Annals of Botany 86: 765±769, 2000 doi:10.1006/anbo.2000.1239, available online at http://www.idealibrary.com on
Cuticles of Vascular Epiphytes: Ecient Barriers for Water Loss after Stomatal Closure?
S. HELBSING{,M.RIEDERER{ and G. ZOTZ*{{ {Lehrstuhl fuÈr Botanik II der UniversitaÈtWuÈrzburg, D-97082, WuÈrzburg, Germany and {Smithsonian Tropical Research Institute, Apdo, 2072, Balboa, Panama
Received: 18 November 1999 Returned for revision: 14 February 2000 Accepted: 16 June 2000 Published electronically: 9 August 2000
Water permeabilities of astomatous, isolated cuticular membranes of 15 vascular epiphyte species (families: Araceae, Orchidaceae and Piperaceae) from the moist lowland forest of Barro Colorado Nature Monument, Panama, were investigated. Permeances determined at 308C ranged from 0.46 Â 10 6 ms 1 (Aspasia principissa (Rchb. f.) R.E. Schult.) to 6.07 Â 10 6 ms 1 (Polystachya foliosa (Hook.) Rchb. f.). Comparison of these data with permeances of plants from other habitats corroborates the notion that cuticular properties re¯ect the climatic demands of the growing sites. The studied group of vascular epiphytes, living in a very drought-prone habitat, showed the lowest cuticular permeances to water recorded to date. # 2000 Annals of Botany Company Key words: Barro Colorado Island, cuticular transpiration, epiphyte, plant water relations, transport barrier.
INTRODUCTION MATERIALS AND METHODS Under severe environmental conditions such as water Fully developed, non-senescent leaves without any visible de®ciency and high temperatures, stomata will frequently signs of herbivory or pathogen attack were collected in the be closed (Lambers et al., 1998). The survival of a plant Republic of Panama in the Barro Colorado Nature may then depend strongly upon the residual amounts of Monument (98100N, 798510W) in April 1997. The vegeta- water loss from leaf surfaces determined, to varying tion of this nature reserve, which is located in Lake Gatun, degrees, by the rates of cuticular transpiration. Crowns of is classi®ed as tropical moist forest, with approx. 200 species tropical rain forest trees represent an extreme habitat where of epiphytic and hemiepiphytic plants (Croat, 1978). The such a scenario frequently occurs (Benzing, 1990). Drought climate is characterized by a pronounced dry season from is particularly pronounced in seasonal forests, but even in mid-December to late April. The mean annual rainfall is wet forests canopy-dwelling plants, such as vascular approx. 2700 mm (Croat, 1978). epiphytes, are only intermittently supplied with moisture The epiphyte species included in this study are listed in from rain, dew or mist. Between these episodic events, high Table 1. Most samples were from plants growing on radiation and strong winds lead to a considerable evapora- Annona glabra L., a tree restricted to the shoreline of tive demand, while the lack of soil reduces available Lake Gatun. Leaves of shade-tolerant species were obtained moisture reservoirs to tissue water. from sites in the forest understorey. We tried to cover a Recently, Schreiber and Riederer (1996) investigated large range of the epiphyte taxa present in Barro Colorado cuticular properties of a large number of species from Nature Monument, but unfortunately many species proved diverse habitats ranging from the temperate to the tropical to be unsuitable for analysis with our method using isolated zone. Although all leaf samples were obtained from plants cuticular membranes. This was not only true for bromeliads growing in the greenhouse, they found a relationship with their epidermal trichomes, but also for other taxa between presumed evaporative demand in the native habitat because the leaves of many epiphyte species are too narrow. and the in vitro permeability of cuticular membranes to In some cases (e.g. Anthurium clavigerum Poepp., Catase- water. Considering the in situ growing conditions of vascular tum viridi¯avum Hook., Polypodium crassifolium L.) enzy- epiphytes, we hypothesized that cuticles of the species matic isolation failed. belonging to this plant group should show extraordinarily Astomatous cuticular membranes were isolated from the low permeabilities for water. Pertinent information, how- upper leaf surfaces using methods described previously ever, is rare (see Benzing and Burt, 1970, for a few bromeliad (SchoÈ nherr and Riederer, 1986; Schreiber and Riederer, species). Therefore, we tested this notion for a number of 1996). The enzymatic solution contained 1 % (v/v) species of dierent taxonomic anity, growing naturally in a Trenolin Super DF (ErbsloÈ h Geisenheim, Geisenheim, moist tropical forest in central Panama. Moreover, we Germany), 1 % (v/v) Celluclast (Novo Nordisk, Bagsvaerd, investigated whether dierences in cuticular properties . Denmark), 0 1 % (v/v) NaN3 in a citric acid buer adjusted might be detectable on a smaller scale, i.e. between those . species growing at less exposed vs. exposed sites of the forest. to pH 3 0 with KOH. Samples were kept in the enzymatic solution for about 7 weeks (range 3±11 weeks). After * For correspondence. Fax (0931) 888-6235, e-mail zotz@botanik. enzymatic isolation, cuticular membranes were stored for uni-wuerzburg.de 4 weeks before measurements. Storage for several weeks is 0305-7364/00/100765+05 $35.00/00 # 2000 Annals of Botany Company 766 Helbsing et al.ÐCuticles of Vascular Epiphytes
TABLE 1. Permeances to water of isolated cuticular membranes (P in m s 1 at 308C) for 14 epiphyte species
Species Family Habitat class n P Â 106
Evergreen Aspasia principissa Rchb. f. Orchidaceae Shade 3 0.5 + 0.10 Anthurium salviniae Hemsl. Araceae Shade 12 0.7 + 0.13 Oncidium ampliatum Lindl. Orchidaceae Sun 13 1.0 + 0.29 Caularthron bilamellatum (Rchb. f.) R.E. Schult. Orchidaceae Exposed 13 1.1 + 0.22 Anthurium brownii Mast. Araceae Sun 10 1.2 + 0.21 Philodendron radiatum Schott. Araceae Sun 14 1.2 + 0.23 Philodendron tripartitum (Jacq.) Schott. Araceae Sun 11 1.1 + 0.35 Notylia pentachne Rchb. f. Orchidaceae Sun 14 1.3 + 0.39 Sobralia suaveolens Rchb. f. Orchidaceae Sun 12 1.7 + 0.30 Epidendrum nocturnum Jacq. Orchidaceae Sun 12 1.8 + 0.49 Trichopilia maculata Rchb. f. Orchidaceae Shade 13 2.2 + 0.48 Sobralia fenzliana Rchb. f. Orchidaceae Shade 12 2.7 + 0.81 Drought-deciduous Peperomia cordulata C. DC. Piperaceae Sun 12 4.6 + 0.83 Polystachya foliosa (Hook.) Rchb. f. Orchidaceae Sun 7 6.1 + 2.32
Means + 95 % con®dence intervals and sample sizes (n) are given. Species names follow nomenclature of d'Arcy (1987). Assignment of the species to dierent habitat classes determined by Croat (1978) and personal observations. necessary to avoid the in¯uence of the isolation procedure Because the water vapour concentration over silica gel is on the permeability of cuticular membranes (Geyer and approximately zero and liquid water was applied in the SchoÈ nherr, 1990). donor chamber, Dc is equal to the saturation water vapour The possibility that dierences in the duration of the concentration at the experimental temperature. enzymatic isolation process could in¯uence subsequent All 14 species were measured at 308C which approximated permeability measurements motivated the following experi- the in situ microclimatic conditions (Zotz and Winter, 1994). ment. Leaf disks from the hemiepiphyte Clusia uvitana Temperature eects on the permeability of cuticular Pittier (Clusiaceae, cultivated in the WuÈ rzburg Botanical membranes were only studied on a subset of ten species. Gardens), which take about 1 week to yield isolated Temperatures used were 208C, and 5 degree increments from cuticular membranes (unpubl. res.), were left in enzymatic 30 to 558C. In other studies, permeances are frequently solution for 1, 2, 4 and 8 weeks, respectively. After each determined at 258C. Although we did not measure time period, 15 cuticular membranes were removed, dried, permeances at this temperature due to technical problems, ¯attened, stored for 4 weeks, and then used for permeability estimated values obtained by interpolation from permeances measurements. Measurements of cuticular membrane at 208C and 308C allow comparison with literature data. In permeabilities were made at 258C and permeances were this report, permeances are based on water concentrations in calculated. No signi®cant dierences between treatments the gas phase, in contrast to the permeances given by . were found (Kruskal±Wallis-ANOVA, d.f. 3, P 4 0 1). Schreiber and Riederer (1996) which were based on liquid This suggests that varying lengths of immersion in water. Conversion of their results to vapour-phase based enzymatic solution have no signi®cant in¯uence on cuticular membrane permeability in thisÐand presumably −10 otherÐspecies. Cuticular permeability was determined following the methods of Schreiber and Riederer (1996). Cuticular −11 membranes were mounted between transport chambers made of stainless steel and the time courses of water loss −12 across the cuticular membranes were measured gravime- In P trically. Linear regression equations were ®tted to plots of −13 total amount of water loss vs. time and the ¯ow rates of water loss were obtained from the slopes of the graphs. −14 Permeances (P;ms 1) were calculated as . 00 . 05 . 10 . 15 . 20 . 25 . 30 . 35 . 40 . 45 3 3 3 3 3 3 3 3 3 3 F P 1 1/T [10−3 K−1] ADc FIG. 1. Arrhenius plot of the permeance to water vapour of the adaxial where F (kg s 1) is the ¯ow rate of water across the cuticular leaf cuticle of Trichopilia maculata (Orchidaceae) with regression lines 2 . 2 (r 0 98) for the high- and low-temperature ranges of temperature membranes, A (m ) is the exposed cuticular membrane area dependence, respectively. Each circle represents the mean of 11±15 . 4 2 3 (1 13 Â 10 m ), and Dc (kg m ) is the gradient of water measurements and error bars are 95 % con®dence intervals. The vapour concentration inside and outside the chamber. intersection is de®ned as the transition point. Helbsing et al.ÐCuticles of Vascular Epiphytes 767 permeances was achieved by multiplying their values with TABLE 2. Transition points (8C) of the temperature depend- the ratio of liquid water densities and water vapour at ence of the permeance to water of isolated cuticular saturation at the temperatures of measurement. membranes from ten species of vascular epiphytes Results were analysed with STATISTICA software (STATISTICA 5.1, StatSoft Inc., Tulsa, OK, USA). In Species Temperature most cases, strong deviation from normality only permitted the use of non-parametric statistics. Phase transitions were Anthurium brownii 38.9 . analysed with breakpoint regressions. Caularthron bilamellatum 38 9 Epidendrum nocturnum 38.9 Sobralia fenzliana 38.9 . RESULTS Trichopilia maculata 38 9 Philodendron radiatum 39.8 The cuticular membranes isolated from 14 vascular epi- Anthurium salviniae 42.3 . phyte species showed water permeances (at 308C) which Philodendron tripartitum 42 3 Sobralia suaveolens 42.3 varied by more than one order of magnitude (Table 1): they Encyclia chimborazoensis 44.9 ranged from 0.46 Â 10 6 ms 1 (Aspasia principissa) to 6.07 Â 10 6 ms 1 (Polystachya foliosa). Interspeci®c dierences did not correlate with the habitat of a given recalculated from Schreiber and Riederer, 1996). Primary species. Comparing the species growing in the forest hemiepiphytes such as Clusia sp. or Ficus sp. go through an understorey or the inner canopy of trees (shade habitat, epiphytic stage with much reduced availability of water Table 1) with those found at more exposed sites (sun (Holbrook and Putz, 1996). It would be interesting to habitat) yielded no signi®cant dierences in permeances determine whether dierences in cuticular properties can be (Mann±Whitney U test, P 1.00). Remarkably, the two found between ontogenetic stages of such species. Most drought-deciduous species (Croat, 1978; pers. obs.) investi- terrestrial plants show much higher permeabilities to water gated had substantially higher cuticular permeances than further indicating the importance of water in the epiphytic the evergreen species included in this study (Table 1). habitat. The permeance to water of isolated cuticular membranes Remarkably, on a lower spatial scale, a correlation increased with temperature. The mean permeances between relative water availability and cuticular properties measured at 558C were up to 28-fold higher than those could not be found. Epiphytes growing at exposed sites did for 308C. Cuticular permeances increased steeply with not feature leaf cuticles with extraordinarily low per- temperature in the high-temperature range, while at lower meability. The species with the lowest permeance (Aspasia temperatures moderate temperature dependencies were principissa) is normally restricted to the shaded understorey observed. The transition points, which were determined (Zimmerman and Aide, 1989). 2 . . by breakpoint regressions (r between 0 95 and 0 99), Many members of the Bromeliaceae have gained a ranged from 39 to 458C for the ten species investigated in certain independence from intermittent water supply this respect (Table 2). by evolving water-impounding foliage, so-called `tanks' (Benzing, 1980; Zotz and Thomas, 1999). Thus, following DISCUSSION 40 Permeances to water of isolated cuticular membranes from 35 n = 7 the adaxial leaf surfaces of vascular epiphytes ranged from . 6 . 6 1 30
0 5 Â 10 to 6 1 Â 10 ms . Values for the subset of ) 1 evergreen epiphyte species ranged from 0.5 Â 10 6 to − 25 . 6 1
2 7 Â 10 ms (Table 1). Thus, the results of the present (ms
6 20 study support the notion put forward by Schreiber and − n = 8 Riederer (1996) that since cuticular properties become 15 n = 12 increasingly important in more arid habitats, there should P x 10 10 be a negative relationship between cuticular permeability n = 5 5 and water supply and/or evaporative demand. Since the n = 7 0 growing sites of most vascular epiphytes are characterized epiphytes hemiepiphytes mediterranean, bromeliads, temperate, by ¯uctuating water supply and frequent drought (Benzing, non-impounding and climbers terrestrial epiphytic terrestrial 1990), low cuticular permeances to water were anticipated Plant groups for the species included in our study. A comparison between plants from a variety of habitats (Fig. 2) indicates FIG. 2. Permeances (P, at 258C) for cuticular membranes of dierent plant groups. Given are means + 95 % con®dence intervals, sample that the permeances reported here for the cuticular size (n) is shown on top of the bars. From left to right: tropical, non- membranes of selected epiphyte species were the lowest impounding epiphytes (this study); tropical hemiepiphytes and found to date (Kruskal±Wallis-ANOVA, d.f. 4, climbing plants (Schreiber and Riederer, 1996); terrestrial plants P 5 0.001). Hemiepiphytes and climbing plants (grown in from mediterranean-type climate (Schreiber and Riederer, 1996); epiphytic bromeliads (Kerstiens, 1996, after Benzing and Burt, 1970); the greenhouse) show only somewhat higher permeances terrestrial plants from temperate climates (Schreiber and Riederer, 6 (e.g. Vanilla planifolia,0.6 Â 10 ; Monstera deliciosa, 1996). Statistical dierences between plant groups were highly 1.4 Â 10 6; Philodendron selloum,2.2 Â 10 6 ms 1;data signi®cant (Kruskal±Wallis-ANOVA, d.f. 4, P 5 0.001). 768 Helbsing et al.ÐCuticles of Vascular Epiphytes
TABLE 3. Model calculations of cuticular water loss in two species of epiphytes
Parameter Species Source
Caularthron bilamellatum Aspasia principissa