Photosynthetic performance of submerged macrophytes from lowland stream and lake habitats with contrasting CO2 availability

Authors


Author for correspondence:

Annette Baattrup-Pedersen

Tel: +45 87158776

Email: abp@dmu.dk

Summary

  • We examine the photosynthetic response of submerged plants from streams and lakes with contrasting free-CO2 and nitrogen (N) availability. We hypothesized that: the photosynthetic capacity of stream plants is higher because of higher N availability; the photosynthetic N-use efficiency (PNUE) is also higher because stream plants are acclimated to higher free-CO2; and PNUE is lower in aquatic compared to terrestrial plants.
  • We tested these hypotheses by measuring tissue-N, photosynthetic capacity and inorganic C extraction capacity in plants collected from streams and lakes and by comparing the PNUE of aquatic plants with previously published PNUE of terrestrial plants.
  • We found that the organic N content was consistently higher in stream (3.8–6.3% w/w) than in lake plants (1.2–4.3% w/w). The photosynthetic capacity correlated positively with tissue-N. The relationships were similar for stream and lake plants, indicating that N allocation patterns were similar despite variability in free-CO2 between the two habitats.
  • The slope of the relationship between photosynthetic capacity and tissue-N was lower than found for terrestrial plants, whereas the compensatory N content for photosynthesis was similar. This suggests that PNUE is lower in aquatic plants, perhaps reflecting that the selection pressure for a high C fixation rate per unit N is reduced as a result of low inorganic C availability in the aquatic environment.

Introduction

The most significant difference between lakes and streams as habitats for submerged macrophytes is the contrasting flow regimes. In lakes flow velocities vary between 0.04 and 4 cm s−1 (Losee & Wetzel, 1993), whereas in streams flow velocities are seldom below 5 cm s−1 but may be higher than 1 m s−1 (Whitton, 1975; Sand-Jensen et al., 1989; Macfarlane & Raven, 1990). Moving water has positive and negative impacts on the performance of submerged macrophytes. Water flow reduces the boundary layer thickness, thereby potentially enhancing the supply rates of dissolved inorganic carbon (C) and nutrients to the shoots at low velocities, which may stimulate photosynthesis and ultimately growth (Westlake, 1967; Madsen & Søndergaard, 1983; Borchardt et al., 1994). The interaction between flowing water and macrophyte performance is not restricted to the transfer of dissolved substances, as flowing water also provides a mechanical force which may break or uproot the plants at high flow velocities and affect plant morphology at intermediate velocities (Kraemer & Chapman, 1991; Madsen et al., 1993; Riis & Biggs, 2003; Sand-Jensen, 2003; Luhar & Nepf, 2011).

In addition to differences in flow regimes, low gradient streams and lakes in agricultural areas may differ in inorganic C and dissolved nutrient concentrations. The concentration of free-CO2, which is the most readily available C source for photosynthesis, generally declines from groundwater, over streams to lakes (Steemann Nielsen, 1975; Rebsdorf et al., 1991; Sand-Jensen & Frost-Christensen, 1998; Sand-Jensen & Stæhr, 2012). In a comparison between groundwater, streams and lake outlets in Denmark, the median concentration of free-CO2 was 460 μM in groundwater, 212 μM in streams and 29 μM in lakes (Sand-Jensen & Stæhr, 2012). The concentrations can vary greatly during the day and over the year from below air saturation during summer to above air saturation in winter. Generally, however, concentrations remain higher in streams compared to lakes throughout the year (Sand-Jensen & Stæhr, 2012). Similarly to the concentration of free-CO2, concentrations of nutrients also vary between streams and lakes (Bjerring et al., 2012; Wiberg-Larsen et al., 2012). The average annual concentration of dissolved nitrogen (N; 2003–2011) in streams that are part of the monitoring programme in Denmark is 5.71 mg N l−1 whereas in lakes the average concentration is 1.85 mg N l−1, although pronounced differences may occur among streams and lakes, reflecting differences in agricultural activities in the catchments and amounts of N and P transported to the aquatic environments (Bjerring et al., 2012; Wiberg-Larsen et al., 2012). Nutrient uptake can also take place from the interstitial water in the sediment that is continually renewed, which may further add to the variability in nutrient availability among streams and lakes (Carignan, 1982; Madsen & Cedergreen, 2002).

Overall, we expect that the higher concentrations of free-CO2 and possibly also dissolved nutrients in streams in concert with reduced transport resistance will improve the supply rates of C and nutrients to plants in streams compared to lakes. Furthermore, we expect that the interactive effects of differences in free-CO2 availability may affect the resource-use efficiency of macrophytes in the two habitat types (Bowes, 1993; Reich et al., 2006). For example, the light-use efficiency for growth may increase in response to elevated CO2 as observed for the submerged aquatic macrophyte Elodea canadensis (Madsen & Sand-Jensen, 1994) and for terrestrial species (Long & Drake, 1991), and also the N-use efficiency for photosynthesis (PNUE) may be influenced by the availability of free-CO2 (Hocking & Meyer, 1991; Polley et al., 1995; Madsen et al., 1998; Ge et al., 2012). Thus, large amounts of N are bound in the carboxylating enzyme ribulose bisphosphate carboxylase-oxygenase (rubisco) that operates more efficiently at elevated CO2, and plants may therefore acclimate to elevated CO2 by reducing the amount of rubisco (Nowak et al., 2004). Only sparse information exists on the interactive N and C effects in natural freshwater habitats, but studies of laboratory grown E. canadensis have revealed that N-use efficiency was improved at high inorganic C availability and that the tissue-N concentration needed to sustain maximum growth declined (Madsen et al., 1998).

In the present study we examine photosynthesis in a range of submerged species from streams and lakes situated in agricultural lowland areas. Specifically, we investigate species photosynthetic capacity and inorganic C extraction capacity in relation to environmental variability, in particular free-CO2 and N, offered by the two habitat types. We hypothesized that: the photosynthetic capacity of stream plants is higher than in lake plants because of higher N availability, N being a strong regulator of photosynthesis in plants; the photosynthetic N-use efficiency (PNUE) is also higher because stream plants are acclimated to higher free-CO2; and PNUE is lower in aquatic plants compared to terrestrial plants. These hypotheses were tested by measuring photosynthetic capacity, inorganic C response and the ability of plants to extract inorganic C in several species collected from streams and lakes with contrasting inorganic C availabilities and by comparing our findings with previously published data on photosynthetic capacity in terrestrial species in relation to tissue-N contents.

Materials and Methods

Plant material

Submerged macrophytes were collected from 13 lowland streams and 10 lakes situated within a 50 km radius in rural areas in Mid-Jutland, Denmark. The sampling sites were representative for lowland stream and lake habitats excluding severely nutrient impacted sites and stream outlets from lakes. A total of nine species were collected and of these seven were collected in both streams and lakes: Elodea canadensis L.C. Rich. (= 14), Fontinalis antipyretica L. (= 2), Myriophyllum alterniflorum DC. (= 4), Potamogeton crispus L. (= 6), Potamogeton perfoliatus L. (= 9), Sparganium emersum Rehman. (= 3), and Callitriche cophocarpa Sendtner (= 4). The two other species were found in only one habitat type: Ranunculus peltatus Schrank (= 4) was collected only from streams when occurring, and Littorella uniflora (L.) Ascherson (= 4) only from lakes. A majority of these species can utilize both CO2 and math formula for photosynthesis, but C. cophocarpa, F. antipyretica, S. emersum and L. uniflora can only utilize CO2 and L. uniflora only by uptake from the sediment and not through the shoots like the other species. Additionally, L. uniflora has the possibility of crassulacean acid metabolism (CAM). Fontinalis antipyretica is an aquatic moss with no roots and nutrient uptake can therefore only take place from the water. After collection, plants were kept in the laboratory for a maximum 3 d before measurements were taken in vigorously aerated tap water with an alkalinity of 1.5 meqv l−1 at 15°C in a growth chamber. Light intensity was 200 μmol m−2 s−1 (PAR) in a 16 h light : 8 h dark cycle.

Photosynthetic capacity

The photosynthetic capacity was determined as oxygen exchange at saturating light and free-CO2 concentrations on apical parts of the shoots or young but fully developed submerged leaves incubated in 30 ml glass bottles (c. 100 mg FW). The growth medium was modified from Smart & Barko (1985) by adding KHCO3 to a final concentration of 1.0 mM dissolved inorganic carbon (DIC) and lowering pH to 6.5 with a CO2-saturated medium giving a free-CO2 concentration of 800 μM. After 10 min of preincubation, the medium was renewed and incubation initiated. The bottles were placed either on a rotating wheel in an incubator at 15°C and 450 μmol m−2 s−1 (PAR) or in a shaking water bath at 15°C and 400 μmol m−2 s−1 (PAR). Rates of photosynthesis were calculated from the increment in O2 concentrations determined by Winkler titrations using automatic deadstop titration (precision 0.01 mg O2 l−1).

Inorganic carbon response curves and bicarbonate uptake capacity

The relationships between photosynthesis and external concentrations of CO2 and bicarbonate (math formula) were determined for E. canadensis collected from five lakes and three streams and P. perfoliatus from four lakes and three streams. These species were selected as they were found in most streams and lakes. Photosynthesis was measured as oxygen exchange of shoots (200–450 mg FW) incubated in a stirred perspex chamber (140 ml) equipped with a Clark-type O2 electrode whose signal was recorded at 10 Hz using a computerized data-acquisition system. The chamber was kept at 15°C (± 0.1°C). The light intensity was 450 μmol m−2 s−1 (PAR) provided by a 150 W, 24 V metal-halide lamp. CO2 response curves were measured in media with low alkalinity (0.010 meqv l−1). The CO2 concentration was increased successively by adding small aliquots of CO2-saturated water with similar alkalinity through an injection port in the chamber lid. The CO2 concentrations were calculated from changes in O2 and added CO2, assuming a molar equivalency between O2 and CO2. math formula response curves were determined by measuring photosynthesis at different math formula concentrations, varying between 60 and 3000 μM calculated from changes in O2 and added math formula, at low CO2 (< 2 μM, obtained by bubbling the medium with air containing c. 45 ppm CO2). To test if the concentration of dissolved inorganic C remained constant in the medium over time, the alkalinity was determined before and after an experimental run on several occasions by Gran-titration (Stumm & Morgan, 1981). In all instances the alkalinity remained unchanged. All gas exchange media were modified from Smart & Barko (1985) by adjusting the amount of inorganic C added. The initial slopes of the CO2 (αCO2) and math formulamath formula) response curves, giving the CO2 and math formula conductance were estimated by linear regressions over the lower linear part of the curves.

Inorganic carbon extraction capacity

The inorganic C extraction capacity of a subset of the plants including those able to use math formula in photosynthesis – namely, E. canadensis, P. perfoliatus, P. crispus, M. alterniflorum and R. peltatus – was measured in pH drift experiments. The apical parts of shoots or young but fully developed leaves (c. 100 mg FW) were incubated in closed 30 ml bottles in a medium modified from Smart & Barko (1985) by adding additional KHCO3 to a final alkalinity of 1.0 meqv l−1. The bottles were placed on a rotating wheel in an incubator at 15°C with a light intensity of 450 μmol m−2 s−1 (PAR). The medium was bubbled with atmospheric air from the laboratory before the experiment. After 24 h of incubation the final pH was measured and the DIC concentration calculated assuming constant alkalinity (Stumm & Morgan, 1981). The carbon extraction capacity is expressed as a CT/Alk value, where CT is the final DIC concentration and Alk the alkalinity. High CT/Alk values close to one indicate limited or no math formula use, whereas lower CT/Alk values indicate an ability to exploit math formula in photosynthesis, the ability increasing with decreasing values of CT/Alk (Allen & Spence, 1981).

Common garden growth experiments

In order to investigate acclimation plasticity, E. canadensis from two lakes and two streams was grown in laboratory cultures for 75 d under identical conditions. Elodea canadensis was chosen because it exhibited the broadest range in tissue-N and measured photosynthetic parameters. Eight apical shoots (5 cm long) were planted in 500 ml pots in sandy sediment. Sufficient inorganic nutrient supply was ensured by adding three fertilizer sticks (containing by weight 11% N, 2% P, and 5% K; weight 1 g) to the sediment twice during the growth period. Two pots were placed in an aquarium containing 15 l of growth medium modified (see Smart & Barko, 1985) by adding NaHCO3 and KHCO3 (1 : 1 on a molar basis) to a final total inorganic C (DIC) of 1.5 mM and micronutrients from a commercially available solution. The medium was vigorously aerated and renewed every second day. Illumination intensity was 250 μmol m−2 s−1 (PAR) in a 16 h light: 8 h dark cycle and the temperature was 15°C. The photosynthetic capacity, the inorganic C extraction capacity, and the initial slopes of the CO2 (αCO2) and math formulamath formula) response curves were determined after the growth period following the procedures described previously.

Dry weight, leaf area and chemical analysis

Dry weight of plant shoots used for photosynthetic measurements was determined after freeze-drying for 24 h. The total organic N content of plant shoots were determined by Kjeldahl digestion followed by titration with 0.05 N HCl (Jackson, 1958). One-sided leaf area was determined on mature leaves of five apical shoots using an area-meter. The alkalinity of water samples from the streams and lakes was determined by titration with 0.01 M HCl by Gran titration. The concentration of free-CO2 was calculated according to Stumm & Morgan (1981).

Statistical analysis

The photosynthetic capacity, the inorganic C extraction capacity, and the initial slopes of the CO2 (αCO2) and math formulamath formula) response curves (for E. canadensis and P. perforliatus) together with the DW : FW ratio and tissue N contents were determined on three replicates of each species from each stream and lake habitat. Mean values for each species from each habitat were then ranked and the distributions of organic tissue-N contents, photosynthetic capacities and inorganic C extraction capacities were compared for stream and lake plants applying Mann–Whitney U tests using a significance level of < 0.05. To analyse the influence of tissue-N contents on the photosynthetic capacity and the inorganic carbon extraction capacity linear regressions were fitted on the whole datasets as well as on species-specific datasets. We used the slope of the relationship between tissue-N and the photosynthetic capacity as an expression of the PNUE. To test if the PNUE varied between stream and lake plants, regression slopes and intercepts for stream and lake plants were compared by ANCOVA using a significance level of < 0.05. Similarly, to test whether the initial slopes of the CO2 and math formula response curves varied between stream and lake plants of E. canadensis and P. perfoliatus, regression slopes and intercepts were compared by ANCOVA also using a significance level of < 0.05.

Results

The alkalinity ranged from 1.0 to 2.5 meqv l−1 in the streams and from 0.4 to 1.8 meqv l−1 in the lakes from which the plants were collected (Table 1). At the time of plant collection free-CO2 concentrations in the streams varied from 125 to 304 μM or c. 2–10 times the concentration measured in the lakes (Table 1).

Table 1. Median values and ranges of pH, alkalinity and free-CO2 in the sampling streams (= 13) and lakes (= 10)
 pHAlkalinity (meqv l−1)Free-CO2 (μM)
Streams7.2 (7.1–7.4)1.8 (1.0–2.5)198 (125–304)
Lakes8.1 (7.5–8.5)1.2 (0.4–1.8)49 (21–79)

The organic N content of the plant tissue was significantly higher for plants collected in streams (median value 3464.1 μmol N g−1 DW; range 2667.6–4503.5 μmol N g−1 DW, corresponding to 3.8–6.3% w/w) than for plants collected in lakes (median value 1991.4 μmol N g−1 DW; range 1109.5–3100.8 μmol N g−1 DW, corresponding to 1.2–4.3% w/w) (Fig. 1a; Mann–Whitney, < 0.05). This difference was not due to interspecific variation as Elodea canadensis covered the full range of tissue-N contents.

Figure 1.

Frequency distribution of organic tissue nitrogen (N) contents (a; μmol N g−1 DW;= 49) and photosynthetic capacities (b; μmol O2 g−1 DW h−1; = 49) of nine species of submerged macrophytes collected in 13 lowland streams (= 25) and 10 lowland lakes (= 24) and inorganic carbon extraction capacities expressed as CT/Alk (c; = 39) of five species of submerged macrophytes from streams (= 16) and lakes (= 23) able to use HCO3 in photosynthesis. CT is the final dissolved inorganic carbon (C) concentration and Alk is the final alkalinity. The frequency distributions of tissue-N contents, photosynthetic capacities and inorganic C extraction capacities were significantly different between plants from the two habitat types (Mann-Whitney U tests; < 0.05).

The photosynthetic capacities were higher for stream plants (median value 877.2 μmol O2 g−1 DW h−1; range 534.4– 1373.5 μmol O2 g−1 DW h−1) than for lake plants (median value for photosynthetic capacity 420.1 μmol O2 g−1 DW h−1; range 179.4–1011.9 μmol O2 g−1 DW h−1; Fig. 1b; Mann– Whitney, < 0.05). The inorganic C extraction capacities were also significantly higher in stream plants (median value for inorganic C extraction capacity 0.519; range 0.431–0.618) than in lake plants (median value 0.614; range 0.499–0.911) (Fig. 1c; Mann–Whitney, < 0.001) although the variability among lake plants was high.

The photosynthetic capacity and carbon extraction capacity correlated with tissue-N contents for both stream and lake plants (Figs 2, 3), and the relationships were not significantly different between stream and lake plants (ANCOVA, > 0.05). For the pooled data the relationship between the photosynthetic capacity (PSmax) and tissue-N was PSmax = 0.28N – 81.0 on a DW basis (Fig. 2a; Adj r2 = 0.59, < 0.001) and PSmax = 0.26N – 0.04 on an area basis (Fig. 2b; Adj r2 = 0.71, < 0.001). For the individual species we also found positive relationships between tissue-N and the photosynthetic capacity (Adj r2 varying between 0.14 and 0.70), although these relationships were not significant for all species. The relationship between the C extraction capacity (CT/Alk) and tissue-N was CT/Alk = −7.2 × 10−5N + 0.78 (Fig. 3; Adj r2 = 0.38, < 0.001).

Figure 2.

Photosynthetic capacity as a function of tissue nitrogen contents in nine submerged species of macrophytes from streams (= 25) and lakes (= 24) expressed on (a) a dry weight basis, regression line ± 95% confidence limits (grey): Y = 0.28X-81.0 (Adj r2 = 0. 59; < 0.001), and (b) an area basis, regression line ± 95% confidence limits (grey): Y = 0.26X-0.02 (Adj r2 = 0.71; < 0.001). Mean ± SD is given (= 3). Solid circle, Elodea canadensis; open circle, Potamogeton crispus; solid box, Callitriche cophocarpa; open box, Potamogeton perfoliatus, open triangle apex down, Myriophyllum alterniflorum; open triangle apex up, Fontinalis antipyretica; solid triangle apex down, Littorella uniflora; solid triangle apex up, Sparganium emersum; rhombus, Ranunculus peltatus.

Figure 3.

Carbon extraction capacity as a function of tissue nitrogen (N) contents (μmol N g−1 DW) in five species of submerged macrophytes from streams and lakes able to use HCO3 in photosynthesis. The relationship is described by = −1.1 × 10−4 + 0.93 (r2 = 0.79; < 0.001). Mean ± SD is given (= 3). Solid circle, Elodea canadensis; open circle, Potamogeton crispus; solid box, Callitriche cophocarpa; open box, Potamogeton perfoliatus, open triangle apex down, Myriophyllum alterniflorum; open triangle apex up, Fontinalis antipyretica; solid triangle apex down, Littorella uniflora; solid triangle apex up, Sparganium emersum; rhombus, Ranunculus peltatus.

The photosynthetic capacity varied significantly between stream and lake plants of E. canadensis (ANOVA, < 0.05; Table 2), but not between stream and lake plants of Potamogeton perfoliatus (ANOVA, P > 0.05; Table 2). No significant difference in math formula uptake capacity was observed between stream and lake for either E. canadensis or P. perfoliatus (ANOVA, > 0.05; Table 2). The initial slopes of the CO2 response curves (αCO2) were significantly higher in stream E. canadensis than in lake E. canadensis (ANCOVA, < 0.05), whereas the initial slopes of the math formula response curve (αmath formula) were similar for stream and lake E. canadensis (ANCOVA, > 0.05; Table 2). For P. perfoliatus both αCO2 and αmath formula were similar for stream and lake plants (Table 2).

Table 2. The photosynthetic capacity, math formula saturated rate of photosynthesis and initial slopes of the CO2 (αCO2) and math formulamath formula) response curves of stream and lake specimens of Elodea canadensis collected from five lakes and three streams and Potamogeton perfoliatus collected from four lakes and three streams
 PS capacity (μmol O2 g−1 DW h−1)PS (math formula) (μmol O2 g−1 DW h−1)αCO2(μmol O2 g−1 DW h−1 (μM CO2)−1)αmath formula (μmol O2 g−1 DW h−1 (μM math formula)−1)
  1. Mean values and ranges are given.

  2. a

    Significantly lower species-specific means between streams and lakes (< 0.05; ANOVA).

E. canadensis
Stream884.1 (766.9-1008.3)302.9 (157.1–431.2)19.3 (14.7–23.1)2.3 (1.7–3.0)
Lake319.5 (190.7–431.3)a145.9 (102.1–142.4)8.6 (4.2–11.9)a1.7 (0.5–3.1)
P. perfoliatus
Stream851.9 (711.7–927.4)225.9 (132.9–341.7)12.2 (11.1–13.0)1.1 (0.8–1.4)
Lake613.5 (420.2–788.2)146.4 (86.3–225.9)19.4 (10.5–31.1)1.7 (1.0–2.9)

The math formula uptake capacity correlated positively with tissue-N in both E. canadensis and P. perfoliatus, and the relationships did not differ between stream and lake plants (> 0.05). For the pooled data the relationship between tissue-N (N) and the math formula uptake capacity (PSmath formula) was PSmath formula = 0.107N-83.4 (Adj r2 = 0.63, < 0.001). Positive couplings between the initial slopes of the CO2 and the math formula response curves and tissue-N were observed for E. canadensis (Adj r2 = 0.88, < 0.001 and Adj r2 = 0.53, < 0.05, respectively; correlations not shown), but not for P. perfoliatus.

Common garden growth experiment

After growth of stream and lake plants of E. canadensis for 75 d under identical conditions, tissue N-content and photosynthetic characteristics were measured. Tissue-N increased significantly in lake plants (ANOVA, P < 0.05) but not in stream plants of E. canadensis during the growth period (ANOVA, P > 0.05). After growth, the photosynthetic capacity and the math formula uptake capacity was similar in stream and lake plants of E. canadensis, whereas αCO2 was lower in lake than stream plants. On the other hand, αmath formula was similar in stream and lake plants of E. canadensis (ANCOVA, > 0.05; Table 3). The overall similarity in the photosynthetic response of E. canadensis following growth under identical conditions indicates that the large differences observed in tissue-N contents and photosynthetic parameters between plants collected in streams and lakes is induced by differences in growth conditions in these habitats, and thus phenotypic plasticity rather than genotypic differences between individuals.

Table 3. Photosynthetic capacity, math formula saturated rate of photosynthesis and initial slopes of the CO2 and math formula response curves (αCO2 and αmath formula) of stream and lake specimens of Elodea canadensis after 75 d growth in the laboratory under identical conditions
 αCO2 (μmol O2 g−1 DW h−1 (μM CO2)−1αmath formula (μmol O2 g−1 DW h−1 (μM math formula)−1PS capacity (μmol O2 g−1 DW h−1)PS (math formula) (μmol O2 g−1 DW h−1)
  1. a

    Significantly lower means (< 0.05; ANOVA). Mean ± SD (= 3).

Stream8.1 ± 0.43.6 ± 0.41184.6 ± 278.0468.3 ± 51.5
Lake5.1 ± 1.4a4.0 ± 2.21084.2 ± 303.5590.5 ± 42.7

Discussion

The higher tissue-N contents in submerged macrophytes from streams (3.8–6.3% w/w) compared to lakes (1.2–4.3% w/w) indicate that the availability of inorganic N is higher in Danish lowland streams than in lakes as was expected from the generally higher external concentrations (Jensen et al., 2011). The tissue-N contents in stream plants found in our study were higher than those reported by Kern-Hansen & Dawson (1978) for submerged plants collected in 19 Danish streams in the late 1970s – 1.6–4.0% w/w – although the availability of inorganic N likely has declined in recent decades (Kristensen et al., 1990; Jensen et al., 2011; Bjerring et al., 2012; Wiberg-Larsen et al., 2012). Therefore, the higher tissue-N contents observed in the present study might reflect differences in N allocation patterns as it was measured only on shoots in the present study, whereas Kern-Hansen & Dawson (1978) measured tissue-N on whole plants. Additionally differences in sampling dates, sites and species may also affect tissue-N contents.

Tissue-N contents in stream and lake plants were higher than contents supposed to saturate growth in terms of biomass yield, that is, 1.3% w/w suggested by Gerloff & Krombholz (1966) for submerged macrophytes. Gerloff & Krombholz (1966) determined the critical N content for growth at elevated concentrations of CO2 (0.5–1%). Given, however, that the concentration of CO2 can affect the nutrient-use efficiency (Bowes, 1993), higher contents in stream and lake plants do not necessarily imply that growth is N saturated in both types of habitats. In wheat, the critical tissue-N content for growth was reduced by 25% as the concentration of CO2 was doubled (Conroy & Hocking, 1993), and comparable alterations in the critical N content in response to variations in free-CO2 may occur for submerged macrophytes as well. In support of this, growth per unit of tissue-N was found to be higher at high than at low CO2 for Elodea canadensis and Callitriche cophocarpa grown in the laboratory under different inorganic N and inorganic C concentrations (Madsen et al., 1998).

The photosynthetic capacity and the C extraction capacity increased linearly with increasing tissue-N contents in stream and lake plants covering a range of species with different inorganic C acquisition systems. This finding indicates that both the CO2 and light saturated rate of photosynthesis and the C extraction capacity are regulated by one or several nitrogenous components. Most N in plant leaves is associated with photosynthesis (Chapin et al., 1987; Evans, 1989) and relationships between tissue-N and photosynthetic parameters may therefore be affected by the N partitioning to photosynthesis and between photosynthetic components. However, the finding that the relationships between tissue-N and photosynthetic parameters were similar for lake and stream plants indicates that the large differences in free-CO2 availability between lakes and streams had little, if any, effect on N partitioning in the plants. This result was surprising. We expected that the photosynthetic capacity per unit N would be higher in stream than lake plants due to an acclimation of the carboxylation capacity with down-regulation of rubisco without negatively impacting potential C acquisition in stream plants (Hocking & Meyer, 1991; Bowes, 1993; Conroy & Hocking, 1993; Long et al., 2004). It may be, though, that the generally higher N availability in stream habitats offsets the CO2 acclimation response of the photosynthetic apparatus. Thus, it has been observed that the acclimation response is greater under low N compared to high N availabilities (Ainsworth & Rogers, 2007), although other findings indicate that this may not always be the case (Lee et al., 2011). Additionally, N allocation into supporting and structural tissue may also vary between stream and lake plants because of differences in flow velocities (Armstrong, 1987; Kraemer & Chapman, 1991), which can also affect the photosynthetic capacity per unit N.

Submerged aquatic species are exposed to an environment widely different from the terrestrial one with respect to factors important for the regulation of photosynthesis and growth (Spence, 1981; Maberly, 1985). However, for both aquatic (this study) and terrestrial species (Field & Mooney, 1986; Evans, 1989; Reich et al., 1994), linear relationships exist between photosynthetic capacity and tissue-N. The abscissa intercept or the compensatory N content for photosynthesis of aquatics was similar to that found for terrestrials (Field & Mooney, 1986), that is, c. 500 μmol N g−1 DW. Assuming a photosynthetic quotient of one (Rosenberg et al., 1995), the weight-based slope of the relationship between tissue-N and photosynthetic capacity was lower in aquatic species (0.28 μmol O2 h−1 (μmol N)−1) than in terrestrial species (0.54 μmol CO2 h−1 (μmol N)−1; Field & Mooney, 1986), whereas the area-based slopes were similar (0.26 μmol O2 h−1 (μmol N)−1 in aquatic species and 0.29 μmol CO2 h−1 (μmol N)−1 in terrestrial species). However, the reduction in the slope when converting from a weight- to an area-based relationship in terrestrials is for the most part caused by the deviating response of sclerophylls. These species have high tissue-N contents per unit area but low photosynthetic capacities, probably due to high diffusional resistance and high N contents in nonphotosynthetic compounds (Field & Mooney, 1986; Evans, 1989; Vitousek et al., 1990). Hence, similar area-based slopes in terrestrials (Field & Mooney, 1986) and aquatics (this study) in the relationship between photosynthetic capacity and tissue-N may not be of general value in a comparison of terrestrials and aquatics, but rather a consequence of the contrasting life form of terrestrial sclerophylls.

The relationship between photosynthetic capacity and tissue N content may be analysed in terms of the constraints on photosynthesis (Evans, 1989). Because the compensatory N content for photosynthesis is similar in aquatics and terrestrials, the lower weight-specific slope of aquatics indicates that the photosynthetic N-use efficiency (PNUE) is lower in aquatic than in terrestrial species. The lower PNUE of aquatic species is probably not a result of proportionally less N invested in photosynthesis relative to that invested by terrestrials; the N requirement in nonphotosynthetic compounds in aquatic species may simply be lower due to the higher specific leaf areas (Field & Mooney, 1986). Instead, the lower PNUE may result from a less efficient use of photosynthetic N by aquatics. A broad comparison of the aquatic and terrestrial environments with respect to inorganic N and C suggests that N availability is similar or higher in the aquatic habitat (Etherington, 1982; Wetzel, 1983), whereas the availability of inorganic C is lower. This is a result of 104 times reduced diffusion rates in water and thick boundary layers (Raven, 1984). Therefore, under otherwise similar conditions, to obtain the same flux of CO2 across the boundary layer in air and water and, hence, the same rate of photosynthesis, the CO2 concentration in water must be > 104 times higher than in air. Given that the concentration of CO2 in water in equilibrium with air is about the same as in air and that the concentration seldom exceeds 25 times equilibrium concentrations in aquatic habitats (Rebsdorf et al., 1991; Sand-Jensen & Stæhr, 2012), the selection pressure for a high C fixation rate per unit of N may be diminished, and with that the PNUE of aquatic plants.

We believe that similar to terrestrial species tissue-N is an important parameter that can be used to predict the photosynthetic performance of aquatic species (Field & Mooney, 1986; Evans, 1989). However, several important differences in the photosynthetic response to tissue-N may occur among species (Evans, 1989; Reich et al., 1994) that may also reflect differences in their inorganic C concentrating mechanisms (e.g. Raven et al., 2011). In the present study, the C extraction capacity appeared to be high in some of the species irrespective of tissue-N contents. Also, a detailed investigation of the photosynthetic performance of E. canadensis and Potamogeton perfoliatus revealed species-specific variations in the initial slopes of the inorganic C response curves to tissue-N responses. For E. canadensis the initial slopes of the CO2 and math formula response curves were related to tissue-N contents, whereas no such relationships were found for P. perfoliatus; a difference that may reflect inter-species dissimilarities in N allocation patterns as well as interference from growth-regulating factors other than N. Thus, both tissue-N and the availability of inorganic C have been found to affect the inorganic C affinity of submerged plants (Sand-Jensen & Gordon, 1986; Madsen, 1993; Madsen & Baattrup-Pedersen, 1995). Studies with laboratory grown E. canadensis have revealed positive correlations between tissue-N contents and CO2 and math formula affinities (Madsen & Baattrup-Pedersen, 1995), but the CO2 and math formula affinities of E. canadensis can also be modified by changing the availability of CO2 during growth (Sand-Jensen & Gordon, 1986).

In conclusion, strong and similar relationships between tissue-N and the photosynthetic and C extraction capacities were observed for submerged macrophytes from streams and lakes despite pronounced differences in free-CO2 between the two habitats. The weight-specific slope of the relationship between photosynthetic capacity and tissue-N was lower for aquatic than for terrestrial species, whereas the compensatory N content for photosynthesis was similar (Field & Mooney, 1986). This suggests that PNUE is lower in aquatic species, which may be explained by a less efficient use of photosynthetic N, perhaps reflecting that the selection pressure for a high C fixation rate per unit of N is reduced as a result of the low inorganic C availability in the aquatic environment.

Acknowledgements

This research was supported by the European Union 7th Framework Project REFRESH under contract no. 244121, the Danish Natural Science Research Council and Aarhus University. We thank Juana Jacobsen for figure layout and Anne Mette Poulsen for manuscript editing.

Ancillary