Midday dew – an overlooked factor enhancing photosynthetic activity of corticolous epiphytes in a wet tropical rain forest


  • Michael Lakatos,

    1. Department of Plant Ecology and Systematics, Faculty of Biology, University of Kaiserslautern, PO Box 3049, D-67653 Kaiserslautern, Germany
    Search for more papers by this author
  • André Obregón,

    1. Laboratory for Climatology and Remote Sensing (LCRS), Department of Geography, University of Marburg, Deutschhausstr. 12, 35032 Marburg, Germany
    Search for more papers by this author
  • Burkhard Büdel,

    1. Department of Plant Ecology and Systematics, Faculty of Biology, University of Kaiserslautern, PO Box 3049, D-67653 Kaiserslautern, Germany
    Search for more papers by this author
  • Jörg Bendix

    1. Laboratory for Climatology and Remote Sensing (LCRS), Department of Geography, University of Marburg, Deutschhausstr. 12, 35032 Marburg, Germany
    Search for more papers by this author

Author for correspondence:
Michael Lakatos
Tel: +49 631 205 4448
Email: lakatos@rhrk.uni-kl.de


  • Additional water supplied by dew formation is an important resource for microbes, plants and animals in precipitation-limited habitats, but has received little attention in tropical forests until now.
  • We evaluated the micro-environmental conditions of tree stem surfaces and their epiphytic organisms in a neotropical forest, and present evidence for a novel mechanism of diurnal dew formation on these surfaces until midday that has physiological implications for corticolous epiphytes such as lichens.
  • In the understorey of a lowland forest in French Guiana, heat storage of stems during the day and delayed radiative loss during the night decreased stem surface temperatures by 6°C in comparison to the dew-point temperature of ambient air. This measured phenomenon induced modelled totals of diurnal dew formation between 0.29 and 0.69 mm d−1 on the surface of the bark and the lichens until early afternoon.
  • Crustose lichens substantially benefit from this dew formation, because it prolongs photosynthetic activity. This previously unrecognized mechanism of midday dew formation contributes to the water supply of most corticolous organisms, and may be a general feature in forest habitats world-wide.


Dew formation has been extensively studied in precipitation-limited ecosystems (Agam & Berliner, 2006). Deposition of dew in these dry habitats is known to provide an essential source of water for biofilms, biological crusts, plants, invertebrates and small vertebrates (Broza, 1979; Cloudsley-Thompson, 2001; Zhang et al., 2009). Dew is also used by humans for drinking water and for increasing water availability for agriculture in arid areas (Kabela et al., 2009). By contrast, dew formation has been almost completely ignored in studies of tropical forests (Andrade, 2003). If regular dew formation on the enormous surfaces of tree stems is taken into account, then a deeper understanding of this source of water will enhance our knowledge of micro-ecology in terms of photosynthesis, hydrology, and ecology in these widespread ecosystems.

Dew formation by radiative cooling of the ground surface is related to a delayed amplitude of outgoing net radiation of the surface compared with that of ambient air. This commonly occurs in open vegetation early in the morning, after clear-sky nights. However, dew formation as a water source in tropical lowland forests has not yet been considered, because the infrared emission of the ground is commonly buffered by the dense tree canopy. Nevertheless, the wood and bark of tree stems have higher thermal conductivities (0.32–0.38 W m−1 K−1) and specific heat capacities (1.7–2.8 kJ kg−1 K−1) in comparison to air (0.026 W m−1 K−1 and 1 kJ kg−1 K−1; Aston, 1985; Moore & Fisch, 1986; Haverd et al., 2007), and should heat and cool more slowly than air. Because of differences in heat storage, the rapid temperature change of the air during the day should be slowly followed by the change of temperature of the stem and bark. If the bark temperature is cooler than the dew-point temperature of the air, vapour condensation can occur. Additionally, the high ambient relative humidity (RH) and calm conditions in the understorey of wet tropical rain forests may facilitate this dew formation. This additional water supply could also have an impact on the water status of the closely attached epiphytic organisms, such as the dominant corticolous crustose lichens.

Wet tropical rain forests, characterized by nearly daily rain, are not normally considered as an environment where water availability might be a limiting factor for lichen activity. It may be seen as surprising that additional dew formation should have an impact on the photosynthetic activity of corticolous lichens. Because rainfall events in the perhumid tropics are known for their brevity and timing in the afternoon hours, the early morning to late midday hours can be dry periods for tree-bark-inhabiting lichens (Zotz & Winter, 1994). Most tropical forests also have seasonal dry periods without rain and even during rain events only 0.4–8% of the net precipitation may reach stem-colonizing organisms via stem water flow (Jordan & Heuveldrop, 1981; Cavelier et al., 1997). In tropical habitats, dew formation might therefore play a role as a water source that can enhance the photosynthetic activity of corticolous organisms.

In this study, we addressed a perceived gap in our knowledge of bark-living (corticolous) organisms with respect to dew exploitation. We consider the general relevance of dew formation on stems extending the metabolic activity of poikilohydric plants in a natural ecosystem, and discuss its potential impacts. We examined the micro-environmental conditions of tree stem surfaces and their epiphytic organisms in a dense lowland tropical rain forest in French Guiana (Lakatos et al., 2006). The impact on water gain and loss, as well as the photosynthesis of tropical corticolous green-algal lichens, as predominant representatives of corticolous epiphytes, was studied following the methods of Lakatos et al. (2006) and a new methodological approach. In this new approach, chlorophyll a fluorescence is used to carefully analyse and model the sensitive water status of poikilohydric epiphytes in situ. We hypothesized that: First, spatiotemporal differences among bark, lichen and air temperatures induce dew formation on bark and lichen surfaces; and that second the physiological processes of corticolous lichens benefit from dew formation.

Materials and Methods

Study area and organisms

An evergreen lowland tropical rain forest was sampled at two stations in Les Nouragues National Park, French Guiana (4°05′N, 52°40′W, 160 m a.s.l.), during January–March and October–November 1999 as well as April 2010 during the dry season. At the study site, on average 55 d are without rain during the year and 67 d have very little rain (< 1 mm; Grimaldi & Riéra, 2001), which barely moistens epiphytes in the understorey via stem water fall. The dry season occurs during August to October, with 48% of days without any rain (J. Bendix & A. Obregón, unpublished data). Detailed descriptions of the study site and the abiotic conditions can be found elsewhere (Grimaldi & Riéra, 2001; Lakatos et al., 2006; Obregon et al., 2011). All studied lichen species are commonly found on several tree species, frequently sharing the same tree trunk, and are predominately stem epiphytes. Their morphology, anatomy and specific thallus parameters are described in Lakatos et al. (2006). Because dry weight could not be measured due to the tight association of thallus and bark, thallus water content (WC) was determined as millimetre ‘precipitation-equivalent’ related to the projected area. Thus, it represents the depth of the water column over an area of a square millimetre (Lange et al., 1994).

Microclimate measurement

Relative humidity, temperature and light intensity were measured to evaluate microclimate conditions in the understorey. All climate sensors were connected to a datalogger (CR10x; Campbell Scientific Ltd, Loughborough, UK) and a multiplexer (AM416; Campbell Scientific Ltd) measuring every 30 s and recording maximum, average and minimum values at an interval of 5 min. Ambient RH and temperature (Ta) were quantified adjacent (c. 2 m distant) to the experimental site at a height of 1.5 m (HMP 35; Campbell Scientific Ltd). To differentiate between the temperature influence of the ambient air and that of the phorophyte (the tree substrate) on the crustose lichens, we used three nickel-chromium/nickel thermocouples per lichen thallus. The three soldering joints (< 0.5 mm) were positioned (1) 2 mm inside the bark (Tb), (2) underneath the lichen thallus surface (Ts), and (3) at a distance of 1 cm above the thallus surface representing the boundary layer (Tbl), all facing east and west. All thermocouples were radiation-shielded and referenced to a thermistor (10 TCRT; Campbell Scientific Ltd). Photosynthetic flux density (PFD; λ: 400–700 nm) was measured at a height of 2 m at the phorophyte using 2π quantum sensors (Li-190SA; Li-Cor Inc., Lincoln, NE, USA). Daytime was taken to be from 06:30 to 18:30 h. The data were analysed using Shapiro–Wilk tests for normality and the effects of temperature differences between the microsites were evaluated using the Wilcoxon test, the Kolmogorov–Smirnov test and Friedmann ANOVA (Statistica 6; StatSoft Inc, Tulsa, OK, USA).

Assessment of photosynthetic parameters

Chlorophyll a fluorescence was used to quantify photosynthetic efficiency (Green et al., 1998). Photosynthesis was measured with a miniaturized pulse-amplitude modulated photosynthetic yield analyser (Mini-PAM; Walz, Effeltrich, Germany) under experimental and natural conditions and with a methodology adapted for lichens (Lakatos et al., 2006). The data were analysed using Shapiro–Wilk tests for normality and were checked with one-way ANOVA followed by Fisher’s least significant difference (LSD) post hoc test. Analyses and curve fitting of data (< 0.001) were performed using the Marquardt algorithm (SigmaPlot 9.0; Systat Software Inc., San Jose, CA, USA) and a nonrectangular hyperbola (Leverenz & Jarvis, 1979; Lakatos et al., 2006).

Photosynthetic performance and thallus water content

A new methodological approach was used to model the WC of poikilohydric lichens in situ. Using noninvasive measurements of chlorophyll a fluorescence with low impact on the sensitive lichens, (1) the effective quantum yield of photosystem II (ΔF/Fm′) inline imagein situ clarified the metabolic activity during the day, and (2) the electron transport rate (ETR) combined with light curves in situ were used to recalculate the WC in a backwards approach determined by (3) ΔF/Fm′ measured ex situ at different known water contents with desiccation curves.

To obtain these measurements, two in situ and one ex situ set-ups were established. (1) Diel continuous photosynthetic performances were measured by analysing ΔF/Fmin situ following a noninvasive methodology with four to seven replicates per lichen type during 3 d without rain events. ΔF/Fm′ is an indicator of physiological activity and is affected by light intensity and the WC of the thallus. Because the natural light intensities intercepted by lichens in the understorey are rather homogeneous (Lakatos et al., 2006), significant change in ΔF/Fm′ is mainly attributable to desiccation processes. The variability of ΔF/Fm′ during the morning and late afternoon might reflect some heterogeneity caused by short sun flecks, because monitored lichens were facing east and west. These east- and west-facing positions prevent potential light fluctuations from around midday until the early afternoon. Continuous measurements were conducted from predawn at 05:00 h until sunset at 19:00 h every 2 h. (2) Basic response patterns to thallus WC were studied ex situ under semi-controlled conditions of 75% RH, using desiccation curves after full hydration of the thalli under both prevailing low light and saturating light intensities of 1 and 150 μmol m−2 s−1, respectively (3) Evaluation of the photosynthetic activities in situ without the availability of direct rain or stem water flow indicated the actual lichen performance and modelled WC under the environmental humidity and dewfall conditions. The lichens were measured at predawn (04:30–06:00 h) and midday (12:00–13:30 h) on randomly chosen days after at least 24 h without rain using rapid light response curves in the field. A saturating flash was given to determine the maximum fluorescence yield after dark adaptation (Fv/Fm). At regular 30-s intervals – sufficient time to reduce desiccation effects caused by the measurements – PFD was increased in eight steps from 20 to 300 μmol m−2 s−1.


Climate in the corticolous micro-environment

Examination of the temperature conditions of lichen thalli revealed fluctuations between 18 and 30°C (mean: 24.5 ± 2.3°C; Supporting Information Fig. S1). Bark temperature correlated closely with lichen temperature. Both lichen and bark temperatures followed ambient air temperature with a delay of 3:07 h (± 1:46 h; n = 6) during warming and cooling periods, leading to significant differences from air temperatures of generally 2–6°C (up to 9°C). This temperature delay in corticolous lichens, caused by buffering of the trunk material, induced colder values during the day (Fig. S2a) and warmer values at night (Fig. S2b) as compared with air temperatures. Furthermore, the maximum ranges and timing of the amplitudes of thallus and air temperature were significantly dissimilar (Kolmogorov–Smirnov test: < 0.001; n = 1729).

The RH in the understorey of the study site (mean: 89.3 ± 12%; minimum: 54%; n = 17 d) was below 80% for 50% of the daytime (Fig. S3). The RH was usually above 95% at night and in the morning until c. 07:00 h. It then declined to a minimum between 14:00 and 16:00 h and thereafter continuously increased again, exceeding 95% at c. 23:00 h at the latest. Variations in this general trend occurred, influenced by precipitation events and irradiance fluctuations.

Dew formation at the lichen’s surface

Nocturnal water droplets, commonly produced by dew formation in open vegetation early in the morning, were not observed on the surface of bark and the lichen thallus in the investigation area. This can be explained by the temperature differences between warmer bark and thallus temperatures compared with colder ambient temperatures during night and the calculated dew-point temperatures (Fig. 1; Notes S1 Eqn. 11). Instead, the calculation of the dew point showed that supersaturation (when the thallus temperature falls below the dew-point temperature), and thus the potential for dew formation, occurred from the morning hours until the early afternoon, in total during 25.3% of the day (Figs 1, 2).

Figure 1.

(a) Relative humidity (RH; regular dotted line), ambient temperature (Ta; bold dotted line), lichen surface temperature (Ts; grey line) and calculated dew-point temperature (Td; black line) in the understorey of the study site shown for four consecutive days. (b) The diel mean temperatures (Ta (open circles), Ts (grey line) and Td (black line)) with standard deviation averaged over six consecutive days (24–29 October).

Figure 2.

Calculated dew formation (in mm per 5 min) on six consecutive days (24–29 October).

Dew formation is dependent on radiation balance and turbulent fluxes at the lichen’s surface covering the bark. Based on an estimation of common wind conditions, we estimated daily dew formation at the lichen surface of up to 0.02 mm per 5 min from 09:00 h until the early afternoon –c. 12:00–15:30 h (Fig. 2). Available daily totals of dew water varied between 0.29 and 0.69 mm d−1 on individual days (Table 1).

Table 1.   Daily sums of dew formation using calculated wind speed and a fixed wind speed of 1 m s−1 on six consecutive and representative days
 Daily sum of dew formation (mm d−1)
24 October25 October26 October27 October28 October29 October
At calculated wind speed0.420.510.370.670.690.30
Fixed wind speed of 1 m s−

Dew calculation strongly depends on atmospheric stability and wind speed. Because the wind speed calculated using the wind extinction coefficient (Notes S1, Eqn 6) implies some uncertainty, we also modelled dew formation with a constant wind speed of 1 m s−1 to test the effect of increasing ventilation, beyond that usually experienced in the understoreys of tropical forests, on dew formation (Table 1; Leigh, 1999; Thomas, 2005). The comparison of dew formation under the calculated more realistic very weak wind (± calm conditions all day; average = 0.006 m s−1) and a fixed wind speed of 1 m s−1 demonstrated that dew formation would have been reduced or absent (29 October) at high turbulence.

Physiological response of epiphytic lichens to water availability

Basic response to water status  Four corticolous green-algal lichen species –Thelotrema alboolivaceum (Nyl.) Hale (Thelotremataceae), Herpothallon rubrocinctum (Ehrenb.: Fr.) Aptroot, Lücking & G. Thor (Arthoniaceae), Phyllopsora cf. corallina (Eschw.) Müll. Arg. (Bacidiaceae) and Coenogonium linkii Ehrenb. (Coenogoniaceae) – were selected as representative epiphytic lichens. These species typify the four common functional lichen types: (1) crustose with cortex, (2) crustose without cortex, (3) squamulose, and (4) filamentous (Lakatos et al., 2006). The first three species belong to the corticolous crustose type and adhere tightly to the bark. The filamentous C. linkii is horizontally exposed, and attached to the tree at an angle of almost 90°.

The basic responses to water status were determined from desiccation curves using chlorophyll a fluorescence at low and saturating light intensities, with the mean optimal WC of the lichen thalli (= 3–6 per species) ranging between 0.2 and 0.3 mm H2O (Fig. 3). The dark-adapted samples revealed potential quantum yields (Fv/Fm) of 0.61 and 0.65 (Table 2). During desiccation, all lichens showed a relatively stable ΔF/Fm′ of c. 0.64–0.5 at a PDF of 1 μmol m−2 s−1 over a broad range of WCs. ΔF/Fm′ and ETR150 decreased rapidly in the last 25% of relative WC (< 0.2 mm H2O) before photosynthesis stopped because of dehydration. The duration of desiccation was on average 3 h 10 min for T. alboolivaceum, 3 h for H. rubrocinctum, 2 h for P. corallina and < 1 h for C. linkii under similar experimental conditions. Maximal ETR150 ranged between 5 and 10 μmol electrons m−2 s−1.

Figure 3.

Photosynthesis ex situ shown as the apparent electron transport rate (ETRapp) and effective quantum yield of photosystem II (PS II) (ΔF/Fm′; n = 3; inset) of desiccating lichens at different water contents (mm H2O) and under both prevailing low light (triangle) and saturating light (circle) intensities of 1 and 150 μmol m−2 s−1, respectively, for (a) Thelotrema alboolivaceum, (b) Herpothallon rubrocinctum, (c) Phyllopsora cf. corallina and (d) Coenogonium linkii. The regression was fitted to ETR at 150 μmol m−2 s−1 (< 0.001) and was used to model the water content of lichens in situ (Table 2).

Table 2.   Cardinal points of mean (± SD) photosynthetic capacity (potential quantum yield (Fv/Fm), electron transport rate at the maximum and at 150 μmol m−2 s−1 (ETRmax, ETR150) as well as water content at optimal conditions and modeled (WCopt, WCmodel). Fv/Fm, ETRmax, ETR150 and water content (WC)), determined using desiccation curves ex situ (Fig. 3) and rapid light curves during predawn and midday periods in situ
 Thelotrema alboolivaceum (with cortex)Herpothallon rubrocinctum (without cortex)Phyllopsora corallina (squamulose)Coenogonium linkii (filamentous)Units
  1. Values were calculated following Lakatos et al. (2006) based on the mean of three (ex situ) and five (in situ) samples per species. The estimated water contents at predawn and midday (WCmodel) were recalculated using the corresponding ETR at 150 μmol m−2 s−1 from desiccation curves. Different letters within a row indicate significant differences (LSD post hoc test; < 0.05).

  2. 1Values were obtained from (Lakatos et al., 2006). WC, water content.

Ex situ
 Fv/Fm0.64 ± 0.02a0.61 ± 0.03b0.61 ± 0.03b0.65 ± 0.02aWithout dimension
 ETRmax110.2 ± 1.27a7.55 ± 0.55a7.52 ± 0.59a14.8 ± 1.36bμmol electrons m−2 s−1
 ETR1508.88 ± 1.19a5.11 ± 1.9b7.54 ± 1.7ab6.24 ± 1.29abμmol electrons m−2 s−1
 WCoptc. 0.2 −0.4> 0.15> 0.25> 0.10mm H2O
In situ predawn
 Fv/Fm0.61 ± 0.03a0.57 ± 0.04a0.59 ± 0.04a0.61 ± 0.04aWithout dimension
 ETRmax8.02 ± 2.81a3.03 ± 1.77b2.28 ± 0.76b1.78 ± 0.73bμmol electrons m−2 s−1
 ETR1506.36 ± 2.71a3.03 ± 1.63b2.03 ± 1.00b1.57 ± 0.99bμmol electrons m−2 s−1
 WCmodelc. 0.199c. 0.092c. 0.126c. 0.025mm H2O
In situ midday
 Fv/Fm0.45 ± 0.12a0.44 ± 0.05a0.42 ± 0.09a0.28 ± 0.09bWithout dimension
 ETRmax2.96 ± 2.09a2.14 ± 2.06ab1.45 ± 0.71b0.93 ± 1.13cμmol electrons m−2s−1
 ETR1502.96 ± 2.37a1.63 ± 2.33ab0.75 ± 0.84b0.31 ± 0.07bμmol electrons m−2 s−1
 WCmodelc. 0.062c. 0.05c. 0.104c. 0.005mm H2O

Diurnal photosynthetic performance  The diel photosynthetic performance analysed using ΔF/Fmin situ measurements indicated continuous metabolic activity and water availability throughout the day, with high ΔF/Fm′ values of between 0.66 and 0.55 from 05:00 to 19:00 h (Fig. 4). Only at midday was desiccation observed; the two lichen types with more air-exposed surfaces started to desiccate between 11:30 and 13:30 h, as indicated by significantly lower ΔF/Fm′ (ANOVA: F11:30(3,57): 4.808; P < 0.01; F13:30(3,70): 6.161; P < 0.001). The approximate water status, calculated on the basis of the photosynthetic response to water availability (Fig. 3), ranged between 0.3 and 0.6 mm H2O for the crustose and squamulose lichen types and between 0.2 and 0.6 mm H2O for the filamentous lichen C. linkii. This performance points to a continuous water supply for adpressed and stem temperature-influenced crustose lichens benefiting from dew formation, while the water status of the air-exposed lichen C. linkii declined to c. 0.2 mm H2O at around midday.

Figure 4.

Diel photosynthetic performance analysed using the effective quantum yield of photosystem II (PS II) (ΔF/Fm′) in situ during 3 d without a rain event. The four lichen types – filamentous (circle), crustose with cortex (triangle), crustose without cortex (square) and squamulose (diamond) – were measured from 05:00 h until 19:00 h (Mean, SD, n = 4–7). All measurements were conducted within a 30-min time period at the indicated times during the day, and are placed next to each other for convenience. Horizontal bars indicate the general periods of available water from rainfall, relative humidity (RH) above 80% and dew formation for activating photosynthesis with the measurements compiled from Figs 1 and 2 (white space with text indicates periods of water availability; hatched bars indicate periods when the resource was not available; asterisks indicate significant differences of yield by ANOVA-test with < 0.05, < 0.01 and < 0.001).

Water availability and photosynthetic performance at predawn and midday  Predawn measurements of chlorophyll fluorescence after at least 24 h without previous rain events demonstrated again that all studied green-algal lichens were metabolically active, with Fv/Fm values of 0.57–0.61. These values were not significantly different from those achieved under optimal conditions ex situ (Table 2) and under continuous measurements in situ (Fig. 4). The studied corticolous lichens can therefore potentially achieve metabolic activity every morning, and are not restricted by rain supply. However, the photosynthetic performance assessed using light response curves in situ differed among species (Table 2). The ETRmax predawn had values of 8.0, 3.0, 2.0 and 1.8 μmol m−2 s−1, corresponding to 78, 40, 30 and 12% of ETRmaxex situ for T. alboolivaceum, H. rubrocinctum, P. corallina and C. linkii, respectively (Table 2). The ETR values measured in the morning thus reflected the fact that the lichens were not at optimal water conditions. A precise estimation of predawn WC in situ can be obtained from the predawn ETR150 corresponding to the WC of ETR150 in the experimental desiccation curves (Fig. 3). Accordingly, the WC at predawn was approximately 0.199 mm H2O for T. alboolivaceum and 0.092 mm H2O for H. rubrocinctum, but only 0.025 mm H2O for C. linkii.

At midday, ΔF/Fm′ values between 0.45 and 0.28 indicated photosynthetic activity with ETRmax values of 2.9, 2.1, 1.5 and 0.9 μmol m−2 s−1, corresponding to 29, 28, 19 and 6% of ETRmaxex situ for T. alboolivaceum, H. rubrocinctum, P. corallina and C. linkii, respectively (Table 2). According to the corresponding WC of ETR150 of the experimental desiccation curves, the WC at midday was between c. 0.10 mm H2O for P. corallina and c. 0.05 mm H2O for H. rubrocinctum, while only 0.005 mm H2O was achieved for C. linkii. These values indicate slower water loss, or additional water uptake for the adpressed lichen growth forms, while the exposed filamentous C. linkii did undergo strong desiccation. This confirms again the beneficial effect of diurnal dew formation for adpressed lichens which are highly influenced by stem temperatures.


Delayed heat storage induces cooler stem compared with air temperatures in the morning

One significant component of the surface and energy budget of a tall forest is the total heat storage (c. 5–10% of net radiation) which is affected by humidity, air temperature and the biomass of the canopy and the stem (e.g. Moore & Fisch, 1986). Approximately 80 W m−2 of maximal absolute hourly biomass heat storage fluxes were obtained for a 35-m-tall Amazonian tropical forest (Moore & Fisch, 1986) and similar values for a 30-m-tall Eucalyptus forest in Western Australia (Silberstein et al., 2001), while 63 W m−2 was reported for a 40-m-tall Eucalyptus forest in Eastern Australia (Haverd et al., 2007). In such cases, the absorption of net radiation by the canopy during the daytime and heat storage by the stem are important. The change in biomass heat storage is driven by radial heat diffusion within the stems and surface heat exchange by convection, insolation and longwave radiation (Haverd et al., 2007). In the early morning, the heat storage of air is maximal, corresponding to a rapid change of temperature. Radiative heat exchange with the environment takes place at the top of the canopy, and radiative energy is transferred slowly to the lower levels of stems because of their thermal properties, with a thermal conductivity 10-fold higher and a specific heat capacity 2-fold higher than that of air (Aston, 1985; Moore & Fisch, 1986; Haverd et al., 2007). More rapidly, the sun evaporates water from the canopy, which is carried down by turbulent air, promoting warm humid air in the understorey (e.g. Moore & Fisch, 1986). This process leads to warmer ambient air temperatures and cooler stem and bark temperatures during the daytime, a phenomenon that continues sometimes until late afternoon. This principle of delayed warming of stems in comparison to ambient air is not restricted only to tall forests (Aston, 1985; Moore & Fisch, 1986; Haverd et al., 2007); it has also been reported for a mixed 19-m-tall forest in Ontario, Canada (McCaughey, 1985) and for a 10-m-tall walnut (Juglands regia) orchard in California, USA (Garai et al., 2010). Thus, a colder stem temperature in the morning seems to be a common phenomenon, and may drive dew formation on the bark and epiphytic organisms under specific environmental conditions world-wide.

Dew formation occurs on bark and corticolous lichens during the daytime

Dew formation occurs as a consequence not only of surfaces being cooler than ambient air, but also as a result of the dew-point temperature being reached and wind speed being low. The delayed changes in microclimatic conditions of the bark and lichens studied here caused temperatures to be warmer during the night and colder during the day compared with the ambient air. Because of the close contact of crustose and adpressed lichens with the bark, the thallus temperatures of the three crustose lichens were tightly linked with the bark temperature. The lichens benefited from the cooling effect of the bark, with temperatures below the dew-point of ambient air for 50% of the daytime, and hence potential dew formation. The estimated amount of 0.3–0.7 mm dew d–1 represents a significant water source for these lichens, because their optimal WC for performing photosynthesis ranges between 0.1 and 0.3 mm H2O. This daily sum of dew formation is not available at once, but our conservative approximation suggests a deposition sum of up to 0.176 mm H2O h–1 on the lichen surface around noon. In view of their photosynthetic capacity, the maximum gain of 0.01 mm H2O (5 min)−1 of additional water around noon (Fig. 2) leads to an increase of c. 10% in ETR at low water contents (Fig. 3). However, atmospheric stability and wind speed strongly influence the approximation of the dew calculation and imply some degree of uncertainty. The wind speed is often < 25 cm s−1 and even in tree fall gaps on Barro Colorado Island rarely exceeds 50 cm s−1 (Leigh, 1999). As a consequence, turbulent mixing is largely negligible in tropical understoreys (e.g. Szarzynski & Anhuf, 2001; Thomas, 2005). Thus, we are confident that our modelled estimates of dew formation are plausible, and are confirmed by our chlorophyll fluorescence activity measurements.

Midday dew as an additional water source to prolong photosynthetic performance

Potential water sources for poikilohydric epiphytes are precipitation, humidity and deposition by dew and fog. On the one hand, it is obvious that the high amount of precipitation in tropical lowland rain forests frequently causes stem water flow. The amount of water supplied by stem flow in tropical forests ranges from 0.4% (Cavelier et al., 1997) to 8% of the net precipitation (Jordan & Heuveldrop, 1981), and should be sufficient to support the water requirements for photosynthesis of the lichen thallus after heavy rain events. For example, in central Guyana, Jetten (1996) showed that stem water flow is only detectable during heavy rainstorms in the rainy season. Based on measurements during rainstorms, he concluded that stem water flow is only 0.63% of total rainfall. This would mean c. 19 mm of stem flow per year for our study site – a neglectable component compared with the estimated annual dew water deposition of c. 179 mm shown in the present study. On the other hand, stem water flow does not occur during light rain events and in the dry season. Rainfall measurements at the study site revealed that 85 d over the year are without rain (Table 3). Moisture conditions are particularly harsh during the dry season when approximately half the days (48%) are lacking rainfall. While 43% of the rainless periods last only 1 d in the study area, there are also considerable numbers of longer dry spells with a duration > 2 d (40%), which is particularly critical regarding desiccation processes (J. Bendix & A. Obregón, unpublished data). Some tropical ecosystems have a more pronounced dry season; for example, 17 wk in Barro Colorado Island, Panama with rainfall mostly < 12 mm (Windsor, 1990). As a consequence, lichens are not rehydrated during such events and experience desiccation during the daytime (Zotz & Winter, 1994). Another issue is that the prevailing light intensities in the understorey are below the light compensation points of the studied lichens. To achieve positive carbon balances, prolonged hydration is necessary to make use of the interception of infrequent sunflecks (Lakatos et al., 2006). Utilizing different water sources to maintain sufficient water content for photosynthesis during the daytime is thus a crucial factor for carbon gain, and at least mitigates respiratory carbon loss.

Table 3.   Days without rain at the Les Nouragues National Park, French Guiana
 Percentage of dry daysDry days per year
  1. Measurements were conducted 6 km south of the study site at Saut Parare during the period August 2007 to April 2010 with indications of dry and rainy seasons.

Dry seasons (15 August to 14 November)4844
Rainy seasons (15 November to 14 August)1541

Two microclimatic processes are described here to be important for maintaining photosynthesis in these lichens at an adequate water supply: water vapour uptake and dew formation. Since the pioneering discovery of Butin (1954), numerous laboratory and field studies have been performed, resulting in the consensus that water vapour above 85% RH can reactivate photosynthesis in green-algal lichens (e.g. Lange et al., 1989, 2001; Green et al., 2002; Lakatos, 2011). By contrast, in lower air humidity of 80% RH the net photosynthesis of very few lichens can be reactivated (Bertsch, 1966; Brock, 1975; Nash et al., 1990). In the present study, RH at night was usually above 80% and could metabolically activate the lichens (Fig. S3). The average course of the day starts after dawn, with high RH supporting metabolic activity of the lichens. At 08:00 h lichen temperatures decline below the dew point and high RH together with potential dew formation facilitates water uptake until c. 11:00 h. Humidity then drops below 80% RH, and for the next 3.5 h potential dew formation alone provides liquid water for metabolic activity until c. 14:30 h, when lichen temperature increase above the dew-point temperature and the thallus starts to desiccate without an additional water supply. Finally, humidity increases above 80% RH again at c. 16:30 h and may reduce desiccation or even reactivate metabolism (Figs 1, 4). Thus, during the daytime, RH above 80% supports metabolic activity for 4 h, while for 6 h potential dew formation provides liquid water for photosynthesis. These observations indicate that green-algal lichens benefit from both dew formation and humidity for more than three-quarters of the daytime. Additionally, dew formation occurs particularly during a time of day (8:00 to 14:30 h) when rainfall probability is low (Fig. S4).

Lichens benefit differently from micro-conditions

Under dawn conditions, crustose adpressed lichens potentially achieved 30–78% of the maximal photosynthetic rate (ETRmax) with a WC of c. 0.1 mm, while the air-exposed C. linkii achieved only 12% of the ETR at an estimated WC of 0.025 mm (Table 2). For photosynthetic performance at dawn, this indicates that adpressed lichens benefit more from the co-action of different water sources during the night. Also, during the course of the day the filamentous and air-exposed C. linkii desiccated to a greater extent than the adpressed lichens (Table 2, Fig. 4). Water uptake and loss depend on the morphological properties of lichens that allow them to absorb and retain water as well as on evaporation rates influenced by air humidity (Pardow et al., 2010). The crustose T. alboolivaceum, for example, profits from a thick cortex enabling it to take up a high amount of water, and full desiccation of this lichen was not observed in the field (Lakatos et al., 2006). In comparison, the crustose H. rubrocinctum, which does not have a cortex, and the squamulose P. corallina, which has a porous upper surface, are exposed to faster water loss, as shown here.

In contrast to the vertically exposed lichens, the horizontally growing C. linkii, with distances of their thallus margins to the bark surface of up to 4 cm, revealed differences in its thallus temperature and light absorption. The boundary temperature measurements at a distance of 1 cm above the bark were correlated with those of ambient air (Figs S1 and S2). This implies that horizontally growing C. linkii and other air-exposed lichens are less influenced by the bark temperature causing dew formation on the surface. Thus, more exposed epiphytes such as C. linkii may not make use of this principle of midday dew formation.

Ecological relevance of dew formation for nonvascular plants

The positive consequences of dew and/or fog for the water status of poikilohydric plants have been demonstrated in temperate grasslands (Csintalan et al., 2000), tropical mountain forests (Leon-Vargas et al., 2006), coastal Mediterranean forests (Hartard et al., 2008), semi-arid and arid deserts (Lange et al., 1992, 2007; Agam & Berliner, 2006; Zhang et al., 2009), and the Antarctic (Büdel et al., 2008). Dew rates in the mid-latitudes are typically c. 0.5 mm per 10-h night (Jacobs et al., 2008). In deserts, dew results in 53–63% of integrated carbon gain in lichens (Lange et al., 2007), with dew formation rates ranging from 0.14 to 0.56 mm d−1 (Zhang et al., 2009). Estimation of the source of thallus hydration in coastal Mediterranean forests revealed that dew formation was responsible for 15% and atmospheric water vapour uptake for 11% of calculated average annual gross CO2 fixation (Matthes-Sears & Nash, 1986). In these ecosystems, the general pattern of metabolic activation through moistening by dew and/or fog without rain can lead to photosynthetic gain in the early morning hours, followed by subsequent dehydration with corresponding inactivation of photosynthesis. For the remaining parts of the day, nonvascular plants stay desiccated until the late afternoon, when they may be rehydrated by water vapour at higher humidity. These common activity patterns in radiation- and wind-exposed microhabitats are different from those in shaded and calm microhabitats, as described here. Thus, here we provide for the first time evidence of diurnal dew formation on bark, at a rate of up to 0.7 mm dew per day, as an ecologically relevant water source for lichens, enabling them to extend photosynthetic activity to the whole day.


The previously unknown mechanism of diurnal dew formation affecting the photosynthetic activity of lichens in the understorey of wet tropical rain forests may also occur in other ecosystems. It may be a general principle, as colder stem vs air temperatures for several hours in the morning have been reported for different types of forests world-wide (Aston, 1985; McCaughey, 1985; Moore & Fisch, 1986; Agam & Berliner, 2006; Haverd et al., 2007; Garai et al., 2010). Moreover, this overlooked mechanism of diurnal dew formation contributes not only to the water supply of lichen fungi and their bacterial and algal symbionts, but also to all corticolous poikilohydric phototrophs, such as free-living cyanobacteria and algae, as well as bryophytes (Lakatos, 2011). This mechanism thus has beneficial effects on primary production within the corticolous food web. These dew effects could also cause adverse conditions for the lichens and trees, as the increased moisture may facilitate growth of harmful bacteria and fungi. In the context of climate and land-use change, mechanisms such as midday dew may become more important for buffering decreasing access to water during increasing drought conditions. Further investigations will reveal if the significance of midday dew in forest environments will have implications for the disciplines of ecology, forestry, hydrology and biogeochemistry.


We are grateful to J. Chave and P. Gaucher for the opportunity to stay at the Les Nouragues field station (USR 3456) in French Guiana. The authors gratefully acknowledge R. Wirth, T. A. G. Green, D. Ackerly and two anonymous reviewers for helpful comments on the manuscript. We thank F. Ulm and U. Rascher for their field assistance. The study was supported by the EC (DG/FJ/042/98), the CNRS (PIR Amazonie grant), the German Research Foundation DFG (LA 1426/6-1; BE 1780/13-1) and the German exchange programme DAAD.