• The flux of ultraviolet (UV)-B radiation to the Earth's surface is increasing, particularly in high latitudes. We studied the sensitivity of some dominant plant species of boreal and subarctic peatlands to this increase.
• Intact peat monoliths with the mosses Sphagnum balticum and Sphagnum papillosum, and cotton grass (Eriophorum vaginatum) were exposed to ambient solar UV-B or ambient solar UV-B supplemented by 30% in a field experiment in central Finland.
• Although the UV-B dose was low during the growing season, owing to frequent cloudiness, both Sphagnum species showed significantly higher membrane permeability under enhanced UV-B. In S. balticum, UV-B tended to decrease the capitulum dry mass and induced a 30–40% increase in the concentration of chlorophyll and carotenoid pigments. Enhanced UV-B had no effects on leaf morphology, chlorophyll fluorescence or stomatal functioning in E. vaginatum.
• The various UV-B responses in the Sphagnum species under investigation indicate that they may be sensitive even to small increases in solar UV-B radiation. By contrast, E. vaginatum appeared to tolerate the UV-B fluxes of the experiment.
Sphagnum mosses cover large areas in the boreal and subarctic regions, especially on peatlands and tundra. They are indispensable for peatlands as they create the conditions that favour carbon accumulation as peat (van Breemen, 1995). Clymo & Hayward (1982) speculated that there might be more carbon stored in living and dead Sphagnum than in any other genus of plants. According to a recent model estimate, northern peatlands harbour 270–370 Pg (1 Pg = 1015g) of carbon, and the accumulation rate is 0.066 Pg C per year (Turunen et al., 2002).
There are few studies on the potential effects of increasing UV-B on Sphagnum. Gehrke (1998) found that UV-B exposure simulating 15% ozone depletion alters shoot morphology and decreases dark respiration of a hummock species Sphagnum fuscum. We have demonstrated that UV-B induces the leakage of ions from the tissues of Sphagnum angustifolium and Sphagnum magellanicum (Niemi et al., 2002). In southern Argentina, near-ambient solar UV-B (90% of ambient UV-B) reduced the height growth of S. magellanicum compared with the plants grown under attenuated UV-B (15–20% of ambient UV-B) for two or more growing seasons (Ballaréet al., 2001).
Cotton grass (Eriophorum vaginatum) is a widespread sedge throughout the circumboreal tundra and nutrient-poor peatlands (Wein, 1973). Sedges enhance the emission of a significant greenhouse gas – methane – by releasing substrates from the root system for methanogenesis and by serving as a gas transfer route within the aerenchymatous space in the roots and shoots (Schütz et al., 1991; Joabsson et al., 1999). Despite its importance to carbon gas cycling in peatlands, E. vaginatum has been neglected in UV-B research. In the sunny summer of 1999, we found that a 30% supplement to ambient UV-B in central Finland reduced the leaf cross-sectional area by 25% and the relative proportion of the gas cavities in the leaf cross-section by 14% (Niemi et al., 2002). However, owing to a lack of physiological measurements, it is not known whether these morphological changes were associated with alterations in the physiology of E. vaginatum.
The purpose of this experiment was to study further the UV-B sensitivity of representative species of peatland vegetation, Sphagnum mosses and E. vaginatum. Our objective was first to determine whether increasing UV-B flux affects the stomatal functioning, chlorophyll fluorescence and leaf anatomy of E. vaginatum. Second, we characterized the UV-B responses of Sphagnum balticum and Sphagnum papillosum with the aid of variables used in previous studies (analyses of pigment concentrations, membrane permeability and capitulum dry mass) supplemented with a new one (an analysis of the carbon isotope ratio), which might reflect the effects of UV-B. Furthermore, we were able to demonstrate year-to-year variation in the UV-B responses of vegetation as a result of differences in the UV-B doses between growing seasons.
Materials and Methods
Peatland cores and UV-B exposure
Peatland cores with a natural vegetation cover were exposed to ambient solar flux of UV-B radiation or to a flux supplemented with lamps to 30% above the ambient UV-B weighted with the erythemal (‘skin-reddening’) action spectrum at an outdoor site in Kuopio, Central Finland (62°13′ N, 27°35′ E, 80 m above sea level (a.s.l.)). Exposure was started on 2 June and terminated on 5 September, 2000. The peatland samples (10.5 cm diameter, 40 cm deep) were cored into polyvinylchloride (PVC) tubes at an Eriophorum lawn at an oligotrophic fen of the Salmisuo mire complex in Ilomantsi, Eastern Finland (62°47′ N, 30°56′ E, 145 m a.s.l), and the tubes were plugged at the bottom. The vegetation was composed of a near-continuous S. balticum (Russ.) C. Jens and S. papillosum H. Lindb. moss matrix. The vascular plants were dominated by E. vaginatum L., with scattered Andromeda polifolia L. and Vaccinium oxycoccus L. The water table of the cores was maintained at 2 cm below the Sphagnum surface by replacing the evaporation loss, as necessary, with distilled water.
The experiment was a randomized complete block design with four blocks (replicates) each containing one lamp array of the supplemental UV-B treatment and one of the ambient control. Each of the eight lamp arrays was formed of an aluminium frame (3.0 × 1.2 m) with six Philips TL40/12 lamps (Philips Lighting, Eindhoven, The Netherlands) positioned at 130 cm above the Sphagnum surface. Four peatland cores were placed in a PVC container filled with circulating lake-water under each lamp array. For the UV-B treatment, the lamps were wrapped in 100-µm cellulose diacetate film (Expopak, Jäminkipohja, Finland), which cuts off the radiation below 290 nm. The ambient control was achieved using nonenergized lamps that provided equal shading to that occurring in the UV-B treatment. The lamp output was modulated by an electronic controller operated by PC software to maintain the UV radiation under the UV-B treatment at a constant 30% elevation above that measured with an erythemally weighted broad-band sensor (Davis Instruments, Hayward, CA, USA) under an ambient control. Similar sensors monitored UV-B under each UV-B treatment array at the level of the vegetation in the centre of the exposed area. The lamps were automatically switched off when irradiation fell below 10 mW m−2 s−1. A UV-A control was not included in the experimental design because we had already shown that there were no clear differences between the ambient and UV-A controls (Niemi et al., 2002).
Analyses of E. vaginatum
The chlorophyll a fluorescence of six intact current-year E. vaginatum leaves per lamp array was measured with a portable pulse-modulated fluorometer (FMS 2; Hansatech, King's Lynn, Norfolk, UK) on 30 June, 24 July, 18 August and 29 August, 2000. After a 30-min dark-adaptation, the leaves were subjected to a weak modulated beam in order to obtain a minimum fluorescence yield (Fo) followed by an exposure to a saturating white-light pulse in order to obtain a maximum fluorescence yield (Fm). The quantum efficiency of photosystem II (PSII) in the dark is Fv/Fm, where Fv = Fm − Fo. The quantum efficiency of PSII in the light (ΦPSII) was determined by another sequence of exposure to a saturating light beam after the leaves were adapted to actinic light (1200 µmol m−2 s−1) produced with a halogen lamp in the fluorometer.
The stomatal conductance of six current-year E. vaginatum leaves beneath each lamp array was measured on July 19 and August 15 with an LI-1600 steady-state porometer (Li-Cor, Lincoln, NE, USA) using a cuvette designed for conifer needles. The leaf area was approximated and the same value was used for all the leaves measured. At the end of the experiment, six leaves per lamp array were sampled for the determination of stomatal density. The top 3–8 cm of the leaf was glued onto a microscope slide with superglue, and subsequently removed so that a replica of the leaf surface remained on the glass. Eriophorum vaginatum leaves are approximately triangular in cross-section, and the narrowest of the sides (abaxial surface) lacks stomata (see Fig. 1). Stomatal density per mm2 adaxial leaf surface was counted from six sites per leaf using sampling frames under a light microscope.
Eight current-year E. vaginatum leaves per lamp array were sampled in August for a structural study. The leaves were cut with a razor blade, 3 cm from the apex, into 1.5-mm long pieces which were immediately immersed in 2.5% glutaraldehyde fixative in a Na-phosphate buffer (0.1 m, pH 7.0) and stored at 4°C over night. The samples were postfixed in a 1% buffered osmium tetroxide (OsO4) solution at 4°C for 5 h, dehydrated in a graded ethanol series and embedded in LX 122 Epon (Ladd Research Industries, Burlington, VT, USA). Five transverse sections per lamp array were cut semithin onto microscope slides, stained with toluidine blue and observed with a Zeiss-Axiolab (Carl Zeiss, Jena, Germany) light microscope. Samples were photographed, and the digital light micrographs were used to measure the area of the leaf cross-section and the relative proportion of aerenchymatous space per leaf cross-section (Fig. 1) with Adobe Photoshop 5.0 (Adobe Systems, San Jose, CA, USA).
Analyses of Sphagnum mosses
On 28 August 2000, four shoots of S. balticum and of S. papillosum were collected per lamp array for membrane permeability tests, and placed in a humidity chamber for 2 h. Next, the top 2 cm of the shoot was cut and immersed in 5 ml deionized water. After 2 h the conductivity of the leachate was measured with a glass platinum electrode of a Philips PW 9505/20 (NV Philips, UK) conductivity meter. Concentrations of K+, Mg2+ and Ca2+ in the leachate were determined with an atomic absorption spectrophotometer (460; Perkin-Elmer, Norwalk, CT, USA) in an air–acetylene flame. Conductivity and ion leakages were calculated per sample dry weight.
At the end of the experiment, on 5 September, eight S. balticum shoots per lamp array were collected, and the capitula and 1.0-cm sections of the stem below the capitulum were cut. The dry mass of each capitulum and a 1.0-cm stem section was determined after 24 h at 60°C.
Apical 1.5-cm pieces of S. balticum and S. papillosum shoots were collected into liquid nitrogen as pooled samples per lamp array on September 5, freeze-dried and analysed as three (pigments) or six (stable carbon isotopes) subsamples each containing several Sphagnum pieces.
Concentrations of chlorophyll a (chla), chlorophyll b (chlb) and carotenoid pigments were analysed using the DMSO (dimethylsulphoxide) method (Barnes et al., 1992). The content of methanol-extractable UV-B-absorbing compounds was determined according to a protocol described by Niemi et al. (2002). The absorbance between 280 nm and 320 nm was scanned with a UV/visible spectrophotometer (Lambda Bio; Perkin-Elmer, Wilton, CT, USA) after the extraction of the samples in methanol–H2O–HCl solution (70 : 29 : 1, v : v : v) and a 10-min centrifugation at 600 g.
The powdered samples (0.3–0.6 mg) collected for determination of stable carbon isotope ratios (δ13C) were combusted on an elemental analyser (NC 2500; CE Instruments, Milan, Italy) and the CO2 produced was measured on a gas isotope ratio mass spectrometer (Delta + XL; Finnigan MAT, Bremen, Germany). The results are given as delta values δ13C = (Rsample/Rstandard − 1)1000, where R is the carbon isotope ratio 13C : 12C for sample and V-PDB standard, respectively. Two laboratory references of cellulose were analysed in parallel with plant samples. Based on repeated analyses of reference and plant samples, the precision of the analyses was ±0.1‰ or better.
The differences between the UV-B exposure and the ambient control within each species were tested with a Student's t-test using the SPSS 10.0 package (SPSS, Chicago, IL, USA). The t-tests were always performed using values of each variable averaged within a lamp array (n = 4). The nonnormality of membrane permeability was corrected, before further analyses, by log-transformation.
Erythemally weighted solar UV-B radiation remained low during the 2000 growing season, owing to frequent cloudy conditions. The daily doses of ambient radiation ranged between 0.29 kJ m−2 d−1 (June 8) and 2.33 kJ m−2 d−1 (July 1). The ambient UV-B dose accumulated over the whole experiment was 21% lower that that of the 1999 field season (Fig. 2; Niemi et al., 2002).
UV-B responses of E. vaginatum
Eriophorum vaginatum showed no significant UV-B responses in the chlorophyll fluorescence parameters (Fo, Fv/Fm, and ΦPSII) or stomatal conductance on any date of measurement (P > 0.1). Results for the measurements in mid-July, when the UV-B fluxes were at their highest, are shown in Table 1. Enhanced UV-B had no effects on the leaf cross-sectional area (P > 0.9), the relative proportion of aerenchymatous space per leaf cross-section (P > 0.7) or stomatal density (P > 0.3; Table 1).
Table 1. Physiological and morphological variables of Eriophorum vaginatum exposed to ambient or enhanced UV-B
Fo, minimum fluorescence yield; Fv/Fm, quantum efficiency of photosystem II in dark; ΦPSII, quantum efficiency of photosystem II in light; Values are means of four Lamp arrays ± SE. The t-tests showed no significant differences between treatments.
24.0 ± 3.1
25.0 ± 2.9
0.89 ± 0.01
0.89 ± 0.01
0.44 ± 0.02
0.47 ± 0.03
Stomatal conductance (cm s−1)
0.94 ± 0.20
0.78 ± 0.09
Stomatal density (mm−2)
183 ± 2
190 ± 6
Leaf cross-section (mm2)
0.31 ± 0.01
0.31 ± 0.02
Aerenchymatous space/leaf cross-section (%)
47.4 ± 1.0
47.8 ± 0.6
UV-B responses of Sphagnum
Membrane permeability, measured as conductivity, was almost doubled by UV-B exposure in both S. balticum and S. papillosum (Fig. 3). The slightly increased leakages of K+ and Ca2+ in S. balticum were not statistically significant (Fig. 3a). In S. papillosum, the leakages of Mg2+ and Ca2+ were 180% and 350% higher under UV-B exposure compared with ambient conditions, while the leakage of K+ was only slightly increased (Fig. 3b).
Enhanced UV-B slightly reduced the dry mass of the S. balticum capitulum compared with the ambient control. The values were 7.88 ± 0.86 mg in the ambient control and 6.20 ± 0.16 mg in the UV-B treatment (mean ± SE; P = 0.11). The dry mass of the 1.0-cm stem section below the capitulum was 5.28 ± 0.33 mg in the ambient control and 4.20 ± 0.54 mg in the UV-B treatment (P = 0.12).
Enhanced UV-B induced a significant increase in the concentration of chl and carotenoid pigments in S. balticum (P < 0.05). This increase was 40% in chla, 29% in chlb, 37% in total chl and 32% in carotenoid concentrations (Fig. 4a). The quotient of chla to chlb was significantly raised by enhanced UV-B from 3.02 ± 0.04 to 3.27 ± 0.09. In S. papillosum, no statistically significant changes were found in the chl or carotenoid concentrations; however, we did note some subtle trends towards an increase (Fig. 4b). The ratio of total chl to carotenoids was slightly higher under enhanced UV-B in both species (values not shown, P = 0.14). The amount of the methanol-extractable UV-B-absorbing compounds was unaffected by UV-B in both Sphagnum species (P > 0.6, data not shown).
No effects of enhanced UV-B were observed on the carbon isotope discrimination in either species (P > 0.4). The δ13C values for S. balticum ranged between −25.4 and −24.2‰ and those for S. papillosum between −26.2 and −24.3‰.
Continuous 30% supplementation of ambient UV-B flux for a growing season that was cloudy and rainy compared with the average (Finnish Meteorological Institute, 2000) altered the pigment content and membrane permeability in Sphagnum mosses. No UV-B responses were observed in the variables measured in Eriophorum vaginatum.
Summers can differ drastically in the solar UV-B doses because of differences in cloud cover. The ambient UV-B dose during the current experiment was 21% lower than that during the previous year and, hence, the current UV-B treatment was nearly the same as the ambient conditions of the preceding growing season. The UV-B exposures in ecological studies follow either a square-wave system in which the constant or step-wise UV-B supplement simulates an ozone depletion scenario under clear-sky conditions, or a modulated system in which the supplement follows a constantly monitored ambient solar UV-B (McLeod, 1997). In this particular growing season a square-wave enhancement would have resulted in a supplement greatly in excess, while our modulated system retained the UV-B enhancement as intended, at 30% above the ambient.
No morphological changes were observed in E. vaginatum during this experiment, in contrast to the experiment in 1999 when UV-B induced a reduction in the leaf cross-sectional area and proportion of the aerenchymatous space (Niemi et al., 2002). No stomatal or chlorophyll fluorescence responses to UV-B were observed in the current experiment either. The lack of UV-B effects may be because the UV-B levels were too low to induce changes in E. vaginatum. Therefore, the changes in the morphology and functioning of E. vaginatum may be surprising if the estimated future increase in UV-B is attenuated by a heavy cloud cover. It is known that several grass species and a sedge (Carex arenaria) contain constitutive concentrations of flavonoids high enough to protect them against ambient and above-ambient UV-B fluxes (van de Staaij et al., 2002). However, our previous results (Niemi et al., 2002) indicate that in clear-skied summers, realistic UV-B increases can lead to alterations in E. vaginatum. Furthermore, the UV-B sensitivity of grasses has been reported before by Gwynn-Jones & Johanson (1996), who found that two Calamagrostis species show several responses to UV-B manipulations, and by Day et al. (1999), who observed an improvement of vegetative growth in the tussock grass (Deschampsia antarctica) following a reduction in the UV flux.
Despite low radiation fluxes, enhanced UV-B altered the pigment content and membrane permeability in Sphagnum, together with a subtle change in dry weight. Higher membrane permeability under enhanced UV-B is in accord with the results of the previous summer, when the UV-B supplement led to greater leakages of Mg2+ and Ca2+ from S. angustifolium and S. magellanicum (Niemi et al., 2002). In S. papillosum, there was a trend towards a parallel shift (Niemi et al., 2002). All the Sphagnum species that were studied here are prone to losing ions from the cells, owing to changes in membrane permeability, even with the present solar UV-B flux. However, it should be noted that in addition to protoplast, Mg2+ and Ca2+ may also originate from cation exchange sites in the cell wall.
The dry mass of the S. balticum capitula was slightly reduced by enhanced UV-B, suggesting alterations in net photosynthesis or carbon allocation. A lower dry mass following outdoor UV-B exposure has been found in the moss Hylocomium splendens (Gehrke, 1999), whereas a greater dry mass has been observed in S. fuscum (Gehrke, 1998). Ultraviolet-B exposure has been observed to have no effects on capitulum dry mass in S. angustifolium (Niemi et al., 2002).
There was no change in the carbon isotope ratios, indicating that UV-B had no effect on the isotopic fractionation by Rubisco (assuming that the water content in Sphagnum and the isotopic composition of source CO2 were equal; Farquhar et al., 1989). However, when the mosses are well-watered, as in this experiment, the hyaline cells surrounding the chlorophyllose cells are full of water and the isotopic fractionation by Rubisco is minor compared with diffusional resistance (Rice & Giles, 1996), and differences in enzyme functioning may be difficult to detect.
Unlike Gehrke (1998), who found a reduction, and Searles et al. (1999) and Niemi et al. (2002), who reported no clear changes, we observed an increase in the concentrations of chla, chlb and carotenoid pigments in S. balticum under enhanced UV-B. In S. papillosum we discovered a trend of parallel change. The observed accumulation indicates that supplementing UV-B flux has either stimulated the synthesis or depressed the degradation of these pigments. The higher pigment concentration may result from a compensatory process that maintains photosynthetic efficiency under enhanced UV-B. The chla, chlb and carotenoid concentrations in S. papillosum under enhanced UV-B in this experiment (1.05, 0.35 and 0.31 mg g−1 dry wt, respectively) were equal to those in the ambient control of the previous summer on approximately the same date (1.01, 0.35 and 0.33 mg g−1 dry wt; Niemi et al., 2002). Since the UV-B dose of these two treatments was of the same magnitude, it appears that a moderate UV-B flux is needed for an induction of pigment accumulation. The flux of ambient UV-B may not have been great enough for this in the current experiment. The slight increase in the quotient of total chl to carotenoids in both species is similar to a finding presented by Gehrke (1998). However, it is not known why the synthesis of chl has been stimulated more by UV-B than that of carotenoids.
The accumulation of UV-B-absorbing compounds has rarely been observed in outdoor UV-B studies with mosses (Gehrke, 1998; Gehrke, 1999; Searles et al., 1999; Niemi et al., 2002). Also, in this experiment, the amount of methanol-extractable UV-B-absorbing compounds remained unaffected by UV-B in both Sphagnum species. As Liakoura et al. (2001) have discussed, the amount of UV-B-absorbing compounds may be underestimated by a methanol-extraction if the major part of the compounds is bound to the cell walls. However, Searles et al. (1999) used a method to extract cell wall-bound pigments and observed no responses in S. magellanicum to the attenuation of ambient UV-B.
The competition for light and water forces Sphagnum shoots to grow in one layer (Hayward & Clymo, 1983); this means that all the capitula are exposed to solar radiation. The sensitivity of Sphagnum mosses to UV-B is to be expected, since the leaves are only one cell-layer thick, and lack any epidermal structures attenuating the penetration of UV-B (hairs, waxes or a cuticle) and do not accumulate significant amounts of UV-B-absorbing compounds. The alterations induced by UV-B in the present study generally tended towards a similar direction in both S. balticum and S. papillosum. Therefore, neither of the species appears more resistant to UV-B than the other.
Although the depletion of stratospheric ozone may follow calculated model estimates, future ground-level UV-B flux will still vary unpredictably according to cloudiness (Blumthaler, 1993). Our results suggest that a vascular plant, E. vaginatum is resistant to UV-B at a dose equivalent to a sunny summer in Central Finland today, but may show significant responses to a 30% increment on those fluxes. This appears to correlate with a meta-analysis of 62 UV-B field experiments (Searles et al., 2001), which showed that UV-B corresponding to 10–20% ozone depletion has only modest effects on vascular plants in outdoor studies. Our study demonstrates that Sphagnum mosses, however, may be sensitive even to slight increases in UV-B flux, based on the observed alterations in membrane permeability, chlorophyll content and growth. If these changes alter the Sphagnum-associated carbon dynamics in northern peatlands, they may have an important atmospheric feedback.
We thank Ulla Kääriäinen and Heikki Venäläinen for help in the field, Jaana Rissanen and Mirja Korhonen for assistance in the laboratory, Timo Oksanen for the UV data, Marion Fields for language revision; and the Academy of Finland (projects 39465 and 48798) and the Kone Foundation for financial support.