Translocation of 15N indicates nitrogen recycling in the mat-forming lichen Cladonia portentosa

Authors


Author for correspondence:P. D. Crittenden Tel: +44 115 9513211 Fax: +44 115 9513251 Email: pdc@nottingham.ac.uk

Summary

  • • Nitrogen translocation was measured in Cladonia portentosa during 2 yr growth in Scottish heathland. Translocation was predicted to occur if N is resorbed from senescent basal tissue and recycled within the thallus.
  • • 15N was introduced into either the lower (TU thalli) or upper (TD thalli) 25 mm of 50-mm-long thalli as 15N-NH4+, 15N-NO3 or 15N-glycine. Labelled thalli were placed within intact lichen cushions, either upright (TU) or inverted (TD). Vertical distribution of label was quantified immediately following labelling and after 1 and 2 yr.
  • • Independently of the form of introduced label, 15N migrated upwards in TU thalli, with new growth being a strong sink. Sink regions for 15N during year 1 (including new growth) became sources of 15N translocated to new growth in year 2. Upward migration into inverted bases was minimal in TD thalli, but was again marked in new growth that developed from inverted apices.
  • • Relocation of N to regions of growth could facilitate internal N recycling, a process postulated to explain the ecological success of mat-forming lichens.

Introduction

Terricolous mat-forming lichens of the genera Alectoria, Cetraria, Cladonia (subgenus Cladina), Flavocetraria and Stereocaulon are among the principal vegetation components on well drained terrain at high latitudes. This includes oligotrophic tundra, forest tundra and subarctic taiga (Kershaw, 1977; Ahti & Oksanen, 1990; Bliss & Matveyeva, 1992). For example, Auclair & Rencz (1982) estimated that 4.4 × 106 km2 of lichen woodland occur in Canada alone. In such habitats, mat-forming lichens make significant contributions to ecosystem function, including primary production, nutrient cycling (Auclair & Rencz, 1982), nitrogen fixation (Crittenden & Kershaw, 1979) and water cycling (Vowinckel & Orvig, 1976; Lafleur & Schreader, 1994). They are consumed by caribou and reindeer as winter forage, and therefore constitute a key component of the trophic structure of circumboreal and circumpolar regions (Arseneault et al., 1997). Mat-forming lichens are also locally abundant in temperate heaths (Rodwell, 1991; Brown et al., 1993) and forests (Watson & Birse, 1991).

The ecological success of mat-forming lichens has been attributed to their specialized mode of growth (Crittenden, 1989, 1991). Mat-forming lichens grow acropetally (at the apices vertically upwards) while older basal regions of thalli senesce, generating an understorey of intact and persistent litter or necromass. A structural consequence of the mat-forming mode of growth is that the thalli of such lichens are only loosely attached to the underlying substrata, the stability of mats being maintained by interlocking between adjacent thalli and between the lichen canopy and the shoots of adjacent plants. The mat- forming growth habit is restricted to c. <0.5% of lichen species; most lichens are either firmly attached to, or intimately associated with, underlying substrata by means of a living mycelium. In contrast, accumulated litter in well developed lichen mats can constitute a large proportion of the standing lichen mass. This recurrent production of basal necromass is a consequence of indeterminate growth; interception by the apical strata of light and nutrients deposited from the atmosphere creates a zone of depletion in lower strata, rendering thallus tissue below a certain depth in the lichen mat expendable (Lechowicz, 1983; Sveinbjörnsson, 1987; Coxson & Lancaster, 1989). Crittenden (1991) hypothesized that the large-scale senescence of older basal tissue in mat-forming lichens might facilitate internal recycling of potentially growth-limiting nutrients such as N and P. Remobilization of nutrients from senescing tissue and their upward translocation to the growing apices might promote apical growth rates in excess of those that could be sustained solely by external sources, that is, from atmospheric deposits. Under steady-state conditions, enhanced growth rates at the apices funded by tight internal recycling would be accompanied by higher rates of litter production and the development of deeper mats. Accordingly, terricolous lichens with a mat-forming growth habit have certain attributes broadly comparable with those of dominant (sensuGrime, 2001) vascular plants in unproductive habitats: deep canopies and production of copious persistent litter (see also Facelli & Pickett, 1991). Implicit in this argument is the suggestion that the development of basal litter, and the capacity to relocate nutrients, are advantageous traits for which there has been positive selection.

The above model for the growth of mat-forming lichens is consistent with available data. First, there are strong vertical gradients in the concentration of elements within lichen mats (Pakarinen, 1981; Hyvärinen & Crittenden, 1998a; Ellis et al., 2003) and in physiological activity (Plakunova & Plakunova, 1984; Sveinbjörnsson, 1987), with maxima in the apices. For example, Moser et al. (1983) observed that 14CO2 incorporation did not occur below a depth of c. 55 mm in physiologically active mats of Cladonia rangiferina and Cladonia stellaris >80 mm deep, and that 70–80% of total respiratory activity occurred in the uppermost 40 mm (Nash et al., 1980). Second, a marked and characteristic pattern of vertical variation in δ15N values of lichen thalli is consistent with putative internal N recycling (Ellis et al., 2003). Third, mat-forming lichens have high capture efficiencies for wet deposited inorganic N and phosphorus (Crittenden, 1989; Hyvärinen & Crittenden, 1998c), while there is no evidence that they acquire significant quantities of N from underlying soil (Ellis et al., 2003, 2004). Fourth, well developed lichen mats cast deep shade (Kershaw & Harris, 1971; Lechowicz, 1983) and are formidable physical barriers to seedling emergence (see Crittenden, 2000 for discussion).

Pivotal to this explanation for the ecological success of mat-forming lichens is a hypothesized capacity for vertical translocation of metabolites in the fungal mycelium of the thallus, from senescing basal tissue to regions of growth, following a source–sink relationship. To test this hypothesis, we used the stable isotope 15N as a tracer, to examine the extent of N translocation in thalli of the mat-forming heathland lichen C. portentosa.

Materials and Methods

Study site

Field work was conducted in Calluna vulgaris–Scirpus cespitosus blanket mire at The Halsary, Caithness (58°25′50″ N, 3°23′37″ W; altitude c. 90 m asl). The common heathland lichen Cladonia portentosa (Dufour) Coem. is an abundant component of the mire vegetation, where it grows as discrete cushions among raised hummocks (cf. Rodwell, 1991).

Initial preparation of lichen samples

Naturally hydrated aggregates of thalli were collected from lichen cushions (powder-free latex gloves were worn at all times when handling lichens). Thalli were cleaned of extraneous debris, cut to a uniform length of 50 mm (measured downwards from the apex), and transferred to the laboratory where they were air-dried overnight. Each thallus was tagged with a preweighed acetate label and weighed. Oven-dry mass was estimated using the air-dry : oven-dry mass ratio of additional thalli oven-dried at 40°C. The mean initial mass of thalli was 159 ± 6.24 mg (± 1 SE) and there were no significant differences in mean mass between the six subsequent treatment combinations (see below): tested using a Kruskal–Wallis one-way anova, H = 5.488, P = 0.358, df = 5 (genstat ver. 7.1, Payne, 2003). Weighed thalli were returned to the field and rehydrated by lightly spraying with deionized water.

Exposure to 15N and growth of thalli

Labelling of lichen thalli with 15N was undertaken in the field under ambient conditions of temperature and indirect irradiance. This minimized the length of time hydrated, metabolically active lichens were subjected to artificial environments, and allowed the experiment to be set up in a remote location without access to sophisticated laboratory facilities such as a controlled environment room.

Rewetted podetia were exposed in the field for 1 h to a solution of simulated rainwater augmented with 10 µg ml−1 15N. Simulated rainwater was prepared to a composition approximating the mean concentration of salts recorded in precipitation at a rainfall sampling station in the UK Acid Deposition Monitoring Network, 7.2 km from the study site (Campbell et al., 1994, data collated from the UK National Air Quality Information Archive): NaCl, 11.2 µg ml−1; MgSO4·H2O, 7.2 µg ml−1; HCl, 7 µg ml−1; CaSO4, 2.4 µg ml−1; K2SO4, 0.44 µg ml−1. The concentrations of NH4+ and NO3 in precipitation at the Halsary are similar (c. 0.33 mg N l−1), and N concentration in the simulated rainwater was therefore c. 30 times greater than that in natural precipitation. Labelled N was added as either NH4SO4, NaNO3 or glycine, each compound enriched in 15N to ≥ 98% (Cambridge Isotope Laboratories, Inc., Andover, MA, USA). Each 50-mm thallus was submerged to a depth of 25 mm in labelled solution, using uncapped 30-ml universal vials containing 5.2 ml simulated rainwater (total N in solution, 52 µg, represents ≤ 20% of total N in a typical thallus weighing 150 mg and having a total [N] value of 2 mg g−1).

Thalli were exposed to labelled rainwater in one of two orientations: top-up (TU) thalli in which the bottom 25 mm of the thallus was submerged, or top-down (TD) thalli in which thalli were inverted and the apical 25 mm submerged. Exposure to label was undertaken in a 0.05 m3 humidified transparent Perspex glove box to reduce evaporation, and vials were agitated gently at frequent intervals by manual rotation. Air temperature in the glove box during exposures was in the range 16–18°C. After 1 h submersion, thalli were raised out of the bathing solutions and excess liquid was removed by contact with absorbent paper. The inclusion of TU and TD thalli was designed to investigate mechanisms of N transport whereby orientation-dependent differences in patterns of 15N migration would indicate the relative importance of extracellular and intracellular transport. For example, if translocation of N was largely extracellular, then upward translocation of 15N might occur irrespective of thallus orientation.

An initial subsample (72 thalli), comprising 12 replicates in each of the six treatment combinations, was returned to the laboratory immediately after exposure to 15N (time zero, t0). The remaining thalli were inserted into protective, cylindrical, stainless steel mesh cages (c. 35 mm diameter × 85 mm tall; 0.4 mm gauge wire with 16 holes cm−2) embedded vertically into otherwise undisturbed cushions of C. portentosa; thalli were positioned in the cages with the apices uppermost (TU thalli) or lowermost (TD thalli). The cages were open at the top and filled to c. 30 mm with compacted lichen litter, providing a substratum permeable to water and raising the experimental thalli in unity with the height of the canopy of the surrounding lichen cushion. The experiment was set up between 12 and 26 June 2000, and samples harvested after c. 12 months (27 and 28 July 2001, t1) and 24 months (20 June 2002, t2). Some loss of replicates occurred and recovery was incomplete (c. 80% at t1, n = 10; c. 50% at t2, n = 6). Additionally, the glycine treatment at t0 was repeated on 20 June 2002 because of instrument failure in the original analysis of 15N.

Preparation of samples for 15N analysis

Thalli harvested at t0, t1 and t2 were cleaned of extraneous debris, oven-dried at 40°C and weighed. Thalli were then rehydrated in water-saturated air (over water in a desiccator at 4°C for 12 h) and cut horizontally into three strata measured downwards from the end of the thallus that was not exposed to 15N: 0–10, 10–25 and 25–50 mm. New growth recorded for TU thalli at t1 and t2 was cut into strata measured at 10 mm intervals, upwards beyond the position of the original apex in 50 mm thalli at t0 (Fig. 1a). New growth recorded for TD thalli at t2 was analysed as a single unit (Fig. 1b). All strata were oven-dried at 40°C, weighed, reduced to powder in a ball grinder and stored in sealed glass vials until analysis.

Figure 1.

Comparison of 2 yr growth in (a) top-up (TU) and (b) top-down (TD) thalli of Cladonia portentosa. Solid horizontal lines, strata selected for analysis; dashed line, new growth in TD thalli.

Determination of N concentration and 15N abundance

Powdered samples were transferred to ultra-clean tin capsules (Elemental Microanalysis Ltd), oven-dried at 40°C and weighed (Cahn C-31 Microbalance). Total N concentration and 15N abundance were determined using a 20–20 continuous flow stable isotope mass spectrometer (ANCA-SL, Europa Scientific, Crewe, UK). Because of the low [N] in C. portentosa tissue, ≥20 mg powdered material was submitted for 15N analysis. Values of 15N are expressed as atom % (excess) 15N, calculated by subtracting the atmospheric background [15N] (0.366%).

Calculation of growth rate, N-uptake rate and N translocation

Relative growth (RG, mg g−1) and relative growth rate (RGR, mg g−1 yr−1; Hunt, 1982) were calculated as:

image(Eqn 1 )
image(Eqn 2 )

where m is mass and t is time. Nitrogen-uptake rate (URN, mg g−1 h−1) was calculated as:

image(Eqn 3 )

where mN is the quantity of N or 15N in tissue.

The relationship between mass and 15N content of new growth in TU and TD thalli was compared by regression analysis (genstat ver. 7; Payne, 2003). Standard models were tested (linear, logarithmic and exponential) and their efficacy based on the relationship between residuals and fitted values, the distribution of residuals and values of adjusted R2. In all cases, one anomalous value created excessive leverage and was omitted from the analysis.

Results

Thallus growth

Thalli of C. portentosa achieved significant mass gain during the course of this experiment. Relative growth in TU thalli during the intervals t0→1 and t0→2 was c. 406 ± 25 and 1001 ± 110 mg g−1, respectively (mean ± 1 SE). Corresponding RGR values were 336 ± 18 and 335 ± 26 mg g−1 yr−1. Top-up thalli increased in height by 35.8 ± 3.18 mm during t0→2. Relative growth rate in TD thalli during t0→1 and t0→2 was 308 ± 48 and 188 ± 14 mg g−1, respectively. New growth in TD thalli originated from younger regions of the thallus in the inverted apices (Fig. 1b).

Nitrogen uptake

Uptake of N in C. portentosa was readily measurable from each source (NH4+, NO3, glycine) and in both thallus orientations (TU, TD). Between 10 and 30% of supplied N was taken up within 1 h exposure, equivalent to mean URN values of between 0.055 ± 0.004 and 0.276 ± 0.012 mg N g−1 h−1 (Table 1). When compared by two-way anova using genstat ver. 7 (Payne, 2003), the results revealed differences in uptake rate between orientations and between N compounds (Table 2). Uptake rates were greater in the apices (TD) than in the lower thallus (TU), although for a given form of N and thallus orientation the absolute uptake increased with mass of the exposed stratum (data not shown).

Table 1.  Mean uptake rates (URN) for different forms of 15N in apical (top-down, TD) and subapical (top-up, TU) regions of the thallus of Cladonia portentosa
Nitrogen formThallus orientationnURN (mg g−1 h−1 ± 1 SE)
GlycineTU120.055 ± 0.004
TD120.091 ± 0.005
NH4+TU 70.220 ± 0.014
TD 70.276 ± 0.012
NO3TU 80.063 ± 0.006
TD110.111 ± 0.059
Table 2.  Results of a two-way anova and Fisher's least significant difference test to compare nitrogen uptake rate (URN) in Cladonia portentosa between orientations (top-up, TU; top-down, TD) and between different N forms
Source of variationdfAdjusted SSAdjusted MSv.r.F. prDifference between means 
  • *

    Significant at 5% level.

Between orientations 10.0280.028 53.64<0.001TU − TD−0.045*
Between N forms 20.2970.148283.67<0.001Glycine − NH4+−0.175*
     Glycine − NO3−0.018*
     NH4 − NO3  0.158*
Treatment × N form 20.0080.004  0.79<0.461  
Residual500.0280.005    
Total550.352     

Vertical variation in total N and 15N concentrations in thalli

At t0, TU thalli of C. portentosa had a pattern of increasing [N] upwards towards the thallus apices, and this pattern was maintained in the new growth recorded at t1 and t2 (Fig. 2). A similar pattern was evident in TD thalli, but where [N] increased downwards towards the inverted apices, although the gradient became weaker with time and modified by high [N] values in new vertical growth (Fig. 3). The distribution of atom % (excess) 15N in TU thalli changed markedly from t0 when values were highest in the exposed basal strata, through t1 to t2 when values became more uniform throughout the thallus (Fig. 2). Initial (t0) values of at % (excess) 15N in TD thalli were also highest in the exposed stratum (inverted apices); a vertical gradient, with values decreasing upwards toward the inverted bases, remained evident at t1 and t2. However, atom % (excess) 15N increased markedly in new vertical growth (Fig. 3). These data indicate 15N uptake and its subsequent movement within the thallus.

Figure 2.

Nitrogen concentration (filled bars) and atom % (excess) 15N (open bars) at t0, t1 and t2 in horizontal strata of top-up thalli of Cladonia portentosa exposed to different forms of 15N-labelled N. Additional strata (+), new growth measured upwards beyond the position of the original apex at 0 mm (cf. Fig. 1a). Plotted values are means ± 1 SE, n = 6–12.

Figure 3.

Nitrogen concentration (filled bars) and atom % (excess) 15N (open bars) at t0, t1 and t2 in horizontal strata of top-down thalli of Cladonia portentosa exposed to different forms of N. Data for new growth indicated as single bars (cf. Fig. 1b). Plotted values are means ± 1 SE, n = 6–12.

Nitrogen transport

Data for [N] and atom % (excess) 15N (Figs 2, 3) were used to calculate the percentage distribution among strata of total 15N taken up (Figs 4, 5). These data show gradients and temporal changes in 15N similar to, but with greater resolution than, those of at % (excess) 15N values. In the 15N-NH4+ treatment, c. 95% of label was located in the lowermost stratum at t0 in both TU and TD thalli while <0.05% was detected in the top 10 mm. By t1 there had been massive export of label out of the lowermost stratum of TU thalli and into the 15–25 and 0–10 mm strata, and especially into new growth. In TD thalli, in contrast, export of label out of the lowermost stratum was more modest, with very little located in the uppermost stratum. During year 2, export of label from the lowermost stratum in TU thalli was much less than in the previous year. However, new growth was now a major sink for 15N, to the extent that all strata that had been sinks for label in year 1 (10–25 mm; 0–10 mm; new growth in year 1) became donors in year 2. In TD thalli, new growth was also a major sink for label in year 2, and there was a commensurate increase in export from the lowermost stratum. Despite this movement into new growth, very little label appeared in the uppermost stratum (the original inverted thallus bases). Movement of label upwards into new growth in TD thalli, while appearing to bypass the uppermost stratum, is consistent with the new growth originating in the middle and/or lowermost strata (upwards from the inverted apices). Redistribution of label introduced as 15N-glycine mirrored that in the 15N-NH4+ treatment remarkably closely. Redistribution of label introduced as 15N-NO3 also followed a similar pattern, but in this case export from the exposed stratum was more evenly distributed between years 1 and 2 in both TU and TD treatments, and larger quantities of label appeared in the inverted bases of TD thalli.

Figure 4.

Percentage distribution of 15N among strata in top-up thalli of Cladonia portentosa exposed to different forms of N at t0 (filled circles); t1 (open circles); and t2 (filled triangles). Mean values (± 1 SE as horizontal error bars, n = 6–12) are plotted in the centre of their corresponding stratum, which is delimited by vertical bars (NG = additional strata in new growth upwards from position of original apex). Horizontal histogram shows difference between years (filled bars, t0t1; open bars, t1t2) in percentage of total thallus 15N contained within equivalent strata.

Figure 5.

Percentage distribution of 15N among strata in top-down thalli of Cladonia portentosa exposed to different forms of N, at t0 (filled circles); t1 (open circles); and t2 (filled triangles). For further explanation see Fig. 6.

Movement of 15N into new growth of both TU and TD thalli was positively related to the mass increment (Fig. 6). A linear relationship would have suggested a constant rate of movement into a sink of invariable strength. By contrast, the increase in percentage 15N was best described logarithmically (Fig. 6).

Figure 6.

Relationship between mass (log10) and 15N content of new growth in top-up (TU, filled circles) and top-down (TD, open circles) thalli of Cladonia portentosa (filled triangles = anomalous value). 15N content expressed as percentage of total thallus 15N. Tested using regression analysis: TU, R2 = 0.634 (F = 78.97, df = 43, P  0.0001); TD, R2 = 0.516 (F = 15.91, df = 12, P = 0.0015).

Nitrogen loss

It should be possible to calculate loss of 15N from labelled thalli during the 2-yr study as a change in the mean total 15N content of equivalent thalli harvested at t0, t1 and t2. However, as different thalli were harvested on each occasion and the mass of exposed strata at t0 is not known in all cases, and because of the effect of mass on absolute uptake (see above), any such calculations are associated with uncertainties. If it is assumed that, within each treatment, the average masses of the exposed stratum in all subsamples (at t0, t1 and t2) were similar, then such calculations suggest that a decrease in total 15N content occurred during nine of the 12 intervals considered (Fig. 7). Of the three anomalous increases in 15N content, two (glycine TU and TD) arise from a repeat labelling (see above) which might have resulted in 15N loadings or mass values of the exposed strata that were atypical of the main experiment. Overall, therefore, there is strong evidence that 15N loss occurred.

Figure 7.

Total quantity of 15N in top-up (TU) and top-down (TD) thalli of Cladonia portentosa exposed to different forms of N at t0, t1 and t2. Filled circles, 15N-glycine; open circles, 15N-NH4+ filled triangles, 15N-NO3. Putative N loss during the course of the experiment indicated by an arrow and quantified as percentage loss yr−1. Plotted values are means ± 1 SE, n = 6–12.

The potential confounding effect of 15N loss on calculations of label distribution change was examined by rerunning the calculations for TU thalli, but with the estimated total losses during years 1 and 2 (Fig. 7) added to the average amount of 15N in the 50–25 mm stratum (t1) or 50–0 mm stratum (t2), respectively. Even in the unlikely event that 15N was lost exclusively from these regions, the recalculated data support 15N movement upwards out of exposed strata (50–25 mm, t0→1) or strata corresponding to the original thallus (50–0 mm, t0→2).

Discussion

Growth, and uptake and retention of 15N

Top-up thalli of C. portentosa made large growth increments during the 2-yr study, with several replicates almost doubling in height (Fig. 1). This growth represents a potentially strong N sink. The mean RGRs of TU thalli during intervals t0→1 and t0→2 were similar (c. 335 mg g−1 yr−1) and compare favourably with mean values in the range 260–431 mg g−1 yr−1 measured in 50-mm-long thalli of the same species at several sites in the British Isles using a similar culture method (Hyvärinen & Crittenden, 1998b). The culture of C. portentosa in mesh cages inserted into otherwise undisturbed lichen cushions was intended to create growth conditions broadly comparable with those experienced in an intact mat. However, the effects on growth of cutting thalli at 50 mm (e.g. breaking the continuum with necromass), and of a modified physical environment in the mesh cages (e.g. increased exposure to irradiance, modified water relations), are not known.

All forms of 15N supplied to thalli of C. portentosa were taken up, consistent with literature documenting the utilization by lichens of NH4+ and NO3 (Smith, 1960a; Lang et al., 1976; Crittenden, 1996, 1998; Dahlman et al., 2002) as well as some forms of organic N (Smith, 1960b; Kielland, 1997; Dahlman et al., 2004). Introduction of label into the target thallus stratum was achieved with relative precision; contamination of unexposed strata appeared to be small, but was greatest in the 15N-NO3 treatment for both TU and TD thalli, perhaps reflecting a greater mobility of this ion in the thallus-free space. The extent of apparent contamination will also have been dependent on the accuracy of the final dissection. Our results demonstrate the differential uptake of N by C. portentosa, in the order NH4+ > NO3 > glycine (Tables 1, 2). This concurs with the results of laboratory studies demonstrating a greater URN for NH4+ than for NO3 (Smith, 1960a; Lang et al., 1976; Dahlman et al., 2004). Passive binding of NH4+ to cell-wall cation-exchange sites might also have contributed to the greater URN-NH4+ (Miller & Brown, 1999); evidence suggests that NH4+ ions are only temporarily held on lichen cell wall-binding sites, and are subsequently assimilated. A higher URN-NH4+ might also reflect a more efficient transport and assimilatory system for this N form.

There was an overriding trend for progressive loss of label during t0→1 and t1→2, although absolute rates (4–33% yr−1, Fig. 7) are associated with uncertainty caused by an incomplete data set for masses of exposed strata. It is of interest to know whether any 15N loss was an artefact of the experiment or whether it was a natural phenomenon. Natural losses might be explained by loss of metabolites during rainfall (Crittenden, 1983); turnover of surface-bound enzymes (e.g. phosphatases: Lane & Puckett, 1979); and loss of thallus branches caused by physical damage. It should be noted that leakage and loss of metabolites hypothesized to occur during the rewetting of dry lichens, as discussed by Smith (1980), did not find support in field observations (Crittenden, 1983, 1989). Further, estimated 15N losses during t0→1 and t1→2 were broadly similar. Therefore it seems improbable that 15N loss represents label not yet taken up into the symplast being flushed from the free space shortly after setting up the experiment.

Transport and recycling of N

Nitrogen introduced into basal strata of C. portentosa migrated vertically upwards within the thallus. The extent of migration was largely independent of the form in which N was supplied, and new growth was shown to be a strong N sink. For example, at t2≥50% of total label in TU thalli was located in new tissue (new strata) (Fig. 4) that had developed during year 2. It is not known whether this observed translocation is a passive extracellular process involving migration of molecules in the free space, perhaps under the influence of evaporation at the apices, or whether it is intracellular and possibly metabolically driven. However, several elements of the results presented here point to intracellular transport. First, in TD thalli only trace (NH4+, glycine) to small (NO3) quantities of label migrated upwards into the inverted bases representing the uppermost stratum, while comparatively large quantities of label moved upwards into new growth originating largely from the inverted apices (Figs 1, 5). Second, the rate of upward migration into new strata was similar (although slightly greater in TD than in TU thalli) (Fig. 6), although evaporative forces are likely to have been smaller around the emerging new apices in TD thalli as these were initially beneath the lichen canopy surface. Third, the strong translocation of label during t1→2 is difficult to explain in terms of migration in the free space because, while it is possible that label remained within cell walls and exposed to passive forces for a period immediately following exposure to label, it is improbable that residual label remained in this compartment during the second year of growth.

Given the long duration (2 yr) of this study, the short distances over which 15N migration occurred (25–75 mm) could probably be accounted for by intracellular diffusion (Jennings, 1995; Boswell et al., 2002). In contrast, translocation in fungal mycelia attributed to active mechanisms is considerably faster, migration distances comparable with those observed in the present study being achieved in 1–2 d (Olsson & Gray, 1998; Jacobs et al., 2004). In TD thalli, label migrated upwards into apical regions of new growth, but only to a limited extent did it migrate upwards into inverted thallus bases. This might be explained by a greater diffusion resistance in older tissue caused by, for example, a greater frequency of vacuolated or evacuated hyphae, or of septation, or greater tortuosity of the diffusion path. However, the pronounced differential migration into new growth might also indicate active transport of N, a potentially limiting resource, to sink regions. There is abundant evidence that cytoplasm can move in fungal mycelia from older hyphae into regions of active growth (Klein & Paschke, 2004). This is thought to be a mechanism by which filamentous fungi can maximize nutrient conservation and maintain growth under oligotrophic conditions (Paustian & Schnürer, 1987a, 1987b). In mat-forming lichens it might be important in the withdrawal of resources from senescing basal tissues, although detailed anatomical studies that might yield data to verify this suggestion remain to be undertaken. However, it seems unlikely that migration of cytoplasm could account for the observed translocation of 15N so far into the apical stratum.

Do the present results provide evidence of internal recycling of N? In TU thalli, new growth during t0→1 was a strong sink for translocated label, but this new tissue in turn became a source of label for new apical tissue developed during t1→2. This recurrent relocation of resource in a source-to-sink relationship is compelling evidence of recycling: of N allocated to structural and metabolically functional components of physiologically active cells within vegetative mycelium (and algal cells), subsequently remobilized and transported via cytoplasmic pathways into younger cells. However, it is possible that much of the labelled N taken up at t0 remained in excess of requirements for growth and/or cell maintenance, and was either allocated to storage or left unassimilated. Translocation of unassimilated N, and/or relocation of stored N, would not constitute recycling in the sense intended in our initial hypothetical model (see Introduction).

Ellis et al. (2003) observed marked vertical variation in δ15N values in C. portentosa with maximum values in the apices (e.g. −1.5) and minimum values at c. 40–60 mm below the apices (e.g. −7.0). This pattern is typical of mat-forming lichens, and might be explained in the light of Taylor et al.'s (1997) observation that, in fungal tissue of basidiocarps, chitin has a negative δ15N signature while cytoplasm is δ15N-positive. Thus declining δ15N values from the apices downwards in C. portentosa could be generated by ontogenetic changes in the ratio of fungal cell walls (containing δ15N-negative chitin) to fungal cytoplasm plus algal cells (δ15N-positive) as observed by Plakunova & Plakunova (1984). However, Ellis et al. (2003) also showed that the difference between the apical δ15N value and the minimum value at 40–60 mm is positively related to thallus age. This observation could be explained partly by isotope discrimination associated with recurrent resorption and recycling of N, consistent with the present findings for N translocation from source to sink regions within the thallus.

Ecological context

Berendse & Aerts (1987) and Aerts (1990) postulated that natural selection in low-nutrient habitats operates on plant traits that maximize mean residence time of nutrients in living tissues. In C. portentosa such traits include low thallus nutrient concentrations; low tissue turnover rates (long ‘leaf’ lifespan, Westoby et al., 2002); and production and accumulation of secondary chemicals (c. 2% usnic acid, c. 1% perlatolic acid on a dry mass basis, Houvinen & Ahti, 1986). A further mechanism by which mean residence time of nutrients can be increased is internal recycling (Aerts & Chapin, 2000); our present results suggest that this mechanism operates in C. portentosa. Data presented above (7 mg N g−1 in apical tissue, c. 2 mg N g−1 in necromass) suggest that the efficiency of N resorption from senescing thallus in C. portentosa might be as high as 70%. However, the pronounced decrease in thallus [N] from the apices downwards is partly, if not largely, a result of thallus differentiation (e.g. fungal cell-wall thickening and decreasing alga : fungus ratio, Plakunova & Plakunova, 1984). The depth in lichen mats at which senescence is initiated, and at which resorption might occur, is not currently known; coupled anatomical and physiological investigations are now needed to clarify this question. We suggest that internal recycling of key nutrients facilitates higher growth rates in C. portentosa, enabling this and other species of mat-forming lichen to coexist with vascular plants where their vigour is sufficiently reduced because of the low soil nutrient status. Accordingly, the rate of thallus elongation in C. portentosa (c. 18 mm yr−1) is relatively high among lichens, broadly equating to a mass increment of c. 220–330 g m−2 in patches with 95% lichen cover. Extension growth in C. vulgaris and Empetrum nigrum, measured by Lindholm (1980) on a raised bog in southern Finland, was in the range 3–45 mm yr−1 (mean = 15.6); Miller (1979) reported shoot productivity of 193–220 g m−2yr−1 in mature C. vulgaris stands with 95% cover on a Scottish grouse moor. The coexistence between C. portentosa and dwarf shrubs is therefore broadly consistent with available growth data. It is also consistent with the proposed mechanisms for maximizing nutrient-use efficiency in mat-forming lichens.

Acknowledgements

The authors gratefully acknowledge funding from the NERC GANE (Global Nitrogen Enrichment) Programme. C.M.S. is supported by the Scottish Executive Environmental and Rural Affairs Department. We thank Forest Enterprise for granting access to The Halsary; Stuart Gibb for the provision of laboratory facilities at the Environmental Research Institute, Thurso; Winnie Stein (SCRI) for technical assistance in the laboratory analysis of [N] and 15N; and three anonymous referees for constructive comments.

Ancillary