Two mire species respond differently to enhanced ultraviolet-B radiation: effects on biomass allocation and root exudation


  • Riikka Rinnan,

    1. Department of Terrestrial Ecology, Institute of Biology, University of Copenhagen, Øster Farimagsgade 2D, DK-1353 Copenhagen K, Denmark;
    2. Department of Ecology and Environmental Science, University of Kuopio, P.O.Box 1627, FI-70211 Kuopio, Finland
    Search for more papers by this author
  • Carola Gehrke,

    1. Department of Terrestrial Ecology, Institute of Biology, University of Copenhagen, Øster Farimagsgade 2D, DK-1353 Copenhagen K, Denmark;
    Search for more papers by this author
  • Anders Michelsen

    1. Department of Terrestrial Ecology, Institute of Biology, University of Copenhagen, Øster Farimagsgade 2D, DK-1353 Copenhagen K, Denmark;
    Search for more papers by this author

Author for correspondence: Riikka Rinnan Tel: +45 35 32 22 44 Fax: +45 35 32 23 21 Email:


  • • Increased ultraviolet-B (UV-B) radiation arising from stratospheric ozone depletion may influence soil microbial communities via effects on plant carbon allocation and root exudation.
  • • Eriophorum angustifolium and Narthecium ossifragum plants, grown in peatland mesocosms consisting of Sphagnum peat, peat pore water and natural microbial communities, were exposed outdoors to enhanced UV-B radiation simulating 15% ozone depletion in southern Scandinavia for 8 wk.
  • • Enhanced UV-B increased rhizome biomass and tended to decrease the biomass of the largest root fraction of N. ossifragum and furthermore decreased dissolved organic carbon (DOC) and monocarboxylic acid concentration, which serves as an estimate of net root exudation, in the pore water of the N. ossifragum mesocosms. Monocarboxylic acid concentration was negatively related to the total carbon concentration of N. ossifragum leaves, which was increased by enhanced UV-B. By contrast, enhanced UV-B tended to increase monocarboxylic acid concentration in the rhizosphere of E. angustifolium and its root : shoot ratio. Microbial biomass carbon was increased by enhanced UV-B in the surface water of the E. angustifolium mesocosms.
  • • Increased UV-B radiation appears to alter below-ground biomass of the mire plants in species-specific patterns, which in turn leads to a change in the net efflux of root exudates.


Ultraviolet-B (UV-B, 280–320 nm) radiation alters plant carbon (C) allocation in several ways, which could lead to changes in root exudation and thereby in the supply of C substrates to rhizospheric microorganisms. Although the possibility of alterations in the quality and quantity of root exudates has been raised in connection with studies that have reported UV-B-induced changes in soil microbial communities (Avery et al., 2003, 2004; Rinnan et al., 2005) or in methane emission from peatlands (Niemi et al., 2002b), we have no knowledge of the potential effects of increased UV-B radiation on root exudation. This study presents the first attempt to estimate the effects of enhanced UV-B radiation on root exudates in parallel with measurements of biomass allocation to different plant parts.

We focused on mire ecosystems, which cover a considerable area of boreal and arctic regions (Gorham, 1991) where stratospheric ozone depletion has led to increased fluxes of UV-B radiation (WMO, 2003). Northern peat-forming wetlands, i.e. mires, have accumulated 270–370 × 1015 g C in peat since the last glaciation (Turunen et al., 2002) and are significant sources of the important greenhouse gas methane (Cicerone & Oremland, 1988; Bartlett & Harriss, 1993), both of which are a result of restricted oxic decomposition caused by the high ground water table. Peat soils also contribute to the export of dissolved organic carbon (DOC) to freshwater systems and oceans (Urban et al., 1989; Aitkenhead et al., 1999; Freeman et al., 2001).

In mire ecosystems, the major part of the plant biomass is situated below ground (Wallén, 1986; Saarinen, 1996). In addition to living below-ground biomass, photosynthates are directed to mycorrhizal symbionts, root exudation and root respiration. In cereals and pasture plants, about half of the below-ground allocation goes into root biomass, one-third into root exudates and respiration and the remainder into soil organic matter and microorganisms (Kuzyakov & Domanski, 2000). Although root exudates comprise only 0.5–5% of the net fixed C (Farrar et al., 2003), they serve as significant sources of C and energy for soil microorganisms. As an example, C substrates for microbial methane production in the water-logged mires are mainly derived from recent photosynthates (King & Reeburgh, 2002). The soil water concentration of organic acids, which consist of root exudates and products of anaerobic degradation of organic matter, correlates highly with wetland methane emission rates (Christensen et al., 2003).

Increased UV-B radiation induces various adaptive morphological changes in plants. These include leaf thickening and shortening, shifts in root : shoot ratio and alterations in reproductive structures (Barnes et al., 1996; Day et al., 2001; Phoenix et al., 2001). Focus has largely been on the above-ground plant parts, because UV-B does not penetrate into the soil to a significant extent. However, recent observations of UV-B-induced alterations in the soil microbial community composition (Johnson et al., 2002; Avery et al., 2003; Rinnan et al., 2005) stress that plant roots as the interface between the plant and the soil should no longer be ignored.

Our aim was to assess whether increasing solar UV-B radiation changes C partitioning to different plant parts and to root exudation, estimated as DOC and organic acid concentrations in the rhizosphere. We did not attempt to determine whether UV-B stress increases root exudation per unit of root per se, but to estimate the net response in the amount of root exudates in the pore water as a result of either altered exudation, unchanged exudation from a changed amount of roots, or altered microbial consumption of exudates.

As UV-B radiation has been previously observed to reduce the root length growth of sedges on a minerotrophic mire (Zaller et al., 2002), we hypothesized that exposure to enhanced UV-B would reduce allocation to the below-ground biomass and root exudation. The experimental design included two vascular plant species representing different growth forms to enable comparison between the widespread deeply rooting sedge Eriophorum angustifolium Honck. and the oceanic herb Narthecium ossifragum (L.) Huds. (Liliaceae). In order to keep the experimental environment close to natural conditions, the plants were grown in mesocosms with Sphagnum peat containing natural peat pore water and microbial communities. Although this limits the possibilities in terms of determining the causes of changes in the pore water constituents resulting from the presence of both producers and consumers, it enables estimation of the effects of the treatments and the plant species under seminatural conditions.

Materials and Methods

Plant material and peat mesocosms

Forty peat monoliths lacking above-ground vascular tissue were excavated from a Sphagnum magellanicum Brid. lawn at the ombrotrophic Åkhult mire in southern Sweden (57°06′ N, 14°33′ E; described in detail by Malmer, 1962) in May 2000. The monoliths, which were 35 cm deep and 25 cm in diameter, were inserted into round watertight containers and left undisturbed until the start of the experiment in June The water table depth was maintained at the in situ level by watering when needed with water collected from the mire. Excess water was released through two holes that had been made in each container at the height of the initial water level.

Shoots of five vascular plant species –E. angustifolium, N. ossifragum, Trichophorum caespitosum (L.) Hartman, Rhynchospora alba (L.) Vahl and Rubus chamaemorus L. – were collected from the same mire. The shoots were carefully pulled out of the peat leaving the root system intact and placed into mire water. The plants were standardized by cutting roots and leaves to the same length. It was of concern that transplanting would affect the growth and behaviour of these plants, and hence interfere with the experimental analyses. To minimize any adverse effects, the plants were pre-grown for 4 wk in plastic containers containing mire water. Of the collected species, only E. angustifolium and N. ossifragum exhibited visible leaf and root growth after the transplantation, and thus they were the species selected for the UV-B exposure experiment.

One week before the start of the experiment, 12 randomly selected peat mesocosms were transplanted with two E. angustifolium shoots in each (henceforth Eriophorum mesocosms) and another 12 with three N. ossifragum shoots in each (Narthecium mesocosms). The corresponding densities of 40 and 60 shoots m−2 for E. angustifolium and N. ossifragum, respectively, are within the shoot densities reported for these species (Phillips, 1954; Summerfield, 1974). In addition, 12 peat monoliths were kept without transplanted vascular plants (control mesocosms). All the mesocosms had similar Sphagnum moss surfaces.

Application of enhanced UV-B radiation

At the experimental garden in Tåstrup, Denmark (55°41′ N, 12°06′ E), 12 plots were established and randomly assigned to control and UV-B treatments, yielding six replicates in total. One Eriophorum mesocosm, one Narthecium mesocosm and one control mesocosm (three in total) were buried into the ground at the centre of each plot. The mesocosms were exposed to control and UV-B treatments from 5 July until 30 August.

A frame (2.5 m × 1.3 m), giving an evenly irradiated area of 1.8 × 0.6 m, was erected at 1.5 m height on each plot. Each frame carried six parallel UV-B fluorescent tubes (UV-B 313; Q-PANEL, Cleveland, OH, USA). In the UV-B treatment, ecologically irrelevant UV-C radiation (< 280 nm) emitted by the tubes was excluded by cellulose diacetate filters (0.13 mm; Courtaulds, Derby, UK) mounted on UV-transmitting plexiglass (Röhm 2458; Röhm GmbH, Darmstadt, Germany) under each tube (Johanson & Zeuthen, 1998). In the control treatment, all UV-B and UV-C wavelengths were cut off by 4-mm-thick window glass placed under the tubes. UV-A radiation (320–400 nm) emitted by the tubes was less than 0.1% of ambient UV-A, and thus considered irrelevant. The tubes were pre-burned for 100 h in order to achieve a stable output, and the filters were pre-solarized and replaced every 40 irradiation hours to maintain constant transmission properties.

As in Johanson & Zeuthen (1998), a modification of Caldwell's generalized plant action spectrum (Caldwell, 1971) as parametrized by Thimijan et al. (1978) and normalized at 300 nm was used to derive the biologically effective UV-B radiation. The extra UV-B dose required to simulate 15% ozone depletion at Tåstrup was calculated with a computer model (Björn & Murphy, 1985; Björn & Teramura, 1993) for clear sky conditions, aerosol level zero and 50% air humidity. The daily irradiation was centred around solar noon and controlled by timers that switch on three lamps at a time in order to apply a stepwise enhancement of UV-B irradiation. Every second week, the daily irradiation duration was adjusted to follow the seasonal change in natural UV-B radiation.

Analyses of pore water chemistry

Sampling  Pore water from the peat mesocosms was sampled from two depths, 1 and 15 cm from the peat surface, on 6 August (after 32 treatment days), 23 August (49 d) and at the end of the experiment on 30 August (56 d). Each 20-mL water sample was drawn with a motorized 10-mL pipette from two sampling holes, which were at a distance of approx. 2 cm from the plants. All samples were filtered through Whatman glass microfibre GF/D filters and stored at −20°C. Concentrations of DOC, phosphate, ammonium and nitrate were analysed for all three sampling occasions. Organic acids and microbial biomass C were analysed only at the end of the experiment.

DOC and nutrients  The DOC concentration of the pore water was analysed on a TOC-5000A total organic C analyser (Shimadzu, Kyoto, Japan). The ammonium concentration was determined spectrophotometrically by the indophenol blue method, the nitrate concentration by the cadmium reduction-sulphanilamide method and the phosphate concentration by the molybdenum blue method (Allen, 1989).

Organic acids  Concentrations of organic acids in the pore water were analysed as in Ström et al. (1994) using an anion exchange high-performance liquid chromatography (HPLC) system equipped with a Dionex column system (Dionex, Sunnyvale, CA, USA), including the analytical column AS11 (4 mm; P/N 044076; Dionex).

Microbial biomass C  The method for estimating microbial biomass C concentration of the pore water was modified from the traditional chloroform fumigation technique used in soil sciences (Jenkinson & Powlson, 1976). In order to release the C present in microbial cells, 0.2 mL of chloroform was added to 10 mL of pore water and incubated for 24 h in the dark. The microbial biomass C was calculated as the difference in the DOC concentration in the chloroform-treated pore water and in the untreated pore water. Carbon added as chloroform was subtracted.

Measurements of plant biomass, morphology and nutrients

At the end of the experiment, the vascular plants were carefully removed from peat, and it was verified that roots and shoots of both the species had grown during the experiment. Leaf lengths were measured and numbers of roots, leaves and tillers were counted. The plant material was sorted into leaves, rhizomes plus stem bases (only N. ossifragum) and roots. The roots of N. ossifragum, which arise from swollen nodal areas of the rhizome (Summerfield, 1974; for an illustration, see Heath & Luckwill, 1938), were further divided into first-, second- and third-node roots counting from the shoot. The samples were oven-dried at 75°C and weighed. Total C, nitrogen (N) and phosphorus (P) (P in E. angustifolium only) concentrations were analysed in the dried and finely ground plant material. Total C and N were determined by dry combustion, and total P by the molybdenum blue method after wet digestion.

Statistical analysis

The experimental design consisted of three fixed factors: UV treatment (ambient and enhanced UV-B), plant species (E. angustifolium, N. ossifragum and control) and peat depth (−1 cm and −15 cm). Initially all the factors and interaction terms were included in the univariate analysis of variance (ANOVA). When an interaction term was significant at P < 0.1, data were split into further ANOVAs within the factors in the interaction term. When the main effect of the plant species was significant, Tukey's honestly significant difference multiple comparison test was used to identify differences between the species. Data with multiple measurement points were subjected to a repeated measures analysis of variance (RM-ANOVA) with Huynh–Feldt adjustment of the degrees of freedom, following the same principle as for included factors in ANOVA. Treatment effects on plant biomass, morphology and nutrients were tested separately for each species using an independent samples t-test. The Pearson correlation test was used to analyse relationships between the plant and peat pore water variables at the last sampling. All statistical analyses were run with SPSS 11.0 for Windows (SPSS, Inc., Chicago, IL, USA).


DOC and nutrients

The DOC concentration of the pore water was highest in the control mesocosms without vascular plants, intermediate in the Narthecium mesocosms, and lowest in the Eriophorum mesocosms (Fig. 1a–c; P < 0.05, Tukey's HSD with RM-ANOVA). The temporal changes in DOC concentration during the sampling period differed between the plant species, but the DOC concentration remained significantly higher at 15-cm depth than in the surface water in all the mesocosms throughout the samplings (Fig. 1a–c; time × species, P < 0.001, RM-ANOVA).

Figure 1.

Dissolved organic carbon (DOC; a–c), ammonium-nitrogen (NH4-N; d–f, solid lines) and phosphate-phosphorus (PO4-P; d–f, dotted lines) concentrations in pore water from a depth of 1 or 15 cm in peat mesocosms during the UV-B exposure experiment. The Eriophorum mesocosms contained two Eriophorum angustifolium shoots, Narthecium mesocosms contained three Narthecium ossifragum shoots and the control mesocosms had no living vascular plants. Data are means of six replicates (± standard error). Repeated measures analysis of variance (ANOVA) was performed separately for each species; statistical significance is shown when P < 0.1.

Enhanced UV-B had no clear effects on the DOC concentration in the Eriophorum or in the control mesocosms (Fig. 1a–c). In contrast, the DOC concentration in the Narthecium mesocosms was on average 22% lower at both depths under enhanced UV-B compared with the ambient UV-B throughout the sampling period (Fig. 1b).

The ammonium concentration of the pore water was significantly higher in the control mesocosms than in the Eriophorum mesocosms (Fig. 1d–f; P < 0.01, Tukey's HSD with RM-ANOVA). There was no clear temporal pattern, and the concentration was always higher at 15-cm depth than in the surface water (Fig. 1d–f; P < 0.001, between-subject effect of depth, RM-ANOVA). Nitrate concentrations were below the detection limit. The phosphate concentration was higher in the Narthecium mesocosms than in the Eriophorum mesocosms (Fig. 1d,e; P < 0.05, Tukey's HSD with RM-ANOVA). After 32 d, the phosphate concentration was higher at 15-cm depth than in the surface water, but this difference had disappeared by the two later samplings (Fig. 1d–f; time × depth, P < 0.001, RM-ANOVA). Enhanced UV-B had no effects on either ammonium or phosphate concentration in the pore water.

Organic acids

The organic acids detected in the pore water consisted for the most part of monocarboxylic acids, mainly acetic and lactic acids (Table 1). Formic, oxalic and citric acids were present in trace amounts.

Table 1.  Organic acid concentrations in pore water from a depth of 1 or 15 cm in peat mesocosms with Eriophorum angustifolium, Narthecium ossifragum or no vascular plants (control) after 56 d of exposure to ambient or enhanced UV-B radiation
 Organic acid concentration (µmol L−1)
  1. Values are means of six replicates (± standard error). The statistical significance for significant effects (P < 0.05) and trends (0.05 < P < 0.15) is shown (analysis of variance with UV treatment, plant species and depth as fixed factors).

  2. ANOVA, analysis of variance; nd, not detected; ns, not significant.

1-cm depth
 E. angustifolium
  Ambient 17.6 ± 8.514.4 ± 6.21.0 ± 0.30.7 ± 0.20.2 ± 0.1
  Enhanced UV-B 10.6 ± 10.223.8 ± 6.11.7 ± 0.41.0 ± 0.10.1 ± 0.1
 N. ossifragum
  Ambient108.6 ± 44.2 4.6 ± 3.11.2 ± 0.41.0 ± 0.1nd
  Enhanced UV-B 20.6 ± 12.1 8.0 ± 3.31.3 ± 0.40.7 ± 0.2nd
  Ambient 56.1 ± 47.270.4 ± 61.41.5 ± 0.31.0 ± 0.10.1 ± 0.1
  Enhanced UV-B 52.5 ± 32.227.9 ± 12.62.7 ± 0.91.1 ± 0.1nd
15-cm depth
 E. angustifolium
  Ambient 43.4 ± 13.6 5.0 ± 2.01.7 ± 0.31.1 ± 0.10.4 ± 0.1
  Enhanced UV-B 87.8 ± 45.1 6.9 ± 4.51.3 ± 0.41.0 ± 0.20.1 ± 0.1
 N. ossifragum
  Ambient100.7 ± 60.910.8 ± 8.70.6 ± 0.40.8 ± 0.2nd
  Enhanced UV-B 35.4 ± 16.8 5.6 ± 2.68.1 ± 6.60.9 ± 0.2nd
  Ambient230.2 ± 97.1 4.2 ± 4.21.1 ± 0.51.0 ± 0.10.1 ± 0.1
  Enhanced UV-B136.8 ± 68.8 5.7 ± 2.82.9 ± 1.21.1 ± 0.2nd
ANOVA P-values
UV × depthnsnsnsnsns
UV × speciesnsnsnsnsns
Depth × speciesns0.109nsnsns

The total concentration of monocarboxylic acids was significantly higher in the control mesocosms than in the Eriophorum and Narthecium mesocosms, especially at 15-cm depth (Fig. 2). In the Narthecium mesocosms, there was no clear difference between depths, whilst in the Eriophorum and control mesocosms, the monocarboxylic acid concentration was considerably higher at 15-cm depth compared with the surface water (Fig. 2). This difference was a result of the (on average) > 3 times higher concentration of acetic acid in the deeper peat (Table 1). However, the concentration of lactic acid was significantly lower in the deeper peat (Table 1).

Figure 2.

Monocarboxylic acid concentration in pore water from a depth of 1 or 15 cm in peat mesocosms after 56 d of exposure to ambient or enhanced UV-B radiation. Data are means of six replicates (± standard error). Statistical significances are shown when P < 0.1 [analysis of variance (ANOVA)].

The main effect of enhanced UV-B on the total monocarboxylic acid concentration was close to significant (P = 0.07). At 15-cm depth in the Eriophorum mesocosms, there was a nonsignificant trend for a higher monocarboxylic acid concentration under enhanced UV-B, whereas in the Narthecium mesocosms the concentration was significantly lower under enhanced UV-B (Fig. 2). The monocarboxylic acid concentration of the control mesocosms showed high variance and a nonsignificant trend for a decrease under enhanced UV-B. These changes were mostly accounted for by the changes in the acetic acid concentrations (Table 1). Formic acid concentration was strongly but nonsignificantly increased by enhanced UV-B in the control mesocosms and at 15-cm depth in the Narthecium mesocosms (Table 1).

The concentration of dicarboxylic oxalic acid was low and showed no clear trends with UV treatment, plant species or depth (Table 1).

Microbial biomass C

The microbial biomass C concentration tended to be lower in the control mesocosms than in the mesocosms with vascular plants (Fig. 3; P = 0.090 for control vs Narthecium; P = 0.135 for control vs Eriophorum, Tukey's HSD with ANOVA). There were no significant depth differences. In the Eriophorum mesocosms, enhanced UV-B increased the concentration of microbial C in the surface water by 20% but tended to decrease it at the depth of 15 cm (Fig. 3; P < 0.05, UV × depth interaction). There were no UV-B effects on the Narthecium or control mesocosms.

Figure 3.

Microbial carbon concentration in pore water from a depth of 1 or 15 cm in peat mesocosms after 56 d of exposure to ambient or enhanced UV-B radiation. Data are means of six replicates (± standard error). *, P < 0.05 [analysis of variance (ANOVA)].

Plant biomass, morphology and nutrients

The above-ground plant biomass was not significantly affected by enhanced UV-B, but there were some alterations in the below-ground parts of N. ossifragum (Table 2). Enhanced UV-B tended to decrease the third-node root biomass – which made up the largest root fraction – while it increased the rhizome plus stem base biomass of this species (Table 2). The biomass of leaves (Table 2), as well as that of first- and second-node roots and tillers, was unaffected by enhanced UV-B (data not shown). The ratio of below- to above-ground biomass was not significantly different between the treatments (Table 2). The total below-ground biomass showed a strong negative relationship with formic acid concentration (P < 0.001, r2 = 0.78, n = 11), and the rhizome biomass showed a weaker negative relationship with ammonium concentration (P < 0.05, r2 = 0.43, n = 11) in the peat pore water from a depth of 1 cm. There were no significant correlations between N. ossifragum biomass and pore water variables at the 15-cm depth.

Table 2.  Dry mass of Eriophorum angustifolium leaves and roots, and Narthecium ossifragum leaves, total roots (third-node root fraction shown separately) and rhizomes plus stem bases, and ratio of below-ground to above-ground biomass after 56 d of exposure to ambient or enhanced UV-B radiation
 E. angustifoliumN. ossifragum
AmbientEnhanced UV-BPAmbientEnhanced UV-BP
  • Values are means of six replicates (± standard error).

  • Statistical significance for significant effects (P < 0.05) and trends (0.05 < P < 0.15) from an independent samples t-test is shown.

  • na, not applicable; ns, not significant.

  • a

    For N. ossifragum (roots + rhizomes)/leaves.

Leaves (mg d. wt) 727 ± 33 715 ± 70ns  55 ± 4  55 ± 3ns
Total roots (mg d. wt)  89 ± 13 107 ± 23ns  52 ± 10  41 ± 7ns
Third-node roots (mg d. wt)nana   30 ± 4  23 ± 20.082
Rhizomes (mg d. wt)nana   80 ± 5  93 ± 30.044
Below-ground:above-grounda0.12 ± 0.010.14 ± 0.20.1492.43 ± 0.142.47 ± 0.20ns

In E. angustifolium, we observed no significant differences in the biomass of any plant parts between the UV-B treatment and the ambient control (Table 2), and no significant correlations between biomass and pore water variables. However, there was a nonsignificant trend for a higher root : shoot ratio under enhanced UV-B compared with the ambient UV-B (Table 2).

There were no UV-B effects on the leaf length or on the number of leaves, roots or tillers in either species (data not shown, P > 0.3). Concentrations of total C, N and P in the plant material were mostly unaffected by the UV-B treatment (data not shown). The sole nearly significant effect was observed in the leaf C concentration of N. ossifragum, which was nearly significantly higher under enhanced UV-B (43.8 ± 0.5%; mean ± standard error) compared with the control (42.6 ± 1.4%; P = 0.067, t-test). The leaf C concentration correlated negatively with monocarboxylic acid concentration in the peat pore water both at 1-cm depth (P < 0.01, r2 = 0.67, n = 9) and at 15-cm depth (Fig. 4).

Figure 4.

The relationship between total carbon concentration of Narthecium ossifragum leaves and monocarboxylic acid concentration in peat pore water at a depth of 15 cm (Pearson correlation test, n = 10).


Exposure of two mire plant species to enhanced UV-B radiation for about 2 months yielded results that suggest that increased UV-B radiation can influence the net efflux of root exudates by altering below-ground biomass. Furthermore, responses appeared to differ between plant species. The weak nature of the responses was not surprising considering the general scarceness of quantitative changes in plant photosynthesis or above-ground biomass under realistically enhanced UV-B radiation in field studies (Searles et al., 2001). In the current experiment, part of the biomass present at the end of the experiment had been produced before the UV-B exposure, so the UV-B effects on the biomass accumulated during the experiment are more difficult to detect. Effects of UV-B on the growth of mire plants have previously been observed to show only nonsignificant tendencies in 1 year, while significant differences have been found with repeated measurements over several years (Robson et al., 2003).

In N. ossifragum, enhanced UV-B radiation had no effects on the green above-ground biomass, although it tended to increase the total C concentration per unit dry weight of the leaves. Higher total C concentration suggests that the UV-B-exposed leaves contained more biomolecules that are rich in C compared with the ambient control. Because leaf chemical analyses other than those of total C, N and P could not be conducted because of low sample mass, we cannot elucidate which compounds could have contributed to this increase. Enhanced UV-B increased biomass allocation to rhizomes and stem bases. The rhizome biomass was negatively correlated with ammonium concentration, which implies higher nutrient acquisition to the storage structures by the increased biomass. This result is in contrast with the reported higher rhizome elongation of the arrow-rush Tetroncium magellanicum under filters reducing ambient UV-B on a southern Argentinean Sphagnum bog (Robson et al., 2003). However, the biomass of the largest root fraction of N. ossifragum, consisting of fine roots (third-node roots), was nearly significantly decreased by enhanced UV-B. Concomitant UV-B-induced reductions in the DOC and organic acid concentrations of the pore water in the Narthecium mesocosms indicate that the lower fine root biomass resulted in less total root exudation. Furthermore, the negative relationship between the pore water organic acid concentration and the total C concentration in N. ossifragum leaves suggests that, when more C was allocated to leaves, less was available for root exudation.

Although the difference was not statistically significant, the sedge E. angustifolium tended to have a higher root : shoot ratio in the UV-B treatment, which suggests slightly increased allocation to below-ground biomass. This is in contrast with the previously observed higher root length production of Carex species under reduced UV-B radiation in southern Argentina (Zaller et al., 2002), although root biomass or root : shoot ratio was not reported in that study. However, more consistent with our results, root : shoot ratio of the Antarctic grass Deschampsia antarctica remained unaffected by UV-B enhancement (van de Staaij et al., 2002). Similar UV-B supplementation as in the present study did not affect the physiology and morphology of Eriophorum vaginatum (Niemi et al., 2002a), although the leaf cross-sectional area was reduced under enhanced UV-B during a growing season of higher UV intensity (Niemi et al., 2002b).

In the current experiment, there was a trend for a higher organic acid concentration in the pore water of the Eriophorum mesocosms at the depth of 15 cm under enhanced UV-B radiation. As the roots of E. angustifolium were abundant at this depth, and as approx. 10% of DOC is derived from recent plant photosynthates in a similar mire type (Olsrud, 2004), the observed increase in organic acid concentration may have resulted from a UV-B-induced increase in photosynthate allocation below ground. The related species Eriophorum scheuchzeri responds to experimental light attenuation by reducing exudation of acetate (Ström et al., 2003), and a similar trend has been observed in E. angustifolium (Joabsson et al., 1999). Saarnio et al. (2004) found predominantly acetate and lactate in percolate samples collected from the rhizosphere of E. vaginatum grown in pots with quartz sand. Citrate, formate and succinate were present in small amounts. A similar composition of organic acids was detected in our samples, which supports the hypothesis that they originated from the plants.

In the control mesocosms without transplanted vascular plants, enhanced UV-B radiation had a tendency to decrease organic acid (but not DOC) concentration in the pore water, although this response was less than in the Narthecium mesocosms and statistically not significant. This alteration was independent of any changes in vascular plant root exudation, and was therefore a result of effects of UV-B radiation on Sphagnum mosses, microbial communities or the peat water.

Input of labile C into soil by plant exudation is usually only discussed in connection with vascular plants, although increased Sphagnum exudation has recently been mentioned as a potential response to CO2 enrichment (Mitchell et al., 2003). Furthermore, Fenner et al. (2004) used 13CO2 tracer to show that, as early as 4 h after labeling, C fixed by Sphagnum mosses comprised 4% of the total DOC in pore water. Hence, we propose that the lower organic acid concentration under enhanced UV-B indicates that UV-B may have decreased exudation of the measured organic acids by Sphagnum. Another possibility is that increased membrane leakage of magnesium and calcium from Sphagnum moss tissue in response to enhanced UV-B radiation (Niemi et al., 2002a,b) would lead to chelation of organic compounds and thereby to a lower concentration in the pore water.

The potential direct effects of UV-B on peat water could only occur in the surface water, with limited influence to a depth of 15 cm via diffusion. Under laboratory conditions, UV-B exposure of Sphagnum bog water has been shown to decrease the DOC concentration and increase the bacterial abundance (De Lange et al., 2003). Similarly, exposure of humic acids to enhanced UV-B has been shown to significantly stimulate microbial utilization of DOC, as shown by increased CH4 production (Bianchi et al., 1996). However, UV photolysis of DOC should lead to a higher concentration of low-molecular-weight organic acids (Wetzel et al., 1995).

The control mesocosms differed from the mesocosms with vascular plants by having slightly less microbial C and higher concentrations of DOC and monocarboxylic acids, especially in the pore water from the 15-cm depth. As the rhizosphere maintains higher microbial biomass and activity than bulk soil, the lack of living vascular plants in the control mesocosms resulted in the lower microbial biomass. This decrease further led to accumulation of organic acids, because microorganisms rapidly use labile C compounds. An earlier study demonstrated that the DOC concentration of pore water was higher in the moss-dominated intertussock tundra than in the E. vaginatum-dominated tussocks or in the wet sedge tundra with Carex species and E. angustifolium (Judd & Kling, 2002). These vegetation effects on DOC concentration are consistent with the effect of vascular plants on the organic acid concentration in the current experiment.

Both the DOC and the organic acid concentrations in the current experiment are well within the range measured in peatland mesocosms (Joabsson et al., 1999) and natural mires (Christensen et al., 2003; Pastor et al., 2003; Ström et al., 2003; Blodau et al., 2004). In accordance with Blodau et al. (2004), DOC concentration in the pore water increased with depth.

In conclusion, increased UV-B radiation appears to alter the below-ground biomass of the mire plants, which in turn leads to a change in the net efflux of root exudates. The responses are species-dependent, which implies that the feedbacks of altered exudation on ecosystem functioning, such as methane emission or DOC discharge, depend on the plant species composition. Use of natural peat monoliths instead of sterile growth media lowers resolution in detecting differences in root exudation and thus the effects of UV-B radiation may be underestimated in this study.


This work was supported by the Carl Trygger Foundation and the STINT Foundation (The Swedish Foundation for International Cooperation in Research and Higher Education). We thank Gosha Sylvester and Karna Heinsen for laboratory assistance.