We tested the hypothesis that lichen species with a photosynthetic CO2-concentrating mechanism (CCM) use nitrogen more efficiently in photosynthesis than species without this mechanism. Total ribulose bisphosphate carboxylase-oxygenase (Rubisco; EC 220.127.116.11) and chitin (the nitrogenous component of fungal cell walls), were quantified and related to photosynthetic capacity in eight lichens. The species represented three modes of CO2 acquisition and two modes of nitrogen acquisition, and included one cyanobacterial (Nostoc) lichen with a CCM and N2 fixation, four green algal (Trebouxia) lichens with a CCM but without N2 fixation and three lichens with green algal primary photobionts (Coccomyxa or Dictyochloropsis) lacking a CCM. The latter have N2-fixing Nostoc in cephalodia. When related to thallus dry weight, total thallus nitrogen varied 20-fold, chitin 40-fold, Chl a 5-fold and Rubisco 4-fold among the species. Total nitrogen was lowest in three of the four Trebouxia lichens and highest in the bipartite cyanobacterial lichen. Lichens with the lowest nitrogen invested a larger proportion of this into photosynthetic components, while the species with high nitrogen made relatively more chitin. As a result, the potential photosynthetic nitrogen use efficiency was negatively correlated to total thallus nitrogen for this range of species. The cyanobacterial lichen had a higher photosynthetic capacity in relation to both Chl a and Rubisco compared with the green algal lichens. For the range of green algal lichens both Chl a and Rubisco contents were linearly related to photosynthetic capacity, so the data did not support the hypothesis of an enhanced photosynthetic nitrogen use efficiency in green-algal lichens with a CCM.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Lichens are the symbiotic phenotype of nutritionally specialized fungi (mycobionts) that derive carbon, and in some cases nitrogen, from algal or cyanobacterial photobionts or both (Honegger 1991). Depending on the species, the photobionts of lichens have different modes of acquiring CO2 for photosynthesis and at least three major groups have been distinguished (reviewed in Palmqvist 1995).
Lichens of the first group have cyanobacteria, such as Nostoc, as their primary photobiont and, similar to free-living cyanobacteria (Badger & Price 1994), these have a photosynthetic CO2-concentrating mechanism (CCM) (Badger et al. 1993). The CCM actively transports and accumulates HCO3– into the cell, resulting in a local elevation of CO2 in the vicinity of ribulose bisphosphate carboxylase-oxygenase (Rubisco). In cyanobacteria, Rubisco is located in carboxysomes, where it is believed that the accumulated HCO3– is exclusively dehydrated through the action of a specific carbonic anhydrase (Badger & Price 1994).
The reason for the development, or loss, of a photosynthetic CCM among the green algal lichen photobionts is not yet understood (Palmqvist et al. 1997). However, the CCM might have an indirect ecological advantage if less nitrogen needs to be invested in Rubisco and photorespiratory enzymes in these organisms (Raven 1991). So far, few studies have tested this hypothesis and none have included lichens (Beardall, Griffiths & Raven 1982; Surif & Raven 1990), so in this study we compared the photosynthetic nitrogen use efficiencies of four green algal (Trebouxia) and one cyanobacterial (Nostoc) lichen with a CCM, and three green algal (Coccomyxa and Dictyochloropsis) lichens without a CCM. In addition to total nitrogen and Rubisco, two other nitrogenous constituents of lichens, Chl a and chitin, were quantified. Chl a is closely correlated with photosynthetic capacity in a broad range of crustose lichens with different photobionts (Tretiach & Pecchiari 1995) and hence can be used as a measure of functional photosynthetic units of lichens, irrespective of CO2 acquisition strategy. Chitin is a polymer of β-1,4-linked N-acetyl-glucosamine residues (C8O6H15N) and is a major constituent of the fungal cell wall (Muzzarelli 1977), thus serving as a relative measure of nitrogen investments into the fungus (Ekblad & Näsholm 1996).
MATERIALS AND METHODS
All lichens except Lasallia pustulata (L.) Mérat were from different localities in the county of Västerbotten, Sweden as specified below. Cetraria islandica (L.) Ach. was collected from blocks of stone exposed to the sky on a dry pine (Pinus sylvatica) heath at Hamptjärn (Umeå); Hypogymnia physodes (L.) Nyl. from pine trunks in a dense mixed Norway spruce (Picea abies)–pine–birch (Betula pubescens) forest stand at Ulterviken (Umeå); Lasallia pustulata from a rock at the shore of lake Vänern, Kålland Island, Västergötland; Lobaria pulmonaria (L.) Hoffm. from Salix caprea trunks in a dense Norway spruce forest stand outside the National Park Björnlandet (Fredrika); Nephroma arcticum (L.) Torss. from a moss- covered area below a sparse canopy just outside Kulbäcksliden experimental forest (Vindeln); Peltigera aphthosa (L.) Willd. from north-facing grass and moss-covered blocks of stone outside Botsmark; Peltigera canina (L.) Willd. from calcite bricks remaining from an old house in Österåker (Kroksjö); and Platismatia glauca (L.) W. Culb & C. Culb from lower branches in a dense Norway spruce forest stand outside Björnlandet (Fredrika). All samples representing each species were collected on the same occasion towards the end of the growing season (August–September) 1995. Before collection, dry thalli were sprayed with water to avoid fragmentation and collection of visibly partly deteriorated material. In the laboratory, the lichens were dried in a cold room (10 °C) and thereafter stored dry in a freezer (– 18 °C) for up to 12 months before experimental use.
Photobionts and modes of CO2 and nitrogen acquisition
C. islandica, H. physodes, L. pustulata and P. glauca have green algal Trebouxia photobionts and possess a photosynthetic CCM (cf. Palmqvist 1995). P. canina also has a CCM (Badger et al. 1993) as it has the cyanobacterium Nostoc as the only photobiont, which also provides the thallus with combined nitrogen through N2 fixation (Rai 1988). L. pulmonaria has the green alga Dictyochloropsis reticulata, which lacks a CCM (Palmqvist et al. 1994), as the primary photobiont, but also bears N2-fixing Nostoc in internal cephalodia (Rai 1988). Both N. arcticum and P. aphthosa have the green alga Coccomyxa, which lacks a CCM (Palmqvist et al. 1994), as their primary photobiont along with N2-fixing Nostoc in cephalodia (Rai 1988).
Measurements of photosynthetic capacity
The lichen thalli were removed from the freezer, sprayed with water and reactivated for 2–3 d under controlled conditions before photosynthesis measurements as described previously (Palmqvist 1993). CO2 gas exchange measurements and gas mixing were also performed as in Palmqvist (1993) in a flow-through gas exchange system. Depending on species, one to four thalli (≈ 20 cm2) were used for each measurement. The thalli were mounted on a wire tray and placed in the gas-exchange cuvette. All measurements were made at 15 °C and under saturating light (300 μmol photons m–2 s–1), with thallus water contents optimal for photosynthesis of the respective species. The same thallus was used for measurements first at ambient (35 Pa) and then at a higher (85 Pa) CO2 concentration. Gross, rather than net, CO2 fixation was used as a measure of photosynthetic capacity because dark respiration varied at least 2–3-fold between species. At each CO2 concentration, measurement began with a 10–15 min dark period followed by up to 30 min in light to attain a steady-state rate of net photosynthesis. Thereafter, dark respiration was again followed for 10–15 min. Gross photosynthesis was calculated from the assumption that dark respiration mainly originates from the fungus and that this activity remains unaffected in light. After measurements, a photocopy was taken of the sample to determine the projected effective area of each thallus in its fully hydrated state. Thereafter, the dry weight and the Chl content (see below) were determined for each thallus and related to each other and to effective area. The same parameters were also determined for the thalli used for the in vitro analyses outlined below and compared with the thalli used in the photosynthesis measurements. Within each species, these characteristics varied only to a minor extent between the thalli used in this study.
At least 2 g dry weight (10–20 cm2) of each species was used for the quantification of nitrogen. Healthy-looking lichen thalli were removed from the freezer, equilibrated to room temperature and lightly sprayed with water to determine their effective area (see above). The samples were thereafter freeze-dried, weighed and ball-milled (Retsch-Mühle, Haan, Germany) to a powder. Sub-samples of these batches were used for Chl and chitin analyses (see below). The nitrogen analyses were carried out by an authorized laboratory (SLU, Röbäcksdalen, Umeå, Sweden).
Chitin was measured using high-pressure liquid chromatography (HPLC) as described in Ekblad & Näsholm (1996). In brief, 10 mg freeze-dried and milled lichen powder was treated with 0·2 mol dm–3 NaOH to remove proteins and amino acids that interfere with glucosamine in the HPLC separation. The chitin chains were hydrolysed with acid (6 mol m–3 HCl), yielding glucosamine residues that were converted to fluorescent derivatives by treatment with 9-fluorenylmethylchloroformate. These were subsequently analysed in the HPLC using an isocratic elution. In the present study, background contamination was reduced by using norvalin instead of homocysteic acid, used in Ekblad & Näsholm (1996), as the internal standard. The glucosamine residues are separated as three peaks in the HPLC; in the present study peak 3 was the best separated and therefore used for quantification (cf. Ekblad & Näsholm 1996).
Chlorophyll preparation and determination
The Chl concentrations of the samples used for CO2-gas exchange, nitrogen and chitin analyses were determined according to Ronen & Galun (1984), by heating the milled material for 40 min in DMSO (60 °C) in the presence of MgCO3 to avoid excessive breakdown of chlorophyll. After extraction, the sample was centrifuged at room temperature (14 000 r.p.m.; 5 min) and the supernatant was analysed spectrophotometrically. Total Chl (Chltot) of the samples used for protein preparation (see below) was also determined after extraction in dimethyl sulphoxide (DMSO). The Chl concentration of the lichen protein extracts (see below) (Chlpe) was determined by transferring 100 mm3 sample to 900 mm3 pure, ice-cold acetone. The sample was put in a freezer (– 18 °C) for 20 min and thereafter centrifuged at 4 °C (14 000 r.p.m.; 5 min). The resulting supernatant was analysed spectrophotometrically. The extinction coefficients of Arnon (1949) were used for both DMSO and acetone extracted Chl pigments.
Protein preparation and determination
Approximately 50 mg of dry lichen material, pooled from subsamples of three thalli, was frozen in liquid nitrogen and ground to a powder in a mortar. Half was used for Chltot determination and the remainder homogenized in 1·5 cm3 solubilization medium (specified below) in a glass homogenizer. Several protein preparation protocols and solubilization media were tested to optimize protein yield, minimize breakdown of Chl and reduce the effects of lichen phenolics. The following solubilization medium was finally chosen as giving the highest yields of both protein and Rubisco [20–70 g protein (g Chl a)–1 depending on species]: 1 mol m–3 EDTA, 1 mol m–3 PMSF, 100 mol m–3 sucrose, 100 mol m–3 NaOH, 5 mol m–3 2-mercaptoethanol and 40 mol m–3 DTT. The latter was added immediately before preparation (9·3 mg to 1·5 cm3 medium). After homogenization, 200 mm3 SDS [10% (kg m–3)] was added [≈2% (kg m–3) final concentration] and the sample was stored on ice for 15 min before room-temperature (25 °C) centrifugation (17 000 r.p.m.; 3 min). The supernatant was transferred to a new tube and 200 mm3 SDS [10% (kg m–3)] added, yielding a 4–5% (kg m–3) final SDS concentration. The sample was heated (75 °C for 5 min), frozen and stored at –18 ° C for a few days before further analysis.
The protein content of each sample was measured according to Bradford (Bradford 1976) by mixing 25 mm3 of sample with 1 cm3 of pure, ice-cold, acetone. The mixture was frozen (–18 °C) for 1 h, centrifuged at 4 ° C (14 000 r.p.m.; 5 min) and the acetone phase was removed. The pellet was dissolved in 100 mm3 NaCl (0·15 mol m–3), 1 cm3 of Bradford reagent added, and the sample analysed spectrophotometrically and quantified against a Bovine Serum Albumin (Pierce) standard curve.
Quantification of Rubisco
The protein samples were separated on linear 10% SDS-polyacrylamide gels using a BIORAD Mini-Protean II (Biorad, USA) and loaded on an equal-Chl basis (0·3 μg Chla+b in each lane) on two identical gels. These were electrophoresed together for 1–1·5 h at 200 mA. One gel was stained with 0·25% (kg m–3) Coomassie Brilliant Blue and destained in 10% (kg m–3) methanol plus 10% (m3 m–3) acetic acid. The proteins on the other gel were blotted to a Millipore Immobilon membrane in a transfer buffer [192 mol m–3 glycine, 25 mol m–3 Tris, 20% methanol (m3 m–3), 0·03% SDS (kg m–3)] for 3 h at 60 V. The Immobilon membrane was washed in blocking buffer [20 mol m–3 Tris-HCl; pH 7·5, 150 mol m–3 NaCl, 0·25% (kg m–3) Triton x-100, 2% (kg m–3) low-fat dried milk powder] at 6 °C overnight and then washed (2 × 5 min) in 50 cm3 washing buffer [20 mol m–3 Tris, 150 mol m–3 NaCl, 0·05% (kg m–3) Tween 20 and 5% (kg m–3) low-fat dried milk powder]. The primary antibody, spinach large subunit (LSU) Rubisco antiserum made in rabbit (provided by Professor Rolf Brändén, Department of Biochemistry and Biophysics, Chalmers University of Technology, Göteborg, Sweden), was diluted (1:1000) in 10 cm3 antibody buffer [20 mol m–3 Tris-HCl, 150 mol m–3 NaCl, 0·25% (kg m–3) Triton X-100 and 2% (kg m–3) low-fat dried milk powder], incubated on a shaker (25 °C; 1 h) and then excess antibody washed off (2 × 5 min) in washing buffer. Secondary antibody, anti-rabbit IgG antibody from donkey conjugated to horse-radish peroxidase (Amersham, UK), was diluted (1:10 000) in 10 cm3 antibody buffer specified above and incubated on a shaker (25 °C; 1 h). After washing (2 × 10 min) the membrane was placed in dH2O and incubated with Enhanced Chemi Luminescent reagent (Amersham) for 1 min and shaken. Thereafter, the membrane was exposed to Cronex 4 film (Sterling Diagnostic Imaging, USA) for 2–8 min.
The spinach Rubisco antibody cross-reacted with Rubisco from all eight lichens and the Rubisco LSU appeared as one distinct band (≈55 kDa) on the gel (Fig. 1a). However, the immunoblots could not be used for quantitative analysis of Rubisco because the apparent affinity of the antibody was lower for the lichen photobionts than for spinach Rubisco (not shown). In addition, the antibody bound more strongly to Trebouxia Rubisco than to the other photobionts (not shown). However, the blots were used for precise detection and identification of the Rubisco LSU (≈55 kDa) polypeptide band on the Coomassie-stained gels. Two polypeptide bands were of significantly higher density than the other bands (Fig. 1b). One of these corresponds to the size of Rubisco LSU (Fig. 1a), and the other (≈15 kDa), to the size of to the small subunit (SSU) of Rubisco (Andrews & Lorimer 1987). As the Rubisco holo-enzyme is made up of eight large and eight small subunits, in both green algae and cyanobacteria (Andrews & Lorimer 1987), the similar densities of the 55 and 15 kDa bands indicate that most protein in these two bands can be ascribed to Rubisco. The density of the 55 kDa band (OD55) relative to loaded Chl or to total density of all bands in the same lane or both (ODtot) (Fig. 1b) could subsequently be used as a measure of Rubisco amount in the lichens. Optical densities were quantified by using a gel-scanner and Imagemaster 1D computer software (Pharmacia, Sweden).
Consistent with the general difficulty of breaking algae and cyanobacteria, the yield of broken photobionts was lower than 100% in six of the eight lichens and also varied significantly between species (Table 1). The yield of chlorophyll in the protein extracts was nevertheless reproducible within each species (Table 1), and we assumed equal yields of Rubisco and Chl. We could then use OD55 relative to ODtot, OD55 relative to Chl and OD55 relative to thallus weight as measures of relative Rubisco amount, which allowed a direct comparison between species and samples.
Table 1. . Yield of broken photobiont cells in protein extracts of the lichens. The yield was calculated from the ratio of total Chl (Chltot) in each lichen to the Chl concentration in the protein extract (Chlpe) obtained as described in Materials and Methods. Vaules are means ±1SE of two to four separate protein preparations, using material from three thalli in each preparation
The nitrogen concentration varied by a factor of about 20 among the investigated lichen species, whether related to thallus weight or to effective area (Table 2) and were similar to those previously reported for the same or closely related species (Rai 1988; Crittenden, Kalucka & Oliver 1994). Three of the Trebouxia lichens had lower nitrogen concentrations [0·14–0·28 mmol N (g DW)–1] than are generally found in the leaves of higher plants, while the other five lichens had 1·2–2·6 mmol N (g DW)–1 and were more similar to leaves in this respect (Field & Mooney 1986; Schulze et al. 1994). Of the lichens with N2-fixing Nostoc as primary or secondary biont, the bipartite species, P. canina, had by far the highest nitrogen concentration, particularly in relation to area (Table 2). The Trebouxia lichen, L. pustulata, had a nitrogen concentration similar to the tripartite, N2-fixing species, in relation to both weight and effective area.
Table 2. . Nitrogen content relative to thallus dry weight and effective area, that is, the area when hydrated and fully expanded. Values are averages for 10–15 thalli of each species that were pooled for the analysis as described in ‘Materials and methods’. Accuracy of the nitrogen analysis was ± 10%
Plant leaves show a clear relationship between photosynthesis and nitrogen content, as one or several nitrogenous leaf components directly limit photosynthetic capacity (Amax) (Field & Mooney 1986; Schulze et al. 1994). This relationship, which is general for a broad range of species and leaf types, can be expressed by a linear equation:
where m has been empirically defined as 149 and b as –76·1 for C3 plants, when expressing Amax as nmol CO2 g–1 s–1 and N as mmol g–1 (Field & Mooney 1986). The investigated lichens also showed a clear relationship between light-saturated gross photosynthesis and nitrogen (Fig. 2), even though m was considerably lower, compared to C3 plants, both at ambient (m = 7·9) (Fig. 2a) and at saturating CO2 (m = 16·8) (Fig. 2b).
The Potential Photosynthetic Nitrogen Use Efficiency (PPNUE), a parameter that allows a direct comparison of the Amax–N relationship between and among species (cf. Field & Mooney 1986), was also calculated for each lichen. As shown in Fig. 3, PPNUE was highest in the lichens with the lowest nitrogen contents and photosynthetic rates (compare with Fig. 2), but decreased to a plateau around 20–25 μmol CO2 (mol N)–1 s–1 for the lichens with the highest nitrogen contents and photosynthetic rates. This pattern was similar when photosynthesis was measured at ambient or saturating CO2 (Fig. 3). As also shown in Fig. 3, the negative relationship between PPNUE and nitrogen displayed by the range of investigated lichens is completely different from the positive relationship of PPNUE to both Amax and nitrogen in higher plants (Field & Mooney 1986).
Nitrogen partitioning between symbionts
There are at least two possibilities that may explain the negative relationship between PPNUE and nitrogen in the lichens. First, lichens with low nitrogen may indeed have a higher photosynthetic capacity per unit nitrogen invested into photosynthetic units than species with higher nitrogen. Alternatively, the negative relationship might result from increasing nitrogen investments into non-photosynthetic constituents with increasing nitrogen availability, with the efficiency of photosynthetic nitrogen investments remaining unaffected. To distinguish between these two possibilities we measured the distribution of nitrogen between photosynthetic and non-photosynthetic components in the eight lichens, using Chl a as a relative measure of photosynthetic units and chitin as a relative measure of nitrogen invested into the fungus.
When related to thallus dry weight, the Chl a content varied by a factor of five and the chitin content varied by a factor of 40, among the eight species (Table 3). As for nitrogen, both Chl a and chitin concentrations were in agreement with previous studies (Crittenden et al. 1994; Tretiach & Pecchiari 1996). The lichens fell into two groups when the fraction of total nitrogen invested into Chl a was related to total thallus nitrogen (Fig. 4a). The three Trebouxia lichens with thallus nitrogen below 0·3 mmol N (g DW)–1 had invested 1–2% of their nitrogen into Chl a, while the other species had invested significantly less (0·2–0·4%). Two groups could also be distinguished with respect to nitrogen investments into chitin. For the four Trebouxia lichens, the relative chitin content decreased with increasing thallus nitrogen, while the four species with N2 fixation had invested 6–8% of their nitrogen into this compound (Fig. 4b). Thus, when taken together, all four Trebouxia lichens had invested a significantly lower amount of their nitrogen into chitin in relation to Chl a, compared with the four N2-fixing species (Table 3).
Table 3. . Chlorophyll a and chitin contents and the ratio Chl a:chitin on the basis of nitrogen (N) equivalents. Each Chl a contains 6·27% N, on the basis of molecular weight, so the N content in Chl a was calculated as [N]in Chl = Chl a (0·0627). Chitin contains 6·33% N and its N content was calculated in the same way. The ratio was rounded to the nearest integer. Both components were measured on samples taken from the same batch of pooled, freeze-dried and milled material (see Materials and Methods). Values are means ± 1SE of at least three subsamples of these batches
Photosynthetic capacity in relation to Chl a and Rubisco
The relative content of Rubisco in relation to extracted protein (OD55/ODtot) was similar for all lichens, irrespective of photobiont species or nitrogen status, being in the range of 4·0 ± 0·4–6·2 ± 0·6% (Table 4). However, when related to thallus weight, the relative Rubisco content varied by a factor of about four between the species, with P. aphthosa having the highest and H. physodes the lowest concentration (Table 4). As for Chl a, the relative Rubisco content per unit nitrogen decreased significantly with increasing thallus nitrogen (Table 2, Table 4) and there seemed to be good correlation between Rubisco and Chl a contents over the whole range of lichens (Fig. 5a). However, when the relative content of Rubisco of each species was related to its Chl a content and then plotted as a function of thallus nitrogen, there was obviously an increase in Rubisco relative to Chl a with increasing thallus nitrogen, with the exception of P. canina (Fig. 5b).
Table 4. . Density of the 55 kDa polypeptide band (Rubisco LSU; OD55) in relation to the density of all other protein (polypeptide) bands (ODtot) separated on Coomassie-stained 10% SDS-polyacrylamide gels. The Rubisco LSU was detected and quantified as described in Materials and Methods. Rubisco relative to thallus DW was calculated using the Chl content of the loaded sample (0·3 μg Chla+b) and the Chl content (Chltot) per thallus DW of each lichen (Table 1) using the relation: (OD55/0·3) × (Chltot). Rubisco relative to nitrogen was obtained by dividing OD55(g DW)–1 with the N content per thallus DW (Table 2). Relative Rubisco contents were rounded to the nearest integer. It was assumed that the extraction yields of Chl and Rubisco were equal in the protein preparations. Values are means ± 1SE of two to four separate protein preparations using material from three thalli in each preparation. Each protein preparation was separated on at least two gels
Figure 6 shows the relationship between light-saturated gross photosynthesis and Chl a or Rubisco, respectively, on the basis of dry weight. At ambient CO2 (35 Pa), there was a linear relationship between Chl a content and photosynthesis for lichens with green algal primary photobionts, while the Nostoc lichen (P. canina) had a somewhat higher rate in relation to Chl a compared with the others (Fig. 6a). For the green algal lichens, there was a linear relationship also between relative Rubisco content and photosynthesis at ambient CO2 (Fig. 6b). P. canina again appeared as an outlier, having the highest rate of photosynthesis in spite of having one of the lowest Rubisco contents (Fig. 6b; Table 4). At the higher CO2 concentration (85 Pa), photosynthesis increased significantly in the three tripartite lichens and the Trebouxia lichen L. pustulata, while H. physodes and P. glauca had been saturated already by ambient CO2. The relationship between Chl a content and photosynthesis was therefore changed to a more curvilinear one, with apparently no further increase in light- and CO2-saturated photosynthesis above ≈1·4 μmol Chl a (g DW)–1 (Fig. 6c). The same type of relationship was found between Rubisco content and light-saturated photosynthesis as between Chl a and photosynthesis, with no further increase in light- and CO2-saturated photosynthesis above a relative Rubisco content of ≈30 (g DW)–1 (Fig. 6d). Below these apparently saturating concentrations of Chl a and Rubisco there was approximately a doubling in light- and CO2-saturated photosynthesis when the amount of the respective compound was doubled (Figs 6c and 6d).
Figure 7 shows the area-based relationship between Chl a and light-saturated net photosynthesis or respiration, respectively. Similar to the gross photosynthesis–Chl a relationship, net photosynthesis also increased linearly with increasing Chl a content at ambient CO2 (Fig. 7a) and slightly curvilinearly at the higher CO2 concentration (Fig. 7b). Respiration also increased with increasing Chl a content, at both of the CO2 concentrations, but to a somewhat lesser extent compared with photosynthesis (Figs 7c and 7d).
Green-algal lichens with a CCM do not have enhanced nitrogen-use efficiency of photosynthesis
For the range of investigated green-algal lichens our data do not support the hypothesis of enhanced nitrogen-use efficiency of photosynthesis in species with a CCM compared with species without this mechanism. In contrast, the three Trebouxia lichens with the lowest concentrations of thallus nitrogen had distributed a much larger fraction of their nitrogen into Chl a and Rubisco than lichens with higher nitrogen (Table 2, Table 4, Fig. 4). Also, at ambient CO2, the green algal lichens together displayed a linear relationship between net or gross photosynthesis and Chl a content as well as between light-saturated gross photosynthesis and relative Rubisco content, without any significant outlier species (Figs 6 and 7). This implies that, irrespective of nitrogen status or CO2 acquisition strategy, the various green algal lichens have a similar efficiency with respect to their nitrogen investments into photosynthetic light harvesting and carbon fixation. The cynaobacterial lichen, P. canina, however, appeared to be somewhat more efficient because its photosynthesis was higher both in relation to Chl a and particularly to Rubisco (Figs 6 a and b) at ambient CO2. However, cyanobacteria have phycobilins instead of Chl for light harvesting, which can explain their higher photosynthetic efficiency in relation to Chl a (cf. Raven et al. 1990). Also, Vmax of the cyanobacterial Rubisco may be up to 3–4-fold higher per unit Rubisco compared with that of green algae (Badger & Andrews 1987), which would explain the exceptionally high photosynthetic capacity in relation to Rubisco of P. canina (Fig. 6b).
All lichens, including the cyanobacterial lichen P. canina, had invested a significant part of all their protein (4–6%) into Rubisco (Table 4). If we assume that 25% of the lichen proteins are of photobiont origin (Rai 1988), and if we correct for the differences in photobiont preparation yield (Table 1), Rubisco may amount to 20–30% of the photobiont protein, irrespective of species. The relative Rubisco contents of the photobionts are thus similar to those of higher crop plants, which generally invest 20–30% of their leaf nitrogen in Rubisco (Makino et al. 1992). Moreover, for the range of green-algal lichens, the ratio of Rubisco to Chl increased with increasing nitrogen (Fig. 5b) which is also generally the case for higher plants (Makino et al. 1992).
As the total protein contents of the lichen samples used for Rubisco analysis were known (see Materials and Methods), we could estimate their absolute Rubisco content, using the relation: [(OD55/ODtot) × protein]. For this analysis we had to add the protein content of the unbroken photobionts to the extracted protein, to account for the low and variable yield of broken cells (Table 1). However, as we assumed a ratio between photobiont and mycobiont protein of 25:75, without having this data for all the species (Rai 1988), our estimate is rather crude. Nevertheless, the ratio of Rubisco to total Chl was between 2 and 4 mmol Rubisco (mol Chl)–1, with the highest ratio in the high nitrogen species, consistent with the data presented in 5Fig. 5b. These ratios are somewhat lower than in leaves of crop plants, which may have 5–12 mmol Rubisco (mol Chl)–1 (Makino et al. 1992), but were similar or slightly higher than for free-living Chlorella cells (Yokota & Canvin 1986).
Nitrogen invested into the fungus increases with nitrogen supply
The positive relationship seen in higher plants between Potential Photosynthetic Nitrogen Use Efficiency (PPNUE) and leaf nitrogen content (Field & Mooney 1986) is the result of an increasing proportion of nitrogen being invested into photosynthetic units, particularly Rubisco, with increasing nitrogen supply (Evans 1983; Makino et al. 1992; Schulze et al. 1994). In clear contrast to this, the range of investigated lichens displayed a negative relationship between PPNUE and thallus nitrogen content (Fig. 3), implying that as the nitrogen supply increases, more nitrogen is invested in the fungus. This is supported by the increased chitin to Chl ratio in lichens with the highest thallus nitrogen contents (Table 3, Fig. 4). This significant difference in nitrogen partitioning between plant leaves and lichen thalli suggests that co-operation between mycobiont and photobiont may be sub-optimal with respect to potential energy gain, as the fungal partner seems to expropriate nitrogen better allocated to photosynthetic capacity. However, chitin is a major structural molecule of the fungal cell wall (Muzzarelli 1977) and, as the mycobiont hyphae make up most of the lichen biomass, nitrogen investments into this non-photosynthetic component are required for thallus expansion. This requirement for non-photosynthetic nitrogen for growth, in combination with the higher energetic costs of acquiring and reducing nitrogen compared with CO2, inevitably result in a lower competitive fitness for light and space of the chitin-walled lichens compared with the cellulose-walled plants (Duchesne & Larson 1989).
This point is further emphasized by comparing data from wheat (Triticum) with that from the lichens. When the nitrogen content of expanding wheat leaves was increased from 80 to 140 mmol m–2, by nitrate fertilization, the total Chl content subsequently increased from 0·36–0·62 mmol m–2 (Evans 1983). This would correspond to an increase in Chl a from ≈0·3 to ≈0·5 mmol m–2, assuming a Chl a to b ratio of 3. Five of the lichens had a significantly higher nitrogen content per area compared with the wheat leaves (177–449 mmol m–2, Table 2). Despite this, the Chl a content was not higher than 0·25 mmol m–2 in any of the investigated lichen samples (Fig. 7). As 500 mg Chl a m–2 (= 0·56 mmol m–2) is required to absorb more than 95% of the incident photosynthetically active radiation (Raven 1992), it appears that investment into chitin is favoured over investments into photosynthetic units, even before the photobiont layer is sufficiently developed to absorb all incident light that can be used for photosynthesis.
In spite of the nitrogen requirement for cell-wall formation and thallus expansion, the proportion of photosynthetic to non-photosynthetic tissue in lichens must still be balanced to favour net energy gain. It may therefore seem surprising that chitin relative to Chl increased with increasing thallus nitrogen (Table 2, Table 3, Fig. 4), as this also suggests a relative increase in respiratory load in the form of fungal and non-photosynthetic tissue. Once made, however, chitin is metabolically inactive, so an increased chitin content may not necessarily result in increased maintenance respiration. In contrast, if the mycobiont is able to invest in more and thicker cell walls when nitrogen is abundant, this might be of competitive advantage for the whole lichen, as this would probably increase the water-holding capacity of the hyphae (Rundel 1988), hence prolonging periods of metabolic activity (cf. Sundberg et al. 1997). Also, in spite of the increasing chitin to Chl ratio with increasing thallus nitrogen, the ratio of photosynthesis to respiration appeared to remain fairly constant, as shown by the apparent correlation between Chl a content and respiration rate across the investigated green algal lichens (Fig. 7c).
Is there a regulation and/or optimization in resource allocation between the lichen bionts?
Even though lichens are rather primitive symbiotic organisms (Honegger 1991; Raven 1992) they apparently maintain a similar balance between energy input and expenditure functions irrespective of species and nitrogen supply. It also appears that when nitrogen is scarce, investment into photosynthetic units is favoured over investment into fungal tissue. Thus, there appears to be a regulation of resource allocation between photobiont and mycobiont, even though the nature of this regulation remains unknown. However, the size of the photobiont population appears to be controlled by the mycobiont (Hill 1993), implying that the photobionts may have little influence on the direction and fate of metabolite flow.
Our study was restricted to rather few lichen species and considered only the thallus nitrogen contents that these achieved over years of growth in the field. Therefore more data are needed to understand if, and how, partitioning of nitrogen, and other resources between energy input and expenditure reactions is regulated in lichens. The investigation could be approached in at least two different ways. First, we need to collect similar data for a broader range of lichen species, both with respect to taxonomic, ecological and morphological affiliation. Second, we also need to manipulate the nitrogen status of individual lichen thalli during prolonged periods of growth to see whether the data presented here are applicable when the nitrogen content is changed towards higher or lower concentrations than appear to be natural for the species.
This investigation was supported by a grant to K. P. from the Swedish Natural Sciences Research Council (NFR Stockholm, Sweden). D. C. was supported by an NFR Post-doctoral Fellowship. We thank M. Zetherström (Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural Sciences, Umeå, Sweden) for technical assistance with the HPLC measurements. Dr S. Jansson (Department of Plant Physiology, Umeå University, Sweden) made valuable comments and suggestions for improvements of our Rubisco quantification protocol.