Introduction

Few studies are known about the physiological tolerances of Halophila johnsonii (Dawes et al. 1989; Durako et al. 2003), probably due to its restricted geographical distribution and low abundance where it appears (Kenworthy 1997). This species is also characterized by having a limited reproductive capacity; no male flowers have been observed, thus, sexual reproduction may not occur. The lack of sexual reproduction and patchy distribution make this species very vulnerable to anthropic and natural disturbances (Durako and Wettstein 1994). However, H. johnsonii has a significant ecological importance; it stabilizes sediments and has a significant function in nutrient cycling, due to its rapid turnover (Kenworthy 1997). For all these reasons it is considered a rare species and it has been the only seagrass catalogued as threatened by the Endangered Species Act (63 FR 49035) in the USA (Durako and Wettstein 1994; Kenworthy 1997).

H. johnsonii inhabits a 200 km segment of the coastal lagoons along the east coast of Florida, between Sebastian Inlet and north Biscayne Bay. It has been found in a wide range of environmental conditions, suggesting that it can be considered more eurybiotic than the other species of the same genera (Kenworthy 1997). Field observations indicate that it occurs in waters with temperatures ranging from 21 to 36°C and salinities between 24.3 and 43 psu (Eiseman and McMillan 1980) while in a laboratory experiment Dawes et al. (1989) reported a maximum photosynthesis for this species at 25–35 psu and 30°C.

Salinity is a constant factor in the open sea, although it can fluctuate in more closed environments (lagoons, bays, estuaries, etc.) due to natural freshwater inputs or human activities, such as the reduction in freshwater inputs within the coastal lagoons of South Florida as a consequence of the freshwater management practices (Montague and Ley 1993; Hicks et al. 1998; Kamermans et al. 1999; Irlandi et al. 2002; Lirman and Cropper 2003), wastewater disposal, or the development of seawater desalination plants (Tomasko et al. 1999). Although most seagrasses can tolerate short-term salinity fluctuations, salinity variations will significantly affect some of the biochemical processes involved in photosynthesis and growth, determining the biomass, distribution, and productivity of these species (Montague and Ley 1993; Hillman et al. 1995; Chesnes and Montague 2001). Experimental studies on seagrass tolerance to salinity changes have shown that most species have optimum productivity at around oceanic salinity (Ogata and Matsui 1965; McMillan and Moseley 1967; Biebl and McRoy 1971; Drysdale and Barbour 1975; Hillman et al. 1995; Doering and Chamberlain 1998), though some species have optima at lower salinities (Kamermans et al. 1999; van Katwijk et al. 1999). These investigations have demonstrated that extreme or suboptimal salinities can produce negative alterations of their photosynthetic rate (Biebl and McRoy 1971; Kerr and Strother 1985; Dawes et al. 1987, 1989), metabolism (van Katwijk et al. 1999), reproduction (Ramage and Schiel 1998), growth (McMillan and Moseley 1967; Walker 1985; Walker and McComb 1990), and survival (Vermaat et al. 2000).

Along with salinity levels, other important environmental factors can vary, such as temperature or pH, also affecting the distribution and growth of several seagrass species (Ogata and Matsui 1965; Biebl and McRoy 1971; Hillman et al. 1995). These variations can be natural or produced by certain human activities, such as discharges from a desalination plant. Actually desalination is a growing industry at Florida but there is little information about its potential adverse environmental impacts. Brine discharge has an elevated salinity, a relatively high temperature, and low pH (Lattemann and Höpner 2003). Temperature can alter the metabolism or cause mortality at extreme values (Biebl and McRoy 1971; Drysdale and Barbour 1975; Drew 1979; Marsh et al. 1986). The principal influence of pH is on the concentration and form of available carbonates (Invers et al. 1997), which affect the photosynthesis of marine angiosperms (Beer and Waisel 1979). Few studies have investigated the effect of interactions or co-variations of these factors on seagrasses (Biebl and McRoy 1971; Dawes et al. 1989; Ralph 1999).

The main objective of the present work was to determine the responses of H. johnsonii to salinity in different laboratory-controlled experiments. Growth and mortality rates of this species were examined for medium-term exposures (15 days), while photosynthesis versus irradiance (PI) behavior and respiration were investigated for short-term salinity exposures (1 or 20 h), including the study of possible interactions with pH and temperature.

Materials and methods

Plant material

Intact vegetative plants of H. johnsonii were collected from a shallow bed at Haulover Park, in northern Biscayne Bay, FL (USA, 25°55′N; 80°07′W) in 2001. Plants were transported to the seawater-supplied greenhouse of the Center for Marine Science at Wilmington, NC and allowed to acclimate to laboratory conditions for 1 year prior to experimentation. During this time, plants were maintained in aquariums with a salinity of 30–33 psu.

In November 2002, H. johnsonii’s responses to salinity were monitored through two types of experiments, a medium-term series (15 days) in which growth and survival rates were evaluated at different salinities and several short-term (1 or 20 h) photosynthetic tests in which plants were subjected to rapid salinity exposures.

Growth experiment

Transplant units, consisting of a rhizome segment containing an apical shoot with three leaf pairs, were selected and placed in individual and open plastic pots of 125 cm3 with sandy-mud sediment. The transplant units were acclimated in a 50 l glass aquarium at 30 psu for 24 h, prior to the start of the experiment. Transplant units were randomly chosen from the acclimation aquarium and placed in seven different salinity treatments (0, 10, 20, 30, 40, 50, and 60 psu) replicated in two well-aerated 50 l glass aquariums (we used a total of 14 aquariums). Four transplant units were used per aquarium. Before the experiments, the plants were allowed to acclimate to each 10 psu variation for 3 days because previous studies suggest that acclimation to salinity variations permits a greater tolerance to osmotic stress than an instantaneous transfer (Ralph 1998).

Control-treatment salinity was 30 psu, considered the average salinity of the waters where H. johnsonii has been found, and it is within the optimal range described for its photosynthetic rates (Eiseman and McMillan 1980; Dawes et al. 1989). Increased salinity treatments were produced by adding Instant Ocean salt (Aquarium Systems, Ohio, USA) free of added nutrients to seawater, while lower salinities were obtained by diluting seawater with demineralized freshwater. Salinity values in each aquarium were controlled and maintained (±1.5 psu) during the exposure time. Temperature during the experiment varied between 25 and 32°C. Irradiance measured at the sediment surface was 500–1400 μmol quanta m−2 s−1 of photosynthetically active radiation (PAR, λ=400–700 nm) during the day, measured with a LiCor 193 SA spherical quantum sensor, with a light regime adjusted to 12 h light and 12 h dark using supplemental natural spectrum halogen lighting.

Net shoot production was estimated by a marking method, consisting of a fine plastic fiber placed on the sediment on the same position of the terminal leaf pair of each rhizome (Short and Duarte 2001). At the end of the experiment (duration 15 days), the number of new leaf pairs produced and the mortality rate were recorded.

Photosynthetic experiments

In order to estimate the effects of rapid salinity exposures, a series of experiments were conducted to determine acute responses (treatment exposures of 1 and 20 h) of photosynthesis and respiration on this seagrass to a wide range of salinities (from 0 to 60 psu). In addition, a second series of experiments were conducted to examine possible interactions between salinity, temperature, and pH. Experimental salinity media were prepared by adding Instant Ocean nutrient-free salt to demineralized water; pH was reduced with H2SO4. Because prolonged times in closed systems can change the seawater pH, it was measured at the beginning and end of each incubation.

Net oxygen evolution was measured with a Hansatech oxygen electrode system (Clark type) calibrated with N2 and air-saturated water. Two randomly selected leaves of H. johnsonii were placed in the closed reaction chamber with 2 ml of stirred treatment solution and allowed to equilibrate in the dark for 10 min. Before each incubation, water in the electrode chamber was sparged with nitrogen to reduce the initial oxygen concentration to about 25% of saturation because photosynthesis can be negatively affected by elevated concentrations of O2 (Beer 1989). After equilibration, an initial oxygen consumption reading over a 2 min interval was made to determine the dark respiration rate. Leaves were then subjected to 15 light levels, in increasing order, ranging from 10 to 823 μmol quanta m−2 s−1 of PAR, with 1 min of equilibration between each 2 min photosynthetic measurement interval. Irradiance was provided by a two-armed fiber-optical halogen light source and measured with a LiCor quantum sensor. Temperature was controlled with a recirculating water bath and maintained at 25°C, the ambient temperature, but was increased (35°C) or decreased (15°C) for the experiments testing interactions between salinity and temperature. Photosynthesis–irradiance curves were performed in triplicate for each treatment using different pairs of leaves. Each replicate was randomly chosen during the mid-day portion of the diurnal cycle (10:00–17:00 h) to minimize endogenous rhythm effects. The PI curve parameters: the initial slope (α), light-saturated gross photosynthesis (P max), light-saturation photosynthesis point (I k), compensation point (I c), and respiration rate were estimated for each PI run according to Durako et al. (1993). Photosynthetic and respiration rates were normalized to the weight of leaves after drying at 70°C for 48 h and expressed in μmol O2 g−1 dw h−1 . I c and I k were expressed in μmol quanta m−2 s−1 and α in (μmol O2 g−1 dw h−1)×(μmol quanta m−2 s−1)−1 .

One photosynthetic experiment was carried out in 310 ml biological oxygen demand (BOD) glass bottles that contained rhizome segments with four pairs of H. johnsonii leaves. Treatments consisted of two different salinities (30 and 40 psu) with four pH values (5, 6, 7, and 8.2). Triplicate bottles containing plants and two control bottles (without plants) were used for each treatment. During the experiment all bottles were placed in a waterbath at 22–25°C. At the beginning and end of the incubation period, seawater-dissolved oxygen (DO) was measured using a YSI DO meter (Model 57) and compared to that of control bottles. Change in DO concentration over a 15 h experimental dark period, coinciding with the night hours, was used to determine dark respiration. Plants were then exposed to light (470 μmol quanta m−2 s−1 PAR) for 5 h to estimate the photosynthetic rate. We used these longer incubation times due to the small amount of plant tissue in each bottle (0.0189–0.0377 g dw). At the end of the experiment plants were removed from the bottles, dried at 60°C for 48 h, and subsequently weighed. Photosynthetic and respiratory rates were expressed in mg O2 g−1 dw h−1 .

Statistical analyses

Growth and mortality rates were evaluated using a two-factor ANOVA after testing for homogeneity of variance by Cochran’s test (Underwood 1997). The experimental design considered salinity to be a fixed factor, with seven treatments (0, 10, 20, 30, 40, 50, and 60 psu) in two different aquariums (a nested and a random factor) and with four plants (replicates) in each aquarium, so the linear model of sources of variance was defined as the following:

$$X_{ijn} = \mu + {\text{Salinity}}_i + {\text{Aquarium (Salinity)}}_{j{\text{(}}i{\text{)}}} + {\text{Residual}}_{n{\text{(}}ij{\text{)}}} $$

Photosynthetic parameters were evaluated using a one-way ANOVA to assess salinity treatment effects (Eq. 1) or a two-factor orthogonal ANOVA to determine possible interactions between salinity and temperature or pH (Eq. 2).

$$X_{ijn} = \mu + {\text{Salinity}}_i + {\text{Residual}}_{n{\text{(}}ij{\text{)}}} $$
(1)
$$X_{ijn} = \mu + {\text{Salinity}}_i + {\text{Factor}}_j + {\text{Salinity}} \times {\text{Factor}}_{ij} + {\text{Residual}}_{n{\text{(}}ij{\text{)}}}$$
(2)

When analysis of variance identified a significant difference for any factor, the post hoc test SNK (Student–Newman–Keuls) was applied to determine specific treatment differences. All calculations were performed using the GMAV.5 program (University of Sydney, Underwood and Chapman 1997).

Results

Leaf production of H. johnsonii was affected by salinity. The relationship between growth and salinity is well described by a hyperbolic curve (Fig. 1), with an optimum value (0.33 leaves plant−1 day−1) at a salinity of 30 psu, a significant reduction in higher and lower salinities (Table 1), and no growth at 0 and 60 psu. Mortality showed an opposite pattern (Fig. 1), with a minimum value (37.5%) at 30 psu. All the plants died when exposed to freshwater (0 psu) and to the highest salinity concentration (60 psu).

Fig. 1
figure 1

Halophila johnsonii growth (leaf production per plant and day) and percent mortality at different salinities. Means with different letters are statistically different at P<0.05 (bars represent standard error)

Full size image
Table 1 Summary of the ANOVA testing the effects of salinity treatments (0–60 psu at 10 psu increments) on growth and mortality of H. johnsonii
Full size table

Salinity also significantly affected the photosynthetic characteristics of this species (Table 2). All photosynthesis–irradiance curves exhibited a typical response (Figs. 2, 3, 4), with a rapid initial linear increase in oxygen release at low irradiances (<100 μmol quanta m−2 s−1), followed by a saturated photosynthetic rate at higher light intensities (>100 μmol quanta m−2 s−1) and, in some cases, a photoinhibition (>200 μmol quanta m−2 s−1). Estimated photosynthetic efficiency at subsaturating light (α), maximum photosynthetic rate (P max), and light-saturation point (I k) showed higher values at salinities between 30 and 50 psu, with a maximum value at 40 psu for α and P max (Fig. 5). P max was totally depressed when plants were subjected to a salinity of 0 psu. Respiratory rates were more elevated at lower salinities (≤30 psu), decreasing at 40–60 psu (Fig. 5). The compensating irradiance (I c) was maximum in freshwater, but no significant differences were found for other salinity treatments (Fig. 5).

Table 2 Summary of the one-way ANOVA testing the effects of salinity treatments (0–60 psu at 10 psu increments) on photosynthetic parameters
Full size table
Fig. 2
figure 2

Photosynthesis versus irradiance responses of Halophila johnsonii leaves exposed to different salinities. Symbols represent the average of three replicates

Full size image
Fig. 3
figure 3

Photosynthesis versus irradiance responses of Halophila johnsonii leaves exposed to different salinities and temperatures. Symbols represent the average of three replicates. Diamonds represent temperature =15°C, circles T=25°C, and squares T=35°C. Filled symbols salinity=30 psu and open symbols salinity=50 psu

Full size image
Fig. 4
figure 4

Photosynthesis versus irradiance responses of Halophila johnsonii leaves exposed to different salinities and pH values. Symbols represent the average of three replicates. Diamonds represent pH=5, circles pH=6, and squares pH=8.2. Filled symbols salinity=30 psu and open symbols salinity=50 psu

Full size image
Fig. 5
figure 5

Photosynthetic parameters at different salinities [α (μmol O2 g−1 dw h−1)×(μmol quanta m−2 s−1)−1, P max and respiration rate (μmol O2 g−1 dw h−1), I c and I k (μmol quanta m−2 s−1)]. Means with different letters are statistically different at P<0.05 (bars represent standard error)

Full size image

No interactions between salinity and temperature were observed (Table 3). P max, α, and I k were significantly different as a function of temperature, with lower values at 15°C compared with those obtained at higher temperatures (Fig. 6). Dark respiration was also significantly affected by temperature, but with a similar response for 15 and 25°C and highest values for 35°C. Although temperature effects on I c were not significant, its response patterns were similar to those of other parameters, increasing with higher temperatures (Fig. 6).

Table 3 Summary of the two-way ANOVA testing the effects of salinity (30 and 50 psu) and temperature (15, 25, and 35°C) treatments on photosynthetic parameters
Full size table
Fig. 6
figure 6

Photosynthetic parameters (average and standard error) at different salinities and temperatures [α (μmol O2 g−1 dw h−1)×(μmol quanta m−2 s−1)−1, P max and respiration rate (μmol O2 g−1 dw h−1), I c and I k (μmol quanta m−2 s−1)]. Means with different letters are statistically different at P<0.05

Full size image

A significant interaction (P<0.05) was observed for P max between salinity and pH (Table 4): P max increased with pH=5 at 30 psu, but not for other photosynthesis–irradiance parameters. α showed a significant increase with low pH treatments. No significant differences were found in dark respiration, I c, and I k, but the trend was to obtain higher values at the lowest pH (Fig. 7). In the BOD bottles experiment, significant interactions between salinity and pH were not detected (Table 5). Although, there were no significant differences in photosynthetic rates, this parameter increased with lower pH values at 30 psu (Fig. 8). At the same salinity, respiration rates decreased significantly with pH reduction, but at 40 psu this pattern was not so clear (Fig. 8).

Table 4 Summary of the two-way ANOVA testing the effects of salinity (30 and 50 psu) and pH treatments (5, 6, and 8.2) on photosynthetic parameters
Full size table
Table 5 Summary of the two-way ANOVA testing the effects of salinity (30 and 50 psu) and pH treatments (5, 6, 7, and 8.2) on photosynthesis and dark respiration rates obtained with BOD bottles
Full size table
Fig. 7
figure 7

Photosynthetic parameters (average and standard error) at different salinities and pH values [α (μmol O2 g−1 dw h−1)×(μmol quanta m−2 s−1)−1, P max and respiration rate (μmol O2 g−1 dw h−1), I c and I k (μmol quanta m−2 s−1)]. Means with different letters are statistically different at P<0.05

Full size image
Fig. 8
figure 8

Photosynthesis and dark respiration (mg O2 g−1 dw h−1) rates (average and standard error) obtained with BOD bottles at different salinities (30 and 40 psu) and pH values (5, 6, 7, and 8.2)

Full size image

Discussion and conclusions

Under laboratory conditions, H. johnsonii growth was similar to values recorded in the field (Kenworthy 1997), but mortality was relatively high in the control treatments (37.5%). This was probably due to the experimental manipulations and the excision of plants into short segments. Despite this, significant differences in growth and mortality with salinity treatments were observed. Maximum growth occurred at 30 psu, but plants also showed positive growth at salinities between 20 and 50 psu. In addition, mortality values were lower for the range of salinities between 20 and 50 psu, while all plants died at 0 and 60 psu. Similar responses, but with different salinity-tolerance ranges, have been observed in several seagrasses; hypo- and hypersalinities have caused significant reductions in growth (McMillan and Moseley 1967; Walker 1985; Walker and McComb 1990; Kamermans et al. 1999) and survival (McMillan and Moseley 1967; Pinnerup 1980; Wortmann et al. 1997; van Katwijk et al. 1999; Vermaat et al. 2000). The cause of these effects may be due to toxicity by salt excess (Zhu 2001), elevated metabolic cost to maintain internal ionic balance (Sibly and Calow 1989), or negative alterations in the photosynthetic and respiratory rates (Ogata and Matsui 1965; Biebl and McRoy 1971; Kraemer et al. 1999).

Effects of salinity on photosynthetic activity were also observed in the PI curves, as reflected by differences in α, P max, and I k, which all had highest values at 40 psu. The greatest decrease in photosynthetic activity occurred in freshwater. This decrease may be due to critical changes in intracellular ionic concentrations following an osmotic shock, in which a loss of ions required as cofactors in photosynthesis occurs (Simon et al. 1999). Minimum respiration rates corresponded with elevated salinities (40–50 psu), but data were highly variable and did not show a clear trend. This is consistent with previous studies on H. johnsonii, Zostera marina, Z. muelleri, and H. ovalis, which have suggested that photosynthetic rates are more sensitive to salinity than are respiration rates (Biebl and McRoy 1971; Kerr and Strother 1985; Dawes et al. 1989; Ralph 1998; Hellblom and Björk 1999). In general, H. johnsonii exhibited low compensation (I c) and saturation points (I k) with elevated photosynthetic efficiencies (α), compared to those observed for other seagrasses (Hemminga and Duarte 2000; Touchette and Burkholder 2000). This could be due not only to a species-specific response but also to the fact that these plants were acclimated to low-irradiance greenhouse conditions for over 1 year prior to the experimentation (Major and Dunton 2000). Moreover, it has been observed that when PI parameters are calculated in laboratory conditions they are significantly different from those obtained in the field; values of α tend to be overestimated and I k and I c are underestimated when using laboratory-incubated plants (Major and Dunton 2000).

H. johnsonii is considered a rapid-growth plant with a reduced storage capacity, which implies that growth and photosynthetic rates should be closely coupled. However, in the present work a discrepancy between Pmax and growth responses to salinity was observed. The highest Pmax was observed at 40 psu while maximum growth was obtained at 30 psu. The differences may be a consequence of the differing exposure times, the different acclimation period used in the growth experiment, or the effect of using detached leaves for the PI curves versus whole plant segments for the growth measurements. They may also be due to toxicity of elevated salinities on meristematic tissues, affecting the growth of these plants (Zhu 2001). Thus, photosynthetic behavior could be a useful acute indicator of the response of a plant to a stress condition. However, a more comprehensive approach to determine the ecological salinity tolerance of a species would be to combine photosynthesis measurements with experiments to estimate growth or survival rates.

Our results suggest that H. johnsonii tolerates hypersaline conditions better than hyposaline conditions, although most seagrasses are thought to be more sensitive to increased salinity (Ogata and Matsui 1965; Biebl and McRoy 1971; Zieman 1975; Adams and Bate 1994; Doering and Chamberlain 1998; Kamermans et al. 1999; van Katwijk et al. 1999). Our results are consistent with Dawes et al. (1989), in which the photosynthetic responses of H. johnsonii and H. decipiens to salinity and temperature variations were compared. They concluded that H. johnsonii showed a broader tolerance than H. decipiens to temperature and salinity fluctuations.

Osmotic stress can change the sensitivity of seagrasses to other environmental conditions, such as temperature increments, elevated light levels, or eutrophication (Biebl and McRoy 1971; Ralph 1999; van Katwijk et al. 1999; Vermaat et al. 2000). Biebl and McRoy (1971) found an increase in heat plasmatic resistance with increasing salinities for the subtidal forms of Z. marina, while Vermaat et al. (2000) observed increased mortality in Z. noltii when the temperature reached 20°C at higher experimental salinities. However, in the present work no interactions were detected in the photosynthetic responses of H. johnsonii in response to combined salinity and temperature treatments.

Temperature is considered an important abiotic factor that can affect the metabolism of marine angiosperms (Drew 1979). Laboratory studies with some seagrasses have detected increasing photosynthetic and respiration rates at higher temperatures, with the maximum photosynthetic rates approximately doubling for every rise of 10°C. For this study, all the PI parameters, except I c, increased significantly with temperature. H. johnsonii showed a photosynthetic optimum at 25–35°C, the same optimal range that has been reported for several temperate seagrasses (Biebl and McRoy 1971; Drew 1979; Bulthuis 1987; Marsh et al. 1986).

In our work a slight interaction between salinity and pH for P max was observed, but not for the remaining PI parameters. P max increased significantly with the lowest pH treatment (pH=5) at 30 psu. Apart from minor direct effects of pH on photosynthesis by affecting the electrochemical gradients and proton flow across plasma membranes (Touchette and Burkholder 2000; Beer et al. 2001), the main effect of pH variation is that it alters the concentrations of the different dissolved inorganic carbon (DIC) forms (Beer et al. 1977). A decrease in pH in a closed system is expected to cause an increase in CO2 concentration and subsequently an increase in the photosynthetic rates of seagrasses (Beer and Waisel 1979; Invers et al. 1997). In the PI experiments, P max and α values increased significantly when pH decreased (pH=5). pH reduction did not affect dark respiration, I c, and I k, similar to other studies (Invers et al. 1997). However, estimates of these parameters were highly variable. The results from the BOD bottle incubations showed that reduction in pH produced a significant decline in dark respiration, and although there were no significant differences in photosynthetic rates, an increase of this parameter with lower pH values at 30 psu was observed. The differing responses observed between the PI curve experiments and the bottle incubations may be due to the different exposures of plants to the treatments. The first experiments were only of 1 h duration and two excised leaves were used, while in the second experiment four paired leaves with corresponding rhizomes were incubated for 20 h. In addition, the closed chambers contained a small volume of media with the seagrass leaves oriented perpendicularly to the incident light, whilst the bottles used more diffuse light conditions in their incubations (Alcoverro et al. 1998).

Despite the potential limitations of these experiments, the results of this study show that the endemic seagrass H. johnsonii could be seriously affected by salinity alterations produced by human activities, such as freshwater management practices or the brine discharges from seawater desalination plants. However, salinity changes did not seem to alter the tolerance of this species to other environmental factors, such as temperature or pH.