submergence-induced plasticity in our experiment (Fig. 4a,b, Table 3), indicating that neither root nor shoot weight can be used as a good predictor for the potential of leaves to elongate in response to submergence. However, the carbohydrate concentration in roots and shoots had a significant effect on submergence-induced plasticity (Fig. 4, Table 3). While this result was qualitatively similar for the different leaf developmental stages, the relationship was strongest for the sixth leaf (Fig. 4, Table 3). The same pattern was apparent for total carbohydrate content (Fig. 3, Table 3). Carbohydrates in roots and shoots had a comparable effect on submergence-induced plasticity (Fig. 4, Table 3), even though the carbohydrate concentration was twice as high in shoots as in roots and the absolute carbohydrate content in shoots was > 400% higher in shoots than in roots (Figs 3, 4, Table 3). Surprisingly, despite the lower carbohydrate content and concentration in roots than in shoots, the carbohydrates stored in roots showed a consistently higher correlation with submergence-induced leaf elongation.

Table 3.   Effect of biomass, carbohydrate concentration and absolute carbohydrate content at the onset of submergence treatments on plastic changes of Rumex palustris leaf length in submerged vs drained conditions after 14 d
 WeightCarbohydrate concentrationAbsolute carbohydrate content
RootShootRootShootRootShoot
  1. The values show the standardized regression coefficients with the level of significance and the r2 of the correlation between biomass or carbohydrates and leaf length plasticity. The relationships for the sixth leaf are shown graphically in Fig. 4. The levels of significance are as follows: ***,  0.001; **,  0.01; *,  0.05; $,  0.1.

4th leaf0.026
r2 = 0.00
−0.208$
r2 = 0.04
0.447 ***
r2 = 0.20
0.344 **
r2 = 0.12
0.389 ***
r2 = 0.15
0.249 *
r2 = 0.06
5th leaf0.250 *
r2 = 0.06
0.133
r2 = 0.02
0.299 *
r2 = 0.09
0.231$
r2 = 0.05
0.345 **
r2 = 0.12
0.261 *
r2 = 0.07
6th leaf0.145
r2 = 0.02
−0.178
r2 = 0.03
0.664 ***
r2 = 0.44
0.540 ***
r2 = 0.29
0.680 ***
r2 = 0.46
0.526 ***
r2 = 0.28

Discussion

Our results show that the phenotype expressed in response to a given cue is not fixed, but depends on the conditions experienced earlier in development. Accumulation of carbohydrates during the pretreatments strongly affected the magnitude of leaf elongation to subsequent flooding. However, depending on the developmental stage of the leaves, submergence-induced leaf elongation was either constrained or stimulated by earlier high-light conditions. Moreover, responses further appeared population-specific, with the relative degree of plasticity depending on the conditions that plants experienced early in life.

Early growth conditions constrain or facilitate phenotypic expression

We found that conditions experienced at early developmental stages affected total leaf length and leaf length plasticity expressed in response to a subsequent environmental trigger. The responses, however, depended on the stage of the leaves when they were subjected to early environmental conditions. Consequently, the effects of early light conditions on subsequent submergence-induced plasticity were opposite for developmentally older and younger leaves. Older leaves responded to shading preconditions by elongation, after which their potential for further elongation under submergence was limited and differences in length induced by the light pretreatments persisted. Leaves that were still very young in the pretreatment phase had hardly elongated under shade and thus had the largest elongation potential under subsequent submergence, and especially so when they had been exposed to high nutrient availability. These results indicate that effects of preconditions on plasticity are complex and depend on an interplay between resource and developmental status.

It is well known that exposure to simultaneously or subsequently acting environmental cues can limit an organism’s ability to respond to each individual cue as a result of contrasting effects of cues (Anten et al., 2009), developmental constraints such as the number of cells or the maximum length a cell can achieve (Tsukaya, 2003; Tsukaya & Beemster, 2006; Huber et al., 2008; Weijschede et al., 2008a), or the sequence in which cues are perceived (Weinig, 2000; Weinig & Delph, 2001). In an extreme case, plant organs may be initiated long before they visibly emerge and develop into their final size. Such preformation, which frequently occurs in temperate trees or alpine herbs, strongly limits the response of these organs to a subsequent alteration of their immediate environment (Geber et al., 1997; Diggle, 2002). Interaction of different cues may also increase the phenotypic response, if, for example, plants produce more cells in response to one cue, which can in turn lead to a faster or greater response to another subsequent cue. Alternatively, if in early conditions cell expansion is limited, there is a relatively large potential for subsequent increased cell expansion in response to subsequent cues. Our results indicate that both constrained and enhanced plasticity resulting from preconditions may occur in the same plant individual, depending on the developmental stage of the elongating leaf.

Positive effects of early conditions on plastic elongation are strongly correlated with carbohydrate accumulation

In our experiment root and shoot carbon concentration and content, rather than root and shoot weights, were positively correlated with subsequent elongation in response to submergence. Submergence-induced plasticity depends on the developmental stage of leaves and the availability of carbohydrates (Groeneveld & Voesenek, 2003). In contrast to the artificially large reduction of stored carbohydrates in the experimental setup by Groeneveld & Voesenek (2003), we subjected the plants to shading and nutrient limitation that might better resemble the natural range of environmental conditions. We found that even these milder manipulations of carbohydrate storage can affect submergence-induced plasticity. Plasticity of developmentally younger leaves was increased if plants were subjected to high-light conditions before submergence, possibly because they built up higher amounts of carbohydrates.

Positive effects of exposure to early conditions on responses to conditions experienced later have been described before and have, in the plant–animal interaction literature, been termed ‘priming’ (Bruce et al., 2007). Being subjected to airborne signals emitted by other individuals of the same species or even different species can stimulate optimal defenses to later herbivore or pathogen attacks (Yi et al., 2009; Heil & Karban, 2010). Other environmental cues, too, such as salt or drought conditions, experienced at early developmental stages have been shown to induce stronger responses to later biotic or abiotic cues (Bruce et al., 2007). The environmental agents eliciting priming may be related to but also very different from the subsequent stressors. The actual mechanisms leading to priming are largely unknown, but changes of the internal plant hormonal status, activation of dormant signaling proteins, chromatin modifications or other epigenetic changes have been hypothesized as potential mechanisms (Bruce et al., 2007; Conrath, 2011).

Our results clearly point to the internal resource status, as affected by the previous environmental conditions, as a mechanism modulating the elongation response to subsequent environmental cues. Previous research has shown that plastic elongation is associated with costs in terms of increased biomass invested into the elongating structures (Huber et al., 1998; Weijschede et al., 2006; Chen et al., 2009). In this study we also found that the relative biomass allocation to the elongating leaves was higher in submerged than in control plants. Under natural conditions these costs associated with elongation can be compensated by the increased resource uptake after elongation has taken place (Huber & Wiggerman, 1997; Weijschede et al., 2006, 2008b; Vermeulen et al., 2008; Chen et al., 2011). Sugars accumulated in early life serve as a source of energy to drive energy-demanding processes during enhanced elongation growth under conditions where energy supply is marginal as a result of reduced rates of photosynthesis. Sugars can also serve to build new cell wall material which is needed to sustain cell growth. This shows that carbohydrates accumulated by plants before submergence, and, in the case of R. palustris, particularly those stored in the root system, are of pivotal importance for plants to adequately respond to submergence and to regain contact with the air and increase light and oxygen uptake. Depending on the temporal pattern of submergence incidence, even relatively small differences in elongation capacity can lead to variation in performance and future evolutionary responses. Our results indicate that selection of different elongation responses can occur via evolutionary changes of the response of carbohydrate accumulation in response to the immediate environment.

Effects of early conditions differ among populations

It is well known that populations can differ in the expression of phenotypic plasticity (Gianoli & Gonzalez-Teuber, 2005; Bell & Galloway, 2008; Anten et al., 2009; Chen et al., 2009). Our study adds another level, as it also shows that the effect of cues experienced at early developmental stages on plastic responses to cues experienced at later developmental stages differs among populations. As a result, the relative ranking of the magnitude of plasticity for different populations may change depending on the preconditions. The high among- and within-population variation in internal resource status, which in turn affects the expression of plasticity, provides a huge potential for selection to act upon. Internal resource status, evolutionary history, and cues perceived at different developmental stages thus interact in determining plastic responses to the environment. It can be argued that variation in the magnitude of plasticity in different environmental settings and the change of the relative ranking in the magnitude of plasticity among populations may impede the evolution of consistent plastic responses. These results indicate that selection on optimal response patterns may differ among populations depending on the prevalent environmental conditions. Alternatively, variability in plasticity may also be a consequence of either weak or inconsistent selection on the effects of early environmental cues on subsequent responses.

Conclusions

Our results clearly show that plasticity cannot be viewed as a stable response that is always expressed to the same extent in response to a given cue (Sultan, 2000), but that the magnitude and direction of plasticity are plastic responses themselves. Such plasticity of plasticity has recently been placed in the context of metaplasticity (Novoplansky, 2009). Our results have consequences for experimental approaches used to study plasticity and the evolutionary consequences of different degrees of plasticity in terms of selection gradients. Such studies are typically done under well-controlled common garden conditions and the results of these experiments are often extrapolated to processes taking place under field conditions. However, as our results show, the specific response may depend on the conditions perceived in early ontogenetic stages. Under glasshouse conditions, and even more so under field conditions, the environment is not constant over time and the expression of plasticity may therefore vary among years and experiments, which may explain inconsistent results across experiments (Sultan, 2000). We do not yet know whether, in addition to the magnitude and direction of plasticity, the effect of a given degree of plasticity on plant performance is also affected by the conditions experienced early in development. Studies comparing plants under controlled-environment and field conditions are needed to show which responses are stable under a wide range of (pre)conditions and which are more variable and subject to environmental modulation.

Acknowledgements

We are grateful to Harry van de Steeg for collecting the seeds. We thank Nils van Rooijen, Sanne van Delft and the glasshouse staff of Nijmegen University for help with the experiment, and Yvonne de Jong-van Berkel for help with the carbohydrate analyses. We are grateful for the comments of Laura Galloway and three anonymous referees. This is part of a project from the Centre for Wetland Ecology, a partnership of The Netherlands Institute of Ecology, Radboud University Nijmegen, Utrecht University and the University of Amsterdam.

Ancillary

Advertisement