No allelopathic effect of the dominant forest-floor lichen Cladonia stellaris on pine seedlings

Authors

  • Minna-Maarit Kytöviita,

    Corresponding author
    1. Department of Biological and Environmental Science, Jyväskylä University, PO Box 35, FIN 40014, Finland; and
      *Correspondence author. E-mail: minna-maarit.kytoviita@jyu.fi
    Search for more papers by this author
  • Sari Stark

    1. Finnish Forest Research Institute, Rovaniemi Station, PO Box 16, FIN-96301, Rovaniemi, Finland
    Search for more papers by this author

*Correspondence author. E-mail: minna-maarit.kytoviita@jyu.fi

Summary

  • 1The lichen Cladonia stellaris forms dense mats and dominates the forest-floor vegetation in late-successional oligotrophic boreal forests. This lichen is rich in polyphenolic secondary compounds, and through allelochemical effects of these compounds the dense mat formed by C. stellaris is assumed to have negative effect on forest regeneration and soil microbial functioning.
  • 2We examined the effect of C. stellaris and its secondary metabolite usnic acid on nutrient uptake and growth of pine (Pinus sylvestris) seedlings in symbiosis with the ectomycorrhizal fungus Suillus variegatus and in non-symbiotic condition.
  • 3Contrary to our expectations, usnic acid had no effect on nitrogen uptake and growth in pine seedlings. The addition of lichen fragments into the growth substrate significantly increased biomass accumulation in mycorrhizal pine seedlings and needle nitrogen acquisition in non-mycorrhizal seedlings.
  • 4These results suggest that lichen mats do not have direct allelopathic effects on pine seedling nitrogen acquisition or growth.

Introduction

Lichens are one of the most important vegetation components in northern boreal and sub-arctic ecosystems. Altogether, mat-forming lichens dominate about 8% of all terrestrial surface on earth (Larson 1987). In sub-arctic conifer forests, such as in the open woodlands in northern Canada, lichens cover 97% of forest-floor and contain about 20% of the total ecosystem biomass (Auclair & Rencz 1982). Similar values are reported for other arctic ecosystems (Shaver & Chapin 1991). Mat-forming lichens in the genus Cladonia are important as fodder to reindeer and caribou. In addition to directly reducing the amount of lichen by consuming them, reindeer and caribou negatively affect the lichen grounds by trampling. Lichens are slow growing organisms. For instance, in the reindeer lichen Cladonia stellaris the highest relative daily growth rate in the maximally producing lichen apices under field conditions is about 4 mg per g lichen tissue (Kytöviita & Crittenden 2007). Consequently, it takes decades for the lichen grounds to recover from destruction by fire (Morneau & Payette 1989) or by overgrazing by reindeer (Klein 1987). In undisturbed conditions, however, the ground floor in late-successional oligotrophic boreal and sub-arctic forests may form long-standing mats of up to 10 to 15 cm thickness composed mainly of C. stellaris (Ahti 1977).

In general models of development of boreal lichen rich forests, the C. stellaris layer is reported to have negative effects on pine seed germination and growth of mycorrhizal fungi (Sedia & Ehrenfeld 2003). There are data that support the concept of a negative effect of the lichen layer on plant and microbial activities. However, although lichens have long been considered to have allelopathic effects on plants and soil microorganisms, several studies show contrasting evidence: Cladonia lichens are reported both to possess and not to possess allelopathic properties. For instance, water extracts from lichens were reported to have no effect (Brown 1967) and to significantly inhibit (Sedia & Ehrenfeld 2003) the germination of pine seeds under laboratory conditions. Pine seedlings grown in greenhouse in pots covered with C. stellaris layer accumulated less nitrogen than control seedlings with pots covered with peat moss (Fischer 1979), but removal of the C. stellaris layer in field decreased the pine seedling growth rate when compared to seedlings growing in soil covered with the lichen (Cowles 1982).

Pine trees are associated with ectomycorrhizal fungi which improve the uptake of nutrients through better exploration of the soil and also through exploiting nutrient sources mainly not available to plant roots (Marschner & Dell 1994). Pines are considered obligatorily mycorrhizal, and under undisturbed natural conditions are thought to acquire most of their nutrients through mycorrhizas. Therefore, any impairment in mycorrhizal functions would reduce pine tree growth and, consequently, ecosystem and forestry productivity. The development and function of ectomycorrhizal fungi in symbiosis with pine under axenic conditions was negatively affected by water extracts of C. stellaris (Brown & Mikola 1974), but natural rainfall percolated through a layer of the lichen C. stellaris had no effect on microbial respiration in the underlying soil (Stark et al. 2007a). In field, ectomycorrhizal colonization was lower in plants growing within lichen mats than in moss mats or bare soil (Sedia & Ehrenfeld 2003). However, C. stellaris -modified rainwater had no negative effects on development of mycorrhizal colonization between the ericaceous shrub Vaccinium microcarpon and the mycorrhizal fungus Hymenoscyphus ericacea grown under axenic conditions (Crittenden 2000). Also V. vitis-idaea seedlings grown in soil collected beneath C. stellaris canopies and beneath birch tree canopies were heavily infected by ericoid mycorrhizas in both cases (Crittenden 1989). Similarly, pine seedling frequencies are reported to be higher in lichen-dominated areas in comparison to nearby areas devoid of lichen cover (den Herder et al. 2003) or to areas covered with moss (Steijlen et al. 1995).

The allelopathic effects of lichens are thought to be mediated by the production of diverse range of secondary chemicals that are unique to lichens. The most abundant secondary metabolite in the dominant lichen species, C. stellaris, is usnic acid (Huovinen & Ahti 1986). Usnic acid is a product of the fungal secondary metabolism in lichen symbioses, and it is deposited mainly at the fungal cell wall (Elix 1996). There are two optical forms of usnic acid; the (−) and (+) forms. Usnic acid has been shown to possess antimicrobial qualities in pharmacological surveys and its medical use has a long history (Ingólfsdóttir 2002; Oksanen 2006). Both (−) and (+) forms of usnic acid inhibit clinical isolates of bacteria (Lauterwein et al. 1995) and under sterile conditions usnic acid from C. arbuscula has been shown to inhibit mycobacteria (Ingólfsdóttir et al. 1998). The effects of the two forms of usnic acid on fungi vary. A range of pathogenic fungi showed antifungal activity of (−)-usnic acid, but not of (+)-usnic acid (Halama et al. 2004). However, (+)-usnic acid had antifungal effects on the plant pathogenic fungus Fusarium moniliforme, but no effects on the yeast Saccharomyces cerevisiae (Cardarelli et al. 1997). Neither isomer of usnic acid affected the pathogenic fungus Gremmenniella (Kaitera et al. 1996).

The physiological effects of dilute concentrations of usnic acid in hydroponic media have been investigated with agricultural crop plants and the algal symbionts in lichens. It has been shown that (+)-usnic acid can pass through the plasmalemma (Vavasseur et al. 1991), and affect physiological processes in plants (Romagni et al. 2000). Additions of (+)-usnic acid into the hydroponic growth media of crop plants reduced photosynthetic and respiration rates (Lascève & Gaugain 1990; Latkowska et al. 2006) as well as plant growth and nutrient uptake (Lechowski et al. 2006). In addition, (+)-usnic acid increases the permeability of the lipid membranes (Bačkor et al. 1997) and results in electrolyte leakage in lettuce cotyledons (Romagni et al. 2000). (+)-usnic acid has antimitotic effect on tobacco seedlings (Cardarelli et al. 1997) and also inhibited the lichen photobiont growth (Bačkor et al. 1998). However, the effect of usnic acid has not been studied on higher plants that would encounter this compound in their natural environment.

In summary, although the often perceived negative ecosystem effect of the lichen mat is commonly considered to be mediated by secondary metabolites (Sedia & Ehrenfeld 2003; van der Wal 2006), the effect of lichen metabolites on soil functions or tree growth have not been elucidated in ecologically relevant studies. The antibiotic effect of lichen metabolites has only been studied under axenic conditions, the only exception being the study by Stark et al. (2007a). Studies focusing on the allelopathic effect of lichens on plants, on the other hand, did not examine the effect of known lichen metabolites, but instead compared systems with or without lichen cover or additions of water extracts of lichens with unknown chemical composition to additions of water. Lichen layer is known to affect the temperature and moisture levels in soils underneath (Broll 2000), and therefore it is not clear whether the effects of the lichen mats were mediated by physical or chemical modification of the soil environment. Furthermore, these experiments did not chemically characterize the water extracts of lichens, and it cannot be known for certain if lichen acids actually caused the treatment effects. Lichens in these experiments were also often suddenly rewetted and crushed before extraction and could release unnatural amounts of solutes compared to natural rainfall percolated through lichen layer. To summarize, the previous investigations conducted under ecologically relevant circumstances cannot unequivocally point lichen acids as causal agents for the observed effects. In this study, we examined the effect of the common mat-forming lichen C. stellaris on the growth and ectomycorrhiza-mediated nutrient acquisition of pine (Pinus sylvestris) seedlings. Pinus sylvestris and C. stellaris commonly co-occur over large areas of Eurasia. Pine seedlings were grown in microcosms in symbiosis with their common ectomycorrhizal symbiont Suillus variegatus or in non-mycorrhizal condition. The microcosm growth substrate was amended with C. stellaris fragments or pure usnic acid in a fully factorial fashion. Based on the current available information, we expected the lichen and the pure usnic acid addition to affect adversely both non-mycorrhizal and mycorrhizal pines.

Materials and methods

Pine seeds (Pinus sylvestris, origin Kittilä, 67°42′ N, 24°50′ E) were surface sterilized in hydrogen peroxide solution and germinated on water agar. When one month old, ectomycorrhizal symbiosis was initiated under sterile conditions with S. variegatus (origin Oulu, 65°01′ N, 25°28′ E) on peat-vermiculite moistened with Melin Norkrans solution (Marx 1969). The seedlings were grown in a growth cabinette with 18 + 6 h light period and 20/18 °C temperature. The light level was 360 µmol s−1 provided by Osram HQI lamps. Non-mycorrhizal seedlings were raised similarly, but without the fungal inoculation. The seedlings (4 months old) were transferred to larger plexiglass plates (15 × 10 × 2 cm) filled with 1 : 2 peat : sterilized sand mixture and fertilized with dilute additions of Ingested nutrient solution (Ingestad 1979). When 6 months old, the seedlings were transferred to experimental root observation boxes (25 × 25 × 2·5 cm). Each experimental box contained 284 g dry weight of 1 : 2 mixture of peat : sterilized sand.

Usnic acid is poorly soluble in water, and only about 9 µmol concentrations can be reached at pH 7 (Dawson et al. 1984). At lower pH values usnic acid is practically insoluble to water (Kristmundsdóttir et al. 2005). This characteristic makes usnic acid unlikely to dissolve in rainwater under natural conditions, and no usnic acid was observed in rainwater that had passed through a C. stellaris layer (Stark et al. 2007a). Even if lichen acids do not dissolve in rainwater, rain droplets could physically bombard the usnic acid crystals from the lichen surfaces and consequently allow the crystals to reach the soil underneath C. stellaris mats. Furthermore, lichens are an important fodder for the semi-domesticated and wild Rangifer tarandus (reindeer and caribou, respectively). Grazing and trampling by these large herbivores considerably modifies the lichen vegetation in boreal and sub-arctic forests throughout the circumpolar area (Helle & Aspi 1983; Manseau et al. 1996). Intensive trampling by reindeer fragments the lichen mat and some usnic acid could end up in soil in form of lichen fragments. In order to address both possible paths for the allelopathic effects of lichens, that is, leaching of usnic acid by the rainwater, and the effects of lichen fragments in soil, we used both additions of commercial pure usnic acid and lichens in our microcosm study. To link the effect of added lichen fragments to the usnic acid in the lichen, we reduced the amount of usnic acid in the C. stellaris fragments by washing with acetone. Twelve ectomycorrhizal and 12 non-mycorrhizal plants were randomly allocated to each of the following treatments:

  • 1)Control;
  • 2)+8·5 g ground C. stellaris (containing 128 mg usnic acid) added to the growth substrate;
  • 3)+8·5 g ground and acetone washed C. stellaris (12 mg usnic acid) added to the growth substrate; and
  • 4)+100 mg pure usnic acid added to the growth substrate.

The C. stellaris material was collected from Rajajooseppi (68°25′ N, 28°30′ E) and cleansed from debris. The material was rehydrated in water saturated atmosphere and the top 50 mm part cut and briefly ground with pestle and mortar when air dry. All lichen material was dried in an oven at 80 °C for 48 h to eliminate possible mycorrhiza-forming spores. Assuming that a mature lichen stand contains 750 g lichen m−2 (Yarranton 1975), and that the depth of the organic layer underneath the lichen mat is 1·5 cm with 0·8 g cm−3 density (Stark et al. 2007b), the amount of added lichen in this experiment (8·5 g per 284 g soil) is equivalent to about 48% of a mature C. stellaris mat getting mixed with the organic layer of soil underneath. The N concentration in the lichens was analysed using the dynamic flash combustion technique (CE Instruments EA 1110 Elemental Analyzers). The lichens contained 2 mg g−1 dry weight N and the addition of 8·5 g of lichens added 17 mg N to each box. The total usnic acid content in experimental lichens was determined by soaking the ground lichens 3 × 30 min in acetone. This procedure should recover most of the usnic acid in the ground lichen material (Solhaug & Gauslaa 2001). The usnic acid in acetone extracts of lichens was detected with HPLC similarly as in acetone extracts of soil samples in the end of the experiment (see below). The lichen material contained 15 mg usnic acid g−1 dry weight which falls within the range reported for C. stellaris previously (Huovinen & Ahti 1986). Consequently, with the 8·5 g ground lichen 128 mg usnic acid was added to each box.

For treatment 3, the lichen acids were removed with acetone from the cell walls by immersing the dry lichens for 10 min in acetone for three times. Removal of secondary compounds by acetone washing has been successfully utilized for studies of the ecological role of secondary compounds in lichens (e.g. Pöykköet al. 2005). It has been shown that rinsing lichens briefly in acetone does not affect the lichen metabolism or chemistry other than removing surface phenolics (Solhaug & Gauslaa 2001; Pöykköet al. 2005). Therefore we used this method to remove phenolic secondary compounds from the lichen C. stellaris. Part of the usnic acid is in an immobile fraction and is not removed by this relatively gentle extraction. The 3 × 10 min acetone-extracted lichen material contained 1·4 mg usnic acid g−1 dry weight when determined after 3 × 30 min extraction. Consequently, with the 8·5 g acetone washed ground lichen, we added 12 mg usnic acid to each box.

In treatment 4, (+)-usnic acid (Sigma U-7876) was used. Microbial inoculum from natural pine forest soil collected at Oulu (65°01′ N, 25°28′ E) was added into the observation boxes in form of filtered (5 µm) water suspension. This ensured the presence of natural soil microbes in the experimental soil substrate, but excluded ectomycorrhizal spores. The experimental boxes were carefully wrapped in aluminium foil such that the pine shoots protruded from a small hole.

The plants were watered with tap water when necessary and the experiment was terminated after 6 months growth in the experimental boxes. The root systems grow on a relatively thin layer of substrate in this type of microcosm, which allowed us to monitor the plant roots at every watering occasion. We noted no signs of ectomycorrhizal contamination in the non-mycorrhizal or mycorrhizal pines. At the end, the needle, stem and root dry weights (60 °C, 48 h) were recorded. All needles of a plant were milled with a ball mill grinder and two samples per plant were analysed with flash combustion method (CE Instruments EA 1110 Elemental Analysers) to give average needle N concentration per plant.

Gravimetric moisture (105 °C, 12 h) and organic matter (OM) content (ashing at 475 °C, 4 h) were determined for the growth substrate. The growth substrate was frozen for later analysis of soluble N. A subsample of c. 6 g fresh weight substrate was extracted with 50 mL of 0·5 m K2SO4 and the NH4-N concentration in the extracts was determined by flow injection analysis (FIA 5012, Tecator). Total extractable N in the extracts was determined by oxidizing all extractable N to inline image (Williams et al. 1995), and then analyzing it as inline image by automated flow injection (FIA 5012, Tecator).

Usnic acid was extracted from 10 g oven dried (60 °C, overnight) growth substrate by soaking the substrate in 10 mL acetone and passing the solute through an MN 619 filter to remove visible particles. The procedure was repeated three times. The filtrate was evaporated to dryness at room temperature, taken up in 2 mL acetone and analysed for usnic acid with high performance liquid chromatography (HPLC) using a RP-18 Lichrocart 125–4 column, a mobile phase of 100% methanol and 0·9% phosphoric acid, a detection wavelength of 270 nm and an injection volume of 20 µL. The standard used in the analysis was commercial (+)-usnic acid (Sigma U-7876).

Statistical analyses were conducted using spss 14·0 for Windows (SPSS Inc.). Effects of the addition of usnic acid, acetone washed lichen or intact lichen (substrate addition treatment) and the mycorrhizal symbiosis (mycorrhiza yes/no) were analysed with a two-factor anova followed by orthogonal contrasts to compare among treatments and treatment combinations. To satisfy the assumptions of anova, (normal distribution and heterogeneity of variance) dependent variables were transformed when necessary.

Results

In contrast to our expectations, adding pure usnic acid to the growth substrate had no effect on total biomass accumulation in pine seedlings (Fig. 1). The average root : shoot ratios ranged between 0·71 (mycorrhizal control) and 1·15 (non-mycorrhizal pure usnic acid added). The substrate addition treatment and mycorrhizal treatment interacted significantly (P < 0·001), and according to contrast analyses, non-mycorrhizal pines exposed to pure usnic acid had significantly higher root : shoot ratios than pines in the other treatments (P < 0·05). This effect resulted from the seemingly (but not statistically significantly) smaller needle masses in non-mycorrhizal pines when pure usnic acid was added, while the root masses remained comparable to other non-mycorrhizal treatments (Fig. 1). Altogether, adding intact or acetone washed lichens had no negative effects on pine growth (Fig. 1). Instead, ectomycorrhizal pines grew better in the intact lichen amended treatment, while the substrate additions did not markedly affect the growth of non-mycorrhizal pines (significant substrate addition treatment × mycorrhiza treatment interaction, Fig. 1).

Figure 1.

Average total biomass, needle mass, stem mass and root mass in non-mycorrhizal and mycorrhizal pine seedlings expressed as mg dry weight per plant. The plants were grown in the four different substrate addition treatments: no additions (control), usnic acid added (UA), acetone-washed lichen added (lichen-w), and intact lichen added to the growth substrate (lichen). Mean values ± SE, n = 12, are shown. The significance values (p) from a two-factor anova are given. Different letters indicate statistically significant differences between the substrate addition and mycorrhiza treatments and their interactions (P < 0·05).

The addition of pure usnic acid or lichens did not significantly affect the needle N content by mycorrhizal pines (Fig. 2). However, the needle N content in the non-mycorrhizal pines was significantly lower in control and pure usnic acid treatments when compared to non-mycorrhizal pines with additions of washed or intact lichens (Fig. 2). Therefore, non-mycorrhizal pines seemed to be able to obtain more N when lichen material was added to the growth substrate in comparison to control and usnic acid added treatments resulting in N acquisition rates comparable to mycorrhizal plants (Fig. 2). Substrate additions did not increase the needle N% in mycorrhizal pines, but adding intact lichens increased N% in non-mycorrhizal pines making it comparable to that in mycorrhizal pines in the same treatment (Fig. 2).

Figure 2.

Average total needle nitrogen content per plant (mg N) and needle nitrogen concentration (percentage N per dry weight, N%) in non-mycorrhizal and mycorrhizal pine seedlings grown in the four different substrate addition treatments: no additions (control), usnic acid added (UA), acetone-washed lichen added (lichen-w), and intact lichen added to the growth substrate (lichen). Mean values ± SE, n = 12, are shown. The significance values (P) from a two-factor anova are given. Different letters indicate statistically significant differences between the substrate addition and mycorrhiza treatments and their interactions (P < 0·05).

Calculated from the known amount of added usnic acid at the start, we can estimate that < 30% of the original addition of usnic acid was present at the end of the experiment. Mycorrhizal symbiosis had no clear effect on the amount of usnic acid at the end of the experiment (Table 1, substrate addition treatment P < 0·001, mycorrhiza treatment P = 0·075, interaction P = 0·161). Addition of lichen fragments significantly increased and mycorrhizal symbiosis significantly decreased the organic matter content of the growth substrate (Table 1, substrate addition treatment P < 0·001, mycorrhiza treatment P < 0·001, interaction P = 0·105). The amount of soluble N in the growth substrate at the end of the experiment was significantly higher in the lichen amended microcosms, but not affected by the mycorrhiza or direct addition of usnic acid (Table 1, substrate addition treatment P < 0·001, mycorrhiza treatment P = 0·306, interaction P = 0·367). Nutrient uptake acidifies growth substrate and in line with the higher average nitrogen uptake in mycorrhizal pines, the growth substrate pH at the end of the experiment was significantly lower in the mycorrhizal than non-mycorrhizal microcosms (Table 1, substrate addition treatment P = 0·014, mycorrhiza treatment P < 0·001, interaction P = 0·104). Substrate pH was overall higher in the lichen-amended microcosms where the lichen material probably ameliorated the generally low pH due to the amount of peat in the growth substrate (Table 1).

Table 1.  Substrate characteristics at the end of the experiment. Mean values ± SE (n = 12) of percent organic matter (OM%), concentration of soluble nitrogen (soluble N), pH and usnic acid concentration in the growth substrate are shown. Different letters indicate statistically significant differences between the substrate addition (no additions, usnic acid added, acetone washed lichen added, intact lichen added) and mycorrhiza (Non-mycorrhizal = NM, Mycorrhizal = M) treatments and their interactions (P < 0·05)
TreatmentOM%soluble N (µg/g dry wt soil)pHUsnic acid (µg/g dry wt soil)
No additions
 NM1·99 ± 0·03b9·99 ± 1·34a4·95 ± 1·58abc0 ± 0
 M1·88 ± 0·03a8·76 ± 1·11a4·88 ± 1·86a0 ± 0
Usnic acid
 NM2·15 ± 0·20b7·67 ± 0·44a4·98 ± 1·90bc87 ± 4b
 M1·80 ± 0·03a9·69 ± 0·87a4·87 ± 1·82a84 ± 8b
Washed lichen
 NM4·67 ± 0·05d12·40 ± 0·81b4·98 ± 1·89bc10 ± 1a
 M4·14 ± 0·08c13·07 ± 1·09b4·96 ± 1·51abc14 ± 1a
Intact lichen
 NM4·54 ± 0·05d11·63 ± 0·93b5·03 ± 1·66c80 ± 4b
 M4·44 ± 0·07c12·95 ± 0·82b4·92 ± 1·66ab86 ± 3b

Discussion

In contrast to previous reports of negative effects of usnic acid on plant physiology (Lascève & Gaugain 1990; Latkowska et al. 2006, Lechowski et al. 2006) in our experiment usnic acid had no clear negative effects on pine seedlings exposed to lichen fragments containing usnic acid or to pure usnic acid. While previous work included crop plants such as lettuce (Romagni et al. 2000), tomato (Latkowska et al. 2006; Lechowski et al. 2006), maize or sunflower (Lascève & Gaugain 1990), we investigated a plant species that naturally grows in ecosystems where lichens are abundant. Boreal forest plants are constantly exposed to phenolic substances in their rooting environment (Muscolo & Sidari 2006) and, unlike agricultural crop species, can be expected to be adapted to phenolics such as usnic acid in soil. Furthermore, the experiments so far have been conducted in hydroponic culture conditions with solution pH set close to pH 7, as usnic acid is practically insoluble at lower pH levels. In our experiment, the soil substrate pH was around 5, and therefore only trace amounts of usnic acid would have been in water soluble form. The insolubility to water, however, does not prevent possible toxicity of usnic acid and many known toxins are lipophilic and water insoluble (e.g. Mahmoud & Croteau 2002; Acamovic & Brooker 2005).

The only evidence for an effect by pure usnic acid addition on the pines was the significantly higher root : shoot ratio in the non-mycorrhizal seedlings exposed to pure usnic acid when compared to other seedlings. The total biomass accumulation or needle N contents were not significantly impaired by pure usnic acid addition, and therefore it is not straightforward to evaluate the significance of the effect on root : shoot ratio. In any case it suggests that, in contrast to mycorrhizal seedlings, the non-mycorrhizal seedlings may be affected by pure usnic acid in the rhizosphere. However, in natural conditions pine seedlings are obligatorily mycorrhizal and non-mycorrhizal pine trees are extremely unlikely to ever occur (Jones et al. 2003). The lack of toxic symptoms in the ectomycorrhizal pine seedlings therefore strongly suggests that under ecologically relevant conditions usnic acid does not impair pine seedling growth or nutrient uptake.

We noted no effects of usnic acid on the mycorrhiza-mediated nitrogen gain by the pine in symbiosis with the ectomycorrhizal fungus S. variegatus, which suggests that the fungus was not negatively affected by usnic acid. This is at variance with previous reports of the antifungal effects by usnic acid in clinical tests (reviewed by e.g. Ingólfsdóttir 2002; Oksanen 2006) and with reports of negative effects of lichen extracts on mycorrhizal fungi grown axenically (Brown & Mikola 1974; Goldner et al. 1986). Although clinical tests have shown clear antimicrobial effects, these tests include microbes that would never encounter usnic acid in their natural growth environment. Furthermore, results obtained under sterile agar and at high pH conditions may not be ecologically applicable to forest soil. In our more natural experimental set-up, we did not detect effects of pure usnic acid on soil soluble N concentrations, suggesting that adding this polyphenol did not influence microbial N mineralization rates.

It seems unlikely that significant quantities of usnic acid leaches out of the lichen layer (Stark et al. 2007a). Should usnic acid enter the soil in lichen fragments, however, it would be degraded in soil either through microbial degradation or chemical breakdown. Various micro-organisms, such as pseudomonad bacteria (Bandoni & Towers 1967; Kutney et al. 1977b), the saprophytic fungi Mortierella (Bandoni & Towers 1967; Kutney et al. 1977a, 1984) and Mucor (Kutney et al. 1984) have the capacity to degrade (+)-usnic acid. The chemical conversion of (+)-usnic acid to (+)-2-acetoxyusnic acid could be achieved by an oxidative process (Kutney et al. 1977b), possibly by peroxidase enzymes and phenol oxidases. The analyses on the amount of usnic acid in the mesocosms showed that usnic acid was degraded to a large extent during our experiment. As usnic acid is known to be degraded by light (Begora & Fahselt 2001), we protected the microcosms from light and therefore expect the degradation due to light to be small in our system. The lack of treatment effects of usnic acid addition nevertheless suggests that neither usnic acid nor the oxidation products of usnic acid are likely to have allelopathic effects on pine seedlings or their mycorrhizal partner. Ectomycorrhizal fungi are known to produce phenol oxidases (Burke & Cairney 2002) that could break down phenolic acids such as usnic acid. However, we did not find any clear effect of presence/absence of the ectomycorrhizal fungus S. variegatus on the amount of usnic acid in the growth substrate at the end of the experiment to suggest that mycorrhizal fungi had a major role in degradation of usnic acid in our experiment.

By contrast to the expectation of allelopathic effects, the addition of lichen fragments to the growth substrate increased biomass accumulation in mycorrhizal pine seedlings and total needle N acquisition in non-mycorrhizal seedlings. The lichen fragments decomposed during the experiment to such a degree that they were not visually discernible at the end of the experiment. It has been shown previously that Cladoniaceae lichens may decompose rapidly in soil (Sedia & Ehrenfeld 2006). Lichen fragments may have fuelled the saprophytic microbial activity in the lichen amended treatments in our experiment and released N from the lichen fragments. Therefore, our treatment of adding lichen fragments probably functioned as an addition of N to the system. This assumption is supported by the higher soluble soil N concentration in the lichen fragment treatment. Enhanced N availability enabled higher N acquisition by the non-mycorrhizal pines but not by the mycorrhizal pines, which had high N acquisition rates across all treatments. The N concentration of the peat added into the growth substrate was 1% and markedly higher than that of the added lichen material (0·2%). Pine seedlings in symbiosis with S. variagatus have been shown to access N bound in peat previously (Sarjala & Potila 2005). We therefore assume that the mycorrhizal pines in our experiment could utilize peat as a source of N more efficiently than the non-mycorrhizal pines. Consequently, the N liberated from decomposing lichen material did not add significantly to the available N for the pines in symbiosis with S. variegatus, but did so in the non-symbiotic pines with limited access the N sources in the peat in our growth medium. Usnic acid did not interfere with the uptake of N by pine seedlings whether it originated in the lichen fragments themselves or in form of pure powder added to the growth substrate.

Although our results do not support an assumption of an allelochemical role of C. stellaris, the lichen mats undeniably have an important role in the northern ecosystems, the most important being mediated through their effects on the soil microclimate. The soil temperatures under the lichen carpet are considerably lower than in the exposed soil, thus reducing rates of microbial processes, but the carpet also stabilizes soil moisture levels, thus creating stable conditions for microbial processes (e.g. Broll 2000). The lichen carpet seems to have ambiguous effects on soil microbial processes. The rates of microbial respiration are consistently lower in areas where lichens are removed by reindeer grazing in comparison to undisturbed lichen vegetation (Väre et al. 1996; Stark et al. 2003). However, nitrogen mineralization rates are sometimes higher (Stark et al. 2000) and sometimes lower (Stark et al. 2003) in soils underneath lichen mats compared to those without lichen mats. Soil moisture is also a critical factor affecting seedling establishment. The positive effects by C. stellaris cover on pine regeneration (Steijlen et al. 1995; den Herder et al. 2003) could possibly be explained by higher soil moisture in lichen covered soil, as in the dry oligotrophic pine forests, moisture is most often the limiting factor in tree productivity. On the other hand, dense lichen layer certainly forms a physical obstacle for pine seedlings, and Cowles (1982) report high pine seedling frequencies at the edge of a lichen mat where the lichen mat can probably provide shelter against soil drought without physically suppressing seedling emergence.

Our results show that the lichen C. stellaris does not seem to have negative effects on pine seedling growth that would be mediated by the secondary metabolite usnic acid. This conclusion is derived from our finding that adding pure usnic acid or lichen fragments did not exert negative effects on pine seedlings. By contrast, adding lichen fragments interacted with the effects of mycorrhizas on pine seedling growth, resulting in positive effect on the seedling N acquisition in non-mycorrhizal seedlings. We conclude that the previously reported negative effects of the lichen layer on pine seedling growth are most probably not mediated by usnic acid, but by other chemicals in the water extractions released from the lichens, or by the physical effects of the lichen mat.

Ancillary