Average values of leaf dimensions from adult leaves collected at the different sites.
Mangrove trees of the salt secreting Avicennia germinans and the non-secreting Rhizophora mangle were investigated at the northern coast of Venezuela at a low salinity site (127 mmol kg−1) and two hypersaline sites (1600–1800 mmol kg−1). Leaf sap osmolality and mass/area ratio of both species were positively correlated, while size was negatively correlated with soil salinity. Leaf sap osmolality was always higher in Avicennia and exceeded soil solution osmolality. Salinity increased the concentration of 1D-1-O-methyl-muco-inositol (OMMI) in Rhizophora and glycinebetaine in Avicennia. The latter could make up to 21% of total leaf nitrogen (N). Nitrogen concentration was higher in Avicennia, but subtracting the N bound in glycinebetaine eliminated interspecific differences. Photosynthetic rates were higher in Avicennia, and they decreased with salinity in both species. Leaf conductance (gl) and light saturated photosynthesis (Asat) were highly correlated, but reduction of gl at the hypersaline sites was more pronounced than Asat increasing water use efficiency in both species. Lower values of 13C discrimination at the hypersaline sites evidenced higher long-term water use efficiency. Apparent quantum yield and carboxylation efficiency decreased with salinity in both species. Rhizophora was more sensitive to high salinity than Avicennia, suggesting that glycinebetaine is a better osmoprotectant than OMMI.
- Rhizophora mangle
- Avicennia germinans
- soil salinity
- compatible solutes
Mangrove species in the neotropics are found along large latitudinal ranges including dry and wet coastal environments [1, 2]. Their distribution along steep salinity gradients provide an opportunity to test in situ the impact of soil salinity on osmolyte accumulation in plant tissues, and analyze under similar light, temperature, and air humidity conditions, on their leaf development and photosynthetic performance.
The structural development and complexity of mangrove communities including height, leaf area index, leaf size, stem diameter, branching, litter production, and productivity are inversely related to interstitial soil water salinity. In neotropical mangroves, these properties have been shown to be strongly correlated [3, 4, 5, 6, 7, 8].
Photosynthesis decreases significantly with salinity of interstitial water in several mangrove species [7, 8, 9, 10, 11]. Some species appear to be more sensitive to soil salinity than others, a characteristic that may be associated with specific metabolic and structural properties such as synthesis of compatible solutes, root permeability, salt excretion, and compartmentation of excess ions [12, 13, 14, 15, 16, 17].
We studied the differentiation in leaf morphology, accumulation of osmotically active solutes and the photosynthetic response of two mangrove species, Rhizophora mangle L. and Avicennia germinans (L.) Stearn, reportedly differing in their salt tolerance [18, 19, 20] (Figure 1). These species coexist in neotropical mangroves and differ in their mechanisms of salt tolerance. Rhizophora mangle is considered as salt excluder, dominant in fringe mangroves throughout the neotropics, whereas A. germinans possesses numerous salt secreting glands in their leaves and typically dominates basin mangrove vegetation [20, 21]. Measurements were conducted under field conditions, in contrasting environments regarding fresh water availability and salt concentration of the soil interstitial water. Our objective was to assess quantitatively the impact of high salinity environments on photosynthetic performance and leaf expansion in association with inorganic and organic osmolyte accumulation.
In this chapter, we present results relating leaf sap osmolality and concentration of compatible solutes (cyclitols and glycinebetaine) to leaf morphology and patterns of gas exchange. The compatible solutes are presumably accumulated in the cytoplasm and counteract the osmotic effect of inorganic ions predominantly accumulated in the vacuole . Accumulation of these compounds requires energy and carbohydrates from photosynthesis, and in the case of glycinebetaine, it needs additional amounts of N. The latter probably affects photosynthesis through the reduction of N availability for synthesis of photosynthetic enzymes.
Our hypotheses for this study were: (1) the reduction of photosynthesis resulting from salt accumulation in leaf cell sap is stronger in the species assumed to have lower salt tolerance,R. mangle; (2) salinity affects photosynthesis through diminished nutrient uptake, such as N, affecting protein synthesis, and phosphorus (P), possibly affecting N use efficiency; and (3) water use efficiency is higher in the high salinity (drier) environment, as a result of the combined effect of increased cell sap and interstitial soil water osmolalities on leaf conductance, leading to a proportionally larger reduction of transpiration compared to photosynthesis.
2. Study sites
Field work was carried out at two locations in the Caribbean coast of northern Venezuela, both in the State of Falcón. The site, further on called Tacuato, is a low stature mangrove stand (<5 m tall) of the species Rhizophora mangle and Avicennia germinans growing in a hypersaline lagoon (salinity >1000 mmol kg−1 ≈ 35 ppt) located south of the village Tacuato on the Paraguaná peninsula (11°41′40″N, 69°49′52″W). The climate is dry (<400 mm rainfall) with one rainy season from September to December. The lagoon has access to the gulf of Tacuato with an average salinity of 45 ppt (1600 mmol kg−1). The water depth of the lagoon in the sampling area varied between 0 and 20 cm, depending on rainfall events and tides. Diurnal air humidity was about 70–80%, whereas day air temperatures ranged from 27 to 37°C. The tallest trees of both species in the middle part of the lagoon reached a height of 5 m. Trees used for measurements were smaller, but mature, as they were flowering and fruiting. The site was divided into two sub-sites: the fringe-region and an inner site nearly 20 m apart from the fringe, differing in their average osmolality of interstitial soil water (1600 and 1800 mmol kg−1, respectively).
The second study site, further on called Ricoa, is located at the fringes of the estuary of the Ricoa River west to the village of Tocópero (11°30′21″N, 69°12′19″W). Annual precipitation is about twice that of Tacuato (970 mm in average) with peaks in May–July and November–December. The soil water salinity averaged 127 mol kg−1 (2–3 ppt), the diurnal air humidity was about 70–80%, like that at Tacuato; day air temperatures were in general lower with highest values around 33°C. Reduced soil salinity was a consequence of higher rainfall and the contribution of the river water run-off. At this site, the trees used for measurements were located at the estuary flood plain, and had approximately the same height as the plants used in Tacuato. Measurements and sample collection of R. mangle and A. germinans (from now on designated by their genus names) were carried out during seven field trips distributed over 9 months (from October to June), thus including dry and rainy seasons at both sites.
3. Materials and methods
3.1. Interstitial water
Interstitial water was sampled by digging 10–20 cm into the mud with a perforated plastic tube. Water salinity was determined in situ with a refractometer (ATAGO) calibrated with distilled water just before the measurement. Osmolality of the sampled water was calculated from salinity (in ppt) as in .
3.2. Gas exchange measurements
Gas exchange measurements were carried out with an open IRGA system of the type LCA 3 (ADC3, Analytical Development Co.) combined with a Parkinson leaf chamber of 6.25 cm2. A photometer and a thermocouple attached to the chamber allowed the measurement of incoming light intensity and leaf temperature. Photosynthetic rates used for correlations with leaf conductance (gl), and the concentrations of N and chlorophyll, were measured under natural conditions at saturating intensities of photosynthetic active radiation (PAR) ≥ 1000 μmol m−2 s−1 (Asat). Leaves were oriented at 90° to the incoming radiation during measurements. To obtain a range of quantum fluxes, leaves were shaded in the field by a set of fine wire nets. The wire nets covered the photosynthesis chamber until gl stabilized (2–3 min). A light response curve was a composite of measurements conducted on four leaves. Curves were fitted to the data using Sigmaplot 2.01 (Jandel Corporation 1994) and the following equation [23, 24]:
where Q is the measured quantum flux and A is the rate of photosynthesis. By this procedure, we obtained the maximum photosynthetic capacity at saturating light intensity (Asat) and the apparent quantum yield (φ).
To obtain different leaf internal CO2 concentrations (ci), the concentration of CO2 in the air entering the leaf chamber was reduced stepwise below ambient by passing a part of the air flow over soda lime. Photosynthetic rates were found to be higher in the second and third leaves below the branch apex, and these leaves were used for all measurements. Photosynthesis was measured during late morning and early afternoon (10–15 hours).
3.3. Chemical analyses of the samples
After gas exchange measurements, leaves were detached and gently cleaned with a wet tissue to remove salt from their surfaces. About 7–10 leaves were used to obtain one sample. Petioles and midribs were removed. Every leaf was cut into halves of which one was put into a plastic syringe (for leaf sap extraction) and the other was put into a plastic bag (for the determination of chlorophyll, N, and P). The samples prepared in that way were immediately frozen on dry ice. Upon returning to the laboratory, they were stored in a freezer at −5°C.
Fresh mass was determined in the field by a battery powered balance (precision ±0.01 g). Samples were dried at approximately 70°C in a ventilated oven until constant weight. Total chlorophyll (a + b) (Chlortot) concentration of leaf disks was measured by spectrophotometry of acetone extracts . Syringes containing the samples were thawed, and leaf sap was squeezed out with a pressure device . Osmolality of the leaf sap was determined with a dew point osmometer (WESCOR 5500). Total P concentration was measured in acid digested dry leaf material following the procedure of Murphy and Riley . Nitrogen concentration was measured using a standard microKjeldahl procedure . These measurements were contrasted with the parallel analysis of calibrated leaf material (peach leaf or citrus leaf, National Institute of Standards and Technology, USA). Sample preparation for organic compounds analyses and the chromatographic determinations of cyclitols in Rhizophora mangle and glycinebetain in Avicennia germinans have been described in detail elsewhere [14, 28, 29]. Measurement of carbon isotope ratios (δ13C values) of leaf material was performed at the Institute of Botany and Microbiology, University of Munich following standard procedures described elsewhere .
3.4. Statistical analysis of data
Significant differences between means of species and sites were tested with a one-way analysis of variance (ANOVA) and a multiple range test after Scheffé. Differences were considered significant when P ≤ 0.05. Differences between means of measured parameters in the two species at the same site were tested with students t-test at the level of P = 0.05. All statistical analyses were done using Statgraphics 5.0.
4.1. Leaf morphology
Leaves of Rhizophora showed large differences between sites. They were thin and green at Ricoa, but showed a leathery texture and a yellowish color at the two hypersaline sites. At the Tacuato sites, leaves were smaller, had their edges bent downward, and showed an angle well above 45° from the horizontal. In Avicennia, differences in leaf morphology were not that obvious, but leaf inclination was also more pronounced in the hypersaline sites. In Avicennia, crystals of secreted salt could be observed on the leaf surface at both Tacuato sites.
Leaf length (L) in both species was reduced in the hypersaline sites, while leaf width (W) was reduced only in Rhizophora (Table 1). As a result, Avicennia leaves from hypersaline sites tended to be rounder than those of the low salinity site (smaller L/W ratio).
|Site||n||Length (cm)||Width (cm)||L/W|
|Ricoa||21||12.1 a,0||5.6 a,0||2.2 a,0|
|Tacuato-fringe||78||9.8 b,0||4.3 b,0||2.3 a,0|
|Tacuato-lagoon||41||8.6 c,0||3.3 c,0||2.6 a,0|
|Ricoa||27||7.9 A,1||2.6 A,1||3.1 A,1|
|Tacuato-fringe||85||5.4 B,1||2.6 A,1||2.1 B,1|
|Tacuato-lagoon||87||5.1 B,1||2.5 A,1||2.1 B,1|
For both species, average area of a single fully expanded leaf was greater at the low salinity site (Table 2). Leaves from the two hypersaline sites differed in size only for Rhizophora. Reduction in leaf area from low to high salinity site was more pronounced for Rhizophora (37–59%) than for Avicennia (26–34%). Leaf dry mass decreased significantly in hypersaline sites only in the case of Rhizophora, but fresh mass decreased in both species. The fresh mass/dry mass ratio was higher for Rhizophora at all sites. Leaf dry mass/area ratios were significantly lower for both species at the Ricoa site, and the differences between species within sites were only significant in the case of Tacuato-lagoon (Table 2).
|Species and Sites||n||Area||Dry mass||Fresh mass||Dry mass/Area||Fresh/Dry mass|
|cm2||g||g||g m−2||g g−1|
|Ricoa||11||50.9 a,0||0.92 1,0||2.89 a,0||180 a,0||3.16 a,0|
|Tacuato-fringe||21||32.1 b,0||0.77 b,0||2.11 b,0||239 b,0||2.73 b,0|
|Tacuato-lagoon||15||21.1 c,0||0.50 c,0||1.28 c,0||237 b,0||2.58 c,0|
|Ricoa||9||14.8 A,1||0.28 A,1||0.75 A,1||187 A,0||2.70 A,1|
|Tacuato-fringe||20||10.9 B,1||0.27 A,1||0.68 AB,1||250 B,0||2.41 B,1|
|Tacuato-lagoon||14||9.7 B,1||0.26 A,1||0.63 B,1||267 B,1||2.38 B,1|
4.2. Osmotic adaptation to salinity of the soil solution
Salinity of interstitial water differed by more than one order of magnitude between the low and high salinity sites (Table 3). Within the high salinity site, the Tacuato-lagoon showed always larger osmolalities than the Tacuato-fringe, because the former was not always in contact with the bay water, so that concentration through evaporation could not be compensated by tides.
|Site||n||Osmolality soil solution||Rhizophora mangle||Avicennia germinans|
|n||Osmolality of leaf sap||∆ sap-soil||n||Osmolality of leaf sap||∆ sap-soil|
|Ricoa||3||127a||11||1037 a,0||914 a,0||9||1226A,1||1103A,1|
|Tacuato-fringe||6||1666b||21||1631 b,0||−63 b,0||20||1859B,1||180B,1|
Leaf sap osmolality was also higher for both species in the high salinity sites, particularly at Tacuato-lagoon, but absolute values were only 1.5–1.8 times higher than at Ricoa. Average data of Table 3 show that osmolality was well above 100 mmol kg−1 higher in Avicennia than in Rhizophora for all values of soil salinity. The leaf sap-soil osmolality difference in Rhizophora decreased from nearly 900 mmol kg−1 in Ricoa to negative values approaching 100 mmol kg−1 in Tacuato. In Avicennia, the reduction of Δ leaf sap-soil was also very strong, but average values were always positive. The sap-soil differences were larger in Avicennia, and the differences between sites were all significant.
4.3. Concentration of compatible solutes
In Rhizophora, the main compatible solute is the cyclitol 1D-1-O-methyl-muco-inositol (OMMI) . Its concentration in the leaf sap increased with osmolality from nearly 80 mmol L−1 in Ricoa to about 160 mmol L−1 at Tacuato (Table 4). The other cyclitols present in Rhizophora (L-quebrachitol, L-chiro-inositol, and D-pinitol) were present only as minor components.
|Ricoa||11||77.2 a||88.1 a|
|Tacuato-fringe||19||125.4 b||141.3 b|
|Tacuato-lagoon||15||159.4 c||172.6 c|
|Avicennia germinans||Glycinebetaine mmol L−1||GB N/Total N %|
|Ricoa||9||120.1 A||14.8 A|
|Tacuato-fringe||20||165.8 B||20.1 B|
|Tacuato-lagoon||14||178.1 B||21.1 B|
The only compatible solute in Avicennia is the quaternary ammonium compound glycinebetaine . It reached concentrations of about 120 mmol L−1 at Ricoa to 180 mmol L−1 at Tacuato. Glycinebetaine contained between 15 and 21% of total leaf N with higher values found at the hypersaline sites (calculated with values from Tables 4 and 5). At all sites, concentrations of glycinebetaine in the leaf sap of Avicennia were higher than those of cyclitols in Rhizophora.
|μmol g−1||μmol g−1||mmol m−2||molar||μmol m−2||mmol mol−1||mmol mol−1|
|RICOA||(5) *||31 a,0||(11)**||910 ab,0||161 a,0||29||328 a,0||2.04a,0|
|TACUATO-fringe||(5)||33 a,0||(21)||817 a,0||195 b,0||25||325 a,0||1.67b,0|
|TACUATO-lagoon||(5)||31 a,0||(15)||1017 b,0||240 b,0||33||383 b,0||1.61b,0|
|RICOA||(4)||50 A,1||(9)||1365 A,1||258 A,1||27||467 A,1||1.82A,1||2.1|
|TACUATO-fringe||(5)||52 A,1||(20)||1208 B,1||304 B,1||23||359 B,1||1.19B,1||1.5|
|TACUATO-lagoon||(5)||50 A,1||(14)||1204 B,1||324 B,1||24||411 c,0||1.29B,1||1.6|
4.4. Total phosphorus, nitrogen, and chlorophyll concentrations
Both total P and N concentrations per unit leaf mass were higher in Avicennia than in Rhizophora (Table 5). In both species, no differences in total P concentration between sites were detected. However, concentrations of P and N per unit leaf area increased with salinity in both species because of the higher leaf mass/area ratios. The N to P molar ratios varied between 23 and 29 in both species, suggesting that P was not a limiting nutrient in these soils.
Concentration of Chltot per leaf area at a given site was always higher in Avicennia than in Rhizophora, but there was not a clear pattern relating chlorophyll concentration with salinity (Table 5). The ratio Chlortot/N was higher in Rhizophora, but the differences disappear if the amount of N invested in glycinebetaine is subtracted, suggesting similar N allocation to photosynthetic structures. In both species, this ratio decreases significantly in hypersaline sites.
4.5. Photosynthetic response to light intensities and internal CO2 concentration
Light response curves for both species showed a clear reduction in light saturated photosynthesis in hypersaline sites, 36% in Rhizophora and 21% in Avicennia (Figure 2). Avicennia had higher rates of Asat than Rhizophora at low and high salinity sites. The reductions in photosynthetic light use efficiency (Φ) reached 45% in Rhizophora and nearly 37% in Avicennia. Light compensation points were not so much affected by salinity, although the number of curves measured (4) does not allow a definitive conclusion.
Photosynthetic response to increasing intercellular CO2 was measured at CO2 concentrations near that of ambient air (≈350 ppm) and below (Figure 3). Hence, the transition from the linear part of the curve to the plateau of CO2 saturation was not reached. The initial slope of the curve, representing carboxylation efficiency, was steeper in Avicennia, and in this species compensation values were lower than in Rhizophora. The carboxylation efficiency was reduced at the hypersaline site by 39% in Rhizophora and 26% in Avicennia, compared to that at the freshwater site, while compensation values were nearly unchanged.
4.6. Average values of gas exchange parameters
Average light intensities and temperature recorded during the measurement of Asat under natural conditions were similar for both species at each site, but the temperature was higher at the hypersaline sites (Figure 4). At any given site, Asat was higher in Avicennia than in Rhizophora. Differences between sites were significant for both species. At Tacuato-fringe, the values of Asat in both species were only about 70% of those measured at Ricoa. At the Tacuato-lagoon site, Asat was even lower, especially in Rhizophora.
Leaf conductance to water vapor (gl) showed a similar pattern to that of Asat; however, relative differences between the low salinity site and the hypersaline sites were more pronounced in the former. Differences between the two hypersaline sites were significant in Rhizophora only (Figure 4). Intrinsic water use efficiency (A/gl) was higher in both species at the hypersaline sites, maximum values corresponding to Rhizophora at Tacuato.
4.7. Carbon isotope discrimination
As expected, values fo δ13C followed a pattern opposite to A/gl (Figure 4). 13C discrimination (Δ) was calculated from leaf δ13C values using formulation of Farquhar et al.  (δ13C air = −8‰). Discrimination values were lower at the hypersaline sites for both species, whereas they showed no significant differences at the hypersaline site, between Tacuato-fringe and Tacuato-lagoon. Differences between species were only significant at the Ricoa site, with Rhizophora having higher ∆ values. This pattern in Ricoa was expected as Rhizophora showed distinctly higher leaf conductance.
4.8. Efficiency of photosynthetic resource use
Mass-based assimilation rate can be used as a measure of the biomass use efficiency in photosynthesis. As previously shown for Asat per unit area, at any given site, Avicennia also showed higher assimilation rates per unit leaf dry mass than Rhizophora (Table 6). Comparing sites, values were higher at the low salinity site and decreased with salinity.
|μmol CO2 g−1 s−1||mmol mol−1||μmol CO2 mol−1 s−1|
|Ricoa||11||51.6 a,0||1.5 a,0||56.7 a,0||—||27.7 a,0|
|Tacuato-fringe||21||28.3 b,0||2.5 b,0||34.6 b,0||—||20.9 B,0|
|Tacuato-lagoon||15||18.6 c,0||2.8 c,0||18.4 c,0||—||11.7 C,0|
|Ricoa||9||68.5A,1||1.7 A,0||49.7 A,0||58.8 A,0||27.2 A,0|
|Tacuato-fringe||20||35.0B,1||2.3 B,1||28.7 B,0||34.6 B,0||24.1 A,1|
|Tacuato-lagoon||14||29.4B,1||2.2 B,1||24.3 B,1||28.1 B,1||18.8 B,1|
Water use efficiency was estimated as the ratio of A to transpiration E (calculated using leaf conductance, ambient relative humidity, and temperature). The ratio A/E showed the same pattern of intrinsic water use efficiency for short-term water use, that is, similar for both species at Ricoa, but higher for Rhizophora in the hypersaline sites.
Nitrogen use efficiency (Asat/N) was lower in Avicennia at all sites, when using total N concentration as basis for calculation. When the amount of N bound in glycinebetaine from total N in Avicennia is subtracted, differences between the species disappeared at Ricoa and Tacuato-fringe, but at Tacuato-lagoon, Rhizophora still showed a lower A/N index.
4.9. Relationships between leaf gas exchange and specific leaf area, and osmotic properties
We calculated correlations for a complete subset of data including Asat on a dry mass basis, leaf conductance, specific leaf area, osmolality, and N. For these calculations, average values of photosynthesis of the leaves pooled for chemical analyses were used. In all cases, higher correlations were found when using dry mass as a reference basis.
In both species, leaf conductance, specific leaf area, and N concentration were positively, whereas osmolality was negatively, and significantly correlated with photosynthetic rate (Table 7). The N-photosynthesis correlation in Avicennia increased from 0.67 to 0.79 when the glycinebetaine-N was subtracted from total N.
|Photosynthesis||Rhizophora mangle||Avicennia germinans|
|μmol CO2 kg−1 dry mass. s−1 vs||(n = 47)||(n = 43)|
|Leaf conductance (mol m−2 s−1)||0.91***||0.90***|
|Specific leaf area (m−2 g−1 dry mass)||0.79***||0.88***|
|Osmolality (mmol kg−1)||−0.79***||−0.81***|
|Nitrogen (μmol g−1 dry mass)||0.78** (n = 11)||0.67***|
|N-glycinebetaine (μmol g−1 dry mass)||—||0.79***|
5.1. Leaf morphology, leaf size, and leaf water content
A conspicuous visual feature observed in the field was that leaves had a high degree of inclination at the hypersaline sites. Ball et al.  showed that the high degree of leaf inclination found in mangrove species in nature effectively reduces the intensity of radiant heat loading. Furthermore, in both species, a marked reduction in leaf area in hypersaline sites was observed. This may also improve the energy balance in saline and dry sites, as in smaller leaves convectional cooling is more effective.
Lin and Sternberg  found a reduction in leaf size in dwarf scrub mangroves in contrast to tall growing fringe mangroves in Florida. Salinity, as well as nutrient level, may cause reductions in leaf area while neither sulfide nor original growth form had an influence .
In the present study, reductions in leaf area and dry mass associated with salinity were, respectively, 59 and 46% in Rhizophora compared to the much lower reductions of 34 and 7% in Avicennia, suggesting a higher salt sensitivity in the former species. In Avicennia, not only leaf size but also leaf shape was affected by high salinity, leaves of high salinity sites being rounder than those of low salinity sites.
Larger leaf dry mass/area ratios at the hypersaline sites have been reported before and attributed to increased succulence [34, 35, 36], although there was also evidence for scleromorphy . We observed in both species a significant decrease in the fresh mass/dry mass ratio that together with the increase in leaf mass/area ratio are rather symptoms of scleromorphy than of succulence. Generally, higher leaf mass/area ratios increase heat capacity and may be of importance in controlling leaf temperature .
5.2. Osmotic adaptation
Increases of leaf sap osmolality with soil salinity in both species counteracted the lower soil water potential at the hypersaline sites. In Rhizophora, differences between leaf sap and soil solution osmolalities were small or negative at the hypersaline sites, whereas in Avicennia differences were always positive. Scholander et al. , and Walter and Steiner  found that the osmotic potential in mangrove leaves exceeded that of the seawater surrounding them. However, in a field study in Venezuela, Rada et al.  found that turgor loss occurred at midday in leaves of Conocarpus erectus and Rhizophora mangle during drought periods. We did not measure water potential in the investigated plants, but both conductance and photosynthetic rates measured did not indicate turgor loss even at the most stressful sites.
At the low salinity site for both species, osmolalities of the leaf sap were found to be about 10 times higher than that of the soil solution, indicating their halophytic (salt accumulating behavior) character [15, 17]. The generally larger values of leaf sap osmolality in Avicennia are consistent with the higher salinity tolerance of species of this genus [13, 14, 19, 20, 37].
Concentrations of compatible solutes were clearly correlated to soil salinity and were within the range reported in other field studies . Glycinebetaine concentrations were higher than those of cyclitols at each site, and in both cases their concentration increased with leaf sap osmolality. This is in accordance with their postulated role in keeping osmotic equilibrium between cytoplasm and vacuole. However, the cyclitol/osmolality ratio was 1.25 in Rhizophora, while the glycinebetaine/osmolality ratio in Avicennia was only 0.7, suggesting a higher osmoprotective efficiency of this compound.
5.3. Phosphorus, nitrogen, and chlorophyll
Studies of nutrient availability in soils under several mangrove stands showed that P availability may be limiting growth, especially under oxidized conditions of well drained soils . Growth limitation by P was confirmed by fertilization studies on dwarf red mangrove in Belize , where an increase in growth was brought about only by application of NPK or P. In the present study, leaf P concentration did not differ between sites but was always higher in Avicennia compared to Rhizophora. Besides, N to P molar ratios were below 35 suggesting that P supply was not limiting mangrove growth in the study sites.
Leaf N concentrations of Avicennia were significantly higher than those of Rhizophora as has been found earlier [20, 41] and reported for Australian species of these genera by Popp et al. . Part of the difference can be explained by the amount of glycinebetaine in Avicennia, representing 15–21% of total leaf N. Differences in N concentration between species disappear when this fraction is subtracted from total N.
Differences in N concentrations between sites were mainly related to differences in leaf mass/area ratio in both species. While N concentration decreased with salinity when calculated on a dry mass basis, they increased based on leaf area. Rhizophora leaves at Tacuato-lagoon, however, showed a higher concentration of N per leaf area as well as per g dry mass, indicating a strong reduction in nitrogen use efficiency.
Chlorophyll concentrations per leaf area were similar or lower than those reported earlier for the same species in dry and wet habitats , or for mangrove species in Australia and India growing on a range soil salinities [42, 43]. The average Chltot/N ratios decrease markedly in hypersaline sites, pointing to a reduction in N investment in photosynthetic structures due to salt stress. Besides, these ratios were low compared with earlier reports on these species , a fact perhaps related to the much lower leaf N/P ratios found in the present study. Values of Chltot/N at a given site did not differ significantly between species, when the amount of N bound in glycinebetaine in Avicennia was subtracted from total N. It seems that under similar salinity stress, both species invest a similar N fraction into the construction of photosynthetic structures.
The fractional investment of leaf N into chloroplast protein-pigment complexes can be calculated using the N to chlorophyll ratio in thylakoids estimated by Evans  (50 mol thylakoid N/mol chlorophyll). Both species had values ranging from 9% of leaf N in low salinity sites to 5–6% in high salinity sites. Rhizophora had always slightly higher values than Avicennia. Those values are about half of the average reported for lowland trees in humid tropical forest (107 species, 23.7 ± 0.8% of leaf N) . The large difference underscores the photosynthetic cost of high salinity tolerance.
5.4. Photosynthetic capacity
Avicennia showed consistently higher assimilation rates than Rhizophora, in accordance with previous reports on other species of the same genera [11, 46, 47]. Our results showed that both species had lower Asat at the hypersaline sites. However, the depression of Asat related to high salinity was more pronounced in Rhizophora.
Light saturated photosynthetic capacity reflects the maximum possible benefits from a given investment in photosynthetic machinery . Zotz and Winter  showed a linear relationship between diurnal carbon gain and maximum rate of CO2 uptake in a range of rainforest canopy species. This would explain the low growth rates and the shrubby stature of the plants at the hypersaline sites. In addition, as constructing and maintaining photosynthetic machinery is energetically expensive, photosynthetic capacity should be tuned to the constraints of the environment . The most prominent factor in mangrove habitats is salinity. In Aegiceras corniculatum and Avicennia marina, photosynthetic capacity was found to decrease with increasing salinity [50, 51], and Asat was negatively related to salinity in a range of mangrove species under field conditions . Other environmental factors such as low nutrient availability  and temporal variation in salinity  are also known to depress maximum assimilation rate of mangroves. Extreme low values of Asat in Rhizophora leaves at Tacuato-lagoon may be related to a combination of salinity with one or more of the latter mentioned factors.
Values of light saturated A calculated from light response curves confirmed that the photosynthetic capacity was generally higher in Avicennia compared to Rhizophora and was reduced in both species at the hypersaline sites.
Quantum yield (φ) on an incident light basis was depressed at the hypersaline sites in both species. Björkman et al.  found that quantum yield in mangrove leaves decreased due to the combination of low leaf water potentials with high irradiance; whereas salinity, and the resulting leaf water deficit, had no negative effect on the quantum yield of mangrove leaves protected from direct sunlight.
The Ci-Asat relationships obtained for CO2 concentrations equal or below ambient were linear, with a slope proportional to the in vivo activity of Rubisco (carboxylation efficiency, CE ). The lower CE of both species at the hypersaline site indicates that decreases of Asat with salinity were not only due to stomatal limitation, but also due to the result of changes in the biochemical properties of photosynthesis.
CO2 compensation points were higher in Rhizophora than in Avicennia, suggesting higher photorespiration rates in the former species. However, species characteristic compensation values were similar at Tacuato and Ricoa. Similar results were obtained by Ball and Farquhar  with mangrove species grown at different salinities.
5.5. Water use and N use efficiency in photosynthesis
At the low salinity site, gl was significantly higher for both species. The generally higher values of gl of Avicennia compared to those in Rhizophora are correlated with their assimilation rates. At the hypersaline sites, gl was reduced to a greater proportion than Asat in both species, but in Rhizophora, the reduction of these parameters was more pronounced indicating the lower salinity tolerance of this species.
Intrinsic water use efficiency (A/gl) evaluates the role of biological components in determining water-carbon exchange relationships [50, 51]. In both species, A/gl was higher at the hypersaline sites, because of the relatively larger reduction of gl compared to Asat. Values of A/gl were similar to those calculated from data from Smith et al.  for Avicennia germinans and Conocarpus erectus at a coastal site in northern Venezuela and to those calculated from data from Lin and Sternberg  for Rhizophora mangle at a site in coastal Florida. In an extensive field study, Clough and Sim  found higher water use efficiency in mangroves with increasing salinity and decreasing air humidity. As air humidity in our study did not differ much between sites, changes in A/gl with salinity were less drastic than in the study mentioned above.
Water use efficiency was higher in Avicennia at the low salinity site, but it was higher in Rhizophora in the high salinity sites. More conservative water use in the latter species at the hypersaline sites is probably related to its non-salt secreting character. Water loss is minimized as salt exclusion mechanisms at the roots impose a large resistance to water flow . The higher water use efficiency in Avicennia at the low salinity site may be related to the association of this species with the more saline soils . However, as in Avicennia, leaf-to-soil osmolality difference was at least 50 mmol kg−1 at the hypersaline sites, and as accumulation of NaCl can be counteracted by salt secretion, and in this species restriction of water loss from leaves under hypersaline conditions was lower than in Rhizophora.
Nitrogen use efficiency in photosynthesis based on total leaf N was higher in Rhizophora at all sites. Similar results were reported by Alongi et al.  in Australian mangrove forests of R. stylosa and A. marina. However, those differences disappear if the amount of N invested in glycinebetain is subtracted from the total amount of N. In both species, the NUE decreases in hypersaline sites. In an experimental study, Cardona-Olarte et al.  did not find differences in WUE based on gas exchange of seedling grown in nutrient solutions with salinities between 10 and 40 ppt; however, PNUE decreased from about 85 μmol/mol N at 10 ppt to nearly 60 at 40 ppt.
5.6. Carbon isotope discrimination
The carbon isotope ratio δ13C is related to a long-term average of ci and can be taken as an indicator of water use efficiency [56, 57]. Values of δ13C ranged between −24.3‰ and −29.4‰ (Δ = 16.7 to 22.1‰), with a larger variation found in Rhizophora. They were well within the range reported for mangroves in the literature [9, 20, 32, 47, 58]. These results confirmed for the long-term, the patterns found for short-term water use efficiency discussed above.
The relationships between the set of physiological properties associated with high salinity stress in both mangrove species studied here can be depicted along two sequences of events operating simultaneously (Figure 5). Increases in interstitial water salinity affect essential nutrients uptake and salt accumulation, determining increases of tissue sap osmolality. Both processes lead to a nutritional limitation of photosynthesis resulting in a strong reduction of nitrogen use efficiency for growth. In addition, the differential soil–plant osmotic potential decreases reducing the amount of water available for transpiration, and inducing an accumulation of compatible solutes that protect cytoplasmic organelles from dehydration and toxic ionic effects. As a result, leaf conductance is reduced to a larger proportion than photosynthesis, thereby increasing leaf temperature and water use efficiency. The connections between the boxes of Figure 5 are not necessarily linear, and processes affected may show differential sensitivity toward interstitial soil salinity (as shown in the ratio of conductance to photosynthesis). In the processes documented in the present chapter, Rhizophora appears to be more sensitive than Avicennia, and we speculate that the ultimate cause of this difference may reside in the higher efficiency of glycinebetaine as an osmoprotectant compared to cyclitols.
To the Austrian Research Council for the fellowship to M. Watzka and the Centro de Estudios Avanzados, IVIC, for providing housing facilities; Prof. Dr. H. Ziegler (University of Munich, Germany) helped with the determination of δ13C values of leaf samples. Analyzes of inorganic and organic osmolytes were conducted at the Center of Ecology of IVIC, in Caracas, and the Institute for Ecology and Environmental Conservation, University of Vienna, Austria. Dr. Ariel Lugo (IITF-Forest Service) provided critical comments to the original manuscript.