- Email: [email protected]
Epiphytic lichen conservation in the Italian Alps: the role of forest type
f u n g a l e c o l o g y 1 1 ( 2 0 1 4 ) 1 6 4 e1 7 2
available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/funeco
Epiphytic lichen conservation in the Italian Alps: the role of forest type Juri NASCIMBENEa,b,*, Pier Luigi NIMISb, Matteo DAINESEc a
Museo di Scienze Naturali dell’Alto Adige, via Bottai 1, 39100 Bolzano, Italy degli Studi di Trieste, Dipartimento di Scienze della Vita, via Giorgieri 10, 34100 Trieste, Italy Universita c degli Studi di Padova, Dipartimento Territorio e Sistemi Agro-forestali, Viale dell’Universita 16, Universita b
I-35020 Legnaro, PD, Italy
article info
abstract
Article history:
Epiphytic lichens are a functionally important and species-rich component of Alpine for-
Received 26 March 2014
ests, including several species of conservation concern. Their dependence on specific host
Revision received 16 May 2014
trees predicts that forests with different tree species composition host different lichen
Accepted 30 May 2014
communities, enhancing lichen diversity in forest landscapes. In this study, we tested for
Available online
the first time the effect of forest type on patterns of epiphytic lichen diversity, in the Italian
Corresponding editor:
Alps. We sampled the main forest types of the South Tyrol, a typical Alpine region of Italy.
Darwyn Coxson
We also assessed the influence of factors related to forest structure and climatic conditions. Our results demonstrate that different forest types host statistically different
Keywords:
lichen communities, suggesting that the conservation of lichen diversity is entrusted to the
Climate
maintenance of forest landscape heterogeneity, including forest types of minor economic
Environmental gradient
value and rural habitats. The highest number of species was found in grazed larch forests
Forest structure
and in high-elevation spruce forests, while the poorest pool was found in low-elevation
Species composition
spruce forests, beech forests and Scots pine (Pinus sylvestris) forests. High-elevation
Species richness
spruce forests also had the highest number of red-listed lichens, as the non-intensive
Water-energy conjecture
management of these forest type allows the establishment of a rich lichen biota. Our results also emphasize the role for lichen conservation of some forest types that are of minor economic importance, such as oak (Quercus pubescens), riparian, and silver-fir (Abies alba) forests. This can also apply to grazed larch (Larix decidua) forests that are maintained by traditional farming, which shape one of the most pleasing aspects of the Italian Alpine landscapes. ª 2014 Elsevier Ltd and The British Mycological Society. All rights reserved.
Introduction The forest landscape of the Alps is dominated by pure spruce (Picea abies) forests that since centuries are mainly managed for timber production (Motta, 2002). Depending on elevation,
climatic, and edaphic conditions, the monotony of this landscape is mitigated by the patchy occurrence of several forest types of lesser economic importance, many of which play a relevant role for biodiversity conservation and are listed among the EU habitats of interest (European Commission,
degli Studi di Trieste, Dipartimento di Scienze della Vita, via Giorgieri 10, 34100 Trieste, Italy. * Corresponding author. Universita Tel.: þ39 (0)43942894. E-mail address: [email protected] (J. Nascimbene). http://dx.doi.org/10.1016/j.funeco.2014.06.006 1754-5048/ª 2014 Elsevier Ltd and The British Mycological Society. All rights reserved.
The role of forest type in epiphytic lichen conservation
2003). Grazed larch (Larix decidua) woodlands, which are anthropogenic formations related with traditional alpine farming, are increasingly recognized to host several species of conservation concern (Fontana et al., 2013). Epiphytic lichens are a functionally important and speciesrich component of alpine forests, including several species of conservation concern (e.g., Nascimbene et al., 2010). Their ly € riado et al., 2009; Kira dependence on specific host trees (Ju et al., 2013) predicts that forests with different tree species composition host different lichen communities (Odor et al., 2013), enhancing lichen diversity in forest landscapes. This pattern was demonstrated, for instance, in Mediterranean regions (Brunialti et al., 2013), where only a few lichen species are shared among different forest types. This highlights the importance of maintaining forest landscape heterogeneity to enhance lichen conservation. The same view could also apply to Alpine forests, as suggested by previous studies (e.g., Dietrich and Scheidegger, 1997). However, to our knowledge, no study has explicitly examined this issue for the main forest types of the Italian Alps along wide ecological and geographical gradients. Besides forest type, forest structure and climate may influence lichen diversity patterns in alpine regions. In managed forests, forest structure reflects management practices that depend on forest type and site conditions. These in turn have an influence on tree density, tree size, and stand age, that are among the main drivers of epiphytic lichen diversity (Nascimbene et al., 2013a). For example, oak (Quercus pubescens) forests are coppiced with relatively short rotation period, while mature trees in spruce forests are harvested within an age span of 80e200 yr (elevation and soil fertility determine growth rates; AA and VV, 2010). In mountain areas, energy patterns are strongly related to elevation, while water availability may also depend on gradients influenced by topography. The poikylohydric nature of lichens provides the basis for their sensitivity to both water and energy, which directly control relevant eco-physiological processes, influencing growth rates and species distributions (Insarov and Schroeter, 2002). In particular, lichen physiology is closely coupled to ambient temperature and moisture conditions (Green et al., 2008) that influence thallus water saturation and desiccation. Increasing ambient temperature may negatively affect lichens, due to increased respiratory carbon loss (Schroeter et al., 2000), especially when it is not counterbalanced by a sufficient water availability. Actually, energy availability could interact with water, i.e., in dry mountains the negative effect of high temperature is stronger than in more rainy mountains (modified conjecture of Hawkins et al., 2003; see also Bhattarai and Vetaas, 2003; McCain, 2007). This study tests the effect of forest type on patterns of epiphytic lichen diversity in a typical Alpine region of Italy, by sampling the ten main forest types occurring in the survey area. We also assessed the influence of factors related to forest structure and climatic conditions. We hypothesized that structural factors controlling light and substratum availability and stability (i.e., forest density, tree size and stand age) could be responsible for differences among forest types. We also expected an effect of climatic conditions, since our sampling encompasses a wide elevation gradient. The analyses were
165
performed in two steps: (1) we built a model with only forest type as the predictor variable, and (2) we built a multiple Poisson regression model also including factors indicative of forest structure and climate. The second model aimed to verify whether the effect of forest type could be attributed to differences in structural parameters and/or climatic conditions. Since we were interested in evaluating the role of forest types for lichen conservation, we contrasted overall species with those of conservation concern.
Materials and methods Study area and forest types The study was carried out in the Alpine region of South Tyrol (N Italy; Fig 1) that has an area of 7 400 km2. The climate is largely influenced by elevation, ranging from humid warmtemperate conditions in the Adige valley area, with mean annual temperature of 11e12 C, to alpine conditions above 1 700 m, with mean annual temperatures of 2e3 C. The amount of precipitation is variable across the region (<600e1 400 mm yr1) according to topography. The landscape between 600 and 2 100 m is dominated by forests that cover an area of 370 000 ha (AA and VV, 2010). Pure coniferous forests are the most widespread throughout the region (88 % of the forest surface). Spruce forests are the main forest type (55 %) between 900 and 1 900 m, representing the most important forest type for economic exploitation. Larchstone pine (Pinus cembra) forests are the second forest type per area (27 %), ranging between 1 900 and 2 100 m, followed by scotch pine forests (11 %) between 900 and 1 600 m, and silver fir forests (1 %) between 900 and 1 600 m. Traditionally managed, grazed larch forests are scattered in small patches between 1 300 and 1 800 m. Angiosperm forests cover only 5 % of the forest area; mainly beech (Fagus sylvatica) forests between 700 and 1 200 m and coppiced oak forests between
Fig 1 e Map of the study area in South Tyrol (Italian Alps), with the sampling plots. Plots belonging to different forest types are marked with different symbols.
166
J. Nascimbene et al.
400 and 700 m. Riparian forests, scattered in small patches between 700 and 1 200 m, are mainly included in small protected biotopes.
their frequency was computed as the number of quadrats in which the species occurred.
Species identification Sampling design The following main forest types were sampled, trying to encompass their whole geographical range in the study area (Fig 1): (1) spruce forests, (2) larch-stone pine forests, (3) silverfir forests, (4) Scots-pine forests, (5) grazed larch forests, (6) riparian forests, (7) oak forests, and (8) beech forests. Since spruce forests span a wide elevation range, they were split into three different subtypes: low elevation forests (900e1 200 m), intermediate elevation forests (1 400e1 600 m), and high elevation forests (1800e1900 m) (Table 1), altogether resulting in 10 different forest types. For each forest type, five independent (minimum distance 2 km) mature stands (ready to be harvested) were selected on the basis of the regional forest database (AA and VV, 2010). In each stand, a 13 m radius plot was randomly placed. For each plot, species identity and circumference of all living trees (diameter >15 cm) were recorded, as well as geographic position, elevation, aspect, and slope. In each plot, five mature tree individuals of the dominant species were randomly selected for the lichen survey, giving a total of 250 trees (10 forest types 5 stands 5 trees). In larch-stone pine forests only larch was considered, while in riparian forests both willow (Salix) and alder (Alnus) were considered. The lichen survey was conducted between 2012 and 2013, according to the European guidelines for lichen monitoring (Asta et al., 2002). Lichen diversity was sampled using four standard frames of 10 50 cm as sampling grids, subdivided into five 10 10 cm quadrats, which were attached to the tree trunk at the cardinal points with the shorter lower side at 100 cm from the ground. All lichen species inside the frames, including sterile crustose lichens, were listed, and
When possible, lichens were identified in the field. However, in most cases species identification was based on the study of specimens (c. 900) collected and stored both in the personal herbarium of JN and in the herbarium of the Natural Sciences Museum of South Tyrol (Bolzano). Crustose lichens, in particular, were identified in the laboratory using a dissecting and a biological microscope. Routine chemical spot tests were performed for most specimens. The identification of sterile crustose lichens (including all Lepraria species; c. 150 specimens) was based on standardized thin-layer chromatography (TLC), following the protocols of White and James (1985) and Orange et al. (2001). Nomenclature of lichens mainly follows Nimis and Martellos (2008).
Explanatory variables Forest type Forest type was treated as a categorical variable summarizing several factors potentially influencing lichen communities, such as tree species composition, management regime, elevation and climatic conditions.
Forest structure Plot-level mean tree circumference was calculated by averaging the circumference of all trees recorded within each plot. Tree density, an indicator of both substratum availability and forest density (i.e., light conditions), was calculated based on the number of trees inventoried in each plot. For each tree selected for the lichen inventory, we also measured the circumference and quantified the age by extracting cores using a
Table 1 e Results of multiple Poisson regression model testing the effect of forest type, climate and stand structure on total species richness and red-listed species richness. Non-significant interactions and main effects were removed with a backward elimination procedure (P < 0.05). Variables were tested using a Wald c2 test Variables Total species richness Forest type Oak forest Beech forest Riparian forest Spruce low elevation forest Silver fir forest Scotch-pine forest Spruce intermediate elevation forest Larch grazed forest Spruce high elevation forest Larch-Stone pine forest Temperature Precipitation Slope Mean tree circumference Temperature Precipitation Red-listed species richness Temperature
Estimate
SE
Wald c2
P-value
91.942
<0.001
R2 84.0 %
2.5010 2.0856 2.7277 1.9579 2.7951 2.4564 3.0447 4.0762 3.5833 2.9762 0.0854 0.0014 0.0125 0.0079 0.0004
0.1474 0.1830 0.1573 0.2213 0.1546 0.2107 0.1484 0.1484 0.3114 0.1742 0.0529 0.0003 0.0054 0.0022 0.0002
4.426 3.711 4.904 11.227 6.341
0.035 0.054 0.0268 <0.001 0.0118
0.4202
0.2025
4.305
0.038
17.1 %
The role of forest type in epiphytic lichen conservation
Pressler-type increment borer at a vertical height of 1.30 m. The mean tree age was calculated by averaging the age of all the cored trees within each plot. Due to sampling constraints, this information is missing for grazed larch forests.
Climate As climatic predictors, we considered mean annual temperature as a measure of available energy, and mean annual precipitation as an indicator of water availability (Table 1). Mean annual temperature (period 1980e2011) was interpolated from 84 metereological stations, evenly scattered throughout the study area between 200 and 2000 m, using ordinary kriging with external drift (Benavides et al., 2007). First, we estimated the fitted temperature from a simple regression with elevation. Then, we interpolated the residuals using ordinary kriging, and summed the interpolated residuals with the fitted temperature from the regression with elevation. Mean annual precipitation was interpolated using ordinary kriging from 88 metereological stations evenly scattered throughout the study area. The geostatistical interpolations were computed using the Kriging Interpolator 3.2 extension for ArcView 3.2 (ESRI).
Statistical analyses Forest type structure Analysis of variance (ANOVA) was applied to test differences in stand structure (age, mean tree circumference, and tree density) among forest types. Tukey contrasts were calculated from the models to test for differences between forest types, using the ‘glht’ function (general linear hypothesis test) in the ‘multcomp’ package in R (Hothorn et al., 2008).
Species composition Compositional differences among the ten forest types were tested by multi response permutation procedures (MRPP) as implemented in PCORD (McCune and Mefford, 1999), using the Sørensen distance measure and rank transformation of the distance matrices. The test statistic “A” in MRPP describes the separation among groups. Comparisons were made between pairs of forest types, as well as for the total, i.e., all the forest types pooled together. The test statistic “A” in MRPP describes the separation among groups. When A ¼ 1, all samples are identical within groups, and when A ¼ 0, the within group heterogeneity equals expectation by chance. In community ecology, an A >0.3 is considered fairly high (McCune and Grace, 2002). The pattern of species composition was also evaluated visually, using non-metric multidimensional scaling (NMDS; McCune and Grace, 2002), as implemented in PC-ORD (McCune and Mefford, 1999), using the “slow and thorough” autopilot mode with the Sørensen distance measure. This procedure performed 40 runs with real dataset compared with 50 randomized runs, each run with 400 iterations. This iterative ordination method is based on ranked distances between sample units in the data matrix, known as “species space” (McCune and Grace, 2002). It does not assume normally distributed data and is, therefore, suited for most ecological data. A final 2-dimensional solution was selected (stress 15.7 %, instability 0.00005). An Indicator Species Analysis (ISA;
167
^ne and Legendre, 1997) was used to determine how Dufre strongly each species was associated with each forest type. For each species, the Indicator Value (INDVAL) ranges from 0 (no indication) to 100 (maximum indication). This index combines the mean abundance of a species and its frequency of occurrence in a cluster. A high indicator value was obtained when a given species had both a high mean abundance in one group compared to that in the other groups (specificity) and occurred in most plots of that group (fidelity). Statistical significance of INDVAL was tested by means of a Monte Carlo test, based on 10 000 randomizations. ISA and Monte Carlo test were performed by PC-ORD (McCune and Mefford, 1999).
Species richness As response variables, plot-level species richness (i.e., all species in each plot) was considered as well as the plot-level number of the species of conservation concern, following the Italian red-list of epiphytic lichens (Nascimbene et al., 2013b). Two different models were built following Schroeder et al. (2011). First, the effect of forest types on species richness was tested using ANOVA. Secondly, a multiple Poisson regression model was built considering forest type, temperature, precipitation, slope, age, and mean tree circumference as predictors, to verify whether the effect of forest type on species richness could be attributed to differences in climatic conditions and/or tree structural parameters. The interaction between temperature and precipitation was also included in the model, to verify the water-energy hypothesis (modified conjecture of Hawkins et al., 2003). In such cases, the effect of forest type may be lower compared to the model with forest type as a single variable, and the other variables may become important to explain species richness patterns. Conversely, the effect of forest type may maintain its significance while the other variables may add no contribution to the models. In this case, this pattern could be related to an interactive effect of stand structure and climate differences among forest types that cannot be disentangled (Nascimbene et al., 2013c). Tree density was omitted from the models, as this was highly correlated with mean tree circumference (r ¼ 0.82). The predictors were tested using a Wald c2 test implemented in the ‘car’ package for R (Fox, 2002). A backward manual deletion procedure was performed using analysis of deviance (P < 0.05), to build a minimum adequate model. We started with a complex model containing all the linear and interaction terms, then the model was simplified by removing one by one the non-significant terms (Crawley, 2007). In the case of overdispersion, a “quasilikelihood” model (quasi-Poisson) was applied.
Results Forest type features Forest types significantly differed in their distribution along the elevation gradient (F ¼ 39.82, P < 0.001), oak forests being in the lowest and larch-stone pine forests in the highest part of the gradient (Fig 2). Significant differences were also found for tree age (F ¼ 10.32, P < 0.001), with the oldest trees in high-altitude coniferous forests, the youngest in riparian and oak forests, where the lowest values of mean tree
168
J. Nascimbene et al.
Fig 2 e Mean ± SE of (A) elevation distribution, (B) age, (C) mean tree circumference, and (D) tree density for each forest type. Different letters indicate significant differences according to the Tukey’s multiple comparison test.
circumference were recorded (F ¼ 35.42, P < 0.001). Tree density had a more homogeneous pattern across forest types, except for riparian forests that had significantly higher values (F ¼ 43.06, P < 0.001).
(right part), while the second roughly corresponds to the elevation gradient. Statistically significant differences in species composition among forest types were also confirmed by ISA,
Species composition Significant differences in species composition were found among forest types (MRPP for the ten forest types pooled together: A ¼ 0.7, P < 0.0001), that were also confirmed by pairwise comparisons (A ranging between 0.2 and 0.45, P < 0.01). Only oak and riparian forests hosted comparable assemblages that broadly overlap (pair-wise comparison: A ¼ 0.09, P ¼ 0.02), as indicated by the small A statistic from MRPP (Berryman and McCune, 2006). The visual interpretation of the NMDS ordination (the two axes represent 70.2 % of the total variation in species composition; 35.4 % axis 1, 34.8 % axis 2), in which plots belonging to a given forest type are plotted in the species space, largely corroborated these results. However, some compositional similarity may also occur between spruce forests of intermediate and high elevation and among Scots pine, larch, stone pine and grazed larch forests (Fig 3). These similarities are related to a group of species that are shared among these coniferous forest types. For example, 15 species are shared among the species pools of Scots pine, larch stone pine and grazed larch forests. In this case, compositional differences are mainly related to differences in species richness (see also below) than to species turnover among forest types. The first axis of the ordination diagram contrasts between coniferous (left part) and angiosperm
Fig 3 e Ordination diagram of plots in the species space based on NMDS results. The ten forest types are indicated by different symbols. The two axes represent 70.2 % of the total variation in species composition (35.4 % axis 1 and 34.8 % axis 2).
The role of forest type in epiphytic lichen conservation
169
Fig 4 e Mean ± SE of (A) total lichen species richness and (B) red-listed species richness for the ten forest types. Different letters indicate significant differences according to the Tukey’s multiple comparison test.
revealing that 36 % of the lichens are overrepresented in one of the ten forest types (see Appendix 1 in Supplementary data), with a higher number of associated species in grazed larch forests (14 species including Cyphelium tigillare, Evernia mesomorpha, and Tuckneraria laureri) and in highelevation spruce forests (13 species including Calicium viride, Cliostomum corrugatum, and Tuckermannopsis chlorophylla), followed by oak forests (Nine species including Candelaria concolor, Eopyrenula leucoplaca, and Normandina pulchella). Six species are overrepresented in riparian (e.g., Catillaria nigroclavata and Lecania naegelii), silver-fir (e.g., Phlyctis argena and Schismatomma pericleum), and larch-stone pine forests (e.g., Hypogymnia austerodes and Letharia vulpina), while three are associated with beech forests (e.g., Graphis scripta) and two with Scots pine forests (e.g., Pycnora sorophora). Only one species was overrepresented in both intermediate and lowelevation spruce forests (Hypogymnia physodes and Dimerella pineti, respectively). 40 % of the species were exclusively found in a single forest type. The higher number of exclusive species was found in riparian and oak forests (14 and 12 species, respectively), followed by silver fir and high-elevation spruce forests (both with 10 species). Larch-stone pine and grazed larch forests had seven exclusive species, while only one species was exclusively found in both intermediate and low elevation spruce forests; Scots pine forests completely lacked exclusive species (Supplementary Appendix 1).
The relative frequencies of red-listed, exclusive, and indicator species significantly differed among the ten forest types (Fig 5). In the multiple regression model (Table 1) testing the effect of forest type, stand structure and climate on total lichen species richness, the minimum adequate model explained a high percentage of the total variance and retained as significant predictors forest type, temperature, precipitation (marginal effect), slope, mean tree circumference and the interaction between temperature and precipitation. Precipitation, mean tree circumference, and the interaction term had a positive relationship with total richness, while temperature and slope had a negative effect. For red-listed lichen richness, the model explained a relatively low percentage of variance, retaining only temperature, with a negative effect.
Species richness A total of 167 species was found, including 15 red-listed lichens (see Appendix 1 in Supplementary data). ANOVA revealed significant differences in total richness (F ¼ 11.78, P < 0.001) and red-listed species richness (F ¼ 4.41, P < 0.001) among the forest types (Fig 4). The highest number of species (33.2 4.8) was in grazed larch forests and in high-elevation spruce forests (29.2 6.7), while the poorest pool was in low-elevation spruce forests (7.2 4.3), beech forests (9.6 2.3) and Scots-pine forests (11 3.8). High-elevation spruce forests also had the highest number of red-listed lichens (3 1.7), followed by intermediate elevation spruce forests (2.2 1.3) and silver fir forests (1.6 1.3).
Fig 5 e Relative frequency of red-list species (RL), exclusive species (EX), and indicators species (IN) in each forest type. The differences in frequency were tested using the G test (G [ 31.27, P [ 0.027). The actual numbers of red-listed, exclusive, and indicator species are reported at the top of the bars.
170
Discussion Our results demonstrate that, in our Alpine region, different forest types host different lichen communities, suggesting that the conservation of lichen diversity relies on the maintenance of forest landscape heterogeneity, including forest types of minor economic value and rural habitats.
Drivers of lichen diversity Differences among forest types largely reflect differences in tree species composition. Our dataset includes both angiosperm and coniferous forests. Angiosperm forests are composed of trees with both (sub-) acidic (e.g., oak, beech) and neutral (-basic) bark (e.g., willow). Coniferous forests include only (sub-) acidic-barked trees that differ in surface stability (i.e., low stability on Scots pine) and roughness (i.e., high roughness on larch). Several studies in different environmental conditions stress the role of tree species identity in driving patterns of lichen species richness and compo ly et al., 2013). This effect € riado et al., 2009; Kira sition (e.g., Ju is mainly related to species-specific differences in the chemical and physical traits of the bark, chiefly pH and € riado texture (e.g., Fritz and Heilmann-Clausen, 2010; Ju et al., 2009). Our results suggest that the effect of forest type can also be attributed to specific structural and climatic factors (Moning et al., 2009; Werth et al., 2005) that, for example, explain differences among similar forest types (i.e., spruce forests at different elevations, larch forests in natural and anthropogenic environments). This result is related to the wide environmental gradient encompassed by our study (McCune et al., 1997; Werth et al., 2005). Tree circumference is the most important structural factor, supporting the view that increasing tree size enhances lichen diversity (Nascimbene et al., 2013a). This can be due to an area effect (higher surface availability), more stable substratum conditions due to reduced growth rate, and changes in bark pH and texture (e.g., Gustafsson and Eriksson, 1995). Tree size is often positively correlated with tree age (as in our study), predicting that old large trees are more effective for lichen conservation than young and/or small-sized trees (Nascimbene et al., 2009). Moreover, larger-older trees might be more likely to be found in stands that are part of a less disturbed landscape that has a larger regional lichen species pool. The negative relationship between temperature and species richness reflects the negative effects of increasing temperature on the main eco-physiological processes, causing an increase of respiratory carbon loss (Schroeter et al., 2000) that may limit the distribution of many lichens. Moreover, an increase in temperature is usually associated with a decrease in relative air humidity, resulting in faster desiccation rates of these poikilohydric organisms that hinders photosynthetic activity (Insarov and Schroeter, 2002). This hypothesis is corroborated by the positive effect of precipitation (Marini et al., 2011) and by the interaction between temperature and precipitation, predicting a stronger effect of energy in dry mountains (modified conjecture of
J. Nascimbene et al.
Hawkins et al., 2003). Besides forest type, temperature appears to be the only determinant for red-listed species richness, indicating that threatened species may be more influenced by climate than by forest structure (Werth et al., 2005). However, the low level of variance explained by our model hinders more detailed interpretations and suggests that both local and landscape factors not included in this study (e.g., microclimatic conditions, management intensity, habitat connectivity, past land use) may influence the occurrence of these species.
The role of forest type for lichen conservation The occurrence of different lichen assemblages indicates that each forest type contributes to improve the regional lichen diversity. However, results reveal that some forest types have a higher rank for lichen conservation, due to higher richness and the occurrence of red-listed or indicator species. This is the case of high-elevation spruce forests that host a rich lichen biota including several red-listed species (e.g., C. corrugatum). Among spruce forests, high-elevation stands are usually less intensively managed, due to reduced growth rates and logistic constraints, allowing the permanence of several old trees (e.g., more than 200 yr),which are important for the conservation of red-listed species (Nascimbene et al., 2009). As indicated by our models, climatic factors may also contribute to determine optimal growth conditions for lichens in mountain forests. Our results also emphasize the role in lichen conservation of some forest types that are of minor economic importance, such as oak, riparian, and silver fir forests, supporting the effectiveness of considering these forest types as EU habitats of interest for biodiversity conservation in the Alps (European Commission, 2003). This can also apply to grazed larch forests that are maintained by traditional farming, which shape one of the most pleasing aspects of Alpine landscapes of Italy (Fontana et al., 2013). Similarly to other European wooded pastures traditionally managed with intermediate disturbance regimes (Paltto et al., 2008, 2011; Thor et al., 2010), Alpine grazed larch forests host a rich lichen biota that benefits from open conditions and the availability of large trees. Despite the fact that several species can be commonly found in other coniferous forests, their lichen biota also includes red-listed species, lichens that are significantly overrepresented in this forest type, and several exclusive species. Beech, Scots pine and low elevation spruce forests seem to play a minor role for lichen conservation. In particular, beech forests lack the typical lichen flora of the Lobarion communities and mainly host crustose-pioneer and shade tolerant species (e.g., Athonia radiata, G. scripta), probably reflecting the fact that in our study region these forests are at their climatic limit compared with the more sub-oceanic conditions of the pre-Alps (e.g., Nascimbene et al., 2013c).
Conclusions Our study indicates the importance of different forest types for epiphyte conservation in alpine environments. In this
The role of forest type in epiphytic lichen conservation
perspective, the EU policies for biodiversity conservation would benefit from similar studies that would include all the European alpine regions. The main conservation-relevant messages of this work are: (1) the importance of maintaining an heterogeneous forest landscape in which timber production couples with biodiversity conservation; (2) since the non-intensive management of spruce forests allows the establishment of a rich lichen biota, a further prolongation of the rotation cycle and/ or the retention of over mature trees is expected to further benefit the potential of these forests for lichen conservation (Nascimbene et al., 2009, 2010); (3) in the Italian Alps oak, riparian and silver fir forests should receive more priority for lichen conservation within the EU policy framework. While riparian forests are usually included in strictly protected areas and left unmanaged, both oak and silver fir forests are mainly managed for timber production, hindering the development of old-growth forest structures that would increase their effectiveness for lichen conservation; (4) traditionally managed, grazed larch forests are an anthropogenic habitat which is worthy for lichen conservation. Unfortunately, this forest type and its associated biodiversity are increasingly threatened by two contrasting processes: intensification of land-use in more favorable sites, and complete abandonment of less favorable sites (Fontana et al., 2013). An excessive canopy closure related to the development of secondary woodlands may reduce lichen diversity (Paltto et al., 2011). On the other hand, the increase of nitrogen inputs related to management intensification is expected to cause the rarefaction of many forest species and the establishment of a few nitrogen-tolerant species (Giordani et al., 2014; Johansson et al., 2012).
Acknowledgments The study was conducted in the framework of the project , biomonitoraggio e conservazione dei licheni “Biodiversita epifiti negli ambienti forestali della provincia di Bolzano”, funded by the Autonomous Province of Bolzano (Ripartizione e Ricerca scientifica). Diritto allo studio, Universita The Forest planning office of the Autonomous Province of Bolzano (project partner) is thanked for providing logistic and technical support. In particular, we are grateful to € nther Unterthiner and his collaborators. Francesco BorGu tignon, Philipp Oberegger, Martin Stecher, and Diego Ivan are thanked for their help during the fieldwork. Daniel Spitale contributed precious effort and discussion for setting up the sampling plane. Helmut Mayrhofer (University of Graz) and his collaborators helped us with lichen species identification, particularly with TLC analyses and critical crustose species.
Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.funeco.2014.06.006.
171
references
AA, VV, 2010. Tipologie forestali dell’Alto Adige. Ripartizione per le Foreste vol. 2. Provincia Autonoma di Bolzano-Alto Adige. Asta, J., Erhardt, W., Ferretti, M., Fornasier, F., Kirschbaum, U., Nimis, P.L., Purvis, W., Pirintsos, S., Scheidegger, C., Van Haluwyn, C., Wirth, V., 2002. Mapping lichen diversity as an indicator of environmental quality. In: Nimis, P.L., Scheidegger, C., Wolseley, P.A. (Eds.), Monitoring With Lichens e Monitoring Lichens, vol. 7. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 273e279. Benavides, R., Montes, F., Rubio, A., Osoro, K., 2007. Geostatistical modelling of air temperature in a mountainous region of northern Spain. Agricultural and Forest Meteorology 146, 173e188. Berryman, S., McCune, B., 2006. Epiphytic lichens along gradients in topography and stand structure in western Oregon, USA. Pacific Northwest Fungi 1, 1e38. Bhattarai, K.R., Vetaas, O.R., 2003. Variation in plant species richness of different life forms along a subtropical elevation gradient in the Himalayas, east Nepal. Global Ecology and Biogeography 12, 327e340. Brunialti, G., Ravera, S., Frati, L., 2013. Mediterranean old-growth forests: the role of forest type in the conservation of epiphytic lichens. Nova Hedwigia 96, 367e381. Crawley, M.J., 2007. The R Book. John Wiley & Sons, Ltd, Chichester, UK. Dietrich, M., Scheidegger, C., 1997. Frequency, diversity and ecological strategies of epiphytic lichens in the Swiss Central Plateau and the Pre-Alps. The Lichenologist 29, 237e258. ^ne, M., Legendre, P., 1997. Species assemblages and Dufre indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67, 345e366. European Commission, 2003. Natura 2000 and Forests: Challenges and Opportunities e Interpretation Guide. Office for Official Publications of the European Communities, p. 101. Fontana, V., Radtke, A., Bossi Fedrigotti, V., Tappeiner, U., Tasser, E., Zerbe, S., Buchholz, T., 2013. Comparing land-use alternatives: using the ecosystem services concept to define a multi-criteria decision analysis. Ecological Economics 93, 128e136. Fox, J., 2002. An R and S-plus Companion to Applied Regression. Sage Publications, Thousand Oaks, California. € Heilmann-Clausen, J., 2010. Rot holes create key Fritz, O., microhabitats for epiphytic lichens and bryophytes on beech (Fagus sylvatica). Biological Conservation 143, 1008e1016. Giordani, P., Calatayud, V., Stofer, S., Seidling, W., Granke, O., Fischer, R., 2014. Detecting the nitrogen critical loads on European forests by means of epiphytic lichens. A signal-tonoise evaluation. Forest Ecology and Management 311, 29e40. Green, T.G.A., Nash III, T.H., Lange, O.L., 2008. Physiological ecology of carbon dioxide exchange. In: Nash III, T.H. (Ed.), Lichen Biology. Cambridge University Press, Cambridge, pp. 152e181. Gustafsson, L., Eriksson, I., 1995. Factors of importance for the epiphytic vegetation of aspen Populus tremula with special emphasis on bark chemistry and soil chemistry. Journal of Applied Ecology 32, 412e424. gan, J.-F., Hawkins, B.A., Field, R., Cornell, H.V., Currie, D.J., Gue Kaufman, D.M., Kerr, J.T., Mittelbach, G.G., Oberdorff, T., O’Brien, E.M., Porter, E.E., Turner, J.R.G., 2003. Energy, water, and broad-scale geographic patterns of species richness. Ecology 84, 3105e3117. Hothorn, T., Bretz, F., Westfall, P., 2008. Simultaneous inference in general parametric models. Biometrical Journal 50, 346e363. Insarov, G., Schroeter, B., 2002. Lichen monitoring and climate change. In: Nimis, P.L., Scheidegger, C., Wolseley, P.A. (Eds.), Monitoring With Lichens e Monitoring Lichens, vol. 7. Kluwer
172
Academic Publishers, Dordrecht, The Netherlands, pp. 183e201. Johansson, O., Palmqvist, K., Olofsson, J., 2012. Nitrogen deposition drives lichen community changes through differential species responses. Global Change Biology 18, 2626e2635. € riado, I., Liira, J., Paal, J., Suija, A., 2009. Tree and stand level Ju variables influencing diversity of lichens on temperate broadleaved trees in boreo-nemoral floodplain forests. Biodiversity and Conservation 18, 105e125. ly, I., Nascimbene, J., Tinya, F., Odor, Kira P., 2013. Factors influencing epiphytic bryophyte and lichen species richness at different spatial scales in managed temperate forests. Biodiversity and Conservation 22, 209e223. Marini, L., Nascimbene, J., Nimis, P.L., 2011. Large-scale patterns of epiphytic lichen species richness: photobiont-dependent response to climate and forest structure. The Science of the Total Environment 409, 4381e4386. McCain, C.M., 2007. Could temperature and water availability drive elevational species richness patterns? A global case study for bats. Global Ecology and Biogeography 16, 1e13. McCune, B., Mefford, M.J., 1999. Multivariate Analysis of Ecological Data. Version 4.25. MJM Software, Gleneden Beach, Oregon. McCune, B., Grace, J.B., 2002. Analysis of Ecological Communities. MJM Software, Gleneden Beach, Oregon. McCune, B., Peck, D.J., Heiman, K., Will-Wolf, S., 1997. Regional gradients in lichen communities of the southeast United States. Bryologist 100, 145e158. € ssler, C., Bradtka, J., Moning, C., Werth, S., Dziock, F., Ba € ller, J., 2009. Lichen diversity in temperate Hothorn, T., Mu montane forests is influenced by forest structure more than climate. Forest Ecology and Management 258, 745e751. Motta, R., 2002. Old-growth forests and silviculture in the Italian Alps: the caseestudy of the strict reserve of Paneveggio (TN). Plant Biosystems 136, 223e232. Nascimbene, J., Marini, L., Motta, R., Nimis, P.L., 2009. Influence of tree age, tree size and crown structure on lichen communities in mature Alpine spruce forests. Biodiversity and Conservation 18, 1519e1522. Nascimbene, J., Marini, L., Nimis, P.L., 2010. Epiphytic lichen diversity in old-growth and managed Picea abies stands in Alpine spruce forests. Forest Ecology and Management 260, 603e609. Nascimbene, J., Thor, G., Nimis, P.L., 2013a. Effects of forest management on epiphytic lichens in temperate deciduous
J. Nascimbene et al.
forests of Europe e a review. Forest Ecology and Management 298, 27e38. Nascimbene, J., Nimis, P.L., Ravera, S., 2013b. Evaluating the conservation status of epiphytic lichens of Italy: a red list. Plant Biosystems 147, 898e904. Nascimbene, J., Dainese, M., Sitzia, T., 2013c. Contrasting responses of epiphytic and dead wood-dwelling lichen diversity to forest management abandonment in silver fir mature woodlands. Forest Ecology and Management 289, 325e332. Nimis, P.L., Martellos, S., 2008. ITALIC e The Information System on Italian Lichens. Version 4.0. University of Trieste, Dept. of Biology. IN4.0/1 http://dbiodbs.univ.trieste.it/. ly, I., Tinya, F., Bortignon, F., Nascimbene, J., 2013. Odor, P., Kira Patterns and drivers of species composition of epiphytic bryophytes and lichens in managed temperate forests. Forest Ecology and Management 298, 27e38. Orange, A., James, P.W., White, F.J., 2001. Microchemical Methods for the Identification of Lichens. British Lichen Society, London. n, B., Go € tmark, F., 2008. Partial cutting as a Paltto, H., Norde conservation alternative for oak (Quercus spp.) foresteResponse of bryophytes and lichens on dead wood. Forest Ecology and Managment 256, 536e547. n, B., Sna € ll, T., 2011. Development Paltto, H., Nordberg, A., Norde of secondary woodland in oak wood pastures reduces the richness of rare epiphytic lichens. PLoS One 6, e24675. Schroeder, L.M., Sahlin, E., Paltto, H., 2011. Retention of aspen (Populus tremula)at final cuttings e the effect of dead wood characteristics on saproxylic beetles. Forest Ecology and Management 262, 853e862. Schroeter, B., Kappen, L., Schulz, F., Sancho, L.G., 2000. Seasonal variation in the carbon balance of lichens in the maritime Antarctic: long-term measurements of photosynthetic activity in Usnea aurantiaco-atra. In: Davison, W., Howard-Williams, C., Broady, P. (Eds.), Antarctic Ecosystems: Models for Wider Ecological Understanding. Caxton Press, Christchurch, pp. 258e262. € nsson, M.T., 2010. Lichen diversity and Thor, G., Johansson, P., Jo red-listed lichen species relationships with tree species and diameter in wooded meadows. Biodiversity and Conservation 19, 2307e2328. Werth, S., Tømmervik, H., Elvebakk, A., 2005. Epiphytic macrolichen communities along regional gradients in northern Norway. Journal of Vegetation Science 16, 199e208. White, F.J., James, P.W., 1985. A new guide to microchemical techniques for the identification of lichen substances. Bulletin of the British Lichen Society 57, 1e41.
available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/funeco
Epiphytic lichen conservation in the Italian Alps: the role of forest type Juri NASCIMBENEa,b,*, Pier Luigi NIMISb, Matteo DAINESEc a
Museo di Scienze Naturali dell’Alto Adige, via Bottai 1, 39100 Bolzano, Italy degli Studi di Trieste, Dipartimento di Scienze della Vita, via Giorgieri 10, 34100 Trieste, Italy Universita c degli Studi di Padova, Dipartimento Territorio e Sistemi Agro-forestali, Viale dell’Universita 16, Universita b
I-35020 Legnaro, PD, Italy
article info
abstract
Article history:
Epiphytic lichens are a functionally important and species-rich component of Alpine for-
Received 26 March 2014
ests, including several species of conservation concern. Their dependence on specific host
Revision received 16 May 2014
trees predicts that forests with different tree species composition host different lichen
Accepted 30 May 2014
communities, enhancing lichen diversity in forest landscapes. In this study, we tested for
Available online
the first time the effect of forest type on patterns of epiphytic lichen diversity, in the Italian
Corresponding editor:
Alps. We sampled the main forest types of the South Tyrol, a typical Alpine region of Italy.
Darwyn Coxson
We also assessed the influence of factors related to forest structure and climatic conditions. Our results demonstrate that different forest types host statistically different
Keywords:
lichen communities, suggesting that the conservation of lichen diversity is entrusted to the
Climate
maintenance of forest landscape heterogeneity, including forest types of minor economic
Environmental gradient
value and rural habitats. The highest number of species was found in grazed larch forests
Forest structure
and in high-elevation spruce forests, while the poorest pool was found in low-elevation
Species composition
spruce forests, beech forests and Scots pine (Pinus sylvestris) forests. High-elevation
Species richness
spruce forests also had the highest number of red-listed lichens, as the non-intensive
Water-energy conjecture
management of these forest type allows the establishment of a rich lichen biota. Our results also emphasize the role for lichen conservation of some forest types that are of minor economic importance, such as oak (Quercus pubescens), riparian, and silver-fir (Abies alba) forests. This can also apply to grazed larch (Larix decidua) forests that are maintained by traditional farming, which shape one of the most pleasing aspects of the Italian Alpine landscapes. ª 2014 Elsevier Ltd and The British Mycological Society. All rights reserved.
Introduction The forest landscape of the Alps is dominated by pure spruce (Picea abies) forests that since centuries are mainly managed for timber production (Motta, 2002). Depending on elevation,
climatic, and edaphic conditions, the monotony of this landscape is mitigated by the patchy occurrence of several forest types of lesser economic importance, many of which play a relevant role for biodiversity conservation and are listed among the EU habitats of interest (European Commission,
degli Studi di Trieste, Dipartimento di Scienze della Vita, via Giorgieri 10, 34100 Trieste, Italy. * Corresponding author. Universita Tel.: þ39 (0)43942894. E-mail address: [email protected] (J. Nascimbene). http://dx.doi.org/10.1016/j.funeco.2014.06.006 1754-5048/ª 2014 Elsevier Ltd and The British Mycological Society. All rights reserved.
The role of forest type in epiphytic lichen conservation
2003). Grazed larch (Larix decidua) woodlands, which are anthropogenic formations related with traditional alpine farming, are increasingly recognized to host several species of conservation concern (Fontana et al., 2013). Epiphytic lichens are a functionally important and speciesrich component of alpine forests, including several species of conservation concern (e.g., Nascimbene et al., 2010). Their ly € riado et al., 2009; Kira dependence on specific host trees (Ju et al., 2013) predicts that forests with different tree species composition host different lichen communities (Odor et al., 2013), enhancing lichen diversity in forest landscapes. This pattern was demonstrated, for instance, in Mediterranean regions (Brunialti et al., 2013), where only a few lichen species are shared among different forest types. This highlights the importance of maintaining forest landscape heterogeneity to enhance lichen conservation. The same view could also apply to Alpine forests, as suggested by previous studies (e.g., Dietrich and Scheidegger, 1997). However, to our knowledge, no study has explicitly examined this issue for the main forest types of the Italian Alps along wide ecological and geographical gradients. Besides forest type, forest structure and climate may influence lichen diversity patterns in alpine regions. In managed forests, forest structure reflects management practices that depend on forest type and site conditions. These in turn have an influence on tree density, tree size, and stand age, that are among the main drivers of epiphytic lichen diversity (Nascimbene et al., 2013a). For example, oak (Quercus pubescens) forests are coppiced with relatively short rotation period, while mature trees in spruce forests are harvested within an age span of 80e200 yr (elevation and soil fertility determine growth rates; AA and VV, 2010). In mountain areas, energy patterns are strongly related to elevation, while water availability may also depend on gradients influenced by topography. The poikylohydric nature of lichens provides the basis for their sensitivity to both water and energy, which directly control relevant eco-physiological processes, influencing growth rates and species distributions (Insarov and Schroeter, 2002). In particular, lichen physiology is closely coupled to ambient temperature and moisture conditions (Green et al., 2008) that influence thallus water saturation and desiccation. Increasing ambient temperature may negatively affect lichens, due to increased respiratory carbon loss (Schroeter et al., 2000), especially when it is not counterbalanced by a sufficient water availability. Actually, energy availability could interact with water, i.e., in dry mountains the negative effect of high temperature is stronger than in more rainy mountains (modified conjecture of Hawkins et al., 2003; see also Bhattarai and Vetaas, 2003; McCain, 2007). This study tests the effect of forest type on patterns of epiphytic lichen diversity in a typical Alpine region of Italy, by sampling the ten main forest types occurring in the survey area. We also assessed the influence of factors related to forest structure and climatic conditions. We hypothesized that structural factors controlling light and substratum availability and stability (i.e., forest density, tree size and stand age) could be responsible for differences among forest types. We also expected an effect of climatic conditions, since our sampling encompasses a wide elevation gradient. The analyses were
165
performed in two steps: (1) we built a model with only forest type as the predictor variable, and (2) we built a multiple Poisson regression model also including factors indicative of forest structure and climate. The second model aimed to verify whether the effect of forest type could be attributed to differences in structural parameters and/or climatic conditions. Since we were interested in evaluating the role of forest types for lichen conservation, we contrasted overall species with those of conservation concern.
Materials and methods Study area and forest types The study was carried out in the Alpine region of South Tyrol (N Italy; Fig 1) that has an area of 7 400 km2. The climate is largely influenced by elevation, ranging from humid warmtemperate conditions in the Adige valley area, with mean annual temperature of 11e12 C, to alpine conditions above 1 700 m, with mean annual temperatures of 2e3 C. The amount of precipitation is variable across the region (<600e1 400 mm yr1) according to topography. The landscape between 600 and 2 100 m is dominated by forests that cover an area of 370 000 ha (AA and VV, 2010). Pure coniferous forests are the most widespread throughout the region (88 % of the forest surface). Spruce forests are the main forest type (55 %) between 900 and 1 900 m, representing the most important forest type for economic exploitation. Larchstone pine (Pinus cembra) forests are the second forest type per area (27 %), ranging between 1 900 and 2 100 m, followed by scotch pine forests (11 %) between 900 and 1 600 m, and silver fir forests (1 %) between 900 and 1 600 m. Traditionally managed, grazed larch forests are scattered in small patches between 1 300 and 1 800 m. Angiosperm forests cover only 5 % of the forest area; mainly beech (Fagus sylvatica) forests between 700 and 1 200 m and coppiced oak forests between
Fig 1 e Map of the study area in South Tyrol (Italian Alps), with the sampling plots. Plots belonging to different forest types are marked with different symbols.
166
J. Nascimbene et al.
400 and 700 m. Riparian forests, scattered in small patches between 700 and 1 200 m, are mainly included in small protected biotopes.
their frequency was computed as the number of quadrats in which the species occurred.
Species identification Sampling design The following main forest types were sampled, trying to encompass their whole geographical range in the study area (Fig 1): (1) spruce forests, (2) larch-stone pine forests, (3) silverfir forests, (4) Scots-pine forests, (5) grazed larch forests, (6) riparian forests, (7) oak forests, and (8) beech forests. Since spruce forests span a wide elevation range, they were split into three different subtypes: low elevation forests (900e1 200 m), intermediate elevation forests (1 400e1 600 m), and high elevation forests (1800e1900 m) (Table 1), altogether resulting in 10 different forest types. For each forest type, five independent (minimum distance 2 km) mature stands (ready to be harvested) were selected on the basis of the regional forest database (AA and VV, 2010). In each stand, a 13 m radius plot was randomly placed. For each plot, species identity and circumference of all living trees (diameter >15 cm) were recorded, as well as geographic position, elevation, aspect, and slope. In each plot, five mature tree individuals of the dominant species were randomly selected for the lichen survey, giving a total of 250 trees (10 forest types 5 stands 5 trees). In larch-stone pine forests only larch was considered, while in riparian forests both willow (Salix) and alder (Alnus) were considered. The lichen survey was conducted between 2012 and 2013, according to the European guidelines for lichen monitoring (Asta et al., 2002). Lichen diversity was sampled using four standard frames of 10 50 cm as sampling grids, subdivided into five 10 10 cm quadrats, which were attached to the tree trunk at the cardinal points with the shorter lower side at 100 cm from the ground. All lichen species inside the frames, including sterile crustose lichens, were listed, and
When possible, lichens were identified in the field. However, in most cases species identification was based on the study of specimens (c. 900) collected and stored both in the personal herbarium of JN and in the herbarium of the Natural Sciences Museum of South Tyrol (Bolzano). Crustose lichens, in particular, were identified in the laboratory using a dissecting and a biological microscope. Routine chemical spot tests were performed for most specimens. The identification of sterile crustose lichens (including all Lepraria species; c. 150 specimens) was based on standardized thin-layer chromatography (TLC), following the protocols of White and James (1985) and Orange et al. (2001). Nomenclature of lichens mainly follows Nimis and Martellos (2008).
Explanatory variables Forest type Forest type was treated as a categorical variable summarizing several factors potentially influencing lichen communities, such as tree species composition, management regime, elevation and climatic conditions.
Forest structure Plot-level mean tree circumference was calculated by averaging the circumference of all trees recorded within each plot. Tree density, an indicator of both substratum availability and forest density (i.e., light conditions), was calculated based on the number of trees inventoried in each plot. For each tree selected for the lichen inventory, we also measured the circumference and quantified the age by extracting cores using a
Table 1 e Results of multiple Poisson regression model testing the effect of forest type, climate and stand structure on total species richness and red-listed species richness. Non-significant interactions and main effects were removed with a backward elimination procedure (P < 0.05). Variables were tested using a Wald c2 test Variables Total species richness Forest type Oak forest Beech forest Riparian forest Spruce low elevation forest Silver fir forest Scotch-pine forest Spruce intermediate elevation forest Larch grazed forest Spruce high elevation forest Larch-Stone pine forest Temperature Precipitation Slope Mean tree circumference Temperature Precipitation Red-listed species richness Temperature
Estimate
SE
Wald c2
P-value
91.942
<0.001
R2 84.0 %
2.5010 2.0856 2.7277 1.9579 2.7951 2.4564 3.0447 4.0762 3.5833 2.9762 0.0854 0.0014 0.0125 0.0079 0.0004
0.1474 0.1830 0.1573 0.2213 0.1546 0.2107 0.1484 0.1484 0.3114 0.1742 0.0529 0.0003 0.0054 0.0022 0.0002
4.426 3.711 4.904 11.227 6.341
0.035 0.054 0.0268 <0.001 0.0118
0.4202
0.2025
4.305
0.038
17.1 %
The role of forest type in epiphytic lichen conservation
Pressler-type increment borer at a vertical height of 1.30 m. The mean tree age was calculated by averaging the age of all the cored trees within each plot. Due to sampling constraints, this information is missing for grazed larch forests.
Climate As climatic predictors, we considered mean annual temperature as a measure of available energy, and mean annual precipitation as an indicator of water availability (Table 1). Mean annual temperature (period 1980e2011) was interpolated from 84 metereological stations, evenly scattered throughout the study area between 200 and 2000 m, using ordinary kriging with external drift (Benavides et al., 2007). First, we estimated the fitted temperature from a simple regression with elevation. Then, we interpolated the residuals using ordinary kriging, and summed the interpolated residuals with the fitted temperature from the regression with elevation. Mean annual precipitation was interpolated using ordinary kriging from 88 metereological stations evenly scattered throughout the study area. The geostatistical interpolations were computed using the Kriging Interpolator 3.2 extension for ArcView 3.2 (ESRI).
Statistical analyses Forest type structure Analysis of variance (ANOVA) was applied to test differences in stand structure (age, mean tree circumference, and tree density) among forest types. Tukey contrasts were calculated from the models to test for differences between forest types, using the ‘glht’ function (general linear hypothesis test) in the ‘multcomp’ package in R (Hothorn et al., 2008).
Species composition Compositional differences among the ten forest types were tested by multi response permutation procedures (MRPP) as implemented in PCORD (McCune and Mefford, 1999), using the Sørensen distance measure and rank transformation of the distance matrices. The test statistic “A” in MRPP describes the separation among groups. Comparisons were made between pairs of forest types, as well as for the total, i.e., all the forest types pooled together. The test statistic “A” in MRPP describes the separation among groups. When A ¼ 1, all samples are identical within groups, and when A ¼ 0, the within group heterogeneity equals expectation by chance. In community ecology, an A >0.3 is considered fairly high (McCune and Grace, 2002). The pattern of species composition was also evaluated visually, using non-metric multidimensional scaling (NMDS; McCune and Grace, 2002), as implemented in PC-ORD (McCune and Mefford, 1999), using the “slow and thorough” autopilot mode with the Sørensen distance measure. This procedure performed 40 runs with real dataset compared with 50 randomized runs, each run with 400 iterations. This iterative ordination method is based on ranked distances between sample units in the data matrix, known as “species space” (McCune and Grace, 2002). It does not assume normally distributed data and is, therefore, suited for most ecological data. A final 2-dimensional solution was selected (stress 15.7 %, instability 0.00005). An Indicator Species Analysis (ISA;
167
^ne and Legendre, 1997) was used to determine how Dufre strongly each species was associated with each forest type. For each species, the Indicator Value (INDVAL) ranges from 0 (no indication) to 100 (maximum indication). This index combines the mean abundance of a species and its frequency of occurrence in a cluster. A high indicator value was obtained when a given species had both a high mean abundance in one group compared to that in the other groups (specificity) and occurred in most plots of that group (fidelity). Statistical significance of INDVAL was tested by means of a Monte Carlo test, based on 10 000 randomizations. ISA and Monte Carlo test were performed by PC-ORD (McCune and Mefford, 1999).
Species richness As response variables, plot-level species richness (i.e., all species in each plot) was considered as well as the plot-level number of the species of conservation concern, following the Italian red-list of epiphytic lichens (Nascimbene et al., 2013b). Two different models were built following Schroeder et al. (2011). First, the effect of forest types on species richness was tested using ANOVA. Secondly, a multiple Poisson regression model was built considering forest type, temperature, precipitation, slope, age, and mean tree circumference as predictors, to verify whether the effect of forest type on species richness could be attributed to differences in climatic conditions and/or tree structural parameters. The interaction between temperature and precipitation was also included in the model, to verify the water-energy hypothesis (modified conjecture of Hawkins et al., 2003). In such cases, the effect of forest type may be lower compared to the model with forest type as a single variable, and the other variables may become important to explain species richness patterns. Conversely, the effect of forest type may maintain its significance while the other variables may add no contribution to the models. In this case, this pattern could be related to an interactive effect of stand structure and climate differences among forest types that cannot be disentangled (Nascimbene et al., 2013c). Tree density was omitted from the models, as this was highly correlated with mean tree circumference (r ¼ 0.82). The predictors were tested using a Wald c2 test implemented in the ‘car’ package for R (Fox, 2002). A backward manual deletion procedure was performed using analysis of deviance (P < 0.05), to build a minimum adequate model. We started with a complex model containing all the linear and interaction terms, then the model was simplified by removing one by one the non-significant terms (Crawley, 2007). In the case of overdispersion, a “quasilikelihood” model (quasi-Poisson) was applied.
Results Forest type features Forest types significantly differed in their distribution along the elevation gradient (F ¼ 39.82, P < 0.001), oak forests being in the lowest and larch-stone pine forests in the highest part of the gradient (Fig 2). Significant differences were also found for tree age (F ¼ 10.32, P < 0.001), with the oldest trees in high-altitude coniferous forests, the youngest in riparian and oak forests, where the lowest values of mean tree
168
J. Nascimbene et al.
Fig 2 e Mean ± SE of (A) elevation distribution, (B) age, (C) mean tree circumference, and (D) tree density for each forest type. Different letters indicate significant differences according to the Tukey’s multiple comparison test.
circumference were recorded (F ¼ 35.42, P < 0.001). Tree density had a more homogeneous pattern across forest types, except for riparian forests that had significantly higher values (F ¼ 43.06, P < 0.001).
(right part), while the second roughly corresponds to the elevation gradient. Statistically significant differences in species composition among forest types were also confirmed by ISA,
Species composition Significant differences in species composition were found among forest types (MRPP for the ten forest types pooled together: A ¼ 0.7, P < 0.0001), that were also confirmed by pairwise comparisons (A ranging between 0.2 and 0.45, P < 0.01). Only oak and riparian forests hosted comparable assemblages that broadly overlap (pair-wise comparison: A ¼ 0.09, P ¼ 0.02), as indicated by the small A statistic from MRPP (Berryman and McCune, 2006). The visual interpretation of the NMDS ordination (the two axes represent 70.2 % of the total variation in species composition; 35.4 % axis 1, 34.8 % axis 2), in which plots belonging to a given forest type are plotted in the species space, largely corroborated these results. However, some compositional similarity may also occur between spruce forests of intermediate and high elevation and among Scots pine, larch, stone pine and grazed larch forests (Fig 3). These similarities are related to a group of species that are shared among these coniferous forest types. For example, 15 species are shared among the species pools of Scots pine, larch stone pine and grazed larch forests. In this case, compositional differences are mainly related to differences in species richness (see also below) than to species turnover among forest types. The first axis of the ordination diagram contrasts between coniferous (left part) and angiosperm
Fig 3 e Ordination diagram of plots in the species space based on NMDS results. The ten forest types are indicated by different symbols. The two axes represent 70.2 % of the total variation in species composition (35.4 % axis 1 and 34.8 % axis 2).
The role of forest type in epiphytic lichen conservation
169
Fig 4 e Mean ± SE of (A) total lichen species richness and (B) red-listed species richness for the ten forest types. Different letters indicate significant differences according to the Tukey’s multiple comparison test.
revealing that 36 % of the lichens are overrepresented in one of the ten forest types (see Appendix 1 in Supplementary data), with a higher number of associated species in grazed larch forests (14 species including Cyphelium tigillare, Evernia mesomorpha, and Tuckneraria laureri) and in highelevation spruce forests (13 species including Calicium viride, Cliostomum corrugatum, and Tuckermannopsis chlorophylla), followed by oak forests (Nine species including Candelaria concolor, Eopyrenula leucoplaca, and Normandina pulchella). Six species are overrepresented in riparian (e.g., Catillaria nigroclavata and Lecania naegelii), silver-fir (e.g., Phlyctis argena and Schismatomma pericleum), and larch-stone pine forests (e.g., Hypogymnia austerodes and Letharia vulpina), while three are associated with beech forests (e.g., Graphis scripta) and two with Scots pine forests (e.g., Pycnora sorophora). Only one species was overrepresented in both intermediate and lowelevation spruce forests (Hypogymnia physodes and Dimerella pineti, respectively). 40 % of the species were exclusively found in a single forest type. The higher number of exclusive species was found in riparian and oak forests (14 and 12 species, respectively), followed by silver fir and high-elevation spruce forests (both with 10 species). Larch-stone pine and grazed larch forests had seven exclusive species, while only one species was exclusively found in both intermediate and low elevation spruce forests; Scots pine forests completely lacked exclusive species (Supplementary Appendix 1).
The relative frequencies of red-listed, exclusive, and indicator species significantly differed among the ten forest types (Fig 5). In the multiple regression model (Table 1) testing the effect of forest type, stand structure and climate on total lichen species richness, the minimum adequate model explained a high percentage of the total variance and retained as significant predictors forest type, temperature, precipitation (marginal effect), slope, mean tree circumference and the interaction between temperature and precipitation. Precipitation, mean tree circumference, and the interaction term had a positive relationship with total richness, while temperature and slope had a negative effect. For red-listed lichen richness, the model explained a relatively low percentage of variance, retaining only temperature, with a negative effect.
Species richness A total of 167 species was found, including 15 red-listed lichens (see Appendix 1 in Supplementary data). ANOVA revealed significant differences in total richness (F ¼ 11.78, P < 0.001) and red-listed species richness (F ¼ 4.41, P < 0.001) among the forest types (Fig 4). The highest number of species (33.2 4.8) was in grazed larch forests and in high-elevation spruce forests (29.2 6.7), while the poorest pool was in low-elevation spruce forests (7.2 4.3), beech forests (9.6 2.3) and Scots-pine forests (11 3.8). High-elevation spruce forests also had the highest number of red-listed lichens (3 1.7), followed by intermediate elevation spruce forests (2.2 1.3) and silver fir forests (1.6 1.3).
Fig 5 e Relative frequency of red-list species (RL), exclusive species (EX), and indicators species (IN) in each forest type. The differences in frequency were tested using the G test (G [ 31.27, P [ 0.027). The actual numbers of red-listed, exclusive, and indicator species are reported at the top of the bars.
170
Discussion Our results demonstrate that, in our Alpine region, different forest types host different lichen communities, suggesting that the conservation of lichen diversity relies on the maintenance of forest landscape heterogeneity, including forest types of minor economic value and rural habitats.
Drivers of lichen diversity Differences among forest types largely reflect differences in tree species composition. Our dataset includes both angiosperm and coniferous forests. Angiosperm forests are composed of trees with both (sub-) acidic (e.g., oak, beech) and neutral (-basic) bark (e.g., willow). Coniferous forests include only (sub-) acidic-barked trees that differ in surface stability (i.e., low stability on Scots pine) and roughness (i.e., high roughness on larch). Several studies in different environmental conditions stress the role of tree species identity in driving patterns of lichen species richness and compo ly et al., 2013). This effect € riado et al., 2009; Kira sition (e.g., Ju is mainly related to species-specific differences in the chemical and physical traits of the bark, chiefly pH and € riado texture (e.g., Fritz and Heilmann-Clausen, 2010; Ju et al., 2009). Our results suggest that the effect of forest type can also be attributed to specific structural and climatic factors (Moning et al., 2009; Werth et al., 2005) that, for example, explain differences among similar forest types (i.e., spruce forests at different elevations, larch forests in natural and anthropogenic environments). This result is related to the wide environmental gradient encompassed by our study (McCune et al., 1997; Werth et al., 2005). Tree circumference is the most important structural factor, supporting the view that increasing tree size enhances lichen diversity (Nascimbene et al., 2013a). This can be due to an area effect (higher surface availability), more stable substratum conditions due to reduced growth rate, and changes in bark pH and texture (e.g., Gustafsson and Eriksson, 1995). Tree size is often positively correlated with tree age (as in our study), predicting that old large trees are more effective for lichen conservation than young and/or small-sized trees (Nascimbene et al., 2009). Moreover, larger-older trees might be more likely to be found in stands that are part of a less disturbed landscape that has a larger regional lichen species pool. The negative relationship between temperature and species richness reflects the negative effects of increasing temperature on the main eco-physiological processes, causing an increase of respiratory carbon loss (Schroeter et al., 2000) that may limit the distribution of many lichens. Moreover, an increase in temperature is usually associated with a decrease in relative air humidity, resulting in faster desiccation rates of these poikilohydric organisms that hinders photosynthetic activity (Insarov and Schroeter, 2002). This hypothesis is corroborated by the positive effect of precipitation (Marini et al., 2011) and by the interaction between temperature and precipitation, predicting a stronger effect of energy in dry mountains (modified conjecture of
J. Nascimbene et al.
Hawkins et al., 2003). Besides forest type, temperature appears to be the only determinant for red-listed species richness, indicating that threatened species may be more influenced by climate than by forest structure (Werth et al., 2005). However, the low level of variance explained by our model hinders more detailed interpretations and suggests that both local and landscape factors not included in this study (e.g., microclimatic conditions, management intensity, habitat connectivity, past land use) may influence the occurrence of these species.
The role of forest type for lichen conservation The occurrence of different lichen assemblages indicates that each forest type contributes to improve the regional lichen diversity. However, results reveal that some forest types have a higher rank for lichen conservation, due to higher richness and the occurrence of red-listed or indicator species. This is the case of high-elevation spruce forests that host a rich lichen biota including several red-listed species (e.g., C. corrugatum). Among spruce forests, high-elevation stands are usually less intensively managed, due to reduced growth rates and logistic constraints, allowing the permanence of several old trees (e.g., more than 200 yr),which are important for the conservation of red-listed species (Nascimbene et al., 2009). As indicated by our models, climatic factors may also contribute to determine optimal growth conditions for lichens in mountain forests. Our results also emphasize the role in lichen conservation of some forest types that are of minor economic importance, such as oak, riparian, and silver fir forests, supporting the effectiveness of considering these forest types as EU habitats of interest for biodiversity conservation in the Alps (European Commission, 2003). This can also apply to grazed larch forests that are maintained by traditional farming, which shape one of the most pleasing aspects of Alpine landscapes of Italy (Fontana et al., 2013). Similarly to other European wooded pastures traditionally managed with intermediate disturbance regimes (Paltto et al., 2008, 2011; Thor et al., 2010), Alpine grazed larch forests host a rich lichen biota that benefits from open conditions and the availability of large trees. Despite the fact that several species can be commonly found in other coniferous forests, their lichen biota also includes red-listed species, lichens that are significantly overrepresented in this forest type, and several exclusive species. Beech, Scots pine and low elevation spruce forests seem to play a minor role for lichen conservation. In particular, beech forests lack the typical lichen flora of the Lobarion communities and mainly host crustose-pioneer and shade tolerant species (e.g., Athonia radiata, G. scripta), probably reflecting the fact that in our study region these forests are at their climatic limit compared with the more sub-oceanic conditions of the pre-Alps (e.g., Nascimbene et al., 2013c).
Conclusions Our study indicates the importance of different forest types for epiphyte conservation in alpine environments. In this
The role of forest type in epiphytic lichen conservation
perspective, the EU policies for biodiversity conservation would benefit from similar studies that would include all the European alpine regions. The main conservation-relevant messages of this work are: (1) the importance of maintaining an heterogeneous forest landscape in which timber production couples with biodiversity conservation; (2) since the non-intensive management of spruce forests allows the establishment of a rich lichen biota, a further prolongation of the rotation cycle and/ or the retention of over mature trees is expected to further benefit the potential of these forests for lichen conservation (Nascimbene et al., 2009, 2010); (3) in the Italian Alps oak, riparian and silver fir forests should receive more priority for lichen conservation within the EU policy framework. While riparian forests are usually included in strictly protected areas and left unmanaged, both oak and silver fir forests are mainly managed for timber production, hindering the development of old-growth forest structures that would increase their effectiveness for lichen conservation; (4) traditionally managed, grazed larch forests are an anthropogenic habitat which is worthy for lichen conservation. Unfortunately, this forest type and its associated biodiversity are increasingly threatened by two contrasting processes: intensification of land-use in more favorable sites, and complete abandonment of less favorable sites (Fontana et al., 2013). An excessive canopy closure related to the development of secondary woodlands may reduce lichen diversity (Paltto et al., 2011). On the other hand, the increase of nitrogen inputs related to management intensification is expected to cause the rarefaction of many forest species and the establishment of a few nitrogen-tolerant species (Giordani et al., 2014; Johansson et al., 2012).
Acknowledgments The study was conducted in the framework of the project , biomonitoraggio e conservazione dei licheni “Biodiversita epifiti negli ambienti forestali della provincia di Bolzano”, funded by the Autonomous Province of Bolzano (Ripartizione e Ricerca scientifica). Diritto allo studio, Universita The Forest planning office of the Autonomous Province of Bolzano (project partner) is thanked for providing logistic and technical support. In particular, we are grateful to € nther Unterthiner and his collaborators. Francesco BorGu tignon, Philipp Oberegger, Martin Stecher, and Diego Ivan are thanked for their help during the fieldwork. Daniel Spitale contributed precious effort and discussion for setting up the sampling plane. Helmut Mayrhofer (University of Graz) and his collaborators helped us with lichen species identification, particularly with TLC analyses and critical crustose species.
Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.funeco.2014.06.006.
171
references
AA, VV, 2010. Tipologie forestali dell’Alto Adige. Ripartizione per le Foreste vol. 2. Provincia Autonoma di Bolzano-Alto Adige. Asta, J., Erhardt, W., Ferretti, M., Fornasier, F., Kirschbaum, U., Nimis, P.L., Purvis, W., Pirintsos, S., Scheidegger, C., Van Haluwyn, C., Wirth, V., 2002. Mapping lichen diversity as an indicator of environmental quality. In: Nimis, P.L., Scheidegger, C., Wolseley, P.A. (Eds.), Monitoring With Lichens e Monitoring Lichens, vol. 7. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 273e279. Benavides, R., Montes, F., Rubio, A., Osoro, K., 2007. Geostatistical modelling of air temperature in a mountainous region of northern Spain. Agricultural and Forest Meteorology 146, 173e188. Berryman, S., McCune, B., 2006. Epiphytic lichens along gradients in topography and stand structure in western Oregon, USA. Pacific Northwest Fungi 1, 1e38. Bhattarai, K.R., Vetaas, O.R., 2003. Variation in plant species richness of different life forms along a subtropical elevation gradient in the Himalayas, east Nepal. Global Ecology and Biogeography 12, 327e340. Brunialti, G., Ravera, S., Frati, L., 2013. Mediterranean old-growth forests: the role of forest type in the conservation of epiphytic lichens. Nova Hedwigia 96, 367e381. Crawley, M.J., 2007. The R Book. John Wiley & Sons, Ltd, Chichester, UK. Dietrich, M., Scheidegger, C., 1997. Frequency, diversity and ecological strategies of epiphytic lichens in the Swiss Central Plateau and the Pre-Alps. The Lichenologist 29, 237e258. ^ne, M., Legendre, P., 1997. Species assemblages and Dufre indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67, 345e366. European Commission, 2003. Natura 2000 and Forests: Challenges and Opportunities e Interpretation Guide. Office for Official Publications of the European Communities, p. 101. Fontana, V., Radtke, A., Bossi Fedrigotti, V., Tappeiner, U., Tasser, E., Zerbe, S., Buchholz, T., 2013. Comparing land-use alternatives: using the ecosystem services concept to define a multi-criteria decision analysis. Ecological Economics 93, 128e136. Fox, J., 2002. An R and S-plus Companion to Applied Regression. Sage Publications, Thousand Oaks, California. € Heilmann-Clausen, J., 2010. Rot holes create key Fritz, O., microhabitats for epiphytic lichens and bryophytes on beech (Fagus sylvatica). Biological Conservation 143, 1008e1016. Giordani, P., Calatayud, V., Stofer, S., Seidling, W., Granke, O., Fischer, R., 2014. Detecting the nitrogen critical loads on European forests by means of epiphytic lichens. A signal-tonoise evaluation. Forest Ecology and Management 311, 29e40. Green, T.G.A., Nash III, T.H., Lange, O.L., 2008. Physiological ecology of carbon dioxide exchange. In: Nash III, T.H. (Ed.), Lichen Biology. Cambridge University Press, Cambridge, pp. 152e181. Gustafsson, L., Eriksson, I., 1995. Factors of importance for the epiphytic vegetation of aspen Populus tremula with special emphasis on bark chemistry and soil chemistry. Journal of Applied Ecology 32, 412e424. gan, J.-F., Hawkins, B.A., Field, R., Cornell, H.V., Currie, D.J., Gue Kaufman, D.M., Kerr, J.T., Mittelbach, G.G., Oberdorff, T., O’Brien, E.M., Porter, E.E., Turner, J.R.G., 2003. Energy, water, and broad-scale geographic patterns of species richness. Ecology 84, 3105e3117. Hothorn, T., Bretz, F., Westfall, P., 2008. Simultaneous inference in general parametric models. Biometrical Journal 50, 346e363. Insarov, G., Schroeter, B., 2002. Lichen monitoring and climate change. In: Nimis, P.L., Scheidegger, C., Wolseley, P.A. (Eds.), Monitoring With Lichens e Monitoring Lichens, vol. 7. Kluwer
172
Academic Publishers, Dordrecht, The Netherlands, pp. 183e201. Johansson, O., Palmqvist, K., Olofsson, J., 2012. Nitrogen deposition drives lichen community changes through differential species responses. Global Change Biology 18, 2626e2635. € riado, I., Liira, J., Paal, J., Suija, A., 2009. Tree and stand level Ju variables influencing diversity of lichens on temperate broadleaved trees in boreo-nemoral floodplain forests. Biodiversity and Conservation 18, 105e125. ly, I., Nascimbene, J., Tinya, F., Odor, Kira P., 2013. Factors influencing epiphytic bryophyte and lichen species richness at different spatial scales in managed temperate forests. Biodiversity and Conservation 22, 209e223. Marini, L., Nascimbene, J., Nimis, P.L., 2011. Large-scale patterns of epiphytic lichen species richness: photobiont-dependent response to climate and forest structure. The Science of the Total Environment 409, 4381e4386. McCain, C.M., 2007. Could temperature and water availability drive elevational species richness patterns? A global case study for bats. Global Ecology and Biogeography 16, 1e13. McCune, B., Mefford, M.J., 1999. Multivariate Analysis of Ecological Data. Version 4.25. MJM Software, Gleneden Beach, Oregon. McCune, B., Grace, J.B., 2002. Analysis of Ecological Communities. MJM Software, Gleneden Beach, Oregon. McCune, B., Peck, D.J., Heiman, K., Will-Wolf, S., 1997. Regional gradients in lichen communities of the southeast United States. Bryologist 100, 145e158. € ssler, C., Bradtka, J., Moning, C., Werth, S., Dziock, F., Ba € ller, J., 2009. Lichen diversity in temperate Hothorn, T., Mu montane forests is influenced by forest structure more than climate. Forest Ecology and Management 258, 745e751. Motta, R., 2002. Old-growth forests and silviculture in the Italian Alps: the caseestudy of the strict reserve of Paneveggio (TN). Plant Biosystems 136, 223e232. Nascimbene, J., Marini, L., Motta, R., Nimis, P.L., 2009. Influence of tree age, tree size and crown structure on lichen communities in mature Alpine spruce forests. Biodiversity and Conservation 18, 1519e1522. Nascimbene, J., Marini, L., Nimis, P.L., 2010. Epiphytic lichen diversity in old-growth and managed Picea abies stands in Alpine spruce forests. Forest Ecology and Management 260, 603e609. Nascimbene, J., Thor, G., Nimis, P.L., 2013a. Effects of forest management on epiphytic lichens in temperate deciduous
J. Nascimbene et al.
forests of Europe e a review. Forest Ecology and Management 298, 27e38. Nascimbene, J., Nimis, P.L., Ravera, S., 2013b. Evaluating the conservation status of epiphytic lichens of Italy: a red list. Plant Biosystems 147, 898e904. Nascimbene, J., Dainese, M., Sitzia, T., 2013c. Contrasting responses of epiphytic and dead wood-dwelling lichen diversity to forest management abandonment in silver fir mature woodlands. Forest Ecology and Management 289, 325e332. Nimis, P.L., Martellos, S., 2008. ITALIC e The Information System on Italian Lichens. Version 4.0. University of Trieste, Dept. of Biology. IN4.0/1 http://dbiodbs.univ.trieste.it/. ly, I., Tinya, F., Bortignon, F., Nascimbene, J., 2013. Odor, P., Kira Patterns and drivers of species composition of epiphytic bryophytes and lichens in managed temperate forests. Forest Ecology and Management 298, 27e38. Orange, A., James, P.W., White, F.J., 2001. Microchemical Methods for the Identification of Lichens. British Lichen Society, London. n, B., Go € tmark, F., 2008. Partial cutting as a Paltto, H., Norde conservation alternative for oak (Quercus spp.) foresteResponse of bryophytes and lichens on dead wood. Forest Ecology and Managment 256, 536e547. n, B., Sna € ll, T., 2011. Development Paltto, H., Nordberg, A., Norde of secondary woodland in oak wood pastures reduces the richness of rare epiphytic lichens. PLoS One 6, e24675. Schroeder, L.M., Sahlin, E., Paltto, H., 2011. Retention of aspen (Populus tremula)at final cuttings e the effect of dead wood characteristics on saproxylic beetles. Forest Ecology and Management 262, 853e862. Schroeter, B., Kappen, L., Schulz, F., Sancho, L.G., 2000. Seasonal variation in the carbon balance of lichens in the maritime Antarctic: long-term measurements of photosynthetic activity in Usnea aurantiaco-atra. In: Davison, W., Howard-Williams, C., Broady, P. (Eds.), Antarctic Ecosystems: Models for Wider Ecological Understanding. Caxton Press, Christchurch, pp. 258e262. € nsson, M.T., 2010. Lichen diversity and Thor, G., Johansson, P., Jo red-listed lichen species relationships with tree species and diameter in wooded meadows. Biodiversity and Conservation 19, 2307e2328. Werth, S., Tømmervik, H., Elvebakk, A., 2005. Epiphytic macrolichen communities along regional gradients in northern Norway. Journal of Vegetation Science 16, 199e208. White, F.J., James, P.W., 1985. A new guide to microchemical techniques for the identification of lichen substances. Bulletin of the British Lichen Society 57, 1e41.