The landscape of the upper Rio Negro basin (North Amazon) exhibits distinctive habitats that are associated with differential soil characteristics and topographical conditions as well as species composition (Herrera et al. Reference HERRERA, JORDAN, KLINGE and MEDINA1978). The mixed forests thrive on well-structured oxisols on slightly more elevated areas. The valleys with sandy podzols are occupied by the ‘Amazon caatinga’ complex with three distinct zones: the bottom valley and the gentle slopes, both of which have closed forests, and the sandy domes with open forests (‘bana’ or sclerophyllous forest; Breimer Reference BREIMER, Breimer, van Kekem and van Reuler1985). From the mixed forest towards the caatinga valley-slope-dome habitats, the leaf δ¹⁵N signatures become increasingly negative, suggesting a trend in N limitation in the same direction (Sobrado Reference SOBRADO2010). Thus, negative leaf δ¹⁵N signatures depleted in ¹⁵N compared with the soil indicate a very tight N cycle in all of the habitats. Water availability follows a similar pattern from the top of the oxisol towards the flooded valley bottom of the caatinga, with extreme water-table fluctuations in the sandy domes (Klinge Reference KLINGE1978). Thus, parallel variation in nutrient and water availabilities exist in this area that are associated with soil characteristics and topography. Under such contrasting habitats, species-specific responses would be linked to particular conditions of the habitat at a local scale (Comita & Engelbrecht Reference COMITA and ENGELBRECHT2009, Engelbrecht et al. Reference ENGELBRECHT, COMITA, CONDIT, KURSAR, TYREE, TURNER and HUBBELL2007). A number of studies in these habitats have shown that this is the case for soil fertility (Coomes Reference COOMES1997, Medina et al. Reference MEDINA, GARCIA and CUEVAS1990, Sobrado Reference SOBRADO2010, Sobrado & Medina Reference SOBRADO and MEDINA1980). Similarly, the hydraulic characteristics and long-term water use are species specific and related to particular conditions of the habitat at the local scale (Sobrado Reference SOBRADO2010). In this report, it was hypothesized that the leaf tissue water relations of species thriving in different habitats may reflect the water availability at the particular sites as well. The leaf tissue water relations of species thriving in the extreme nutrient and water-supply conditions of the sandy domes from the caatinga complex have been previously studied in detail (Sobrado Reference SOBRADO2009a). However, these data are currently not available for the species that thrive in the surrounding area of the closed forests, and importantly, such information would allow for a comparison across habitats. Therefore, the present study assessed the minimum leaf water potential (midday) under field conditions as well as the leaf tissue water relations by using pressure-volume analysis of dominant tree species in the top canopy of these high-stature forests.

The study area is located 4 km north-east of San Carlos de Rio Negro (Amazonas, Venezuela; 1°54′N; 67°03′W) within the upper Rio Negro basin. The zone features a mean temperature of 26.2 °C and an annual rainfall of 3500 mm. No month receives less than 100 mm of precipitation. Similar to previous studies conducted in these habitats (Sobrado Reference SOBRADO2010, Reference SOBRADO2011), the experiments assessed Caryocar glabrum (Aubl.) Pers. (Caryocaraceae) and Ocotea aciphylla (Nees & Mart. ex Nees) Mez (Lauraceae) from the mixed forest (oxisol soil) and Eperua leucanta Benth. (Caesalpiniaceae) and Micranda sprucei (Müll.Arg.) R. E. Schultes (Euphorbiaceae) from the caatinga complex (podzol soil). Eperua leucantha is the dominant species of the bottom valley of the caatinga forest (‘yaguacana’ forest), and M. sprucei is the dominant species in the ecotone (‘cunuri’ forest) between the bottom valley and the sandy mounds, which is where the sclerophyllous forest is located (‘bana’; Breimer Reference BREIMER, Breimer, van Kekem and van Reuler1985, Klinge Reference KLINGE1978). In all species, top-canopy leaves are typical sun leaves that possess a dorsiventral anatomy with bundle-sheath extensions (heterobaric) and a chlorophyll a to b ratio of approximately 3:1 (Sobrado Reference SOBRADO2011). However M. sprucei from the caatinga slopes possesses the thickest leaves, lowest-density tissue, lowest palisade to spongy parenchyma ratio, highest sclerophylly index, highest carbon-to-nitrogen ratio, and most negative δ¹⁵N of all of the species in this study (Sobrado Reference SOBRADO2010). The characteristics of M. sprucei are comparable to those found in the sclerophyllous forest.

For all measurements, fully exposed top-canopy branches were selected from three trees per species that were tagged for the experiments during August 2011. Samples were collected by climbing each tree to reach the canopy top. Twenty leaf punches with a 3-cm diameter were collected from three twigs, respectively, at predawn in each tree species and their fresh weight was measured. Subsequently, the dry mass was determined after the leaves were oven-dried at 60 °C. The data were used to calculate the water content and expressed based upon the leaf area (W_a) as well as fresh (W_f) and dry mass (W_d). The minimum tree water status was characterized by measuring the midday leaf water potential (Ψ, MPa) under rainy conditions and after six rainless days that occurred during August 2011. Ψ was measured with a pressure chamber (PMS, Model 1000, Corvallis, Oregon, USA) in two–three twig samples per tree species. Pressure–volume (P–V) curves were determined for two twigs from each tagged sample collected in the field at predawn for a total of six samples per species. Sampled twigs were allowed to dehydrate naturally on the laboratory bench and P–V curves were generated. During dehydration, the weight loss was determined on an analytical balance and Ψ was immediately measured using the pressure chamber. The reciprocal of Ψ (1/Ψ) was plotted as a function of relative water content (RWC). These plots were used to calculate the osmotic potential at full (Ψ₍₁₀₀₎) and zero (Ψ₍₀₎) turgor, RWC at zero turgor (RWC₀), and apoplastic (A) water content (Tyree & Hammel Reference TYREE and HAMMEL1982). The volumetric elastic modulus (ɛ) was calculated as the slope of the relationship of turgor pressure (Ψ_p) above zero as a function of RWC. One-way analysis of variance (ANOVA) followed by the Holm–Sidak test was used to compare data across species if the data were normally distributed, and equal variance was measured. ANOVA-on-ranks test was used if data normality and/or equal variance tests failed, followed by a Tukey test. The significance level was set at P < 0.05. All analyses were performed using SigmaStats 3.1 software for Windows (Systat Software, Inc, Chicago, Illinois, USA).

The leaf W_a was significantly higher in the thicker leaves of M. sprucei compared with C. glabrum, O. aciphylla and E. leucantha (Table 1; Sobrado Reference SOBRADO2010). Moreover, W_f and W_d were species specific and not differentially associated with forest types (Table 1). Values of W_f were comparable to those of the species from the open sclerophyllous forest (0.58 ± 0.02 g g⁻¹; Sobrado Reference SOBRADO2008). On rainy days, the minimum Ψ averaged −0.98 ± 0.05 MPa, which was significantly higher than that found during rainless days (−1.2 ± 0.05 MPa) (Table 1). The osmotic potential at full (Ψ_π(100)) and zero (Ψ_π(0)) turgor was not statistically different in C. glabrum, O aciphylla and E. leucantha and averaged −1.19 ± 0.04 and −1.38 ± 0.06 MPa, respectively (Table 1, Figure 1a–c). However, the values of these parameters for M. sprucei were significantly higher (−0.75 ± 0.06 and −1.12 ± 0.12 MPa, respectively; Table 1, Figure 1d). Therefore, M. sprucei develops lower leaf turgor (Ψ_p) and loses it at the highest Ψ compared with the other species (Figure 1e–h). Conversely, M. sprucei had the highest elasticity (< ɛ; Table 1), which resulted in the highest leaf dehydration among the species (Figure 1a–d), while still maintaining Ψ within the range of positive turgor (Figure 1e–h). Thus, high elastic cells of M. sprucei allow for the partial compensation of unfavourable high Ψ_π. Turgor maintenance as Ψ declines can be achieved through low Ψ_π and high cell elasticity. However, variation in Ψ across species for turgor loss arises predominantly from differences in Ψ_π rather than from differences in ɛ (Lenz et al. Reference LENZ, WRIGHT and WESTOBY2006). The apoplastic water content (A) was comparable across species and averaged 27.7 ± 1.84% (Table 1). Therefore, values of Ψ_π(100) and Ψ_π(0) were higher (less negative) in all of the species than values found in the open sclerophyllous forest (−1.48 ± 0.05 and −1.92 ± 0.05 MPa, respectively) (Sobrado Reference SOBRADO2009a). These results implied that dominant species in the mixed forest as well as in the valley and slope of the caatinga maintain turgor pressure at considerably higher Ψ than those located on top of the sandy mounds. Importantly, the values of ɛ and A found in these species overlapped with those found in the open forests on the top of the sandy domes except in M. sprucei with the lowest ɛ. In the open sclerophyllous forests, the range of ɛ and A are 6–16 MPa and 18–43%, respectively (Sobrado Reference SOBRADO2009a).

Figure 1. Leaf water potential (Ψ) and osmotic potential (Ψ_π) as a function of relative water content (a–d) as well as leaf turgor pressure (Ψ_p) as a function of leaf water potential (e–h) in two species of the mixed forest (Caryocar glabrum and Ocotea aciphylla) and two species of the Amazon caatinga forest (Eperua leucantha and Micranda sprucei). Plots were derived from six pressure–volume (P–V) curves for each species. Linear regressions are shown with the following equations: Ψ = 0.15 RWC – 14.9 (r = 0.97; P < 0.001), Ψ_π = 0.03 RWC – 4.25 (r = 0.77; P < 0.05) and Ψ_p = 0.85 Ψ + 1.17 (r = 0.98; P < 0.001) for C. glabrum, Ψ = 0.14 RWC – 14.3 (r = 0.99; P < 0.001), Ψ_π = 0.03 RWC – 3.96 (r = 0.80; P < 0.05) and Ψ_p = 0.93 Ψ +1.32 (r = 0.99; P < 0.001) for O. aciphylla, Ψ = 0.09 RWC – 10.2 (r = 0.95; P < 0.01), Ψ_π = 0.02 RWC – 2.75 (r = 0.83; P < 0.05) and Ψ_p = 0.79 Ψ +1.16 (r = 0.99; P < 0.001) for E. leucantha and Ψ = 0.05 RWC – 4.84 (r = 0.94; P < 0.001), Ψ_π = 0.02 RWC – 2.66 (r = 0.82; P < 0.001) and Ψ_p = 0.61 Ψ +0.76 (r = 0.96; P < 0.001) for M. sprucei.

Table 1. Leaf water content based on surface area (W_a), fresh weight (W_f) and dry mass (W_d), midday leaf water potential (Ψ) during rainy (RY) and rainless (RL) days, as well as the following pressure-volume parameters: osmotic potential at full (Ψ) and zero turgor (Ψ), relative water content at zero turgor (RWC_o), apoplastic water content (A) and elastic modulus (ɛ). All measurements were taken in Caryocar glabrum and Ocotea aciphylla from the mixed forest and Eperua leucantha and Micranda sprucei from the Amazon caatinga. Each value is the mean ± SE of measurements taken from three trees per species. For each parameter, the mean followed by a letter represent statistical differences at P < 0.05.

In conclusion, the top-canopy leaves of the species assessed in this study experienced a very mild decrease in leaf Ψ (~ −1.1 to −1.5 MPa) during the rainless period. Furthermore, the analysis of tissue water relations showed that these species were unable to develop very low Ψ within the range of turgor maintenance. Thus, drought resistance does not seem to be a highly selected trait in these particular habitats that have relatively advantageous topographical and hydrological conditions. Intriguingly, M. sprucei from the caatinga slopes, which possesses sclerophyllous leaves, was particularly susceptible to a loss in turgor pressure at high leaf Ψ (~1.1 MPa) compared with the other species. Indeed, the mechanical constitution of sclerophyll leaves confers resistance to stress and provides protection from herbivory to enhance leaf longevity and long-term carbon gain (Read et al. Reference READ, SANA, GARINE-WICHATITSKY and JAFFRÉ2006, Turner Reference TURNER1994). However, previous findings do not support the drought tolerance of sclerophyllous species (Read et al. Reference READ, SANA, GARINE-WICHATITSKY and JAFFRÉ2006, Salleo et al. Reference SALLEO, NARDINI and LO GULLO1997). Thus, sclerophyll species, such as M. sprucei and those in the open sclerophyllous forests at the mounds of the caatinga, seemed to have been selected due to the low fertility of both sites, as previously stated (Sobrado Reference SOBRADO2009b, Reference SOBRADO2010). Nevertheless, the open sclerophyllous forests (sandy domes) suffer unpredictable drought spells as well, which would favour the establishment of species that are able to tolerate both nutrient and water deficiencies. Therefore, these species have a lower minimum leaf Ψ (~ 1.5 MPa), tolerate a lower water status, and maintain turgor pressure (Ψ ~ 2 MPa; Sobrado Reference SOBRADO2009a). Conversely, M. sprucei, with sclerophyllous leaves, is confined to the slopes of the caatinga (ecotone), but is never found in the open sandy mounds, which is possibly due to its relatively high drought sensitivity. Therefore, this region, with a high rainfall pattern, possesses well-defined habitats that have contrasting nutrient and water availabilities, which seems to influence the local distribution of species.