Size-dependent growth of two old-growth associated macrolichen species

Authors


Author for correspondence:
Yngvar Gauslaa
Tel:+47 64 96 57 84
Fax:+47 64 96 58 01
Email: yngvar.gauslaa@umb.no

Summary

  • • Relationships between thallus size and growth variables were analysed for the foliose Lobaria pulmonaria and the pendulous Usnea longissima with the aim of elucidating their morphogenesis and the factors determining thallus area (A) versus biomass (dry weight (DW) gain.
  • • Size and growth data originated from a factorial transplantation experiment that included three boreal climate zones (Atlantic, suboceanic and continental), each with three successional forest stands (clear-cut, young and old).
  • • When A was replaced by the estimated photobiont layer area in an area–DW scatterplot including all thalli (n = 1080), the two separate species clusters merged into one, suggesting similar allocation patterns between photobionts and mycobionts across growth forms. During transplantation, stand-specific water availability boosted area gain in foliose transplants, consistent with a positive role of water in fungal expansion. In pendulous lichens, A gain greatly exceeded DW gain, particularly in small transplants. The A gain in U. longissima increased with increasing DW:A ratio, consistent with a reallocation of carbon, presumably mobilized from the dense central chord.
  • • Pendulous lichens with cylindrical photobiont layers harvest light from all sides. Rapid and flexible three-dimensional A gain allows the colonization of spaces between canopy branches to utilize temporary windows of light in a growing canopy. Foliose lichens with a two-dimensional photobiont layer have more coupled A and DW gains.

Introduction

A lichen is a symbiotic association between one or two autotrophic photobionts (green algae and/or cyanobacteria) and one heterotrophic mycobiont. None of these partners belongs to the kingdom of plants. Nevertheless, lichens function to optimize light harvesting in photosynthesis (Palmqvist, 2000). Thereby, they balance resource investments between photobiont and mycobiont tissues (Palmqvist et al., 2002) to obtain a long-term positive net carbon gain (Lange, 2002, 2003a,b). As lichens desiccate rapidly in direct sunlight, photosynthesis often occurs at lower irradiances in lichens than in most plants. Low light during rainy weather and before sunrise after nocturnal dew formation increases the need for efficient light utilization to compensate for the proportionally longer hydration periods in darkness with net respiratory losses. During a long evolutionary history lichens have developed light-harvesting structures and traits analogous to those in plants. For example, foliose lichens form flat thalli resembling leaves with photobiont cells organized in a light-harvesting horizontal layer enclosed in protecting and supporting nonphotosynthetic fungal tissues (Gauslaa & Solhaug, 2001, 2004). By contrast, many fruticose and pendulous lichens form cylindrical structures by locating their photobiont cells in a sleeve beneath a protecting cortex, resembling plants of arid habitats with their main photosynthetic tissues located in stems. Lichen growth forms correspond to functional groups of lichens inhabiting specific ecological niches (McCune, 1993; Clement & Shaw, 1999; Berryman & McCune, 2006). Despite the fact that growth form is a frequently used categorical variable in ecological studies, we still lack a good understanding of how different lichen growth forms function. Indirect evidence suggests that the ratio of active photobiont versus mycobiont biomass may be relatively similar across taxa and functional groups (Palmqvist et al., 2002). Our first aim was to quantify and compare the lichen biomass per unit area of estimated photobiont layer in two contrasting lichen growth forms collected from the canopy of old-growth forests; the fruticose pendulous beard lichen Usnea longissima Ach. and the foliose lichen Lobaria pulmonaria (L.) Hoffm. A fixed allocation pattern across growth forms may imply a functional balance between carbon input (depending on photobiont area) and carbon loss through respiration (depending on the mycobiont, which comprises most of the biomass).

Lichen growth has traditionally been assessed as radial growth or lobe elongation, measured in mm per year (as reviewed by Palmqvist, 2000). Recent growth studies have shown that lichen area (A) and dry weight (DW) gains are not always tightly coupled (Dahlman & Palmqvist, 2003; Gauslaa et al., 2006). Therefore, simultaneous measurement of A and DW gains in relation to already existing area and mass would be required to understand the mechanisms behind lichen growth processes (cf. Palmqvist, 2000; Palmqvist et al., 2008). After transplantation to a new environment, different responses of the DW: A ratio may occur. (1) The lichen maintains an internally controlled species-specific ratio regardless of environmental conditions. Such lichens are probably confined to specific environmental conditions. (2) The lichen finds a new ratio or balance that suits the new environment, thus being able to acclimate to altered conditions. The second aim of this study was to identify and analyse the response types of the lichens investigated to elucidate how internal and external factors determine growth. To do so, we used a large data set containing the growth rates of many U. longissima and L. pulmonaria transplants. Parts of this data set dealing with species-specific mean growth rates across climatic and successional forest gradients have already been reported (Gauslaa et al., 2007). From its initiation, this data set was also collected for analysis of intraspecific variation in growth, which is our focus in this paper.

In addition to growth form, size matters in ecology (Lawton, 1999). Size variability is a fundamental property of natural populations (Weiner, 1990; Filin & Ovadia, 2007) as size variation translates into variation in growth, survival and reproduction (Harper, 1977; Sugiyama & Bazzaz, 1998). For example, allocation patterns often depend on size in vascular plants (Shipley & Meziane, 2002; McCarthy & Enquist, 2007; Niklas et al., 2007), and much of the intraspecific variation in reproductive effort relates to intrinsic size effects rather than external factors such as population density or nutrient availability (Samson & Werk, 1986; Ohlson, 1988; Shipley & Dion, 1992). There is a significant lack of knowledge as regards the size dependence of lichen growth measured as A and DW gain, as well as that of allometric changes in the A:DW ratio. The size dependence of radial lichen growth is, however, known to be species dependent (McCarthy & Smith, 1995; Bradwell & Armstrong, 2007; Benedict, 2008). To date, radial growth has mainly been studied in flat saxicolous lichens forming circular colonies. Our final aim was to search for allometric growth relationships to document and analyse the size dependence of gains in A, DW and the A:DW ratio of two contrasting lichen growth forms in old forests.

Materials and Methods

Lichen material

The foliose Lobaria pulmonaria (L.) Hoffm. was collected on 6 May 2005 in old boreal Picea abies-dominated Atlantic rainforests in western Norway in Overhalla (64°27′N, 11°53′E, 15 m asl) and Namsos (64°25′N, 11°25′E, 20 m asl). Thalli were sampled on twigs of P. abies and scattered stems of Salix caprea. Thalli from both locations were mixed in one batch. One to five branched lobes (5–23 cm2; 50–315 mg DW; n = 540) were cut from each collected lichen thallus. The pendulous Usnea longissima Ach. was collected on 18 May 2005 in old P. abies stands in eastern Norway in Toten (60°35′N, 11°02′E, 600–700 m asl). One to five thallus fragments (1–35 cm2; 50–650 mg DW; n = 540) were taken from each thallus.

Both species were sampled at sites with large populations. All sampled thalli were air-dried in the laboratory on the collection day. During the short period between collection and transplantation, thalli were stored air-dry at 4°C. Attached tree bark and bryophytes were removed. After randomizing all thalli, pre-transplantation measurements were completed within 4 d. All measurements followed the randomized sequence regardless of treatment.

Transplantation experiment

Four thalli of L. pulmonaria were fastened with a thin thread onto a nylon net on each of 135 square frames (15.9 cm × 15.9 cm). Lobaria pulmonaria was horizontally located in a way that allowed curling during natural hydration–desiccation cycles (Barták et al., 2006). The position simulates its typical orientation on P. abies, in which it is restricted to branches and twigs. For the pendulous U. longissima, one thallus was fastened with a thin thread to each of four vertical plastic sticks fastened to each of 135 additional frames. The sticks allowed a natural hanging position of the thalli above the horizontally located frames. One frame for each species was placed in every frame holder; the holders were placed on poles 1 m above the ground. Forty-five frame holders were placed in each of three selected climatic zones, with 15 frame holders randomly placed in each of three successional stands (for details, see ‘Transportation locations’ in the Materials and Methods section). Transplants were left in the field from 11 June to 30 September (the Atlantic location; 110 d), from 13 June to 1 October (the suboceanic location; 111 d), and from 18 June to 4 October (the continental location; 109 d) in 2005. After harvest, all transplants were desiccated at room temperatures in the laboratory on the harvest day.

Transplantation locations

All locations, like the source habitats, represented boreal forests dominated by P. abies. The Atlantic location (Namsos, western Norway; 64°25′N, 11°27′E, 30 m asl) comprised rainforests with few scattered Betula pubescens and Alnus incana trees. The suboceanic location (Siljan, south-eastern Norway; 59°22′N, 9°45′E, 500 m asl) had a few stems of B. pubescens and Sorbus aucuparia. The continental location (Umeå, northern Sweden; 64°10′N, 19°40′E, 250 m asl) had scattered stems of Pinus sylvestris, B. pubescens, Salix caprea, S. aucuparia and Populus tremula.

In each climatic zone, three successional stands (clear-cut, young and old forest) were selected within short distances and in similar topographic and edaphic situations. The indirect site factor (ISF) (Anderson, 1964) computed from hemispherical photographs (e.g. Englund et al., 2000), which is an estimator of canopy openness and of light availability, showed a consistent pattern in all locations. Young forests were shaded (ISF = 0.141–0.224), clear-cut forests were sun-exposed (ISF = 0.647–0.840) and old forests were intermediate (ISF = 0.229–0.278; see Gauslaa et al., 2007).

The rainfall during the transplantation period varied between 440 mm (81 rainy days) in the Atlantic location and 276 mm (54 rainy days) in the continental location, with the suboceanic location (309 mm; 43 rainy days) being intermediate (Gauslaa et al., 2007). The mean daily temperature for the transplantation period was 12°C in the two forested stands of all three climatic zones; the clear-cut sites in the suboceanic and the continental zones had 1.2°C higher total means. The diurnal amplitudes were lowest in the Atlantic location because of cloudy weather.

Lichen weight, area and growth measurements

The hydrated thallus area was measured with a leaf area meter (LI3100; Li-Cor, Lincoln, NE, USA). The measuring device records the projection of the thallus area placed in a horizontal position. The thallus area measured in this way is abbreviated as ‘A’. Thereafter, lichens were desiccated for 48 h at 20°C before the DW was recorded. Ten additional thalli treated similarly were repeatedly weighed. No significant shift in DW occurred, showing that the air humidity during weighing was constant. These ten thalli were then placed in a desiccator for 4 d before their real DW was recorded, and the weight of transplants was corrected by the weight reduction factor for the sacrificed thalli. DW and A were measured before and after transplantation. Before transplantation, DW (as well as A) of L. pulmonaria (136 ± 2 mg; mean ± SE, n = 540) and U. longissima (185 ± 4 mg; n = 540) did not differ (P > 0.05) between subsamples selected for the forest stands (two-way ANOVA; data not shown), meaning that the data set was well suited for detecting effects of treatments. The growth parameters were as follows.

  • • Thallus dry weight (DW) gain (%) = (DWend −  DWstart) × 100/DWstart
  • • Thallus area (A) gain (%) = (Aend − Astart) × 100/Astart
  • • DW per A before transplantation = DWstart/Astart
  • • DW per A after transplantation = DWend/Aend
  • • Change in DW per A (%) = (DWend/Aend − DWstart/Astart) × 100/(DWstart/Astart)

Transplants with a DW and/or A reduction > 10% were classified as fragmented and excluded from the assessment of growth (L. pulmonaria: 6% lost and fragmented transplants; U. longissima: 11% lost).

Estimation of photobiont layer area (PhA)

The photobiont layer was organized in different ways in the two growth forms, as can be seen in simple cross-sections. The foliose lichen has a photobiont layer that follows the lower part of the upper cortex, whereas photobiont cells in U. longissima form a continuous cylindrical sleeve inside the entire outer cortex of thalli. Therefore, the measured projected area (A) in U. longissima underestimated the PhA by a factor close to 3.14 = ∑(π × 2r × length)/∑(2r × length), where ∑(2r × length) for all main axes and fibrils in one thallus corresponds to A. Lobaria pulmonaria has a complex geometry with ridges and pits. If all pits were shaped like the inside of a hemispherical bowl and densely packed over the entire surface, A should underestimate the PhA by a factor of 2 (2πr2r2). However, 100% coverage of a surface with circles is not possible. Furthermore, the pits have larger widths than depths, and pits are irregularly distributed with flat portions in between. Thus, the real correction factor is well below 2. A rough estimation is a mean correction factor of 1.3. The PhA was computed for each species by using these two species-specific correction factors.

Statistical analyses

Regression analyses were computed in SigmaStat Version 1.0 (SYSTAT Software Inc., Richmond, CA, USA). Differences in regression slopes were tested using an ANCOVA with Systat version 10 (SYSTAT Software Inc.).

Results

Before transplantation, A and DW were tightly correlated in U. longissima (r2 = 0.869) and L. pulmonaria (r2 = 0.745), respectively (Fig. 1a). The morphology of the lichens was presumably in equilibrium with their natural canopy environment. The scatter plot showed no overlap between the two growth forms because of a substantially higher DW: A ratio in the pendulous lichen (Fig. 1a). Thus, the interspecific regression model showed a substantially poorer fit (r2 = 0.448) than the species-specific models. The computed photobiont area (PhA) accounts for the fact that foliose lichens harvest light from one side, whereas pendulous lichens utilize light from all sides. Replacing A by PhA in both growth forms caused the two species-specific clusters (Fig. 1a) to merge into one common cluster with a strong correlation (r2 = 0.788; Fig. 1b). However, the slopes of the two species-specific regression lines differed (P < 0.0001; ANCOVA).

Figure 1.

(a) The relationship between thallus area, measured as projected area and thallus dry weight for Usnea longissima (closed circles) and Lobaria pulmonaria (open circles) sampled in natural habitats. (b) The relationship between the estimated area of the photobiont layer in the two species (correction factors given in the insert) and dry weight.

During the 110-d transplantation in new environments, the mean DW gain across all locations and habitats was 21.2 ± 0.5% (mean ± SE; n ~ 500) in L. pulmonaria and 16.3 ± 0.6% in U. longissima (Fig. 2). The mean A gain showed a highly divergent pattern, with an increase of 13.8 ± 0.5% in L. pulmonaria and as much as 64.8 ± 1.9% in U. longissima. Lobaria pulmonaria had a slightly stronger DW gain than A gain during transplantation, resulting in a higher DW:A ratio, whereas the completely opposite response occurred in U. longissima. In order to analyse the growth responses, DW and A gains were plotted for all thalli (Fig. 3). In L. pulmonaria, DW and A gains were strongly correlated (r2 = 0.725; P < 0.0001), and the tendency for a reduction in the DW:A ratio is clear (regression slope: 0.758). Only the few thalli situated above the 1 : 1 line (Fig. 3a) showed the reverse response. Individual regression coefficients (P < 0.0001) and regression slopes for all combinations of climatic zones and successional stands are shown for L. pulmonaria in Fig. 3a (insert). All slopes were less than 1, showing that the A gain progressively lagged behind the DW gain with increasing growth rate. The slope was highest in the wettest climate (Atlantic; pooled slope for the three stands: 0.812; n = 168) and in the shadiest successional stand with least evaporation (young forests; pooled slope for the three climate zones: 0.877; n = 173), and lowest in the driest climate (continental; slope: 0.535; n = 174) and in the most sun-exposed and driest stands (clear-cut; slope: 0.660; n = 165). An ANCOVA showed that slopes differed highly significantly among the nine stands, particularly with respect to climatic zone (Table 1).

Figure 2.

Total mean dry matter (closed columns) and area (open columns) gain during 110 d of transplantation for Lobaria pulmonaria and Usnea longissima. Vertical bars, +1 SE (n ~ 500).

Figure 3.

The relationship between area (A) gain and dry matter (DW) gain in (a) Lobaria pulmonaria and (b) Usnea longissima during a 110-d transplantation period in various Picea abies stands as specified in the insert. The insert also shows individual regressions (r2 and slope) for all stands. The straight line represents the 1 : 1 line. CC, clear-cut; OF, old forest; YF, young forest.

Table 1.  ANCOVA analyses testing the regression slopes for the thallus area (A) gain versus biomass (DW) gain in Lobaria pulmonaria for all nine sites representing all combinations of climate zones and successional stands (r2 = 0.771; n = 506) (Fig. 3a)
SourceSSdfMSF-ratioP
Site515.8 864.482.10.030
DW gain17015.3 117015.3566.1< 0.000001
Site × DW gain1325.3 8165.675.50.000001
Error14666.748830.05  

DW gain and A gain during the transplantation were poorly correlated in U. longissima (Fig. 3b), so it is difficult to quantify effects of moisture availability for this lichen in our experiment. The strongest relationship occurred in the Atlantic rainforest clear-cut site (r2 = 0.298; P < 0.0001) followed by the suboceanic clear-cut site (r2 = 0.134; P = 0.0023). Nearly all thalli had a substantially higher gain in A than in DW. The change in DW:A ratio for U. longissima was not related to climate and/or successional stand (two-way ANOVA; data not shown), suggesting a low degree of environmental regulation. However, thallus size before transplantation was found to play a significant role. In Fig. 4, various growth variables (y-axes) are plotted against measured size variables at the start of the experiment (x-axes). Area gain (Fig. 4a–c) decreased significantly with increased A (Fig. 4a) and DW (Fig. 4b) at the start of the experiment. However, the DW: A ratio at the start of the experiment was the best predictor of A gain (Fig. 4c). A high biomass per area strongly boosted the subsequent A gain. In contrast, DW gain was not significantly related to any measures of transplant size (Fig. 4d–f), but was strongly influenced by the successional stage (denoted by symbol colours in Fig. 4). The strongest relationships occurred between the change in DW per A and variables measured at the start of the experiment (Fig. 4g–i). The small thalli showed the strongest decline in DW:A ratio, whereas the large thalli more or less maintained their initial DW:A ratio (Fig. 4g–h). Hardly any U. longissima thalli experienced a positive change in DW:A ratio during transplantation. Again, A (Fig. 4g) was clearly a better predictor of changes in DW:A ratio than DW (Fig. 4h), although even the latter relationship was highly significant. The strongest relationship was the linear decline in gain in DW:A ratio with increasing DW:A ratio at the start of the experiment (r2 = 0.794; Fig. 4i).

Figure 4.

Size-dependent growth in transplants of Usnea longissima. (a–c) Area (A) gain; (d–f) dry weight (DW) gain; (g–i) gain in DW:A ratio.

For L. pulmonaria, DW and A gain did not depend on transplant size (Fig. 5). DW (Fig. 5f) and A gain (Fig. 5c) slightly declined with increasing DW:A ratio at the start of the experiment, but > 91% of the variation in regression models did not relate to size or DW:A ratio. Regression analyses were also performed for subsets comprising each forest, but the relationships were weak and nonsignificant. Therefore, the transplant size range used for L. pulmonaria (5–23 cm2; 50–315 mg DW) had little or no influence on subsequent gains in DW and A.

Figure 5.

Size-dependent growth in transplants of Lobaria pulmonaria. (a–c) Area (A) gain; (d–f) dry weight (DW) gain; (g–i) gain in DW:A ratio. See Fig. 4(g) for symbol definitions.

Discussion

Lack of light limits lichen growth in old and open P. abies canopies (Gauslaa et al., 2006, 2007). Forest lichens may have experienced long-term selection for traits that maximize light-harvesting efficiency and net carbon gain during periods of hydration (see Palmqvist & Sundberg, 2000). Thus, the overlap in PhA–DW responses in pendulous and foliose lichens (Fig. 1b) may reflect a common resource optimization across growth forms in humid boreal forests between the carbon gain of the photobiont and the carbon loss of the mycobiont in respiration (Palmqvist et al., 2002). The regression slopes differed between the two growth forms, but small L. pulmonaria thalli were flatter than large thalli with deep pits. Thus, specimen-specific correction factors for computing the PhA in L. pulmonaria would probably reduce the difference in slope between the two lichen species. In the lack of specimen-specific correction factors we cannot reject the hypothesis of an interspecific PhA–DW relationship.

Area gain is not necessarily strongly coupled to DW gain (Fig. 3). The DW:A ratio can be influenced by external factors related to sun exposure and/or available humidity (e.g. Green & Lange, 1991; Maguas & Brugnoli, 1996). Acclimation on temporal (MacKenzie et al., 2001; Gauslaa & McEvoy, 2005; Vráblikováet al., 2006) and spatial scales (McEvoy et al., 2007) may change lichen morphology (Gauslaa et al., 2006). Increased thallus thickness may imply higher solar radiation screening (Snelgar & Green, 1981) and increased water-holding capacity (Gauslaa & Solhaug, 1998). Our results show that habitat-specific A–DW gain regression slopes varied highly significantly depending on variables related to moisture availability (insert in Fig. 3a; Table 1). For such reasons, A and DW gains are likely to be regulated in slightly different ways. In free-living fungi, hyphal extension depends on osmotically active carbon compounds (Gadd, 1995) that together with sufficient water create the turgor pressure needed for cell expansion (Gow, 1995). In lichenized fungi, photosynthesis determines the DW gain, but also the pool of osmotically active carbon compounds (as reviewed by Palmqvist, 2000). According to a review by Taiz & Zeiger (2006; p. 50), plant cell expansion stops before photosynthesis starts to decline with falling water potentials during progressive drought stress. The lichen response seems to be similar to the plant response. Our results (Fig. 3a insert and Table 1) are consistent with the hypothesis that water availability influences fungal expansion and thereby stimulates the A gain more than the DW gain also in lichens. It is a common field observation that lichens at sites with high water availability (moist air and frequent rain) have wider and thinner lobes than those at arid sites. Such patterns are consistent with an area-expanding effect of moisture. However, for U. longissima transplants (Fig. 3b) other factors dominate the A gain–DW gain relationship (see later), so it is difficult to quantify effects of moisture availability for this lichen in our experiment.

Severe high-light-induced chlorophyll degradation occurred in foliose transplants at clear-cut sites, whereas acclimation was strong in forest transplants (Gauslaa et al., 2007). A flat foliose growth form maximizes light harvesting in the shade at the expense of a high risk of photoinhibitory damage in excess light (Gauslaa & Solhaug, 1996, 1999). Furthermore, reproductive efforts reduce growth in L. pulmonaria (Gauslaa, 2006). Marginal soralia formation disturbs zones of young meristematic tissue, retards apical growth and thus causes a trade-off between A gain and diaspore production. The A gain suffers more than the DW gain (Gauslaa, 2006). We believe that variations in damage and reproduction blur the effects of external factors on growth in L. pulmonaria and cause poor separation of stand types in scatter plots (Figs 3a, 5a–i).

The pendulous U. longissima responded differently. No photoinhibitory damage was detected, even in sun-exposed clear-cut transplants (Gauslaa et al., 2007). As a result of a lack of variation in chlorophyll degradation among U. longissima transplants, transplantation caused a consistent stand-specific (light regime-specific) increase in DW gain (Figs 3b and 4d–f). The substantial increase in A gain in thalli with no DW gain can probably be explained by internal reallocation of carbon. The genus Usnea is unique in forming a compact central chord that may function to store carbon. It contributes a higher percentage of main axis mass than of fibril mass, as main axes in U. longissima lack a cortex and photobiont layer (Nybakken & Gauslaa, 2007). The A gain increased with increasing DW:A ratio at the start of the experiment (Fig. 4c), which presumably is an indicator of the carbon storage pool. We hypothesize that a high DW:A ratio is associated with a high proportion of the central chord compared with the other parts of the thallus. Mobilization of this specimen-specific carbon store causes the strong linear decline in the change in DW per A, as shown in Fig. 4(i).

Unlike the DW gain (Fig. 4d–f), the A gain strongly depends on thallus size in U. longissima (Fig. 4a–c). The A gain decreased with increasing size. Usnea longissima is unique in the sense that it is one of very few epiphytic lichens that lacks a permanent holdfast (Gauslaa, 1997). This vagrant canopy lichen adheres to its supporting canopy branch by wrapping around it. Its main method of dispersal is by means of coarse thallus fragments with sizes similar to or smaller than the transplant sizes (Esseen, 1985; Gauslaa, 1997). The vagrant strategy probably depends on a fast A gain at an early stage to secure newly detached fragments within the canopy, in this species with mainly downward dispersal (Gauslaa, 1997). Transplantation in U. longissima mimics natural dispersal. The smallest thalli respond with the strongest A gain, whereas thalli that are large enough to become twisted around tree branches maintain their original DW:A ratio. In this way, the size-dependent response may facilitate successful establishment of dispersed fragments of U. longissima.

In general, fruticose lichens grow in three dimensions and harvest light from all sides. They may grow rapidly to occupy spaces between canopy branches and utilize windows of light in a growing and changing canopy. By simple growth, U. longissima is capable of migrating directly from one branch to another without the need to pass through a juvenile stage. A foliose lichen, such as L. pulmonaria, has two-dimensional growth. It grows slowly along one tree branch, and generally spreads from one branch to another by diaspores, so there is a delay in spread between branches caused by the slow establishment of the dispersed diaspores. From the perspective of foliose canopy lichens, the optimal strategy may be to develop traits to maximize light harvesting and thereby carbon gain at the decreasing irradiances caused by tree canopy growth above the lichen site. As a result, L. pulmonaria is more high-light susceptible than U. longissima.

In conclusion, lichens benefit from optimizing their morphology in relation to light availability. A common overall allocation pattern of photobiont area and total DW occurs across the growth forms of the old-forest canopy lichens studied. However, transplantation to new habitats shows that subsequent DW and A gains in lichens are controlled by different combinations of external and internal factors. DW gain responds mainly to environmental and internal variables that determine photosynthetic carbon fixation. Photosynthesis is important also for area extension, but the A gain also depends on water availability for cell expansion. Sometimes A gain may also occur by reallocation of available internal resources. The pendulous lichen (U. longissima) differs from the foliose lichen (L. pulmonaria) in having strong size-dependent growth responses.

Acknowledgements

We thank Ann Sehlstedt, Emma Lindberg, Otilia Johansson, John Gunnar Dokk and Annie Aasen for assistance in the field and/or in the laboratory. We also thank Lise Cats Myhre for practical assistance at the Atlantic transplantation site. Thanks to Ørjan Totland for help with the ANCOVA analyses. The study was funded by the Research Council of Norway (project 154442/720), and was also supported by a grant from FORMAS, Sweden; 24.0795/97 to KP.

Ancillary