Can differences in phosphorus uptake kinetics explain the distribution of cattail and sawgrass in the Florida Everglades?

BACKGROUND: Cattail (Typha domingensis) has been spreading in phosphorus (P) enriched areas of the oligotrophic Florida Everglades at the expense of sawgrass (Cladium mariscus spp. jamaicense). Abundant evidence in the literature explains how the opportunistic features of Typha might lead to a complete dominance in P-enriched areas. Less clear is how Typha can grow and acquire P at extremely low P levels, which prevail in the unimpacted areas of the Everglades.
RESULTS: Apparent P uptake kinetics were measured for intact plants of Cladium and Typha acclimated to low and high P at two levels of oxygen in hydroponic culture. The saturated rate of P uptake was higher in Typha than in Cladium and higher in low-P acclimated plants than in high-P acclimated plants. The affinity for P uptake was two-fold higher in Typha than in Cladium, and two- to three-fold higher for low-P acclimated plants compared to high-P acclimated plants. As Cladium had a greater proportion of its biomass allocated to roots, the overall uptake capacity of the two species at high P did not differ. At low P availability, Typha increased biomass allocation to roots more than Cladium. Both species also adjusted their P uptake kinetics, but Typha more so than Cladium. The adjustment of the P uptake system and increased biomass allocation to roots resulted in a five-fold higher uptake per plant for Cladium and a ten-fold higher uptake for Typha.
CONCLUSIONS: Both Cladium and Typha adjust P uptake kinetics in relation to plant demand when P availability is high. When P concentrations are low, however, Typha adjusts P uptake kinetics and also increases allocation to roots more so than Cladium, thereby improving both efficiency and capacity of P uptake. Cladium has less need to adjust P uptake kinetics because it is already efficient at acquiring P from peat soils (e.g., through secretion of phosphatases, symbiosis with arbuscular mycorrhizal fungi, nutrient conservation growth traits). Thus, although Cladium and Typha have qualitatively similar strategies to improve P-uptake efficiency and capacity under low P-conditions, Typha shows a quantitatively greater response, possibly due to a lesser expression of these mechanisms than Cladium. This difference between the two species helps to explain why an opportunistic species such as Typha is able to grow side by side with Cladium in the P-deficient Everglades.

Background

The wetland species, Cladium mariscus ssp. pan> class="Species">jamaicense (L.) Pohl (Crantz) K. Kenth (sawgrass; hereafter Cladium) and Typha domingensis Pers. (cattail; hereafter Typha) are both native to the Florida Everglades and occupy similar habitats [1]. Cladium was the dominant plant species in the historical freshwater Everglades, whereas Typha was a minor species occurring in small and scattered patches throughout the Everglades [2]. However, during the past decades Typha has expanded rapidly and replaced thousands of hectares of Cladium marshes and aquatic slough areas in the northern part of the Everglades [3-6]. Numerous studies have been conducted to assess the causes and the consequences of this change in vegetation and community structure [7-20], and the driving force for the change appears to be nutrient enrichment, particularly phosphorus (P), from agricultural runoff and Lake Okeechobee outflow [21].

Cladium and Typha are both large, clonal species that can form monospecific communities in freshwater habitats. The two species differ, however, in morphology, growth, anpan>d life history characteristics [10,15,22]. Cladium exhibits manpan>y characteristics of adaptation to infertile enpan>vironmenpan>ts, such as slow growth rate, long leaf longevity, low capacity for nutrienpan>t uptake, low leaf nutrienpan>t concenpan>trations anpan>d a relatively inflexible partitioning of biomass in response to increased nutrienpan>t availability [23,24]. Typha, on the other hanpan>d, has traits of anpan> opportunpan>istic species from nutrienpan>t-rich habitats with high growth rates, short leaf longevity, high capacity for nutrienpan>t uptake, high leaf nutrienpan>t concenpan>trations anpan>d flexible biomass partitioning [8,25]. Both species are adapted to grow in waterlogged soils by virtue of a well-developed aerenchyma system, but convective gas flow has been documented only in Typha and not in Cladium [26-29]. Furthermore, Cladium has lower root porosity and generally higher alcoholic fermentation rates, indicating lower capacity for root aeration than Typha [30]. These inherently different traits are considered the main explanation for the rapid spread and competitive success of Typha in the P-enriched areas of the Florida Everglades.

Cladium and Typha also co-exist in the oligotrophic areas of the Florida Everglades where P availability is extremely low. In the interior of Water Conservation Area 2A, anpan> impounpan>ded area in the northern Everglades, Cladium anpan>d Typha grow together despite soluble P concenpan>trations of less thanpan> 4 μg l-1 in the porewaters throughout the soil profile [31]. Typha is much less abundant than Cladium and has slow growth rates in these areas [32], and although nutrient enrichment and disturbance around alligator holes have been suggested to favour the proliferation of Typha locally [33], the traits that allow the growth of a high resource-adapted plant like Typha in this low P environment are not understood.

Studies at high P availability have demonstrated that Typha has a greater relative growth rate, a greater allocation of biomass to leaves, and a lower P-use efficiency than Cladium [10,15,16]. In fertile habitats, a high nutrient uptake capacity per unit of root biomass and a high growth rate and biomass allocation to leaves increase the capability to compete for light and reduce the need for a high root biomass. However, these traits are not advantageous for growth in a nutrient deficient environment where plants must acquire nutrients at low availability and minimize nutrient losses [34]. In such conditions, optimal features would include an extensive root system for soil exploration, a high root surface area (long, thin roots and/or root hairs) for acquisition of nutrients, and efficient mechanisms to capture nutrient ions at low external concentrations [35-37].

The main research question we address here is: Which characteristics of Cladium and Typha allow the species to grow in the n class="Disease">oligotrophic P-deficient inpan>terior of the Florida Everglades, and at the same time explainpan> why Typha out-competes Cladium unpan>der P-enpan>riched conpan>ditionpan>s? As to the seconpan>d part of the questionpan>, abunpan>dant evidenpan>ce inpan> the literature explainpan>s how the opportunpan>istic features of Typha can lead to complete dominpan>ance inpan> P-enpan>riched areas [e.g. [7,8,13,15]]. Less clear is how Typha can grow and acquire P at the extreme low P-levels prevailinpan>g inpan> the unpan>impacted areas of the Everglades.

We hypothesized that Typha has a more plastic P uptake system than Cladium in relation to P availability, and this strategy will allow adequate uptake of P and better competitive ability over a wide range of external P concentrations. Furthermore, we hypothesized that n class="Chemical">oxygen-deficienpan>t conditions will affect the uptake kinetics of Cladium more thanpan> that of Typha, as the latter species has a more efficienpan>t system for root aeration (via aerenpan>chyma anpan>d internal convective gas flow). These hypotheses were tested in a series of P uptake experimenpan>ts designed to distinguish differenpan>ces in P uptake kinetics betweenpan> the two species.

Apparent P uptake kinetics were measured for whole plants of Cladium and Typha grown from seeds and acclimated to identical, steady state conditions, in a factorial treatment arrangement with two levels of P (5 and 500 μg P l-1) and two levels of n class="Chemical">oxygen (8.0 anpan>d <0.5 mg pan> class="Chemical">O2 l-1) in hydroponic culture solutions (n = 4-8). The P uptake kinetic parameters were estimated using a modified Michaelis-Menten model [38-41].

Results

Plant characteristics

The plants used in the uptake studies all appeared healthy, with no visual signs of nutrient deficiencies. However, long acclimation periods in the various culture combinations (up to 4 months in the low P treatments) inevitably created plants with different biomass, allocation patterns, and tissue nutrient concentrations. Overall the shoot height (average 0.95 m) and root length (average 0.37 m) of the two species varied little across the treatments, but the plant weights and biomass allocation differed significantly (table 1). Low-P acclimated plants had less biomass than high-P acclimated plants, and significantly (P < 0.001) more biomass allocated to roots. Typha in particular allocated less biomass to roots in the high P treatment (table 1). Plants in the low n class="Chemical">oxygen cultures had 25% shorter root systems thanpan> those in aerated cultures (P < 0.001), but the pan> class="Chemical">oxygen treatment did not affect the proportion of biomass allocated to roots (P > 0.05).

Table 1

Plant size

Species	Oxygen	P	Leaf length(m)	Root length(m)	Plant weight(g DW)	Root fraction(%)
Cladium	High oxygen	Low	1.01 ± 0.05	0.30 ± 0.03	5.6 ± 0.7	17.4 ± 1.9
		High	0.98 ± 0.09	0.50 ± 0.10	7.0 ± 1.9	16.1 ± 1.6
	Low oxygen	Low	1.00 ± 0.12	0.32 ± 0.04	7.0 ± 1.9	18.3 ± 2.2
		High	0.95 ± 0.08	0.30 ± 0.06	12.8 ± 2.9	14.1 ± 2.2
Typha	High oxygen	Low	0.69 ± 0.12	0.48 ± 0.05	5.8 ± 0.5	22.3 ± 1.9
		High	0.96 ± 0.05	0.42 ± 0.06	11.6 ± 1.4	8.7 ± 1.6
	Low oxygen	Low	0.96 ± 0.06	0.39 ± 0.03	9.1 ± 1.3	15.1 ± 2.0
		High	1.08 ± 0.09	0.29 ± 0.04	23.6 ± 4.3	7.7 ± 0.6

Average (± SE, n = 4-8) leaf and root length, total plant weight and percentage of biomass allocated to roots of the Cladium mariscus spp. jamaicense and Typha domingensis plants acclimated to a low and a high P level (5 and 500 μg P l-1) and to aerated and low oxygen conditions (8.0 and <0.5 mg O2 l-1) in hydroponic culture solutions

Average (± SE, n = 4-8) leaf and root length, total plant weight and percentage of biomass allocated to roots of the n class="Species">Cladium mariscus spp. n class="Species">jamaicense and Typha domingensis plants acclimated to a low and a high P level (5 and 500 μg P l-1) and to aerated and low oxygen conditions (8.0 and <0.5 mg O2 l-1) in hydroponic culture solutions

The tissue N concentrations in Cladium were similar (average 10.4 mg g-1 dry weight) in all treatments, and lower overall than in Typha (approximately 17 mg g-1 dry weight), except in the aerated, low-P treatment where plants were smaller (figure 1). The tissue P concentrations were, as expected, more variable across treatments, and were generally higher (P < 0.001) in the high P treatment (average 3.1 mg g-1 dry weight) than in the low P treatment (average 1.1 mg g-1 dry weight). Moreover, the tissue P concentrations were consistently higher in the low n class="Chemical">oxygen treatmenpan>ts thanpan> in the corresponding aerated treatmenpan>ts (figure 1). However, tissue P anpan>alyses were carried out on tissues harvested after the uptake kinetic studies, in which planpan>ts were exposed to high P concenpan>trations (up to 500 μg l-1) for several hours. Uptake during the uptake studies may have contributed to enpan>hanpan>ce the tissue concenpan>tration by 0.4-0.7 mg P g-1 dry weight. Henpan>ce, the tissue P concenpan>trations in the low-P acclimated planpan>ts presenpan>ted here are likely higher thanpan> they would have beenpan> if planpan>ts were collected for anpan>alysis before the uptake kinetic studies.

Figure 1

Plant tissue N and P concentrations. Average (± SE) concentrations of nitrogen (N) and phosphorus (P) (mg g-1 dry weight) in whole plants of Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low (5 μg l-1) and high (500 μg l-1) P concentrations and high (+O2: 8 mg l-1) and low (-O2: <0.5 mg l-1) oxygen concentration in the culture solutions.

Plant tissue N and P concentrations. Average (± SE) concentrations of n class="Chemical">nitrogen (N) and n class="Chemical">phosphorus (P) (mg g-1 dry weight) in whole plants of Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low (5 μg l-1) and high (500 μg l-1) P concentrations and high (+O2: 8 mg l-1) and low (-O2: <0.5 mg l-1) oxygen concentration in the culture solutions.

Phosphorus uptake kinetics

The relationships between solution P concenpan>trations anpan>d the net rate of P uptake of selected Cladium anpan>d Typha planpan>ts acclimated to low P anpan>d to high P levels are shownpan> in figure 2. Genpan>erally, the modified Michaelis-Menpan>tenpan> model fitted the relationships anpan>d estimated the kinetic uptake parameters with high statistical confidenpan>ce. Vmax was enpan>tered as a fixed parameter in the model anpan>d was estimated as the average uptake rate at pan> class="Chemical">solution P concentrations where uptake appeared to be saturated. At the low concentration range, where P uptake increases nearly linearly with solution P concentration, the model fitted the data very well, but the uncertainty associated with estimating Cmin for high-P acclimated plants was relatively large. The Cmin values presented for these conditions are therefore uncertain and should be interpreted accordingly.

Figure 2

P uptake kinetic curves. Examples of relationships between the concentration of inorganic phosphorus (P) in the root chambers and the rate of P uptake of single specimens of (A) Cladium mariscus spp. jamaicense and (B) Typha domingensis acclimated to low (5 μg l-1) P concentrations (solid lines and closed symbols) and to high (500 μg l-1) P concentrations (dashed lines and open symbols). Plants were grown at low (<0.5 mg l-1) oxygen concentration in culture solutions. Uptake was measured by P depletions of root solutions followed by stepwise increments of the P concentrations. The relationship was fitted to a modified Michaelis-Menten model (curves).

P uptake kinetic curves. Examples of relationships between the concentration of inorganic phosphorus (P) in the root chambers anpan>d the rate of P uptake of single specimenpan>s of (A) pan> class="Species">Cladium mariscus spp. jamaicense and (B) Typha domingensis acclimated to low (5 μg l-1) P concentrations (solid lines and closed symbols) and to high (500 μg l-1) P concentrations (dashed lines and open symbols). Plants were grown at low (<0.5 mg l-1) oxygen concentration in culture solutions. Uptake was measured by P depletions of root solutions followed by stepwise increments of the P concentrations. The relationship was fitted to a modified Michaelis-Menten model (curves).

P uptake capacity

The saturated rate of P uptake (Vmax) differed significantly between species and was also affected by both P treatment and oxygen level, but the effects of pan> class="Chemical">oxygen differed between P treatments as shown by a significant interaction in the ANOVA (table 2). Across treatments the average Vmax was 38% higher in Typha than in Cladium and the Vmax of low-P acclimated plants was overall more than three-fold higher than the Vmax of high-P acclimated plants (figure 3a and table 2). Vmax was generally more than two-fold higher in the low oxygen treatment compared to the high oxygen treatment, except for Typha from the high P treatment, where rates were equal.

Table 2

Results of ANOVA for uptake kinetics

Source of variation	Vmax	K0.5	Cmin	Vmax/K0.5	α
A (species)	5.72*	17.38***	0.07	12.10**	23.05***
B (P level)	76.67***	10.19**	37.66***	56.85***	30.90***
C (O2 level)	25.67***	0.53	3.74	8.88**	18.98***
A × B	0.30	9.03**	1.87	0.03	0.01
A × C	0.75	2.62	10.35**	0.17	3.53
B × C	11.50**	0.60	12.03**	6.13*	12.14**

Results of analyses of variance (F(1,33)-ratios) for P uptake kinetic parameters of Cladium mariscus spp. jamaicense and Typha domingensis against species, P level (5 and 500 μg P l-1) and oxygen level (8.0 and <0.5 mg O2 l-1) in the hydroponic culture solution. Significant effects are in bold and indicated by *(P < 0.05), **(P < 0.01) and ***(P < 0.001).

Figure 3

P uptake kinetic parameters. Average (± SE) phosphorus (P) uptake kinetic parameters for Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low (5 μg l-1) and high (500 μg l-1) P concentrations and high (+O2: 8 mg l-1) and low (-O2: <0.5 mg l-1) oxygen concentration in the culture solutions. Uptake was measured by P depletions of root solutions followed by stepwise increments of the P concentrations. The kinetic parameters: (A) Vmax (maximum uptake at saturating P concentration), (B) K0.5 (half saturation constant), and (C) Cmin (P concentration where there is no net uptake) were estimated by fitting to a modified Michaelis-Menten model.

Results of analyses of variance (F(1,33)-ratios) for P uptake kinetic parameters of n class="Species">Cladium mariscus spp. n class="Species">jamaicense and Typha domingensis against species, P level (5 and 500 μg P l-1) and oxygen level (8.0 and <0.5 mg O2 l-1) in the hydroponic culture solution. Significant effects are in bold and indicated by *(P < 0.05), **(P < 0.01) and ***(P < 0.001).

P uptake kinetic parameters. Average (± SE) phosphorus (P) uptake kinetic parameters for pan> class="Species">Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low (5 μg l-1) and high (500 μg l-1) P concentrations and high (+O2: 8 mg l-1) and low (-O2: <0.5 mg l-1) oxygen concentration in the culture solutions. Uptake was measured by P depletions of root solutions followed by stepwise increments of the P concentrations. The kinetic parameters: (A) Vmax (maximum uptake at saturating P concentration), (B) K0.5 (half saturation constant), and (C) Cmin (P concentration where there is no net uptake) were estimated by fitting to a modified Michaelis-Menten model.

Half saturation constant

The half saturation constant (K0.5) differed significantly between the two species and was also affected by the P treatment, but the effects of P treatment differed between species as shown by the significant interactions in the ANOVA (table 2). Across treatments the half saturation constants were approximately 70% higher in Cladium than in Typha, indicating that Cladium overall has a lower affinity for P uptake than Typha (figure 3b). In Typha the half saturation constants did not differ much across treatments, but in Cladium the half saturation constants were 1.5-2.5 times higher in the high-P acclimated plants than in low P plants. n class="Chemical">Oxygen did not significanpan>tly affect K0.5.

Minimum P concentration

The n class="Chemical">solution P conpan>cenpan>trationpan> at which there was no net uptake, Cminpan>, was significantly affected by the treatmenpan>ts as shownpan> by the significant inpan>teractionpan>s inpan> the ANOVA (table 2). Onpan> average, low-P acclimated Cladium and Typha plants had a Cminpan> of 3.5 and 9.9 μg P l-1, whereas high-P acclimated plants had a Cminpan> of 43 and 25 μg P l-1, respectively (figure 3c). However, inpan> the low P-aerated treatmenpan>ts both species had a low Cminpan> (1.2 μg P l-1). For high-P acclimated plants Cminpan> was significantly higher (18-64 μg l-1) and the effects of n class="Chemical">oxygen differed between the species (figure 3c).

Affinity for P-uptake

The slope of the initial linear part of the uptake curve at low P solution concentrations, α, as well as the ratio between Vmax and K0.5, are measures of the uptake affinity. Both affinity measures were significantly affected by plant species, P acclimation and oxygen level, anpan>d effects of pan> class="Chemical">oxygen level differed between the P treatments (table 2). Overall the affinity for P uptake by Typha was two-fold higher than that of Cladium, and affinities were 2 to 3 times higher for low-P acclimated plants compared to high-P acclimated plants (table 3). Both species had higher P uptake affinities in low oxygen treatments, and the effects of oxygen were greatest for low-P acclimated plants. The highest affinity was found for low P and low oxygen acclimated Typha plants. Numerically, the affinities derived from the initial slope of the curves (α) were higher than the affinity measures derived from the Michaelis-Menten model (Vmax/K0.5), but the overall treatment effects were alike.

Table 3

Affinity for P uptake

Species	Oxygen	P	Vmax/K0.5(l g-1DW h-1)	α(l g-1DW h-1)
Cladium	High oxygen	Low	2.43 ± 0.40	0.75 ± 0.17
		High	0.40 ± 0.22	0.28 ± 0.20
	Low oxygen	Low	3.40 ± 0.27	1.44 ± 0.15
		High	0.86 ± 0.20	0.32 ± 0.07
Typha	High oxygen	Low	2.51 ± 0.35	0.84 ± 0.14
		High	1.63 ± 0.29	0.90 ± 0.25
	Low oxygen	Low	7.25 ± 1.77	4.08 ± 0.50
		High	1.17 ± 0.13	1.05 ± 0.34

Average (± SE, n = 4-8) ratio between Vmax and K0.5 as estimated by the modified Michaelis-menten model (Vmax/K0.5) and the slope of the initial part of the uptake curve (α) as a measure of the affinity for P uptake of Cladium mariscus spp. jamaicense and Typha domingensis plants acclimated to low and high P level (5 and 500 μg P l-1) and to aerated and low oxygen conditions (8.0 and <0.5 mg O2 l-1) in hydroponic culture solutions

Average (± SE, n = 4-8) ratio between Vmax and K0.5 as estimated by the modified Michaelis-menten model (Vmax/K0.5) and the slope of the initial part of the uptake curve (α) as a measure of the affinity for P uptake of n class="Species">Cladium mariscus spp. n class="Species">jamaicense and Typha domingensis plants acclimated to low and high P level (5 and 500 μg P l-1) and to aerated and low oxygen conditions (8.0 and <0.5 mg O2 l-1) in hydroponic culture solutions

Discussion

Ecophysiological studies on nutrient uptake kinetics must be conducted using hydroponically grown plants rather than in soil. Although it is possible to mimic the porewater composition of wetlanpan>d soils in terms of major nutrienpan>t ions anpan>d pH, the growth conditions in hydroponic cultures differ significanpan>tly from those of wetlanpan>d soils, particularly pan> class="Chemical">oxygen and redox conditions [42-44]. In wetland soils, porewaters are nearly always oxygen-free and may contain variable concentrations of reduced ions and organic compounds resulting in low redox potentials depending on soil organic content, nutrient status and other factors. In hydroponic plant culture, the solution is usually aerated to ensure a good oxygen supply for roots. However, in wetland soils, the root oxygen is delivered from the atmosphere via internal transport through the aerenchyma, and so oxygen supply to support aerobic metabolism within root cells potentially differs considerably [27]. Our experimental conditions mimicked the low oxygen conditions in wetland soils by flushing culture solutions with gaseous N2. This treatment maintained oxygen in culture solutions at levels less than 0.5 mg l-1 and so provided a largely anoxic, but not highly reducing, root environment. The growth of the plants was little affected by the oxygen treatments, except for Typha root length, which was shorter in the low oxygen treatments.

Except for Typha at high P, tissue P concentrations were higher in the low oxygen treatmenpan>t (figure 1), anpan>d uptake kinetics were also significanpan>tly affected by pan> class="Chemical">oxygen (figure 4). In waterlogged, anoxic soils, plants have been reported to produce less root biomass with a greater P uptake per unit of root mass than in drained soils [45]. This response has been ascribed to a combination of increased P availability in waterlogged soils and a higher physiological capacity of roots to absorb P, although the mechanisms are not known [45-47]. In the present study, oxygen treatment likely did not affect P availability in the solution cultures. Also, root morphology did not change in response to the oxygen treatments. We, therefore, suggest that the observed differences in uptake characteristics were related to the effects of oxygen on the function and/or expression of the high affinity ion transporters in the root plasma membranes. Plants acquire P in the form of phosphate anions, mostly H2PO4 -, from the soil solution, and uptake occurs against a steep concentration gradient (three orders of magnitude or greater) across the plasmamembrane [34,48]. Many uptake models have been proposed to explain the ability of plants to acquire P both under deficiency and sufficiency conditions [34,35,48,49]. A dual uptake model involving both a high affinity uptake system (HATS) operating at low (<200 μM) concentrations, and a low affinity uptake system (LATS) operating at high (>1 mM) concentrations is widely used to explain the concentration-dependent uptake of nutrient ions [48]. In P-deficient soils only the HATS is operating, apparently with multiple phosphate transporters located in the plasmamembrane through which the energy-mediated uptake occurs [49]. In general, plants respond to nutrient deprivation by increasing uptake affinity as well as uptake capacity [50]. The lack of response to oxygen of high-P acclimated plants, however, indicates that the high affinity transporters are regulated more by P demand than oxygen.

Figure 4

P uptake curves. Average estimated uptake of phosphorus (μg P g-1 root dry weight h-1) as a function of solution inorganic P concentration for Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low P (5 μg l-1) and high P (500 μg l-1) concentrations and high oxygen (8 mg l-1; dashed lines) and low oxygen (<0.5 mg l-1; solid lines) concentration in the culture solutions. The curves were generated by the modified Michaelis-Menten model using the average kinetic parameters for the two species in the different treatments.

P uptake curves. Average estimated uptake of n class="Chemical">phosphorus (μg P g-1 root dry weight h-1) as a funpan>ctionpan> of solutionpan> n class="Chemical">inorganic P concentration for Cladium mariscus spp. jamaicense and Typha domingensis acclimated to low P (5 μg l-1) and high P (500 μg l-1) concentrations and high oxygen (8 mg l-1; dashed lines) and low oxygen (<0.5 mg l-1; solid lines) concentration in the culture solutions. The curves were generated by the modified Michaelis-Menten model using the average kinetic parameters for the two species in the different treatments.

We hypothesized that oxygen-deficienpan>t conditions would affect the uptake kinetics of Cladium more thanpan> that of Typha because of differenpan>ces in their inherenpan>t ability to tranpan>sport pan> class="Chemical">oxygen to the roots. This hypothesis could not be verified, because of the confounding effects of P availability. At low P, oxygen affected uptake kinetics more for Typha than for Cladium whereas at high P the effects were small and mostly on Cladium. In earlier studies [[14], and unpublished], low oxygen and particularly reducing (Eh-150 mV) soil conditions significantly reduced growth and performance of both Cladium and Typha when P availability was low, but the effects could be largely ameliorated by high P availability. A similar tendency was observed in the present study: effects of oxygen were most pronounced at low P, and nearly disappeared at high P.

Our primary aim was to assess whether differences in P uptake kinetics could explain the observed growth characteristics of the two species in the Everglades. To achieve this goal, we have focused this discussion on plant responses under low oxygen, as this resembles the enpan>vironmenpan>tal conditions in the Everglades soils better thanpan> the aerated culture solutions. Under low pan> class="Chemical">oxygen and high P availability, Vmax did not differ much between the species, but the affinity for P (Vmax/K0.5 and α) was significantly lower for Cladium than Typha. However, Cladium had a greater proportion of the biomass allocated to roots, so overall uptake capacity of the two species at high P did not differ much. Root P uptake was largely controlled by the plant P demand. Although we did not investigate ion transport regulation, uptake kinetics may have adjusted via regulation of the membrane-bound high affinity ion transporters [51,52]. These results imply that under P-sufficient conditions, uptake kinetics does not influence competition between the two species. With sufficient P, the plants only need to adjust their uptake system to meet their respective P demands.

At an external concentration of inorganic P of only 5 μg l-1, P is certainly growth limiting for both species, as has beenpan> shownpan> in manpan>y previous studies [14,15]. Planpan>ts may respond to such pan> class="Disease">P-deficient conditions in several ways, including allocation of greater mass to roots relative to shoots, and the production of thinner and longer roots enhancing total surface area for nutrient acquisition [53]. In this study, the proportion of biomass allocation to roots increased for both Cladium (from 14 to 18%) and Typha (from 8 to 15%), but there were no changes in root morphology to increase absorptive area. Both species also adjusted their P uptake system in an attempt to obtain adequate P, but Typha more so than Cladium. The maximum uptake velocity, Vmax, increased by a factor of 5.3 and 3.1; the half saturation constant, K0.5, decreased by a factor 0.14 and 0.54; and the P uptake affinity, Vmax/K0.5 and α, increased by a factor 3.9-6.2 and 4.0-4.5 for Typha and Cladium, respectively. The estimated levels of Cmin (about 6 and 19 μg l-1 in the low-P treatment for Cladium and Typha, respectively) are obviously too high, as plants were growing and taking up P at 5 μg l-1. The depletion methodology used in this study, which included spiking with P to relatively high levels, may have resulted in the build-up of internal pools of P that interfered with the estimation of the true Cmin values. When Cladium and Typha plants were grown for 30 days in solution cultures with a low supply of P, both species maintained concentrations of 2-3 μg P l-1 in the solutions (unpublished results). We therefore suggest that the true Cmin level for the two species is in this range.

The adjustment of the P uptake system under low P conditions for Cladium and Typha increased the uptake velocity per unit of root mass 4 to 5-fold compared to the velocities for high-P acclimated plants. Adding the simultaneous increased biomass allocation to roots, the adjustment would result in a 5-fold higher P uptake per plant for Cladium and a 10-fold higher uptake for Typha. The combined effect of these adjustments is that the P uptake per plant would be alike for low-P acclimated plants at a solution concentration of ~35 μg P l-1 and for high-P acclimated plants at a solution concentration of ~500 μg P l-1 for both species. Actual plant uptake at 5 μg l-1 was of course much lower, as plants grew slower and had lower tissue P concentrations.

Adjustment of the uptake system, particularly Vmax, when plants are exposed to n class="Disease">nutrient deficiency is a commonpan> responpan>se observed by many species and many nutrienpan>t ionpan>s. The affinpan>ity for nutrienpan>t uptake as expressed by K0.5 is commonpan>ly assumed to be less plastic and less affected by plant growth conpan>ditionpan>s than Vmax [34,45,52]. However, the slope of the uptake curve at low P conpan>cenpan>trationpan>s, α, and the ratio betweenpan> Vmax and K0.5 (Vmax/K0.5) are better measures of affinpan>ity than K0.5. Inpan> the presenpan>t study, these affinpan>ity measures were clearly affected by P availability inpan> a manner similar to Vmax.

A high uptake affinity becomes increasingly important in P-deficient soils, where P uptake is controlled largely by the rate of diffusion to the depleted zones arounpan>d the roots [54]. The fact that Typha adjusts the affinity for P uptake anpan>d root biomass more thanpan> Cladium anpan>d that Typha has thinner root laterals thanpan> Cladium [15], increases this species' capacity to extract P from low P solutions relative to Cladium. However, despite anpan> apparenpan>tly less efficienpan>t uptake system, Cladium outperforms Typha during prolonged growth unpan>der pan> class="Disease">P-deficient conditions (unpublished). Increased allocation of biomass by Typha to roots may reduce its capacity to maintain a balanced acquisition of C and other resources and result in poor growth at P-deficient conditions. Hence, an efficient P uptake system alone does not ensure good performance at persistent low P availability.

Besides optimising the physiology of root P uptake, plants may also increase the availability of P in the rhizosphere by releasing specialized enzymes, known as phosphatases, which hydrolyse soluble organic P derived from soil organic matter to ortho-P for plant uptake [55]. Both Cladium and Typha secrete phosphatases at low P availability, but the rate of secretion is higher in Cladium, enhancing hydrolysis of organic P compounds [56]. Another means of P acquisition from n class="Disease">P-deficient soils is symbiosis with arbuscular mycorrhizal funpan>gi [57]. Inpan>oculation of Cladium with arbuscular mycorrhizal funpan>gi in a greenpan>house pot experimenpan>t increased growth anpan>d P uptake of Cladium significanpan>tly [58]. These additional capabilities are clearly importanpan>t for acquiring sufficienpan>t P from the peat-based, low-P soils of the Everglades, where P is stored primarily as organpan>ic P with anpan> additional componenpan>t of Ca-bounpan>d P [59,60]. The ability of Cladium to access organpan>ic P fractions, in concert with its slow growth rate, long tissue life time anpan>d high P-use efficienpan>cy, likely explains why this species is prolific in the pan> class="Disease">P-deficient Everglades soils.

Conclusions

A main finding of our study was that Typha has a more plastic P uptake system than Cladium that allows uptake of P over a wide range of external P concentrations and promotes high growth rates with a relatively low investment in root mass at high P levels. Both species adjust P uptake kinetics in relation to plant demand when P availability is high, but because of its opportunistic traits, Typha is more likely to outcompete Cladium in P-enriched areas. Under n class="Disease">P-deficient conditions Typha adjusts P uptake kinetics anpan>d biomass allocation to roots more thanpan> Cladium, anpan>d thereby achieves very efficienpan>t acquisition of P at low P levels. Inpan> contrast, Cladium has less need to adjust its P uptake kinetics in response to low P conditions probably because it is already efficienpan>t at acquiring P from peat-based soils (e.g., through secretion of phosphatases, symbiosis with arbuscular mycorrhizal funpan>gi, anpan>d efficienpan>t nutrienpan>t conservation growth traits). These findings suggest that differenpan>tial expression of a similar strategy by Cladium anpan>d Typha unpan>der low-P conditions explains why anpan> opportunpan>istic species like Typha is able to grow side by side with Cladium in the persistenpan>tly pan> class="Disease">P-deficient Everglades.

Methods

Experimental setup

Phosphorus uptake kinetics were measured for whole planpan>ts of Cladium anpan>d Typha acclimated in a factorial setup with two levels of P (5 anpan>d 500 μg P l-1) anpan>d two levels of pan> class="Chemical">oxygen (8.0 and <0.5 mg O2 l-1) in hydroponic culture solutions. Since P availability and internal pools of P in the plant tissues are known to affect plant development and physiology, and hence P uptake characteristics, special care was taken to ensure that plants were germinated and propagated at the desired P treatment concentrations. This was achieved by propagating plants from seeds hydroponically in growth cabinets with efficient control of the nutrient composition of the culture solutions. The P uptake kinetic parameters were estimated for whole individual plants of the two species in a controlled environment, the "PhytoNutriTron" (PNT), which is a computer controlled growth facility with four independent steady state hydroponic rhizotrons built into growth cabinets [61].

Seeds and germination

Seeds of Typha and Cladium were collected from populations of the two species in the oligotrophic interior of Water Conservation Area 2A in the northern Everglades anpan>d germinated on vermiculite at a 14:10 h day:night photoperiod anpan>d a 25:10°C thermoperiod, a climatic regime shownpan> to be optimal for germination of the two species [62]. The seedlings were pan> class="Chemical">watered with a basic nutrient solution (table 4) at pH 6.5 with additional phosphorus added as K2HPO4 to obtain the two P levels (5 and 500 μg l-1). The composition of the nutrient solution was developed to resemble the porewater concentrations of the major nutrient ions in the interior oligotrophic area of Water Conservation Area 2A of the Everglades. When plants had developed a root system that was large enough to allow the mounting of plants in hydroponic culture (after 3 to 5 months), the seedlings were transferred to a hydroponic nursery culture system.

Table 4

Hydroponic nutrient solutions

	Basic nutrient solution	P addition solution	Major adjustment
Condition
pH	6.5	6.5	NaOH, H2SO4
Conductivity	1 mS cm-1		Intermediate renewal
Temperature	27°C
Oxygen	<0.5 or 8 mg l-1
Element
Phosphorus	5 or 500 μg l-1	50 mg l-1	KH2PO4
Nitrogen	2.4 mg l-1		(NH4)2SO4
Potassium	3.4 mg l-1	802 mg l-1	K2SO4, KH2SO4
Calcium	130 mg l-1	291 mg l-1	CaSO4, CaCl2
Sulphur	98 mg l-1	690 mg l-1	(NH4)2SO4, CaSO4, MgSO4
Magnesium	41 mg l-1	114 mg l-1	MgSO4
Sodium	50 mg l-1	12 mg l-1	NaCl
Chloride	216 mg l-1	18 mg l-1	NaCl, CaCl2
Silicium	351 mg l-1		Na2SiO3
Boron	27 μg l-1	5.4 mg l-1	H3BO3
Manganese	11 μg l-1	2.2 mg l-1	MnSO4
Zinc	13 μg l-1	2.6 mg l-1	ZnSO4
Copper	13 μg l-1	2.6 mg l-1	CuSO4
Molybdenum	4.8 μg l-1	1.0 mg l-1	Na2MoO4
Iron	112 μg l-1		FeSO4

Composition of the hydroponic nutrient solutions. The basic nutrient solution was developed to resemble the pore water concentration of the major nutrient ions in the interior oligotrophic area of Water Conservation Area 2A of the Everglades. The P addition solution compensated for the uptake of nutrients by plants during growth and was added relative to the depletion of phosphorus by the plants.

Composition of the hydroponic nutrient solutions. The basic nutrient solution was developed to resemble the pore water concenpan>tration of the major nutrienpan>t ions in the interior oligotrophic area of pan> class="Chemical">Water Conservation Area 2A of the Everglades. The P addition solution compensated for the uptake of nutrients by plants during growth and was added relative to the depletion of phosphorus by the plants.

Plant acclimation in nursery system

The plants were acclimated to hydroponic growth at low and high P levels (5 and 500 μg l-1) in a nursery system that was set up in a growth chamber operated at a 15 h light/9 h dark cycle, a 30:25°C thermocycle and a 85:90% relative air humidity day:night cycle. Light was provided by a combination of inflorescent light tubes and metal halide bulbs at a photon flux denpan>sity of 350 μmol m-2 s-1 (PAR) at the base of the planpan>ts. Betweenpan> day anpan>d night, the climatic parameters were chanpan>ged gradually over a one hour tranpan>sition period. The nursery contained four indepenpan>denpan>t hydroponic growth unpan>its, each consisting of one or two 30 litre aerated growth tanpan>ks with up to 22 planpan>ts. The tanpan>ks of each growth unpan>it were connected to a 360 litre external nutrienpan>t solution reservoir. The culture solution was recirculated betweenpan> the reservoir anpan>d the growth tanpan>ks by pumps delivering 6 litre min-1 of solution to each growth tanpan>k. The culture solution consisted of the basic nutrienpan>t solution, anpan>d pan> class="Chemical">phosphorus was adjusted to the experimental levels using the P addition stock solution (table 4). The NH4 + level was adjusted with a solution of (NH4)2SO4. Changes in nutrient concentrations in the culture solutions were minimized through daily monitoring and adjustment of concentration levels. On weekdays, pH was adjusted to pH 6.5, and 112 μg Fe l-1 (FeSO4) was added to each unit. Temperature and conductivity were registered and the concentrations of NH4 + and PO4 3- were analysed using standard colorimetric methods (Lachat Instruments, Milwaukee, WI, USA). Orthophosphate detection was based on the ascorbic acid method (Method EPA-600/4-79-020, 1983, U.S. Environmental Protection Agency) and NH4 + was analysed using the salicylate method (Ammonia in waters 1981, London, Her Majesty's Stationary Office). Changes in conductivity during operation of the nursery were minimized by intermediate renewal of approximately 75% of the culture solutions when conductivity reached 2 mS cm-1. After 2 to 4 months, depending on species and treatment, when the plants started to produce rhizomes and ramets, individual plants were transferred to the controlled environment of the PNT.

Controlled environment

The uptake experiments were carried out in the PNT, which is a computer controlled hydroponic growth facility for experiments with whole plants [61]. The hydroponic rhizotron system of the PNT consisted of four independent growth units each containing eight root vessels built into a controlled growth chamber in a block design. The growth chamber regulated air temperature, humidity and light intensity (maximum 1200 μmol m-2 s-1 PAR at the base of the plants) in day:night cycles similar to those of the nursery. Each of the four growth units was connected to a separate steady state, temperature (27°C), pH (6.5) and oxygen controlled reservoir (180 l) through which the culture solution was recirculated. The reservoirs were equipped with UV-sterilization unpan>its, anpan>d the concenpan>trations of pan> class="Chemical">NH4 + and PO4 3- were monitored continuously by an auto-analyzer using standard colorimetric methods (Lachat Instruments, Milwaukee, WI, USA). The nutrient concentrations were maintained at constant levels through computer-mediated feedback regulation of peristaltic pumps that delivered the P nutrient stock solution and a (NH4)2SO4 stock solution to the reservoirs. The nutrients were supplied continuously at rates equivalent to their depletion in the culture solutions. The reservoirs and the root vessels were sealed from the atmosphere and flushed with either N2 gas or atmospheric air to control solution oxygen at the desired levels (<0.5 and 8 mg l-1, respectively). Each root vessel (height 700 mm, diameter 80 mm) had a lid with two openings for plants. The culture solution was circulated through each vessel at a rate of 4 litre min-1. The use of the P nutrient stock solution and partial replacement of the culture solutions (approximately 60% of the volume) ensured that concentrations of the major nutrients were maintained within ± 10% of desired set point level during the acclimation periods in PNT.

The four growth units of the PNT were used in sequence to create the 8 different treatment combinations (2 P levels × 2 species × 2 n class="Chemical">O2 levels). Inpan> order to optimize the control system for P detection, experimenpan>ts with low anpan>d high-P acclimated planpan>ts were carried out separately. Betweenpan> four anpan>d eight planpan>ts of each species for each treatmenpan>t were selected at ranpan>dom from the stock of planpan>ts in the nursery unpan>it. New ramets, rhizomes anpan>d senpan>escenpan>t planpan>t parts were removed from the planpan>ts before they were mounpan>ted in the root vessels of the PNT. The planpan>ts were acclimated to the steady state nutrienpan>t anpan>d pan> class="Chemical">oxygen levels in the controlled rhizotrons for at least one month prior to measurement of nutrient uptake.

Nutrient uptake kinetics

Plant P uptake was measured by the rate of depletion from nutrient solutions using an automated flow injection analyzer with four PO4 3- chanpan>nels (Lachat Inpan>strumenpan>ts, Milwaukee, WI, USA). pan> class="Chemical">Orthophosphate detection was based on the ascorbic acid method (Quickchem method no. 10-115-01-1-A (low sensitivity) and 10-115-01-1-B (high sensitivity), Lachat Instruments). At least 16 h before uptake kinetic studies were initiated, four plants were selected at random from the plants in the PNT. The roots were carefully rinsed in nutrient solution, and depending on plant size, transferred to root chambers with a volume between 0.7 and 12 litre (figure 5). The solution levels in the root chambers were marked and the chambers were placed in one of the growth cabinets of the PNT. Each root chamber was placed in a 10 litre tank containing 7 litre of nutrient solution. Centrifugal pumps recirculated water between the tanks and the root chambers at a rate of 1 litre min-1. Magnetic stirrers ensured mixing of the nutrient solutions in the root chambers, and depending on treatment, the chambers were flushed with atmospheric air or N2 gas at a rate 0.5 litre min-1.

Figure 5

Experimental chamber used for uptake studies. Schematic diagram of the experimental system used to estimate the plant P uptake showing the connection between the plant chamber and the flow injection analyzer (FIA) used for PO4 3- detection. 1: Acrylic root chamber with diffuser for flushing with air or N2; 2: Collar and seal for plant mounting; 3: Magnetic stirring bar protected by a net; 4: High speed peristaltic pump for recirculation of solution in the root chamber; 5: Tee pieces; 6: Three-way valves for calibration of FIA; 7: Peristaltic pump of the FIA; 8: Six port valve of a FIA analyzer channel.

Experimental chamber used for uptake studies. Schematic diagram of the experimental system used to estimate the plant P uptake showing the connection between the plant chamber and the flow injection analyzer (FIA) used for PO4 3- detection. 1: Acrylic root chamber with diffuser for pan> class="Disease">flushing with air or N2; 2: Collar and seal for plant mounting; 3: Magnetic stirring bar protected by a net; 4: High speed peristaltic pump for recirculation of solution in the root chamber; 5: Tee pieces; 6: Three-way valves for calibration of FIA; 7: Peristaltic pump of the FIA; 8: Six port valve of a FIA analyzer channel.

One hour before the initiation of the uptake kinetics studies, the circulation between the root chambers and the solution in the external tank was stopped, and the root chambers were flushed five times with a nutrient solution containing 5 μg P l-1. Thereafter, uptake studies were initiated by measuring a series of P depletion rates at stepwise increasing levels of P. The concentration of P was increased from 5 to 500 μg P l-1 in steps of 5 to 50 μg P l-1 followed by periods where the rates of depletion were measured. The root chamber was connected to the flow injection analyzer (FIA) through tubing (figure 5). A high-speed peristaltic pump (figure 5, item 4) recirculated solution through two tee-pieces connected to the FIA. The peristaltic pump of the FIA pumped the solution through a three-way valve connected to the recycling flow and to calibration solutions and back into the recycling flow line through the six-port valve of the analyzer. A three-way valve connected the return flow line from the FIA to the drain in order to collect sample water during instrumenpan>t calibration. The setup enpan>sured that no solution was lost during sampling, except for the volume extracted by the FIA for P anpan>alysis. During measuremenpan>t of P depletion the sample rate was 20 per hour with a sample volume of 780 μl. The extracted volume was replaced with a complete nutrienpan>t solution but without P. Calibration of the anpan>alyzer chanpan>nels was performed at regular intervals during the depletion series. The loss of test volume during the measuremenpan>ts due to evapotranpan>spiration, was replaced with distilled pan> class="Chemical">water. pH was maintained at 6.5 ± 0.5 by addition of NaOH when necessary.

Plant tissue measurements

At the end of the depletion experiments, the length of the root and shoot system was measured and plants were separated into roots, leaves, rhizomes and ramets, rinsed in deionised water, anpan>d the fresh weights (FW) were recorded before drying in a forced venpan>tilated ovenpan> for 48 hours at 80°C for dry weight (DW) determination. The dried planpan>t material was thenpan> finely grounpan>d anpan>d anpan>alyzed for N using a N-protein anpan>alyzer (Na2000, Carlo Erba, Milanpan>, Italy). The concenpan>trations of P in the planpan>t fractions were anpan>alyzed using anpan> ICP-AES (Plasma II, Perkin-Elmer, CT, USA) after pan> class="Chemical">HNO3 and H2O2 digestion [63]. The concentration of N and P in the plant fractions and the weight proportions of the fraction were used to calculate the nutrient concentration on a plant basis.

Estimation of uptake kinetics parameters

The P uptake rates were estimated by linear regression analysis of 3 to 5 consecutive samples of P concentrations measured during depletion. The slopes of the regression lines were corrected for the loss due to sampling by the flow injection analyzer, and divided by root dry weight to obtain the P uptake rate (μg P g-1 root DW h-1). Data collected during the initial 8 to 12 minutes after a shift in P concentration in the root chambers were omitted from the analyses in order to ensure that equilibrium between apoplastic and the external P concentrations was achieved.

The relationship between n class="Chemical">solution P conpan>cenpan>trationpan> (C) and uptake rates (V) was plotted and fitted by a modified Michaelis-Menpan>tenpan> model [38-41]:

The kinetic constants in this model are: Vmax, the maximum uptake velocity at saturating ion concentration; K0.5 (= K + Cmin), the half-saturation constant where uptake is 50% of Vmax; and Cmin, the ion concentration at which there is no net inflow. Vmax is a measure of the uptake capacity, and K0.5 as well as the ratio between Vmax and K0.5 are measures of the affinity of the uptake system. Also the slope, α, of the initial linear part of the uptake curve is a measure of affinity. This model has been successfully used in many studies to assess the underlying mechanisms responsible for up-regulating and down-regulating influx of different nutrient ions into whole plants, as well as the acclimations of the uptake system to different environmental conditions [39,40,61,64-67]. Vmax was entered into the model as a fixed parameter, and was estimated prior to the fitting as the average of the maximum uptake rates registered in a run. The modified Michaelis-Menten equation was fitted to the experimental data by means of a non-linear regression analysis using the Marquardt method (Statgraphics ver. 3, Manugistics, Inc., Maryland, USA). Two measures for P uptake affinity was calculated: (1) the ratio between Vmax and K0.5 (Vmax/K0.5); and (2) the slope, α, of the initial linear part of the uptake curve as estimated by linear regression analysis.

Statistics

All statistics were carried out using the software Statgraphics ver. 3 (Manugistics, Inc., Maryland, USA). The difference in uptake kinetic parameters between the two species and the effects of growth P level and n class="Chemical">oxygen treatmenpan>t were assessed by a factorial ANOVA model usinpan>g type III sum of squares. Three-way inpan>teractionpan>s were ignored inpan> the model. The data were tested for normality and variance homogenpan>eity usinpan>g Cochran's C-test, and data were log-transformed whenpan> necessary. Multiple levels of mainpan> effects were compared by multiple range tests usinpan>g the Tukeys HSD procedure at the 5% significance level.

Authors' contributions

BL carried out the controlled growth experiments and uptake studies and performed most of the statistical analyses. HB and BL drafted the manuscript. IAM, KLM and SLM participated in the design of the study and helped to draft the manuscript. All authors read and approved the final manuscript.