Studies on Spatial and Temporal Distributions of Epiphytic Lichens

Linköping Studies in Science and Technology Dissertation thesis No. 1471

Studies on spatial and temporal distributions of epiphytic lichens

Håkan Lättman

School of Life Science

Södertörn University SE-141 89 Huddinge, Sweden

Department of Physics, Chemistry and Biology Division of Ecology

Linköping University SE-581 83 Linköping, Sweden

Linköping, October 2012

Linköping Studies in Science and Technology ISBN 978–91–7519–810–1 ISSN 0345–7524

Printed by LiU-Tryck Linköping, Sweden, 2012

II Table of contents

LIST OF PAPERS IV MY CONTRIBUTIONS TO THE PAPERS IV ABSTRACT V INTRODUCTION 1 WHAT IS A LICHEN? 2 DISPERSAL STRATEGIES 2 ENDURING HARSH ENVIRONMENTS, YET ALSO SENSITIVE 4 SUNLIGHT 4 THE AIR 5 CLIMATE CHANGE: TEMPERATURE AND MOISTURE 6 SUBSTRATE 7 AIMS OF THE THESIS 8 FURTHER BACKGROUND AND THE INCLUDED PAPERS 9 GROWTH 9 GENERATION TIME OF LICHENS 10 Paper I 11 IS SUBSTRATE OR DISPERSAL LIMITING? 12 Phorophyte and stand 12 Spore dispersal 12 Paper II 13 LARGE-SCALE DYNAMIC OF LICHENS 16 Dynamics of lichen thalli 16 Lichens on the move 17 Paper III 17 LICHENS IN THE URBAN ENVIRONMENT 19 Trees: an important urban element 20 Urban effects on lichens 21 Paper IV 21 CONCLUDING REMARKS 25 POPULÄRVETENSKAPLIG SAMMANFATTNING 26 VAD ÄR EN LAV? 26 RESULTAT FRÅN MIN FORSKNING 26 ACKNOWLEDGEMENT 28 REFERENCES 29

III List of papers

The following papers are included in the thesis and are referred in the text by their Roman numerals:

Paper I Lättman H, Brand A, Hedlund J, Krikorev M, Olsson N, Robeck A, Rönnmark F & Mattsson J-E. (2009) Generation time estimated to be 25–30 years in Cliostomum corrugatum (Ach.) Fr. The Lichenologist 41: 557–559.

Paper II Lättman H, Lindblom L, Mattsson J-E, Milberg P, Skage M & Ekman S (2009) Estimating the dispersal capacity of the rare lichen Cliostomum corrugatum. Biological Conservation 142: 1870–1878.

Paper III Lättman H, Milberg P, Palmer MW & Mattsson J-E (2009) Changes in the distribution of epiphytic lichens in southern Sweden using a new statistical method. Nordic Journal of Botany 27: 413–418.

Paper IV Lättman H, Bergman K-O, Rapp M, Tälle M, Westerberg L & Milberg P. Biodiversity in the wake of urban sprawl: loss among epiphytic lichens on large oaks. Submitted manuscript.

Published papers are reproduced with kind permission from the publishers.

My contributions to the papers I have, together with the co-authors, designed all field work for Paper I–II and IV. I performed all field work by myself for Paper II, about half in III–IV and in collaboration with the other authors for Paper I. I also made DNA extractions, PCR amplification and sequencing for Paper II as well as editing and alignment. I made all the statistical analyses in Paper I as well as parts of the analyses in Paper II and IV. I have been writing most of Paper III–IV, and contributed to Paper I–II.

IV Abstract

Lättman, H. 2012. Studies on spatial and temporal distributions of epiphytic lichens Doctoral dissertation

Lichens are an important group of organisms in terms of environmental issues, conservation biology and biodiversity, principally due to their sensitivity to changes in their environment. Therefore it is important that we develop our understanding of the factors that affect lichen distribution. In this thesis, both spatial and temporal distributions of epiphytic lichens at different scales have been studied in southern Sweden. Generation time of the red-listed lichen Cliostomum corrugatum was examined using Bjärka-Säby as the study site. The results showed that the average age of an individual of C. corrugatum is 25–30 years at the onset of spore production. The rarity of C. corrugatum was also examined. DNA analysis of an intron from 85 samples, collected at five sites in Östergötland, yielded 11 haplotypes. Results from coalescent analysis, mantel test and AMOVA indicated that C. corrugatum have a high ability to disperse. The study concluded that its rarity is most likely connected with the low amount of available habitat, old Quercus robur. The changes in the distribution of epiphytic lichens in southern Sweden, between 1986 and 2003, were also compared. For each year a centroid was calculated on all combinations of tree and lichen species. The three significant cases showed that the centroid movement pointed toward a north-east or north-north-east direction. Finally differences in species richness and cover of lichens on large Q. robur were examined between urban and rural environment. The results demonstrated that species number and percent cover was significantly higher on oaks standing rural compared to oaks standing urban. Effects of urban sprawl showed a decline in species richness and cover with increasing age of the surrounding buildings.

Keywords: centroid, Cliostomum corrugatum, direction, dispersal, generation time, global change, habitat availability, lichen, movement, Quercus robur, range shift, urban

Authors address: Håkan Lättman, School of Life Sciences, Södertörn University, SE-141 89 HUDDINGE, Sweden; IFM Division of Biology, Linköping University, SE-581 83 LINKÖPING, Sweden. E-mail: [email protected]; [email protected]

ISSN 1652–7399 ISBN 978–91–7519–810–1 ISSN 0345–7524

Introduction

Lichens are amazing, fascinating and slightly peculiar organisms. They can be described as small ecosystems in their own right, in which several groups of organisms live together in the same body (Bates et al. 2011). This symbiosis is sensitive to changes in the external environment and therefore is an important model in providing answers to many of our questions concerning the environment. Throughout the history of the Earth the environment has constantly been changing in response to various causes. Undoubtedly, today humans have had the greatest impact on the environment (Vitousek et al. 1997, Foley et al. 2005) and species diversity (Jenkins 2003). The development of human civilisation has resulted in a widespread exploitation of nature with significant degradation effects, including recent global climatic changes (Vitousek et al. 1997). Due to mankind’s large and unprecedented impact on our surrounding, it has been suggested that the current geological epoch Holocene has come to an end and that we are now entering Antropocene (Zalasiewicz et al. 2008). Due to our actions, more and more of the Earth’s surface is exploited, resulting in an increasing habitat loss and fragmentation of the remaining habitats. This has led to the decline in abundance and distribution of many species, and also their extinction in several cases (Gonzalez et al. 1998). In order to understand and predict how species will respond to human activities in natural communities, basic knowledge about species behaviour is vital. It is also important to study their different requirements i.e. sunlight, chemical composition of the atmosphere, temperature, humidity and the choice of substrate in order to conserve biodiversity. How environmental changes will affect individual species are difficult to predict. Many species’ environmental requirements are not fully understood and therefore it is important to be able to draw general conclusions. The environmental impact on a species can in turn affect other species to form a chain reaction where more and more species will be affected either positively or negatively. In Europe especially, the broad-leaved forests have been affected by human disturbance (Hannah et al. 1995). Many lichens, insects, and fungi are dependent on these forests and are unable to extend their range to other habitats. Globally, lichens are a group of organisms that have been less studied than other comparable multi-cellular organisms. Thus, there is a gap in the scientific knowledge concerning lichen species’ dispersal capacity and establishment on different substrates, their habitat requirements, and population structure. Our lack of knowledge of lichens is probably explained by their inconspicuousness and their small thalli, which may make them difficult to identify. This might also explain why Carl von Linné (1707–1778) effectively ignored lichens. Fortunately, his protégé, Erik Acharius (1757–1819), made great progress by identifying many lichen species, estimating them to comprise of more than 300 taxa (Krempelhuber 1867). In Sweden today there are more than 2400 known taxa (Feuerer 2009). Hale (1974) reported the worldwide number of lichen species to be approximately 17000. Ten years later Hawksworth and

Hill (1984) reported the number to be 13500. At present there are 18803 described lichen species (Feuerer 2009). The reported number of lichens occurring on Earth is probably underestimated. Swedish lichens and flora are well-studied in comparison with other countries, and contains a large proportion of the worlds lichens (Table 1). Shown in Table 1 is the total number of species of some groups of organisms in Sweden (Gärdenfors 2010) and worldwide (Chapman 2009) and the proportion in Sweden. It is almost certainly an exaggeration that 13% of the earth lichen species exist in Sweden, and is an artefact of this extensive local analysis. It is worth noticing the large proportion of moss and mushroom species that are also present in Sweden which are also probably due to a large number of undescribed species worldwide.

Table 1. Total number of species in Sweden and worldwide for ten major groups of organisms. Groups of organisms Sweden Worldwide Proportion of species (%) Lichens 2419 18803¤ 12.86 Mosses 1049 16236 6.46 Mushrooms ~5000 98998 5.05 Birds 253 9990 2.53 Insects 23900 ~1000000 2.39 Arachnids 1821 102248 1.78 Mammals 63 5487 1.15 Vascular plants 1556 281621 0.55 Fishes 142 31153 0.45 Amphibians & reptilians 19 15249 0.12 ¤ Number taken from Feuerer (2009).

What is a lichen? Lichen symbiosis always consists of a mycobiont and photobiont. The mycobiont is a fungus, mostly an ascomycete, but in some lichens it is a basidiomycete. A photobiont that is capable of photosynthesis is an algae or a cyanobacterium. In some lichens, both an algae and a cyanobacterium are present together with the fungus. There is one lichen described in which the algae belong to phaeophyceae (Sanders et al. 2004). The different organisms belong to different kingdoms and include two domains. The nature of the symbiotic relationship is not trivial. It appears that the fungus sometime acts like a parasite on the photobiont (Brodo et al. 2001). In other lichens, however, the relationship should be considered mutualistic. Both the mycobiont and photobiont can be obligately or facultatively associated with the symbiosis. When the association is obligate, the mycobiont and photobiont can only occur in the lichenized stage which is the stage that describes the symbiosis while in the case of facultative association the organisms may either be free living or a part of the symbiosis (Nash III 1996).

Dispersal strategies The fungal partner in lichens reproduces sexually with spores like other fungi. After fertilization between two different ascomycete mating types, ascospore production is established. When talking of sexual reproduction of lichens, it is only the fungal ascospores that act and function as dispersal units, this is referred to as mycobiont dispersal. Lichens furthermore have the ability to disperse asexually and in those cases fungi and photobiont disperse together in a process called vegetative dispersal. Soredia

and isidia are two examples of asexually produced dispersal units. Soredia are microscopic globule-shaped units that usually originate from the algae layer and consist of singular algal cells surrounded by some fungal hyphae. Isidia are small outgrowths on the cortex of the lichen that have a similar internal structure as the thallus. Moreover, the thallus in some lichens gets easily fragmented and small pieces of the thallus may disperse from one location to another. Dispersal is when sexually and asexually produced units move from one place to another often away from their place of origin. Successful dispersal is when vegetative dispersal units or spores from the mycobiont succeed in finding a suitable algae at the new site and establish a relichenization. Thus, dispersal is never successful unless spores or vegetative diaspores (e.g. isidia, fragments of the thallus or soredia) spread and establish functional thalli on uncolonized patches. The spores and vegetative diaspores of lichens disperse from one site to another by means of both biotic and abiotic vectors. Ants and oribatide mite are examples of biotic vectors (Bailey 1970, Stubbs 1995), while the wind is an abiotic vector (Hansson et al. 1992). The total numbers of dispersed ascospores increase with decreased ascospore size, i.e., small spores often disperse over a greater distance than large spores. Usually, most lichen ascospores are small, approximately 1–30 µm. The ascospores of lichens are often assumed to have unlimited dispersal over great distances (Hansson et al. 1992). Large ascospores and most vegetative diaspores are supposed to disperse over a shorter distance. Thus, they mainly contribute to population turnover at a site rather than to dispersal of the species over a longer distance (Hansson et al. 1992). Long distance dispersal (LDD) of ascospores and eventually vegetative diaspores of lichens is necessary in order to expand a species distribution; Nathan (2006) discussed LDD for plants and claimed the importance of extreme weather events to ensure LDD; this may also apply to lichens. Dispersal of lichens is the activity when the spores or vegetative diaspores of an individual move from one place to another. To establish offspring in new habitats, fast dispersal over long distances may be promoted by easily spread diaspores. This dispersal may be passive or active. The passive dispersal refers to a situation when a vector, e.g., wind, water or animals carries the diaspore. Active dispersal does not occur in lichens but it describes the process when the individual itself actively moves to another site. Conditions such as a high density and high competition are known to influence individuals of certain species to engage in acts of active dispersal (Johst & Brandl 1997, Bowler & Benton 2005). The success of the dispersal is related to the abundance of acceptable habitats and substrates. The results of studies examining the distance that different lichen species are able to disperse differ. Tapper (1976), Armstrong (1987, 1990), Heinken (1999) and Lorentsson and Mattsson (1999) have shown that dispersal can range up to a few hundred meters. Dispersal limitation has also been reported for lichens on trees, e.g., within tree stands, between compact tree stands, or between trees up to a few kilometres apart (Dettki et al. 2000, Sillett et al. 2000, Hilmo & Såstad 2001, Johansson & Ehrlén 2003, Walser 2004, Öckinger et al. 2005). However, genetic studies of Xanthoria parietina and Lobaria pulmonaria suggest that they indeed have efficient dispersal within a few kilometres radius and furthermore, there was no sign of any of them being dispersal-restricted (Lindblom & Ekman 2006, 2007, Wagner et al. 2006, Werth et al. 2006a, 2006b). At a large spatial scale, for populations separated by hundreds of kilometres or more, genetic studies of lichen populations have revealed considerable gene flow (Printzen et al. 2003, Palice & Printzen 2004, Walser et al. 2005), whereas

studies relying on biogeographic patterns (Munoz et al. 2004), trapping of lichen fragments in the atmosphere (Harmata & Olech 1991) and observations of lichen fragments on bird feet (Coppins & James 1979) concluded that effective dispersal was frequent. Finally, the small size and weight of the ascospores has been taken as indirect evidence that lichens are able to disperse widely (Nordén & Appelqvist 2001).

Enduring harsh environments, yet also sensitive Lichens can give us answers to many questions regarding changes in the environment caused by humans since they are sensitive and respond more or less instantly. One explanation for this sensitivity to human activities is the complex structure of the symbiosis. However, somewhat paradoxically, this is also the key to their survival in very harsh environments. The association between the mycobiont and the photobiont in lichens has been a successful relationship lasting at least 4 × 108 years (Taylor 1995). In partnerships, the two organisms are able to withstand harsh environments that they could not withstand individually. Together they are tough and can survive inhospitable environments where many other organisms have difficulties to survive. For example lichens occur at high altitude and latitude, where other groups of organisms have difficulties to survive. Furthermore, some species live inside stones. With their hyphae they can penetrate even granite, with the only externally visible structure the fruiting body (Brodo et al. 2001). Another clear example of their ability to withstand harsh environments has been demonstrated by the lichens Rhizocarpon geographicum and Xanthoria elegans. The two species have survived in outer space (Sancho et al. 2007a). A prerequisite to cope with these harsh environments is that the lichens have had the time to dehydrate. In this dry state, called cryptobiosis, the metabolic rate drops dramatically and the life- sustaining activities in the cells are very low. Once they have entered this state, lichens, as just mentioned, can survive in extreme environments. On the other hand, lichens are sensitive to other changes in abiotic and biotic factors, for instance, polluted air in urban areas. One reason for this is, compared with spermatophytes, the lack of a protective cuticle. A cuticle is a waxy outer layer on, for instance, the leaves that prevents harmful airborne compounds from entering and damaging the internal tissue. Lichens do not have roots for their uptake of water like spermatophytes; instead they satisfy their need of water from precipitation and moisture in the air. In spermatophytes, the casparian strip in the roots is also a barrier that helps the plants avoid harmful substances. Thus, it is not the difference in absorption capacity that makes lichens sensitive compare to spermatophytes. It is when the precipitation and surrounding air contains harmful and toxic substances that uptake of water becomes a problem as they are unable to control what enters the thallus. The lack of a cuticle allows the substances to enter the tissues and cells and injure or kill the lichen. Thus, lichens can withstand extremely harsh environments, yet they are sensitive to certain types of changes in the environment, especially those related to an increase in harmful substances in the atmosphere and to changes in air humidity. Hence, in addition to the general threats to biodiversity – habitat loss, fragmentation and global warming – lichens are also (locally) threatened by air pollution (Giordani 2007, Affum 2011).

Sunlight Lichens are dependent on sunlight, since they are photosynthetic organisms. The amount of light or electromagnetic radiation needed for photosynthesis and how much

each lichen species can withstand varies among species (Demmig-Adams et al. 1990, Green & Lange 1991, Gauslaa & Solhaug 1996, Kappen et al. 1998). The lichen Verrucaria rubrocincata, which is endolithic and lives in desert environments, has been shown to tolerate a high degree of electromagnetic radiation; even at 2600 μmol/m2 s–1 during a hot summer day the lichen still continues to photosynthesise (Garvie et al. 2008). For the epiphytic forest lichen Lobaria pulmonaria, Gauslaa and Solhaug (2000) showed that the electromagnetic radiation never exceeded 610 μmol/m2 s–1 on the Quercus trunk where the lichen occurred. These authors also transplanted L. pulmonaria to a neighbouring tree trunk where the lichen did not occurred and appeared to have similar light conditions. The results showed that the neighbouring Quercus trunk had 6 hours more radiation above 1000 μmol/m2 s–1 during early spring and peaked with 2000 μmol/m2 s–1. The transplanted specimen showed extensive bleaching and damage on the chlorophyll. Forestry practices in Sweden have changed during recent centuries including clear-cut. Lichens that have adapted to live in a forest, relatively shaded compared with open habitat, may be permanently damaged due to high electromagnetic radiation (Gauslaa & Solhaug 1999, Gauslaa et al. 2006). Out of all land areas in Sweden (41.3 × 106 ha), 23 × 106 ha consists of production forest (NBF 2007). The human impact on the production forest is significant and the forests are managed, primarily to promote economic interests. Prevailing forest management is unfortunately often incompatible with the necessary life conditions for lichens, dependent as they are on old tree trunks and standing and prostate dead wood. Such habitats have become scarce. Thus, the change in the demography of forests towards a greater proportion of younger trees, and the practice of clear-cutting, are negative for many lichen species.

The air Soon after the industrial revolution in Europe in the late 18th century, air quality started to change. Independent observers in Manchester, England (Grindon 1859), Munich, Germany (cf. Gries 1997) and Paris, France (Nylander 1866) documented in the mid 19th century that lichens were disappearing from the cities. Fifty years later, in the beginning of the 20th century, the same pattern was recognised across the whole of Europe (Erisman & Draaijers 1995, Mylona 1996). The cause for the decline of lichens in cities was first suggested to be the dust from coal, but later it was realised that sulphur dioxide (SO2) was the main toxic agent. This is one of the most lethal substances for lichens and responsible for most of the modern decrease of lichens (Gilbert 1968). Hawksworth and Rose (1970) showed that lichens could be used to monitor the SO2 content in the air. They used an ordinal scale with teen categories of lichens with increasing sensitivity to SO2. International co-operation with the goal to reduce the effects of air pollutants on the environment has been successful, and a dramatic decrease in SO2 has been observed in recent decades (UNECE 1999). Schopp et al. (2003) predicted a continued decrease of SO2 in Europe up until the end of 2030. Recolonization of lichens in London, United Kingdom, has been shown by Hawksworth and McManus (1989) and has been attributed to the decrease in sulphur dioxide. Another example of successful urban recolonisation is the rare Parmelina tiliacea that was found in central Malmö, southern Sweden, ten years ago (Kärnefelt & Lättman 2001). At that time, this species was red- listed in both Denmark and Sweden (critically endangered and vulnerable, respectively). Most remarkable was that the specimen was healthy-looking and well-developed, and apparently unaffected by the urban environment. This was in stark contrast to the other

lichens present on the tree that were not fully healthy-looking, and thereby of typical appearance of lichens in a larger city. Another indication of the improved SO2- conditions for this lichen is that subsequently, P. tiliacea has been removed from the red list.

Climate change: Temperature and moisture Anthropogenic greenhouse gases are the main cause of climate change that occurs today (Rosenzweig et al. 2008). Climate is a description of weather for an extended period of time that includes variables such as temperature, precipitation, humidity, atmospheric pressure and wind. During the last decades, many studies of climate change have revealed its enormous impact on numerous different organisms including lichens (Vitousek 1994, Hughes 2000, Saxe et al. 2001, Walther et al. 2002, Parmesan & Yohe 2003, Root et al. 2003, Sanz-Elorza et al. 2003, Perry et al. 2005, Thuiller et al. 2005, Parmesan 2006, Maclean et al. 2008, Allen et al. 2010, Dawson et al. 2011, Gosling et al. 2011). It is difficult to separate and determine which of the various abiotic factors are most important for species’ distribution and abundance. For example, Warren et al. (2001) expected a positive response to climate warming in their study on butterflies since they benefit from warm conditions. Unexpectedly three-quarters of the surveyed species decreased. The authors concluded that the positive response to climate warming had been overshadowed by the negative response to habitat loss. Climate warming is probably the major contributor to changing the range boundaries of terrestrial and freshwater habitats (Thomas 2010). Different groups of organisms will have variable successes in meeting the challenges of necessary dispersal. Malcolm et al. (2002) used several vegetation models to determine whether species are able to move as fast as climate zones change. The results showed that changes of climate zones will, in many cases, exceed species capability to migrate. In all vegetation models, high migration rates ≥ 1000 m per year were relatively common. Species that do not have the ability to move and establish on new sites fast enough will face a rapidly changing environment and extinction of some of these species will be inevitable. Thomas et al. (2004) modelled extinction risk for some species in 20% of Earths terrestrial environments and estimated that by the year 2050, 15–37% of the species will be committed to extinction. Discussions about global warming and whether humans affect the process has been and is still under debate. However, according to a survey among climatologists, 96% believe that global average temperatures the past 100 years have increased and 97% that it is induced by man (Doran & Zimmerman 2009). Among the weather variables, it is temperature and water in different forms, e.g., precipitation, fog and dew that has most impact on lichens. For instance photosynthesis is only activated when the lichen thallus is moist, and it stops when the lichen thallus is dehydrated. In the dehydrated state, the lichen is highly resistant to extreme external environment as discussed earlier, but this does not apply for the moist condition where the thallus is sensitive. Lichens can withstand a wide range of temperature when dehydrated, with temperatures ranging from 90°C (or even higher) to –196°C for several days and –60°C for several years (Nash III 1996). However the exposure of a moist thallus of Evernia prunastri to 80°C showed to have a strong negative effect (Pisani et al. 2007). Furthermore some species of the genus Cladonia have proved to be even more sensitive when moist. Kappen and Smith (1980) examined how close Cladonia oceanica could grow a hot steam area of Hawaii. The results of maximum

temperature were 27.2°C in stunted form and 23°C in ramified growth form. Grüninger (1988) sampled ten specimens of Hypogymnia physodes in Reutlingen, West Germany and enveloped the thallus in paper. Two days later he transplanted the lichen thallus on tree trunks on the campus of the University, San José, Costa Rica. Three thalli had died after one week and after ten months they were all dead. The reason for this could be due to the large differences in temperature between Germany and Costa Rica. When a thallus is in a moist state and photosynthesis is activated lichens are most heat sensitive (Rogers 1971, Kappen & Smith 1980). Photosynthesis requires water and the thallus can absorb large amounts. The thallus of green algae lichens can absorb and maintain a water content of 250–400% whilst in cyanolichens this figure can be as high as 600– 2000% and some even higher on a weight basis. When water is available, the thallus reservoir is filled very quickly. In a few seconds 60–70% of the thallus becomes saturated and full saturation is reached within minutes. Photosynthesis has its maximum speed, for green algae lichens, when the thallus has a water content of 70–150% and corresponding value for cyanolichens is 300–600% (Nash III 1996).

Substrate Corticolous (growing on bark), lignicolous (growing on wood), saxicolous (growing on rock) and terricolous (growing on soil) are convenient terms to describe the substrate preferences of lichens. Some species have the ability to grow on several kinds of surfaces while others are limited to just one or a few. Some epiphytic lichens are able to grow on a variety of tree species, while others only grow on one or just a few. Furthermore, some are demanding in terms of size or age of the tree and prefer a large, old tree. Wedin et al. (2004) made a remarkable discovery regarding the genera Stictis and Conotrema. They showed that a fungal species can adopt two different lifestyles depending on the circumstances. If the fungal spore ends up on the substrate wood, the fungus adopts a non-lichenized, saprophytic lifestyle and has been called Stictis. On the other hand, if the fungal spore ends up on bark – and the needed photobiont is available – the fungus adopts a lichenized lifestyle and has been called Conotrema. Hence, two different genera proved to be the same fungal species but with different lifestyles: lichenized or non-lichenized. The fungus’ plasticity in terms of lifestyle and the frequency with which it occurs is unknown but may be common (Hawksworth 2005). Thus, requirements on the substrate quality for different lichen species vary from generalists to highly specialised lichen species. As our human population increases, we use more and more land area: we claim more space. As a consequence, the amount of available habitat reduces for most other organisms. This unceasing demand to exploit new areas and their resources for our benefit has created severe situations for many other non-human organisms, particularly those that require large areas for their survival. Furthermore, remaining habitats become fragmented and thus divided from each other into patches often surrounded by areas affected by different kinds of intense human activity (e.g. agriculture, urban areas, roads and railway tracks). Organisms with a limited ability to disperse face a severe situation in fragmented areas. Fragmentation and its edge effects have been shown to have a negative impact on lichen biodiversity (Turner 1996) and abundance (Esseen & Renhorn 1998).

Aims of the thesis

The overall goal of this thesis is to increase our knowledge regarding the changes in time and space in the occurrence of epiphytic lichen species and their communities with the use of existing or new methods. The complexity of these factors and their synergistic effects make it necessary to undertake studies on different spatial and organisational levels. These range from genetic, individual and population levels through to community and biotope levels. Thus, different studies with different objectives and methods were designed in order to target the overall aims of the thesis. Estimating the generation time of the red-listed crustose epiphytic lichen Cliostomum corrugatum was the objective of Paper I. A method for assessing generation times, from meiospore to meiospore, is often necessary in order to understand population dynamics as well as to describe evolutionary history both in the short as well as in a long run perspective. The results were used in the following study (Paper II). Paper II makes an attempt to determine whether the rarity of C. corrugatum is due to difficulties with dispersal or if it is its habitat – old and often large Quercus robur – that is limiting. Using the analysis of a nuclear ribosomal RNA gene, three different methods to analyze the pattern, the lichens ability to disperse were tested. Paper III examines the problems of describing changes in the lichen flora on a regional scale. From field surveys of common epiphytic lichens in southern Sweden, conducted in 1986 and 2003, the change in position of the centroids of these species over time was assessed. A centroid of a species is the mean position of its sites in an area, calculated from the coordinates of sampling sites. Finally, in Paper IV the focus is to investigate differences in species richness and cover of some common and rare epiphytic lichens on Q. robur standing in urban and rural environments. The effect of urban sprawl was also examined on species richness and cover on common and rare epiphytic lichens on Q. robur. Two methods were used to measure the degree of urbanization, one of which took into account the average age of five adjacent buildings and the second the area of nearby buildings at six radii centred on a visited tree. In the rest of this thesis, each of the papers is presented with a general background introducing the specific study.

Further background and the included papers

Growth Lichens have been studied for a long time and knowledge of their biology has accumulated. However in the light of for example vascular plants we have just begun to understand how lichens function, with this field much less studied. Crustose lichens growth and growth rate has been studied and especially so in the lichen Rhizocarpon geographicum (Armstrong 1983, Proctor 1983, Haworth 1986, Armstrong 2002, 2006, Hansen 2010). In general, lichens are slow-growing and have a reputation for not only this but also for being long-lived (Hale 1973, Matthews & Trenbirth 2011). However, there are both fast- and slow growing lichens. Benedict (2008) reported an annual radial growth rate for R. superficiale to be as little as 0.006 mm per year. An example of a fast growing lichen is Usnea longissima, which showed a maximum growth of 18.4 cm in a single year (Keon & Muir 2002). The variation in growth within a single species may also be large. Hill (1981) reported annual radial growth rates (RGR) of Lecanora muralis to be 0.03–0.55 mm per year while Seaward (1976) showed the RGR to be 2.84–6.05 mm per year. Several methods for study growth rate have been developed (Platt & Amsler 1955, Farrar 1974, Honegger et al. 1996). Lichenometry is a frequently used method where measurement of the lichen thallus RGR is central. The method is used for example to date moraines and the retreat of glaciers (Karlén & Black 2002). A lichen’s growth curve varies depending on growth form. The growth curve of the crustose lichen R. geographicum can be divided into three parts: 1) where RGR gradually increases to a maximum; 2) maximum speed is kept for a short time period; and 3) the speed of RGR decreases (Armstrong 2005, Armstrong & Bradwell 2010). There is no evidence that foliose lichens would have a phase where there is a reduction of RGR as there is in crustose forms. The growth of lichens and their population dynamics is, of course, also influenced by abiotic and biotic factors, but these affects lichens differently depending on the species. The abiotic weather factors that mainly affect lichens are temperature, humidity and sun exposure. Regardless of season there is a growth all year round in R. geographicum but predominantly during the summer months (Armstrong 2006). Other abiotic factors that affect lichens are nutrients. Gauslaa et al. (2006) performed growth experiments with Lobaria pulmonaria. Some of the thallus was sprayed with clean water and others with nutrients added to the water. The results showed that water with added nutrients only slightly increased the growth of L. pulmonaria. Results with stronger support for the importance of nutrients were presented by McCune and Caldwell (2009), where L. pulmonaria thalli were immersed in a bath of phosphorus for twenty minutes. After one year the biomass was doubled compared to the control group. Thus, growth and growth rate is relatively well studied at least for the lichens R. geographicum and L. pulmonaria.

Another population dynamic aspect which is also, at least in some areas, relatively well studied is mortality. Most studies on mortality have dealt with pollution such as emission from industries and urban areas to determine the highest concentrations of harmful substances lichens can withstand before they disappear. The knowledge gap in mortality is in knowing the dynamics under normal conditions. For instance, at what age (or size) does a lichen thallus die a natural death or what is the lifespan of a lichen thallus? There is some knowledge about the well studied lichen R. geographicum which can reach an age of about 1000 year (Matthews & Trenbirth 2011) but the maximum age is likely to be very much shorter in epiphytic lichens. In fact, Hypogymnia physodes never reach such an age. Studies have shown that H. physodes has a growth rate of 3–4 mm per year (Gorbach & Kobzar 1981). Since the thallus rarely exceeds 5 cm in diameter this means that the individuals of H. physodes reach approximately 6 to 8 years of age before they die. Furthermore Mattsson et al. (2006) have shown that H. physodes has a rapid turnover in the sense of appearance/disappearance at sites. By using the growth rate and lichen thallus diameter, it is possible to estimate the age on an individual thallus. However there are many unanswered questions: what is the population dynamics in terms of mortality, at what age do they die and begin to reproduce sexually and asexually? Hence, to better understand the population dynamics of lichens, we certainly need much more basic field research.

Generation time of lichens Generation time can be explained as the time span from a given point in the parent life cycle to the same given point in the offspring. Two different types of generation time can be distinguished i.e. fundamental and realized generation time. The fundamental generation time is based on the shortest possible time (age) of reproduction for an individual of that particular species while realized time is the average parental age at reproduction under natural conditions. Generation length is sometimes used synonymously with the word generation time. Generation time varies for different eukaryotic organisms, for instance, the oriental latrine fly, Chrysomya megacephala have a short generation time in only 20.7 days (Gabre et al. 2005), in the same way human generation time is approximately 25 years and Japanese timber bamboo, Phyllostachys bambusoides, have a generation time on about 120 years (Janzen 1976). The consequences of different generation time are that the genetic material will evolve at different rates depending on the species. The time span between meiosis events is important to estimate as these events have a potential for genetic recombination, while the vegetative phase of organisms is more inert at the genetic level. Thus, knowledge of species’ generation time is essential for calculations of the speed of evolutionary changes. The relative importance of sexual versus asexual reproduction depends on the species. The spores in mycobiont-dispersed species have undergone genetic recombinations that may increase genetic variation and the spores are often small and may be dispersed far away. The downside is that the mycobiont must find a suitable photobiont before being able to become lichenized. Species that disperse with vegetative diaspores has the advantage that both partners are spread along together but there is no recombination as they are clonal (Nash III TH & Gries 2002).

Paper I In Paper I, the generation time of Cliostomum corrugatum was studied in Bjärka-Säby, Östergötland, Sweden. The largest thallus area and largest diameter on apothecia were recorded on Quercus robur. Only large trees were included in the study since the occurrence of C. corrugatum is low on trees with a small circumference at breast height (CBH) (e.g. Ranius et al. 2008, Johansson et al. 2009). By plotting thallus area or apothecia diameter against oak diameter and extrapolate the regression line it was possible to identify the age of Q. robur when C. corrugatum colonised it and at what age it becomes fertile. Quercus robur CBH were translated into oak age and the estimated time it takes for C. corrugatum to become fertile (fundamental generation time) for C. corrugatum were found to be 25–30 years (Figure 1). The fertile thallus may then continue to produce spores for many years ahead.

Figure 1. The epiphytic lichen Cliostomum corrugatum becomes fertile at an age of 25– 30 years.

It is surprising that sexual maturity takes such a long time to form in this lichen, especially considering that by being an epiphyte, its substrate has a limited life span. To my knowledge, this is the first time that anyone has determined the generation time of lichens. It would be interesting to estimate this in other suitable lichen species. The question of whether the generation time is longer among rare and red-listed species than among common lichens appears to be particularly pertinent. Clearly more research is required for a comprehensive picture of lichen generation times to emerge.

Is substrate or dispersal limiting?

Phorophyte and stand Epiphytes are plants that grow on other plants, principally trees and shrubs but also dwarf shrubs (e.g. Calluna and Vaccinium sp.) and even leaves (in the tropics) can be used as a substrate. The common name for the various substrates is called phorophyte. The utility of the bark substrate for various lichen species may vary considerably. The common lichen Hypogymnia physodes does not have high demands on the substrate but can grow on a wide range i.e. corticolous, lignicolous or sometimes saxicolous but also man-made substances such as rubber and steel (personal field observations) (Figure 2).

Figure 2. The lichen Hypogymnia physodes is common in Sweden and grows on many different substrates. It inhabits mostly bark, wood or stones but sometimes also man- made substrates such as rubber and steel.

What may be important for the epiphyte is to what species the phorophyte belongs, and its age (or dimension of the trunk). Several phorophytes together form a stand that is scattered in the landscape in different ways making it more or less suitable (such as sun exposure, pH and structure of the bark), influencing accessibility for the establishment and also how beneficial these stands are for long term survival of the lichen. In general, areas designated for forestry have a lower value than protected stands since production areas are more homogeneous and often lack old and large trees and also are low in dead wood.

Spore dispersal Present lichen distribution is in part a result of their ability to spread from one place to another. Lichen ascospores are relatively small ranging in size from 2–3 μm for Chaenotheca furfuracea to 150 μm for Phlyctis argena with a few other species having even larger spores (Foucard 2001). The production of spores is large and one fruit body may contain more than 1 × 106 spores, which can be equated with 12–18 × 106 spores in one square centimetre (Tibell 1994). The main vector for their dispersal is wind which probably is important for long distance dispersal (LDD). The air at ground level contains a large amount of spores. Gregory (1978) measured the levels of spores, from various cryptogams, during five summer months in Rothamsted, England, and showed average concentrations of 12000 spores m-3 but for short periods as many as 1 × 106 may be present. The concentration decreases with increasing altitude with 10000 spores m-3 one kilometre above the ground and hundreds of spores three kilometres up (Hirst et al. 1967). In contrast to vascular plants, many lichens have very large geographic distributions and if their distribution is not cosmopolitan, it may be pantemperate,

pantropical (Lücking 2003) or amphitropical (Søchting & Olech 1995, Myllys et al. 2003). The lichen Porpidia flavicunda has a circumpolar distribution and Buschbom (2007) showed that the gene flow was high among the four surveyed sites. Furthermore, the gene flow occurred in almost all possible directions and the lichen has had several repeated LDD of vegetative diaspores between the sites. Högberg et al. (2002) made an exciting discovery on Letharia vulpina. They found that in North America, this lichen dispersed sexually by spores but in Europe the populations spread clonally by soredia and/or isidioid soredia. Long distance dispersal with spores of species from the genus Umbilicaria has on several independent occasions traveled to Antarctica from surrounding temperate areas. This does not only apply to the spores but also to the algal cells. Thus despite the hostile environment prevailing in Antarctica, successful relichenizations have been established, on multiple occasions, between fungi and algal cells (Romeike et al. 2002). Several papers have argued that some lichen have limited dispersal abilities. However, knowledge is often lacking about whether dispersal or substrate availability is the limiting factor for a specific lichen population to survive in the long run, both locally and regionally. To fully understand the importance of these different factors in isolation or in combination, it is necessary to study both dispersal efficiency and the impact of substrate abundance and microhabitat characteristics.

Paper II In Paper II, I investigated whether the limited occurrence of the lichen C. corrugatum is due to limitation by dispersal or limitation by habitat availability. The investigation was conducted in Östergötland, south-eastern Sweden, at five sites. The laboratory methods involved DNA extraction, PCR amplification and sequencing of a group 1 intron at the end of the small subunit (SSU) nuclear ribosomal RNA gene. Attempts were made using other genomic regions (ITS and IGS) but the variability at these regions were too low for our purposes. Out of the 96 collected samples of C. corrugatum, 85 were successfully extracted and shown to represent 11 haplotypes (Figure 3).

Figure 3. An unrooted haplotype network of the epiphytic lichen Cliostomum corrugatum. The most common haplotypes 1 (N = 30) and 2 (N = 46) are in the centre of the network and are most likely the oldest. The terminal haplotypes are rare (N = 1) and have most likely derived from haplotypes 1 or 2.

Several statistical methods were used to analyse the genetic variation and to make inference about the lichen’s ability to disperse. Firstly, a coalescent simulation showed that the gene flow was considerable between the five investigated sites. Secondly, a mantel test showed that there were no significant correlation between the genetic distance and the geographic distance matrices. Thirdly, an AMOVA test showed that 0.4% of the variation was between the populations and 99.6% of the variation was within the populations. All three tests indicate that C. corrugatum does not seem to have any difficulties dispersing from one place to another. In addition our results indicate that the five sites behave more or less as a single, sexual interbreeding population, i.e. a panmictic population. Consequently, C. corrugatum rarity is likely to be connected with the limited amounts of the suitable habitat, old oaks. The distribution of Q. robur is more or less the whole of Europe, but unfortunately large, old oaks are relatively scarce. During several hundred years an oak trunk serves as a suitable substrate for this lichen, a time during which it can, with exception of the first 25 years, disperse spores (Paper I). However, genetic methods have unfortunately its limitations. The result of the genetic information reflects, to some extent, the population of C. corrugatum historical status and not the lichens situation today. Another context in which to view these results regards the peculiar Quaternary history of the study area that involves land uplift. Since the end of the last glaciation, all sites have gone from being submerged to terrestrial (Figure 4).

Figure 4. The two maps show a snapshot over the five studied sites during and after the glacial retreat. a) Twelve thousand years ago the glacier edge (solid blue line) stretched in a WSW/ENE direction and all five sites were submerged. b) Ten thousand years ago the glacier had retreated and the three sites in the west have become terrestrial.

At 10000 BP the three western sites had become terrestrial (Figure 4b). The remaining two sites in the east (Stegeborg and Bråborg) had to wait, approximately, another 5000 years until they emerged from the sea. In southern Scandinavia Quercus invaded early after the last glaciations and they have been present in the study area for at least 6000 years (Bradshaw 2000, Rasmussen 2005). It is important to note that a site becoming terrestrial is just one prerequisite for oak colonisation. Another is the edafic conditions and a third is whether oaks were present in the area at the time. Notably, the three sites in the west, i.e. those with the longest history as terrestrial, also had greater haplotype diversity (Figure 5). What this means for our understanding of C. corrugatum ability to disperse and the evolution is difficult to evaluate. Have the rare haplotypes evolved at these sites, or have they accumulated here through migration over time due to their relatively long history? Should the lack of rare haplotypes in Bråborg and Stegeborg be interpreted as an indication that the species is limited by dispersal?

y = 0.0271x + 2.2536 Solberga 5 R2 = 0.8259 Bjärka-Säby

4 Orräng Bråborg 3

2 Stegeborg

Number of haplotypes of Number 1

0 0 20 40 60 80 100 Height above see level (m)

Figure 5. The three sites in the west are higher above the sea level and have more haplotypes than the two sites in the east.

Large-scale dynamic of lichens

Dynamics of lichen thalli Lichens can be divided into crustose (resembling a crust), foliose (resembling a leaf) and fruticose (resembling a shrub). In the group crustose lichens, many are slow- growing (see above) and some have been used in lichenometry, which is a method for dating exposed rock. I believe this is the major reason why the general picture of lichens is that they grow slowly and have a slow population turnover. In this context, slow population turnover means that the rate of mortality and nativity are low, i.e. individual thallus remain attached to the same site year after year. In comparison with other spore- producing organisms such as bryophytes, most crustose lichens have a slow turnover

(Pharo & Beattie 1997, Pharo et al. 1999), but this slow turnover does not apply to foliose and fruticose lichens, at least not all of them. Gustafsson and Milberg (2008) demonstrated that the foliose lichen Lobaria pulmonaria had a high turnover in south- eastern Sweden (permanently marked thalii). Mattsson et al. (2006) showed that several common species, e.g. the fruticose lichen Hypogymnia physodes in southern Sweden, has a high turnover (presence/absence at sites). In California, USA, Boucher and Nash III (1990) estimated the annual turnover of biomass of the common fruticose epiphytic lichen Ramalina menziesii to be 29% where the annual turnover of biomass is the sum of all fallen thallus. This implies that R. menziesii contribute substantially to the nutrient turnover in the ecosystem. It also means that lichen thalli are constantly replaced by new ones. Therefore, even though it might look like it is the same lichen thallus that sits on the tree trunk every year, it may be a new one.

Lichens on the move Being able to move from one place to another is generally fundamental for most organisms, and this also holds true also for lichens. This has become especially important in recent decades with rapid climate change. Since lichens have been shown to be sensitive to changes in their environment, they are a useful group of organisms to help detect the biological response to global warming (Pisani et al. 2007, Sancho et al. 2007b). The change in the amount of lichen biomass will ultimately affect the distribution pattern of individual species. In a climate that is changing it is expected that lichen species will alter their distribution pattern. Long-term empirical studies in alpine and arctic environment have shown conflicting results in terms of an increase or decrease in the amount of lichen biomass. A decrease has been shown by Kari (2008) and Hudson and Henry (2010) that contradicts findings by Hollister et al. (2005) who found an increase of lichen cover across time. Unchanged amount of lichen biomass have also been reported (Hudson & Henry 2009). Hauck (2009) proposes an alternative cause for the decline in lichens in alpine and arctic environment. He argues that changes in the use of land and to high th atmospheric SO2 levels in the mid-20 century better correlated with the decline of lichens than an annual increase in average temperature. Models incorporating different types of scenarios have been used to predict future distribution and abundance of different lichen species (Ellis et al. 2007a, Ellis et al. 2007b). To date, research has not yet demonstrated that lichens have been seriously threatened because of climate change but rather that a decrease or increase of species distribution will occur. Aptroot and van Herk (2007) have shown that lichens with green algae Trentepohlia sp. as a photobiont are increasing their distribution. There are areas though which could ultimately be threatened by climate change. Aptroot (2009) highlights areas and habitats likely to experience problems e.g. low-level islands with endemic lichens, arctic and tundra regions and high ground in the tropics. It is important that we have methods to quantify changes in the distribution of different lichens so we are able to take appropriate actions and develop new and improve existing types of lands use, e.g. in forestry.

Paper III This brings us to Paper III, where studies on lichens movements were investigated. In this paper, movement of the, centre of distribution (centroid) within southern Sweden, of some common epiphytic lichens were studied, based on a small repeated sampling. The study was conducted at 64 sites and the inventory in the field was carried out in

1986 and 2003. Fifty-six epiphytic lichens and 22 tree species (phorophytes) were included in the study. Thirty cases were possible to analyze, out of which three showed a significant shift in the centroid. The centroid movements of the lichens Hypogymnia physodes and Vulpicida pinastri on the tree species Juniperus communis were 50 km and 151 km (p-value 0.0258, 0.0002) with the direction 27° and 48°, respectively. The movement of the centroids of H. physodes on Pinus sylvestris was 41 km (p-value 0.0066) with the direction 30° (Figure 6). All three significant cases had moved in a north-east or a north north-east direction.

Figure 6. The three arrows indicate direction and distance that the centroid had moved between the years 1986 and 2003 of Hypogymnia physodes and Vulpicida pinastri on the tree species Juniperus communis but also H. physodes on the tree species Pinus sylvestris. Squares represent the study sites.

The data set was fairly small with only presence information of epiphytic macrolichens on different substrates recorded. The sites were only roughly described, without information on tree size and tree species abundance and trees without lichens were not recorded. Hence, the statistical power of the analyses was low. Of the two species that turned out significant at least H. physodes has a large ecological amplitude, which

should provide a strong resistance against small changes in environmental conditions. Nevertheless, some significant changes were recorded on the movement of the centroid. This indicates a greater impact of global warming on the epiphytic lichen flora than previously presumed. The temperature of the planet is increasing and has done so at least during the past 40 years (Rosenzweig et al. 2008). In our study area, the current trend appears to be the same. On 29 meteorological stations in the study area the average temperature has increased by 0.056°C year –1 during the period 1986 to 2003 (Figure 7) (SMHI 1986– 2003). Another explanation why some lichens have moved their centroid may be due to large scale forest structure changes in recent years. Hedwall et al. (2012) used data from the Swedish National Forest Inventory to compare the field- and tree layer in boreal and temperate Sweden between 1994 and 2010. They found that the canopy has become denser and the distribution of species abundance on the forest floor has changed as it has become darker. The lichens could respond in a similar way to a darker environment.

9.0

8.0

7.0

6.0

5.0 y = 0.056x + 6.804 R² = 0.112 4.0

3.0

2.0

The average temperature average The 1.0

0.0 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03

Year Figure 7. The average temperature over 29 meteorological stations in southern Sweden during the period 1986 to 2003. Lichens in the urban environment Urbanization transforms a natural, semi-natural or agricultural landscape into an environment with buildings. In an ecological perspective, the organisms and the environment they live in are, to a varying degree, interconnected and dependent on each other. As an example, trees are of paramount importance for epiphytic lichens and many tree species are, more or less, dependent on mycorrhizal fungi. There are many reasons to maintain a high biodiversity in urban areas especially in highly urbanized areas. In a study in Flanders in Belgium, an area where the proportion of forest is only 10%, Cornelis and Hermy (2004) showed that their 15 surveyed city parks contained about 30%, 50%, 40% and 60% of all wild plant species, nesting birds, butterflies and

amphibians, respectively, of the total national number. This highlights the magnitude of biodiversity even in cities. Urbanization is a major cause of a homogenization of biotic factors (McKinney 2006), but the goal of biodiversity conservation should be towards diversity. In Sheffield, United Kingdom, Gaston et al. (2005) tested various methods to increase biodiversity. They added nests for various insects, ponds for birds, dead wood for fungi and patches of Urtica dioica for butterflies. Since urban areas are largely composed of private and residential gardens, the investigation was focused on these spaces. Some of the tested methods were found to boost biodiversity, indicating the potential for retaining biodiversity within city borders. Ranta and Viljanen (2011) list the causes of the relatively high biodiversity of vascular plants in Finnish cities as spacious urban structure, small human populations, late urbanisation, and abundant remnant natural vegetation (forest). The investigations demonstrated that much can be done to increase biodiversity in urban environments and there is certainly room for improvement. Hahs et al. (2009) examined the factors affecting the rate of plant extinction from urban environment. They concluded that it largely has to do with the city’s history in combination with current proportion of native vegetation. They also claim that the transformation of the landscape to an urban area is likely to involve an extinction debt, i.e. that there is a time delay in the loss of biodiversity from our cities when they grow.

Trees: an important urban element Trees in cities are in a stressful environment because usually they stand in a soil with poor quality and their roots often have a limited ability to spread. In addition, roots are frequently damaged during ground works i.e. when various types of pipes and cables should be buried (Jim 1998). Increased runoff from buildings, hard surfaces and drainage (Leopold 1968) which reduces the impact of a positive long-term effect of moister from the precipitation in cities and the smog in some cities can cause tissue damage on the trees (Middleton et al. 1958, Tripathi & Gautam 2007, Honour et al. 2009). Nevertheless, trees are often an important feature of cities. Hence, trees have been planted, or retained from native or rural setting during urban sprawl, for a number of reasons: wellbeing of people, economic value, aesthetics, the production of shade and, in some situation, even for fire prevention purposes. Lohr et al. (2004) surveyed residents in the United States concerning advantages and disadvantages associated with trees in a city. Results from their study showed that the public evaluated trees ability to shade and cool the surroundings highest and secondly it helped people feel calmer. Highest ranked disadvantages were that they can cause allergies and block store signs. Lohr et al. (2004) concluded that most people clearly appreciated the value of urban trees in their lives. Several studies have shown that trees in cities are highly economically valued. For instance Donovan and Butry (2011) showed that an increased number of trees in a residential garden increased the monthly rent by USD5.62 and trees in the public right of way increasing the rent even more. Not surprisingly, the sale prices of buildings are affected by trees. An increase in the number of trees also added an extra USD8870 to a house in Portland, Oregon and reduced the time on the market by 1.7 days (Donovan & Butry 2010). Nowak et al. (2002) calculated the value of urban forests using tree valuation methods and field data. The evaluated total compensatory value was 101 × 106 and 5.2 × 109 USD in Jersey City and New York, respectively. Donovan et al.

(2011) showed a reduced risk of poor birth outcomes by 1.42 per 1000 births if canopy cover were increased by 10% within 50 m of residential buildings. Clearly is that those who can afford, like to have trees around. The studies above give us an indication that even in the future; trees will continue to be highly valued in cities, although increasing land prices adds a threat towards retained trees. On balance, the prospect for continued survival of epiphytic organisms in cities seems brighter than for many other types of organisms.

Urban effects on lichens Lichens have different tolerance or preferences concerning substrates. This means that some lichens grow on different substrates, even some man-made. Hence, there are lichens that grow on either coniferous or deciduous trees or others confined to a particular tree species (Washburn & Culley 2006, Spier et al. 2010). Other lichens can be restricted to a particular size of a tree e.g. large tree (Washburn & Culley 2006) that usually has a coarse bark structure. Such specialized lichens have declined since the supply of substrate is less in urban environments (Shukla & Upreti 2011). Large trees of Quercus have proved to be very rich in species that also include lichens (Rose 1974, Hultengren 1995), but few studies of lichens have been conducted in an urban setting, an exception being Larsen et al. (2007) who investigated lichen distribution on Quercus robur and Q. petraea in London, England in relation to air pollution. The authors were able to distinguish three zones where lichen species increase in number from the inner to the outer zones. Lichens have been used to estimate air quality in urban areas. Wielgolaski (1975) inventoried several groups of organisms such as vascular plants, bryophytes, lichens, microorganisms, invertebrates, fish and plankton on their biological value to use as a tool for evaluating air, freshwater and marine quality. The author argued that most foliose and fruticose lichen in a wider sense can be used for evaluating the air quality. Hawksworth and Rose (1970) well-known work was more fine-tuned and they also used crustose lichens and algae. They were positioning a selection of lichens (and algae), on a ten-point scale reflecting various sensitivities, and then used it to estimate the air quality with particular attention to SO2. During the 1970s a number of studies showed that lichens are adversely affected by road traffic emissions of SO2 (e.g. Brawn & Ogden III 1977). Even the air quality and quantity of heavy metals has been investigated by means of lichens (Pandey et al. 2002, Montero Alvarez et al. 2006). Even though there are positive signs that the air quality is beginning to improve in urban environments (Lisowska 2011), SO2 is still, 30 years later, considered the main limiting factor for lichens in urban settings (Giordani 2007). Trees surrounding an expanding city will be enclosed with buildings. Lichens on these trees will be isolated and activities such as dispersal and establishment may be negatively affected. Mobile organisms like animals can escape while trees will inevitably be incorporated and trapped into the city or cut down during urban sprawl. The fate of epiphytic lichens will parallel that of the trees.

Paper IV In Paper IV, I examined species number and cover of epiphytic lichens on remnant oaks in urban and rural environment. In Linköping, County of Östergötland, Sweden, 105 urban and 109 rural oak trees were surveyed for 17 selected lichen species. Trees with a CBH > 250 cm were selected from a database developed by the County Administrative

Board in Östergötland. The majority of the available urban trees were selected since there were only a limited number. I then selected a population of rural trees, aiming for (i) a population of similar circumferences and (ii) similar densities of oaks in the surrounding. I calculated densities with a radius of 302 m, as a previous study had identified this as the appropriate scale for lichen richness in the study area (Muhammadi 2011). Nine of the lichens were common and eight were rare. During the field-work, CBH, depth of bark crevices and sun exposure were documented per tree. The density of oaks in the vicinity of target trees was calculated within radii of 150, 250, 350, 500, 700 and 1000 m. Two variables were also constructed as a measure of degree and age of urbanisation. Firstly, the density of buildings around each Q. robur within a radius of 150, 250, 350, 500, 700 and 1000 m was calculated. Secondly, the average age of the five closest buildings were also used as a measure of the age since urbanization. Lichens richness and cover was higher in the rural environment than in the urban environment there were, however, one exception: Lepraria incana (Figure 8).

Figure 8. a) to c) Fourteen investigated lichen species and their occurrence and cover between urban and rural environment on tree trunks of Quercus robur are compared. a) Shows the proportion of trees with lichen occurrence. b) Shows the average percentage cover on trees with occurrence. c) Shows the average percentage cover of all trees (also those lacking the species in question).

Our results recorded a clear effect of urban sprawl. Both species richness and cover of lichens were significantly higher in younger compared to older parts. When including all the external factors, it was possible to analyse eight out of the 14 lichens. Input factors include: age of urbanization, tree circumference (held constant), bark crevices, sun exposure, density of oaks (held constant) and the cover of buildings. Several species were negatively affected by the age of the urbanization process but one was positively affected. Age of urbanization affected five species negatively while cover of buildings affected three species negatively, for the radii 500, 700 and 1000 m and one positive for radius 350 m. The tree circumference and the bark crevices were positively affected except for one species where the whole model showed to be insignificant. Cities are artificial environments but, despite this, they can be compared to natural ones. If we think of the buildings as small hills and the tiles and asphalt on the ground as rocks, there are certain similarities to natural habitats. But what makes them different? In the city there are several factors that make them different from rural areas, i.e. pollution (more harmful particles in the air), temperature (higher), moisture, disturbance and habitat configuration (McDonnell et al., 1997), but also habitat loss and fragmentation (Niemelä, 1999, McKinney, 2006). Thus, the reason for the decrease in species number and amount of lichens in the urban area is probably a mix of several factors, for example the lichens dispersal mode, a higher temperature and lower humidity.

Concluding remarks

All studies in this thesis have dealt with epiphytic lichens and their movement through space and time. Noteworthy findings that contribute to the lichenology are, in my view, the following. First, an epiphytic lichen must be relatively old before they start producing spores, 25–30 years for the red-listed species C. corrugatum. Whether generation time is a distinguishing feature for rare lichens and how it relates to common epiphytes is an open question. Second, this rare species does not seem to be limited by dispersal, at least as judged by the examined gene haplotype distribution. This finding contributes to the ongoing debate of dispersal vs. habitat availability as the cause for rarity, and to our view of epiphytic lichens as more mobile then generally thought. Third, I was able to detect shifts in the distribution of some epiphytic lichens within just a few decades using crude field data. This point to the benefits of using lichens for monitoring large-scale changes in biodiversity. Fourth, trees that are integrated in a growing city do not provide the same epiphytic lichen abundance or diversity as their rural counterparts. Furthermore, age of urbanization affected species occurrences negatively, which means that we can expect further extinctions during urban sprawl if habitats for epiphytic lichens are surrounded by buildings. Overall, I hope this thesis will contribute to our understanding of epiphytic lichens as highly mobile organisms and their great value for monitoring changes, as well as fascinating and unpredictable in their own right.

Populärvetenskaplig sammanfattning

Vad är en lav? En lav består av en svamp och en alg som lever i symbios vilket betyder ungefär; att leva tillsammans. Ibland finns även en cyanobakterie i symbiosen. Svampen kallas för mykobiont och är för det mesta en sporsäckssvamp, men hos vissa lavar är det en basidsvamp. Algen är för det mesta en grönalg eller cyanobakterie och kallas för fotobiont. Dessutom finns ett fall beskrivet där algen är en brunalg. I Sverige finns det drygt 2100 lavarter som är vetenskapligt beskrivna och i hela världen nästan 19000. Lavar som växer på träd kallas för epifyter. Det kommer från grekiskan där epi betyder ”på” och phyton (fyt) betyder ”växt”. Lavar är en viktig organismgrupp att studera eftersom de kan ge oss svar på många frågor när det gäller förändringar i vår miljö. De är nämligen känsliga för yttre förändringar samtidigt som de är tuffa och kan leva i ogästvänliga miljöer där många andra organismer har svårt att överleva. Kartlav har visat på deras förmåga att överleva i extrema miljöer. Under 14 dagar vistades laven i yttre rymden. En förutsättning för att de ska klara av de här extrema miljöerna är att vävnaderna har haft en chans att torka upp och få ett lågt fuktinnehåll. Å andra sidan så är de känsliga för t.ex. luftens föroreningar i urbana miljöer. Orsaken är att de saknar en skyddande kutikula. Växter däremot har en kutikula, det består av ett vaxartat yttre lager som bl.a. täcker bladen och hindrar skadliga luftburna föreningar från att tränga in och skada cellerna. Lavar har inte några rötter utan tar upp vatten från det regn som faller och från den fukt som luften innehåller. Vad som är problematiskt för laven är när nederbörden och luften innehåller skadliga och giftiga ämnen. En avsaknad av en kutikula gör att de skadliga och giftiga ämnena tränger rakt in i vävnader och celler och skadar och dödar. Så de kan alltså uthärda hårda miljöer men samtidigt är de känsliga för föroreningar och förändringar i miljön. I den här doktorsavhandlingen har jag studerat utbredning och spridning av epifytiska lavar i tid och rum i södra Sverige.

Resultat från min forskning I Bjärka-Säby, Östergötland, sökte jag på flera ekstammar efter den sällsynta laven gul dropplav för att mäta diametern på det apothecium som var det största men även mäta och registrera arean på den största bålen (Papper I). Apothecier kallas lavens fruktkroppar, de bildas vid sexuell reproduktion. Gul dropplav är epifyt på framförallt stora gamla ekar men kan även förekomma på alm och asp. Den är trefärgad; ljusgrå bål, ljusgula apothecier och svarta pyknid och relativt lätt att känna igen i fält (pyknid är asexuellt bildade fruktkroppar som många gånger har formen av en liten urna). Undersökningens syfte var att ta reda på hur många år det tar för en spor av gul dropplav att växa upp och själv bilda egna sporer. En spor är lavar, mossor och

svampars motsvarighet till de blommande växternas frön. Resultaten visade att gul dropplav bildar sporer först i 25–30 årsåldern. Med hjälp av genetiska metoder försökte jag ta reda på varför gul dropplav är sällsynt (Papper II). Jag hade som hypotes att det antingen kunde bero på att den har svårt för att sprida sig från en plats till en annan. Eller så kunde det bero på att stora gamla ekar är sällsynta. Är stora gamla ekar sällsynta så kommer också laven att bli sällsynt. Fem lokaler i Östergötland med ekar inventerades på gul dropplav och undersöktes genetiskt. Samtliga tester antydde att gul dropplav inte har svårt för att sprida sig från en plats till en annan, utan det verkade vara den låga tillgången på stora och gamla ekar som gör gul dropplav sällsynt. Genetiska metoder har tyvärr sina begränsningar. Resultaten från den genetiska informationen avspeglar, till viss grad, gul dropplavs status bakåt i tiden och inte lavens situation idag. I en annan studie undersökte jag skillnader i förekomsten hos några vanliga epifytiska lavar åren 1986 och 2003 hos (Papper III). Studien genomförds i södra Sverige på 64 lokaler. För vart och ett av åren beräknades en medelpunkt, en så kallad centroid, för varje kombination av lokal, träd- och lavart. Utav alla tänkbara kombinationer var det 30 fall som kunde analyseras, varav 3 uppvisade en signifikant förflyttning av centroiden, samtliga i en nordöstlig eller nord-nordöstlig riktning. Vår tolkning är att detta speglar den globala uppvärmningen. De två lavarna har en nordlig utbredning så riktningen av centroid-förflyttningen mot norr på grund av ett varmare klimat är inte osannolikt. I en fjärde studie (Papper IV) studerade jag epifytiska lavars antal och täckningsgrad på ek där hälften stod i stadsmiljö i Linköping och andra hälften i närliggande landsbygd. Jag valde träd så att de två grupperna skulle få så lika omkrets som möjligt (> 250 cm) och så lika täthet av ekar i omgivningen som möjligt. Det fanns betydligt fler lavarter och större täckning av lavar på landsbygden än i staden. Samtliga ekar är stora, och fanns på platsen innan urbaniseringen började. Jag beräknade tidpunkten för urbaniseringen som medelvärdet på byggnadsåret för de fem närmaste byggnaderna runt varje ek. Även här var resultaten tydliga: antal lavarter och täckning av lavarna minskar med tiden en ek stått omgiven av byggnader. Anledningen till att artantalet och mängden lavar är lägre i staden beror troligtvis på flera faktorer och säkert är de samverkande. I staden är det högre halter föroreningar i luften, en högre temperatur, ändrade fuktighetsförhållanden i luften, förlust av habitat och fragmentering. Samtliga studier har behandlat lavar och deras förflyttning i tid och rum. Undersökningarna har visat att lavar måste vara relativt gamla innan de själva bildar sporer. De har visat sig bra på att sprida sig från en plats till en annan. Att de troligtvis svarar på förändringar i klimatet. Och, trots förbättrade luftkvalitet, är de mindre förkommande i urbana miljöer än på landsbygden.

Acknowledgement

I am indebted to so many people and I am sure that I am neither the first nor the last with this debt of gratitude. There are so many people I would like to thank who meant a lot to me during my time as a PhD. In order not to exclude anyone by mistake I: collectively want to thank friends and colleagues who have contributed and supported me to a doctoral thesis.

However, I cannot help but mention a few who has been particularly important to me.

Great thanks to my supervisors Jan-Eric Mattsson and Per Milberg. Thanks to unofficial stand in co-supervisors and co-authors Patrik Dinnétz and Karl-Olof Bergman.

Thanks to my loving mother Ingegerd and my sister Eva for all your support and encouragement. Last but not least, I want to thank my family Åsa, Elin, Gustav, Linn, Kim and David for your presence and support.

References

Affum HA, Oduro-Afriyie K, Nartey VK, Addo MA, Nyarko BJB, Mumuni II, Adu PS, Appiah GK & Coleman A (2007) Lichens as a monitor for atmospheric manganese pollution. Research Journal of Environmental and Earth Sciences 3: 534–540. Allen CD, Macalady AK, Chenchouni H, Bachelet D, McDowell N, Vennetier M, Kitzberger T, Rigling A, Breshears DD, Hogg (Ted) EH, Gonzalez P, Fensham R, Zhang Z, Castro J, Demidova N, Lim J-H, Allard G, Running SW, Semerci A & Cobb N (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management 259: 660–684. Aptroot A (2009) Lichens as an indicator of climate and global change. In: Letcher TM (Ed.) Climate change: Observed impacts on planet Earth. Elsevier, Oxford. Aptroot A & van Herk CM (2007) Further evidence of the effects of global warming on lichens, particularly those with Trentepohlia phycobionts. Environmental Pollution 146: 293–298. Armstrong RA (1983) Growth curve of the lichen Rhizocarpon geographicum. New Phytologist 94: 619–622. Armstrong RA (1987) Dispersal in a population of the lichen Hypogymnia physodes. Environmental and Experimental Botany 27: 357–363. Armstrong RA (1990) Dispersal, establishment and survival of soredia and fragments of the lichen, Hypogymnia physodes (L) Nyl. New Phytologist 114: 239–245. Armstrong RA (2002) The effect of rock surface aspect on growth, size structure and competition in the lichen Rhizocarpon geographicum. Environmental and Experimental Botany 48: 187–194. Armstrong RA (2005) Growth curves of four crustose lichens. Symbiosis 38: 45–57. Armstrong RA (2006) Seasonal growth of the crustose lichen Rhizocarpon geographicum (L.) DC. in south Gwynedd, Wales. Symbiosis 41: 97–102. Armstrong RA & Bradwell T (2010) Growth of crustose lichens: A review. Geografiska Annaler 92 A: 3–17. Bailey RH (1970) Field and study notes: Animals and the dispersal of soredia from Lecanora conizaeoides Nyl. ex Cromb. The Lichenologist 4: 256. Bates ST, Cropsey GWG, Gregory Caporaso J, Knight R & Fierer N (2011) Bacterial communities associated with the lichen symbiosis. Applied and Environmental Microbiology 77: 1309–1314. Benedict JB (2008) Experiments on lichen growth III. The shape of the age-size curve. Arctic, Antarctic and Alpine Research 40: 15–26. Boucher VL & Nash III TH (1990) The role of the fruticose lichen Ramalina menziesii in the annual turnover of biomass and macronutrients in a blue oak woodland. Botanical Gazette 151: 114–118.

Bowler DE & Benton TG (2005) Causes and consequences of animal dispersal strategies: relating individual behaviour to spatial dynamics. Biological Reviews 80: 205–225. Bradshaw RHW, Holmqvist BH, Cowling SA & Sykes MT (2000) The effects of climate change on the distribution and management of Picea abies in southern Scandinavia. Canadian Journal of Forest Research 30: 1992–1998. Brawn K & Ogden III JG (1977) Lichen diversity and abundance as affected by traffic volume in an urban environment. Urban Ecology 2: 235–244. Brodo IM, Sharnoff SD & Sharnoff S (2001) Lichens of North America. Yale University Press, New Haven. Buschbom J (2007) Migration between continents: Geographical structure and long- distance gene flow in Porpidia flavicunda (lichen-forming Ascomycota). Molecular Ecology 16: 1835–1846. Chapman AD (2009) Numbers of living species in Australia and the world. Report for the Australian Biological Resources Study, Australian Biological Resources Study, Canberra. Coppins BJ & James PW (1979) Birds and the dispersal of lichen propagules. The Lichenologist 11: 105–106. Cornelis J & Hermy M (2004) Biodiversity relationships in urban and suburban parks in Flanders. Landscape and Urban Planning 69: 385–401. Dawson TP, Jackson ST, House JI, Prentice IC & Mace GM (2011) Beyond predictions: Biodiversity conservation in a changing climate. Science 332: 53–58. Dettki H, Klintberg P & Esseen PA (2000) Are epiphytic lichens in young forests limited by local dispersal? Ecoscience 7: 317–325. Donovan GH & Butry DT (2010) Trees in the city: Valuing street trees in Portland, Oregon. Landscape and Urban Planning 94: 77–83. Donovan GH & Butry DT (2011) The effect of urban trees on the rental price of single- family homes in Portland, Oregon. Urban Forestry & Urban Greening 10: 163–168. Donovan GH, Michael YL, Butry DT, Sullivan AD & Chase JM (2011) Urban trees and the risk of poor birth outcomes. Health & Place 17: 390–393. Doran PT & Zimmerman MK (2009) Examining the scientific consensus on climate change. EOS, Transactions of the American Geophysical Union 90: 22–23. Ellis CJ, Coppins BJ & Dawson TP (2007a) Predicted response of the lichen epiphyte Lecanora populicola to climate change scenarios in a clean-air region of Northern Britain. Biological Conservation 135: 396–404. Ellis CJ, Coppins BJ, Dawson TP & Seaward MRD (2007b) Response of British lichens to climate change scenarios: Trends and uncertainties in the projected impact for contrasting biogeographic groups. Biological Conservation 140: 217–235. Erisman JW & Draaijers GPJ (1995) Atmospheric deposition in relation to acidification and eutrophication, Amsterdam, Oxford, Elsevier. Esseen P-A & Renhorn K-E (1998) Edge effects on an epiphytic lichen in fragmented forests. Conservation Biology 12: 1307–1317. Farrar JF (1974) A method for investigating lichen growth rates and succession. The Lichenologist 6: 151–155. Feuerer T (ed.) (2009) Checklist of lichens and lichenicolous fungi, Version 1 September 2009. http:www.checklist.de Foucard T (2001) Svenska skorplavar och svampar som växer på dem. Interpublishing, Stockholm.

Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe MT, Daily GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda C, Patz JA, Prentice C, Ramankutty N & Snyder PK (2005) Global consequences of land use. Science 309: 570–574. Gabre RM, Adham FK & Chi H (2005) Life table of Chrysomya megacephala (Fabricius) (Diptera: Calliphoridae). Acta Oecologica 27: 179–183. Gaston KJ, Smith RM, Thompson K & Warren PH (2005) Urban domestic gardens (II): experimental tests of methods for increasing biodiversity. Biodiversity and Conservation 14: 395–413. Gauslaa Y, Lie M, Solhaug KA & Ohlson M (2006) Growth and ecophysiological acclimation of the foliose lichen Lobaria pulmonaria in forests with contrasting light climates. Oecologia 147: 406–416. Gilbert OL (1968) Bryophytes as indicators of air pollution in the Tyne valley. New Phytologist 67: 15–30. Gilbert OL (1990) The lichen flora of urban wasteland. The Lichenologist 22: 87–101. Giordani P (2007) Is the diversity of epiphytic lichens a reliable indicator of air pollution? A case study from Italy. Environmental Pollution 146: 317–323. Gonzales A, Lawton JH, Gilbert FS, Blackburn TM & Evans-Freke I (1998) Metapopulation dynamics, abundance, and distribution in a microecosystem. Science 281: 2045–2047. Gorbach NV & Kobzar NN (1981) Determination of the age of epiphyte lichen Hypogymnia physodes (L.) Nyl. Doklady Akademii Nauk BSSR [Minsk] 25: 848– 851. Gosling SN, Warren R, Arnell NW, Good P, Caesar J, Bernie D, Lowe JA, van der Linden P, O’Hanley JR & Smith SM (2011) A review of recent developments in climate change science. Part II: The global-scale impacts of climate change. Progress in Physical Geography 35: 443–464. Gregory PH (1978) Distribution of airborne pollen and spores and their long distance transport. Pure and Applied Geophysics 116: 309–315. Gries C (1997) Lichens as indicators of air pollution. In: Nash III TH (Ed.) Lichen biology. University Press, Cambridge. Grindon LH (1859) The Manchester flora. W. White, London. Grüninger W (1988) Use of the temperate lichen Hypogymnia physodes (Parmeliaceae) to evaluate air pollution in the Tropics. Revista de Biologia Tropical 36: 545–547. Gustafsson A & Milberg P (2008) Changes in the abundance of Lobaria pulmonaria in south-eastern Sweden from 1994 to 2007. Graphis Scripta 20: 44–51. Gärdenfors U (2010) Rödlistade arter i Sverige 2010. The 2010 red list of Swedish species. ArtDatabanken, Uppsala. Hahs AK, McDonnell MJ, McCarthy MA, Vesk PA, Corlett RT, Norton BA, Clemants SE, Duncan RP, Thompson K, Schwartz MW & Williams NSG (2009) A global synthesis of plant extinction rates in urban areas. Ecology Letters 12: 1165–1173. Hale ME (1973) Growth. In: Ahmadjian V & Hale ME (Eds.) The Lichens. Academic Press, London. Hale ME Jr (1974) The biology of lichens, 2nd Edn. A series of student texts in contemporary biology. Edward Arnold, London. Hannah L, Carr JL & Lankerani A (1995) Human disturbance and natural habitat: a biome level analysis of a global data set. Biodiversity and Conservation 4: 128– 155.

Hansen ES (2010) A review of lichen growth and applied lichenometry in southwest and southeast Greenland. Geografiska Annaler 92 A: 65–79. Hansson L, Söderström L & Solbreck C (1992) The ecology of dispersal in relation to conservation. In: Hansson L (Ed.) The ecological principles of nature conservation. Elsevier Applied Science, London. Harmata K & Olech M (1991) Transect for aerobiological studies from Antarctica to Poland. Grana 30: 458–463. Hauck M (2009) Global warming and alternative causes of decline in arctic-alpine and boreal-montane lichens in North-Western Central Europe. Global Change Biology 15: 2653–2661. Hawksworth DL & Rose F (1970) Qualitative scale for estimating sulphur air pollution in England and Wales using epiphytic lichens. Nature 227: 145–148. Hawksworth DL & Hill DJ (1984) The lichen-forming fungi. Blackie, Glasgow, United Kingdom. Hawksworth DL & McManus PM (1989) Lichen recolonization in London under conditions of rapidly falling sulphur-dioxide levels, and the concept of zone skipping. Botanical Journal of the Linnean Society 100: 99–109. Hawksworth D (2005) To be or not to be a lichen. Nature 433: 468. Haworth LA, Calkin PE & Ellis JM (1986) Direct measurement of lichen growth in the central brooks range, Alaska, U.S.A., and its application to lichenometric dating. Arctic and Alpine Research 18: 289–296. Hedwall P-O, Brunet J, Nordin A & Bergh J (2012) Changes in the abundance of keystone forest floor species in response to changes of forest structure. Journal of Vegetation Science in press. Heinken T (1999) Dispersal patterns of terricolous lichens by thallus fragments. The Lichenologist 31: 603–612. Hill D (1981) The growth of lichens with special reference to the modelling of circular thalli. The Lichenologist 13: 265–287. Hilmo O & Såstad SM (2001) Colonization of old-forest lichens in a young and an old boreal Picea abies forest: An experimental approach. Biological Conservation 102: 251–259. Hirst JM, Stedman OJ & Hogg WH (1967) Long-distance spore transport: Methods of measurement, vertical spore profiles and the detection of immigrant spores. Journal of General Microbiology 48: 329–355. Hollister RD, Webber PJ & Tweedie CE (2005) The response of Alaskan arctic tundra to experimental warming: Differences between short- and long-term responses. Global Change Biology 11: 525–536. Honegger R, Conconi S & Kutasi V (1996) Field studies on growth and regenerative capacity in the foliose macrolichen Xanthoria parietina (Teloschistales, Ascomycotina). Botanica Acta 109: 187–193. Honour SL, Bell JNB, Ashenden TW, Cape JN & Power SA (2009) Responses of herbaceous plants to urban air pollution: Effects on growth, phenology and leaf surface characteristics. Environmental Pollution 157: 1279–1286. Hudson JMG & Henry GHR (2009) Increased plant biomass in a High Arctic heath community from 1981 to 2008. Ecology 90: 2657–2663. Hudson JMG & Henry GHR (2010) High Arctic plant community resists 15 years of experimental warming. Journal of Ecology 98: 1035–1041.

Hughes L (2000) Biological consequences of global warming: Is the signal already apparent? Trends in Ecology and Evolution 15: 56–61. Hultengren S (1995) Något om lavfloran på en västsvensk ek. Svensk Botanisk Tidskrift 89: 165–170. Högberg N, Kroken S, Thor G & Taylor JW (2002) Reproductive mode and genetic variation suggest a North American origin of European Letharia vulpina. Molecular Ecology 11: 1191–1196. Janzen DH (1976) Why bamboos wait so long to flower. Annual Review of Ecology and Systematics 7: 347–391. Jenkins M (2003) Prospects for biodiversity. Science 302: 1175–1177. Jim CY (1998) Impacts of intensive urbanization on trees in Hong Kong. Environmental Conservation 25: 146–159. Johansson P & Ehrlen J (2003) Influence of habitat quantity, quality and isolation on the distribution and abundance of two epiphytic lichens. Journal of Ecology 91: 213– 221. Johansson V, Bergman K-O, Lättman H & Milberg P (2009) Tree and site quality preferences of six epiphytic lichens growing on oaks in southeastern Sweden. Annales Botanici Fennici 46: 496–506. Johst K & Brandl R (1997) The effect of dispersal on local population dynamics. Ecological Modelling 104: 87–101. Kappen L & Smith CW (1980) Heat tolerance of two Cladonia species and Campylopus praemorsus in a hot steam vent area of Hawaii. Oecologia 47: 184–189. Kari K (2008) Species-specific responses of an alpine plant community under simulated environmental change. Journal of Vegetation Science 19: 363–372. Karlén W & Black JL (2002) Estimates of lichen growth-rate in northern Sweden. Geografiska Annaler 84 A: 225–232. Keon DB & Muir PS (2002) Growth of Usnea longissima across a variety of habitats in the Oregon coast range. The Bryologist 105: 233–242. Krempelhuber A (1867) Geschichte und litteratur der lichenologie von den ältesten zeit bis zum schlusse des jahres 1865. Selbstverlag, München. Kärnefelt I & Lättman H (2001) An unusual discovery of Parmelina tiliacea in central Malmö, Sweden. Graphis Scripta 11: 9–12. Larsen RS, Bell JNB, James PW, Chimonides PJ, Rumsey FJ, Tremper A & Purvis OW (2007) Lichen and bryophyte distribution on oak in London in relation to air pollution and bark acidity. Environmental Pollution 146: 332–340. Leopold LB (1968) Hydrology for urban land planning ― A guidebook of the hydrologic effects of urban land use. U.S. Geological Survey Circular 554, Washington, D.C. Lindblom L & Ekman S (2006) Genetic variation and population differentiation in the lichen-forming ascomycete Xanthoria parietina on the island Storfosna, central Norway. Molecular Ecology 15: 1545–1559. Lindblom L & Ekman S (2007) New evidence corroborates population differentiation in Xanthoria parietina. The Lichenologist 39: 259–271. Lisowska M (2011) Lichen recolonisation in an urban-industrial area of southern Poland as a result of air quality improvement. Environmental Monitoring Assessment 179: 177–190.

Lohr VI, Pearson-Mims CH, Tarnai J & Dillman DA (2004) How urban residents rate and rank the benefits and problems associated with trees in cities. Journal of Arboriculture 30: 28–35. Lorentsson S & Mattsson J-E (1999) New reports of soredia dispersed by ants, Formica cunicularia. The Lichenologist 31: 204–207. Lücking R (2003) Takhtajan's floristic regions and foliicolous lichen biogeography: A compatibility analysis. The Lichenologist 35: 33–54. Maclean IMD, Austin GE, Rehfisch MM, Blew J, Crowe O, Delany S, Devos K, Deceuninck B, Günther K, Laursen K, van Roomen M & Whal J (2008) Climate change causes rapid changes in the distribution and site abundance of birds in winter. Global Change Biology 14: 2489–2500. Malcolm JR, Markham A, Neilson RP & Garaci M (2002) Estimated migration rates under scenarios of global climate change. Journal of Biogeography 29: 835–849. Matthews JA & Trenbirth HE (2011) Growth rate of a very large crustose lichen (Rhizocarpon subgenus) and its implications for lichenometry. Geografiska Annaler 93 A: 27–39. Mattsson J-E, Lättman H & Milberg P (2006) Rapid changes in the epiphytic macrolichen flora on sites in southern Sweden. The Lichenologist 38: 323–329. McCune B & Caldwell BA (2009) A single phosphorus treatment doubles growth of cyanobacterial lichen transplants. Ecology 90: 567–570. McKinney ML (2006). Urbanization as a major cause of biotic homogenization. Biological Conservation 127: 247–260. Middleton JT, Darley EF & Brewer RF (1958) Damage to vegetation from polluted atmospheres. Journal of the Air Pollution Control Association 8: 9–15. Montero Alvarez A, Estévez Alvarez JR, Iglesias Brito H, Pérez Arriba O, López Sánchez D & Wolterbeek HT (2006) Lichen based biomonitoring of air quality in Havana City west side. Journal of Radioanalytical and Nuclear Chemistry 270: 63– 67. Muhammadi UH (2011) Spatial pattern of occurrence of eleven epiphytic lichen species in a heterogeneous landscape. Master thesis, Linköping University, LiTH-IFM-Ex-- 11/2441--SE. Munoz J, Felicisimo AM, Cabezas F, Burgaz AR & Martinez I (2004) Wind as a long- distance dispersal vehicle in the southern hemisphere. Science 304: 1144–1147. Myllys L, Stenroos S, Thell A & Ahti T (2003) Phylogeny of bipolar Cladonia arbuscula and Cladonia mitis (Lecanorales, Euascomycetes). Molecular Phylogenetics and Evolution 27: 58–69. Mylona S (1996) Sulphur dioxide emissions in Europe 1880–1991 and their effect on sulphur concentrations and depositions. Tellus Series B-Chemical and Physical Meteorology 48: 662–689. Nash III TH (ed.) (1996): Lichen Biology. 1st Ed. – Cambridge University Press, Cambridge. Nash III TH & Gries C (2002) Introduction. In: Nash III TH, Ryan BD, Gries C & Bungartz F (Eds.) Lichen flora of the Greater Sonoran Desert Region. Vol. I, pp. 1– 53. Lichens Unlimited, Arizona State University. Nathan R (2006) Long-distance dispersal of plants. Science 313: 786–788. NBF (2007) Statistical Yearbook of Forestry 2007. NBF, Jönköping, Sweden. Nordén B & Appelqvist T (2001) Conceptual problems of ecological continuity and its bioindicators. Biodiversity and Conservation 10: 779–791.

Nowak DJ, Crane DE & Dwyer JF (2002) Compensatory value of urban trees in the United States. Journal of Arboriculture 28: 194–199. Nylander W (1866) Les lichens du Jardin du Luxembourg. Bulletin de la Société Botanique de France 13: 364–372. Palice Z & Printzen C (2004) Genetic variability in tropical and temperate populations of Trapeliopsis glaucolepidea: Evidence against long-range dispersal in a lichen with disjunct distribution. Mycotaxon 90: 43–54. Pandey V, Upreti DK, Pathak R & Pal A (2002) Heavy metal accumulation in lichens from the Hetauda industrial area Narayani zone Makwanpur District, Nepal. Environmental Monitoring and Assessment 73: 221–228. Parmesan C (2006) Ecological and evolutionary responses to recent climate change. Annual Review of Ecology Evolution and Systematics 37: 637–669. Parmesan C & Yohe G (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42. Perry AL, Low PJ, Ellis JR & Reynolds JD (2005) Climate change and distribution shifts in marine fishes. Science 308: 1912–1915. Pharo EJ & Beattie AJ (1997) Bryophyte and lichen diversity: A comparative study. Australian Journal of Ecology 22: 151–162. Pharo EJ, Beattie AJ & Binns D (1999) Vascular plant diversity as a surrogate for bryophyte and lichen diversity. Conservation Biology 13: 282–292. Pisani T, Paoli L, Gaggi C, Pirintsos SA & Loppi S (2007) Effects of high temperature on epiphytic lichens: Issues for consideration in a changing climate scenario. Plant Biosystems 141: 164–169. Platt RB & Amsler FP (1955) A basic method for the immediate study of lichen growth rates and succession. Journal of the Tennessee Academy of Science 30: 177–183. Printzen C, Ekman S & Tønsberg T (2003) Phylogeography of Cavernularia hultenii: Evidence of slow genetic drift in a widely disjunct lichen. Molecular Ecology 12: 1473–1486. Proctor MCF (1983) Sizes and growth-rates of thalli of the lichen Rhizocarpon geographicum on the moraines of the Glacier de Valsorey, Valais, Switzerland. The Lichenologist 15: 249–261. Ranius T, Johansson P, Berg N & Niklasson M (2008) The influence of tree age and microhabitat quality on the occurrence of crustose lichens associated with old oaks. Journal of Vegetation Science 19: 653–662. Ranta P & Viljanen V (2011) Vascular plants along an urban-rural gradient in the city of Tampere, Finland. Urban Ecosystems 14: 361–376. Rasmussen P (2005) Mid- to late-Holocene land-use change and lake development at Dallund Sø, Denmark: vegetation and land-use history inferred from pollen data. The Holocene 15: 1116–1129. Rogers RW (1971) Distribution of the lichen Chondropsis semiviridis in relation to its heat and drought resistance. New Phytologist 7: 1069–1077. Romeike J, Friedl T, Helms G & Ott S (2002) Genetic diversity of algal and fungal partners in four species of Umbilicaria (lichenized Ascomycetes) along a transect of the Antarctic peninsula. Molecular Biology and Evolution 19: 1209–1217. Root TL, Price JT, Hall KR, Schneider SH, Rosenzweig C & Pounds JA (2003) Fingerprints of global warming on wild animals and plants. Nature 421: 57–60. Rose F (1974) The epiphytes of oak. In Morris MG & Perring FH (Eds.) The British oak: its history and natural history, Faringdon, Classey.

Rosenzweig C, Karoly D, Vicarelli M, Neofotis P, Wu Q, Casassa G, Menzel A, Root TL, Estrella N, Seguin B, Tryjanowski P, Liu C, Rawlins S & Imeson A (2008) Attributing physical and biological impacts to anthropogenic climate change. Nature 453: 353–358. Sancho LG, de la Torre R, Horneck G, Ascaso C, de Los Rios A, Pintado A, Wierzchos J & Schuster M (2007a) Lichens survive in space: Results from the 2005 LICHENS experiment. Astrobiology 3: 443–454. Sancho LG, Green TGA & Pintado A (2007b) Slowest to fastest: Extreme range in lichen growth rates supports their use as an indicator of climate change in Antarctica. Flora 202: 667–673. Sanders WB, Moe RL & Ascaso C (2004) The intertidal marine lichen formed by the pyrenomycete fungus Verrucaria tavaresiae (Ascomycotina) and the brown alga Petroderma maculiforme (Phaeophyceae): Thallus organization and symbiont interaction. American Journal of Botany 91: 511–522. Sanz-Elorza M, Dana ED, Gonzalez A & Sobrino E (2003) Changes in the high- mountain vegetation of the central Iberian peninsula as a probable sign of global warming. Annals of Botany 92: 273–280. Saxe H, Cannell MGR, Johnsen B, Ryan MG & Vourlitis G (2001) Tree and forest functioning in response to global warming. New Phytologist 149: 369–399. Schopp W, Posch M, Mylona S & Johansson M (2003) Long-term development of acid deposition (1880–2030) in sensitive freshwater regions in Europe. Hydrology and Earth System Sciences 7: 436–446. Seaward MRD (1976) Performance of Lecanora muralis in an urban environment. In: Brown DH, Hawksworth DL & Bailey RH (Eds.) Progress and problems in lichenology. Academic Press, London. Shukla V & Upreti DK (2011) Changing lichen diversity in and around urban settlements of Garhwal Himalayas due to increasing anthropogenic activities. Environmental Monitoring Assessment 174: 439–444. Sillett SC, McCune B, Peck JE, Rambo TR & Ruchty A (2000) Dispersal limitations of epiphytic lichens result in species dependent on old-growth forests. Ecological Applications 10: 789–799. SMHI (1986–2003) Väder och vatten. -Swedish Meteorological and Hydrological Institute, Norrköping, Sweden. Spier L, van Dobben H & van Dort K (2010) Is bark pH more important than tree species in determining the composition of nitrophytic or acidophytic lichen floras? Environmental Pollution 158: 3607–3611. Stubbs CS (1995) Dispersal of soredia by the oribatid mite, Humerobates arborea. Mycologia 87: 454–458. Søchting U & Olech M (1995) The lichen genus Caloplaca in polar regions. The Lichenologist 27: 453–471. Tapper R (1976) Dispersal and changes in the local distributions of Evernia prunastri and Ramalina farinacea. New Phytologist 77: 725–734. Taylor TN (1995) The oldest fossil lichen. Nature 378: 244. The World Conservation Union (2010) IUCN Red list of threatened species. Summary Statistics for Globally Threatened Species. Table 1: Numbers of threatened species by major groups of organisms (1996–2010). Tibell LB (1994) Distribution patterns and dispersal strategies of Caliciales. Botanical Journal of the Linnean Society 116: 159–202.

Thomas CD (2010) Climate, climate change and range boundaries. Diversity and Distributions 16: 488–495. Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BFN, Ferreira de Siqueira M, Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Townsend Peterson A, Phillips OL & Williams SE (2004) Extinction risk from climate change. Nature 427: 145–148. Thuiller W, Lavorel S, Araujo MB, Sykes MT & Prentice IC (2005) Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America 102: 8245–8250. Tripathi AK & Gautam M (2007) Biochemical parameters of plants as indicators of air pollution. Journal of Environmental Biology 28: 127–132. Turner IM (1996) Species loss in fragments of tropical rain forest: A review of the evidence. The Journal of Applied Ecology 2: 200–209. UNECE (1999) Protocol to the 1979 convention on long-range transboundary air pollution to abate acidification, eutrophication, and ground-level ozone. United Nations Economic Commission for Europe, Geneva, Switzerland. Vitousek PM (1994) Beyond global warming: Ecology and global change. Ecology 75: 1861–1876. Vitousek PM, Mooney HA, Lubchenco J & Melillo JM (1997) Human domination of earth’s ecosystems. Science 277: 494–499. Wagner HH, Werth S, Kalwij JM, Bolli JC & Scheidegger C (2006) Modelling forest recolonization by an epiphytic lichen using a landscape genetic approach. Landscape Ecology 21: 849–865. Walser JC (2004) Molecular evidence for limited dispersal of vegetative propagules in the epiphytic lichen Lobaria pulmonaria. American Journal of Botany 91: 1273– 1276. Walser JC, Holderegger R, Gugerli F, Hoebee SE & Scheidegger C (2005) Microsatellites reveal regional population differentiation and isolation in Lobaria pulmonaria, an epiphytic lichen. Molecular Ecology 14: 457–467. Walther G-R, Post E, Convey P, Menzel A, Parmesank C, Beebee TJC, Fromentin J-M, Hoegh-Guldberg O & Bairlein F (2002) Ecological responses to recent climate change. Nature 416: 389–395. Warren MS, Hill JK, Thomas JA, Asher J, Fox R, Huntley B, Roy DB, Telfer MG, Jeffcoate S, Harding P, Jeffcoate G, Willis SG, Greatorex-Davies JN, Moss D & Thomas CD (2001) Rapid responses of British butterflies to opposing forces of climate and habitat change. Nature 414: 65–69. Washburn SJ & Culley TM (2006) Epiphytic macrolichens of the greater Cincinnati metropolitan area―Part II: Distribution, diversity and urban mycology. The Bryologist 109: 516–526. Wedin M, Döring H & Gilenstam G (2004) Saprotrophy and lichenization as options for the same fungal species on different substrata: Environmental plasticity and fungal lifestyles in the Stictis-Conotrema complex. New Phytologist 164: 459–465. Werth S, Wagner HH, Gugerli F, Holderegger R, Csencsics D, Kalwij JM & Scheidegger C (2006a) Quantifying dispersal and establishment limitation in a population of an epiphytic lichen. Ecology 87: 2037–2046.

Werth S, Wagner HH, Holderegger R, Kalwij JM & Scheidegger C (2006b) Effect of disturbances on the genetic diversity of an old-forest associated lichen. Molecular Ecology 15: 911–921. Wielgolaski FE (1975) Biological indicators on pollution. Urban Ecology 1: 63–79. Zalasiewicz J, Williams M, Smith A, Barry TL, Coe AL, Bown PR, Brenchley P, Cantrill D, Gale A, Gibbard P, Gregory FJ, Hounslow MW, Kerr AC, Pearson P, Knox R, Powell J, Waters C, Marshall J, Oates M, Rawson P & Stone P (2008) Are we now living in the Anthropocene? Geological Society of America 2: 4–7. Öckinger E, Niklasson M & Nilsson SG (2005) Is local distribution of the epiphytic lichen Lobaria pulmonaria limited by dispersal capacity or habitat quality? Biodiversity and Conservation 14: 759–773.

PAPER I

Short Communication

Generation time estimated to be 25–30 years in Cliostomum corrugatum (Ach.) Fr.

Knowledge of spore to spore generation time semi-open forests or parklands (Thor & is extremely important for several reasons. As Arvidsson 1999). It occurs mainly on the flat it is the shortest generation time, it indicates outer parts of the bark but is occasionally the maximum nucleotide substitution rate found on the coarse bark of other deciduous over time and provides a rate limit for the trees and on decorticated stumps, old wood, evolution of a species. In population genetics and twigs of Picea abies L. In northern most calculations involving time use ‘genera- Europe it is rare and red-listed, for example, tions’ as the unit of measurement and in in Sweden, Near Threatened; in Norway, order to convert these ‘generations’ into Critically Endangered, and in Finland, Near ‘years’, knowledge of generation time is Threatened (Lättman et al. 2009). Cliosto- needed but rarely available. Knowledge of mum corrugatum produces sexual spores but generation time may also be essential for not vegetative diaspores, such as soredia or conservation purposes and assessments of isidia, and we assume that new thalli estab- migration history. This knowledge also lish from sexually produced spores. Based on makes it possible to estimate both the age of a results of Lättman et al. (2009) we assume no population and also to determine to what dispersal limitation at the spatial scale extent a population represents the genetic studied. Suitable habitats, i.e. old oaks will diversity of a species (Rosenberg & Nordborg probably be colonized as soon as their bark is 2002). In this paper we present a method for coarse enough to be a suitable habitat. assessing generation length for lichens using Cliostomum corrugatum (Ach.) Fr. as an On a site in the province of Östergötland, (south- example. This lichen was selected for inves- central Sweden) all Quercus robur trunks >200 cm cir- tigation because it is restricted to forests with cumference were examined in April 2006 from ground level to 2 m in a search for Cliostomum corrugatum. long temporal continuity (Lättman et al. Preliminary studies had shown a very low occurrence of 2009) and information on generation time is C. corrugatum on trees with a circumference < 200 cm in essential to estimate the rate of dispersal at agreement with later studies (Johansson et al. 2009; the landscape level. Ranius et al. 2008). The occurrence and the size of the largest C. corrugatum thalli with apothecia and the size of Cliostomum corrugatum is an epiphytic the largest apothecia were recorded. Data for the largest crustose lichen with distinctive morphology thallus on each trunk was used to determine the smallest and clearly delimited habitats making it rela- trunk size colonized by C. corrugatum. The relationship tively easy to find and identify in the field. It between the square root of the thallus area and trunk has a greyish thallus which bears conspicu- circumference was examined using linear regression analysis in R 2.7.2 (R development core team, 2005) ous, light yellow to light brown, apothecia, and the average circumference at the time of spore 0·5–1·2 mm wide, as well as black pycnidia, germination was determined using the point at which 0·2–0·5 mm wide. Its geographical distri- the regression line intercepts the x-axis. The minimum bution in Sweden largely reflects that of circumference value was compared to that of the smallest trunk with fertile lichens. The time (in years) Quercus robur L. as it is chiefly confined to the between spore germination and spore production coarse bark of old trunks in old stands of was subsequently calculated by dividing the difference this tree species, often in relatively dry and (in mm) between this circumference value and the 558 THE LICHENOLOGIST Vol. 41 circumference of the thinnest trunk carrying lichen by 2 multiplied by the annual radial growth. Carlsson (2004) found an annual radial growth of 1·6–2·4 mm of oaks at a single location with a circumference of 200 cm. Compared to this Carbonnier (1975) gives a value of 2·4 mm on good soils in southern Swe- den and an assessment based on data in Tree-ring Data- base, NOOA Paleoclimatology (http://www.ncdc.noaa. gov/paleo/paleo.html) gives a value of 1.6 mm, with a standard deviation of 1.0 mm (28088 measurements from 9 data sets referring to Europe north of latitude 54 after the year 1700). As the soil conditions on the studied site are good it seems reasonable to assume an annual radial growth of 2 mm.

Seventeen large Quercus robur trunks were F. 1. The relationship of the size of the largest thallus examined. Trunks with a circumference of Cliostomum corrugatum on each Quercus robur and the >350 cm were not included because cracks in circumference of the trunk. the bark that limit the growth of thalli in some directions, are most frequent on older trees. In addition, as the life span of lichens is Small thalli with well-developed apothecia limited, older trees may support thalli repre- were not encountered, indicating a corre- senting later colonizations. Also, the com- lation between thallus size and formation of paratively small thallus (40 cm2) on a trunk apothecia. Further, as apothecia are absent with a perimeter of 280 cm was omitted as it on the smallest trees this also indicates rela- probably represents a late colonization event. tively rapid colonization when the trees The linear regression analysis gave a mini- develop a suitable bark structure for coloniz- mum value for oak circumference at which ation. All these observations suggest that it colonization by C. corrugatum occurs of would be appropriate to apply the results to a 215 ± 16 cm, P=0·001, r2=0·826 (Fig. 1). larger geographic area. The smallest circumference of trunks with The method described here uses observa- fertile specimens of C. corrugatum was 240 tions on the smallest/youngest tree with ster- cm. During the time from colonization to ile thalli, thus identifying the shortest formation of apothecia an oak will increase possible generation length, rather than aver- its perimeter by 25 ± 16 cm. The time span age spore to spore generation length which is from spore to spore may thus be calculated to probably signiﬁcantly longer. Within popu- approximately 20 years (250/(2 × 2) = 20). lation genetics the latter value would be more If the standard error of the average circum- useful. It seems reasonable to assume an ference (16 cm) is taken into account the average length of about 25–30 years. On the time span will be 7–25 years. other hand, a minimum value also indicates The observations supported the assump- possibilities. It shows the maximum rate of tion of a correlation between thallus age, as evolutionary processes and in this case also indicated by thallus size, and the formation indicates a fairly high average rate as all thalli of apothecia because all fertile thalli were at on bigger oaks are fertile. least 30 cm2. There was also a correlation The importance of other means of disper- between tree age and occurrence of thalli sal, for example, vegetative diaspores, should indicating good dispersal properties to ap- be considered. Since no soredia or isidia have propriate substrata of a certain age, although been reported in C. corrugatum, it seems one tree was thought to have been colonized reasonable to assume meiospore dispersal as late. Of course, other factors, such as micro- all larger thalli possess apothecia. There is of climatic conditions might affect dispersal so course a possibility of production of non- that speciﬁc values (e.g. of minimum circum- viable spores in the apothecia observed but it ference) might only be relevant to the site. seems reasonable to assume functionality. 2009 Short Communication 559

The pycnidia of the thallus produce mi- Johansson, V., Bergman, K.-O., Lättman, H. & totic conidia that may also function as sper- Milberg, P. (2009) Tree and quality preferences of six epiphytic lichens growing on oaks in south east- matia, fungal diaspores, or maybe both, as ern Sweden. Annales Botanici Fennici (in press). has been known from studies since the 19th Lättman, H., Lindblom, L., Mattsson, J.-E., Millberg, century (Vobis 1980). As far as we know P., Skage, M. & Ekman, S. (2009) Estimating the none of these studies concerns C. corrugatum, dispersal of the rare lichen Cliostomum corrugatum. but the possibility of dispersal by pycnoco- Biological Conservation (in press). R development core team. (2005) R: a language and nidia should not be neglected. If pycnoco- environment for statistical computing. http://cran.r- nidia act as dispersal units the shortest project.org generation time from spore to spore may be Ranius, T., Johansson, P., Berg, N. & Niklasson, M. underestimated by our method. (2008) The inﬂuence of tree age and microhabitat quality on the occurrence of crustose lichens associ- This study is part of the multi-disciplinary project “Eco- ated with old oaks. Journal of Vegetation Science 19: logical and Societal Systems in Interaction” a multidis- 653–662. ciplinary project aiming at identifying and describing Rosenberg, N. A. & Nordborg, M. (2002) Genealogical processes behind environmental and conservation poli- trees, coalescent theory and the analysis of genetic cies of public authorities at different levels in the Baltic polymorphisms. Nature Reviews Genetics 3: 380– Sea area. The project is hosted at and is a part of the 390. strategic development of environmental sciences at Thor, G. & Arvidsson, L. (1999) Swedish Red Data Book Södertörn University and is generously funded by of Lichens. Uppsala: ArtDatabanken, SLU. Östersjöstiftelsen. Economic support from the regional Vobis, G. (1980) Bau und Entwicklung der Flechten- county administration (Stockholm läns landsting) is also Pycnidien und ihrer Conidien. Bibliotheca Licheno- acknowledged. The leader of the course Field Studies of logica 14. the Biological program of Södertörn University, Mikael Lönn, made student participation and inclusion of this Håkan Lättman, Anneli Brand, project in the course possible. Finally, we are also grate- ful to Jörg Brunet for help with references and to Stefan Johanna Hedlund, Mikael Krikorev, Ekman for the calculations of the average annual growth Niklas Olsson, Alexandra Robeck, from NOOA data. Together with two unknown referees, Fredrik Rönnmark and the latter also contributed useful comments which Jan-Eric Mattsson greatly improved the manuscript. Finally we thank Anthony Wright for linguistic revision of the text. H. Lättman: School of Life Sciences, Södertörn Univer-  R sity, SE-141 89 Huddinge, Sweden and . IFM Division Carbonnier, C. (1975) Produktionen i kulturbestånd av of Ecology, Linköping University, SE-581 83 ek i södra Sverige. Studie Forestalia Suecica 125. Linköping, Sweden. Skogshögskolan, Stockholm. A. Brand, J. Hedlund, M. Krikorev, N. Olsson, A. Carlsson, J. (2004) Betydelsen av ålder och tillväxt hos Robeck, F. Rönnmark and J.-E. Mattsson (correspond- ek (Quercus robur) för förekomst av rödlistade arter i ing author): School of Life Sciences, Södertörn Univer- Östergötland. LiU-IFM-Biol-Ex-1137. Student sity, SE-141 89 Huddinge, Sweden. Email: jan-eric. dissertation, Linköping University. [email protected]

PAPER II

Biological Conservation 142 (2009) 1870–1878

Contents lists available at ScienceDirect

Biological Conservation

journal homepage: www.elsevier.com/locate/biocon

Estimating the dispersal capacity of the rare lichen Cliostomum corrugatum

Håkan Lättman a,b,LouiseLindblomc,f, Jan-Eric Mattsson a, Per Milberg b,g, Morten Skage c,d,StefanEkmanc,e,* a School of Life Sciences, Södertörn University College, SE-181 49 Huddinge, Sweden b IFM Division of Ecology, Linköping University, SE-581 83 Linköping, Sweden c Department of Biology, University of Bergen, P.O. Box 7800, NO-5020 Bergen, Norway d Department of Biology, University of Oslo, P.O. Box 1066, NO-0316 Oslo, Norway e Museum of Evolution, Evolutionary Biology Centre, Uppsala University, Norbyvägen 16, SE-752 36 Uppsala, Sweden f Museum of Natural History, University of Bergen, P.O. Box 7800, NO-5020 Bergen, Norway g Department of Crop Production Ecology, SLU, Box 7043, 750 07 Uppsala, Sweden article info abstract

Article history: The objective of this study was to estimate the dispersal rate in an organism assumed to be confined to Received 31 October 2008 tree stands with unbroken continuity. We used the lichen-forming ascomycete Cliostomum corrugatum, Received in revised form 6 March 2009 which is largely confined to old oak stands. Five populations, with pairwise distances ranging from 6.5 Accepted 21 March 2009 to 83 km, were sampled in Östergötland, south-eastern Sweden. DNA sequence data from an intron in Available online 5 May 2009 the small subunit nuclear ribosomal RNA gene was obtained from 85 samples. Nearly all molecular variance (99.6%) was found within populations and there were no signs of isolation-by-distance. The abso- Keywords: lute number of immigrants per population per generation (estimated to 30 years), inferred by Bayesian Dispersal MCMC, was found to be between 1 and 5. Altogether, evidence suggests abundant gene flow in the history Establishment Ecological continuity of our sample. A simulation procedure demonstrated that we cannot know whether effective dispersal is Old-growth forests ongoing or if it ceased at the time when oaks started to decrease dramatically around 400 years BP. How- Quercus ever, a scenario where effective dispersal ceased already at the time when the postglacial reinvasion of Ascomycete oak had reached the region around 6000 years BP is unlikely. Vegetation history suggests that the habitat of C. corrugatum was patchily distributed in the landscape since the early Holocene. Combined with the high dispersal rate estimate, this suggests that the species has been successful at frequently crossing distances of at least several kilometres and possibly that it has primarily been limited by the availability of habitat rather than by dispersal. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction has been criticized for often being vaguely defined on spatial and temporal scales (Gauslaa and Ohlson, 1997; Nordén and Appelq- Many organisms, belonging to a variety of taxonomic groups vist, 2001; Sverdrup-Thygeson and Lindenmayer, 2003). In most like wood-decay fungi, lichens, bryophytes, vascular plants, and in- real applications, EC implicitly refers to the forest stand level. sects, seem to be confined to habitat patches that have persisted Using indicator species to assess EC is also problematic, for several presumably unchanged over an extended period of time (Berg reasons: the group of species claimed to indicate EC probably in- et al., 1994; see Nordén and Appelqvist, 2001, p. 781 for references cludes species with poor dispersal capacity as well as species with to specific organism groups). The concept ‘ecological continuity’ particular microhabitat requirements. Their dispersal capacity and (EC), coined by Rose (1974), has been used to refer to the tempo- habitat requirements are often poorly or not at all understood rally unbroken continuity of such habitat, often assumed to be pri- (Nordén and Appelqvist, 2001; Rolstad et al., 2002). Forest history meval or old-growth forests. It has also been proposed that certain and dynamics is poorly known and often judged from anecdotal species can be used as indicators of EC (e.g., Rose, 1974; Tibell, evidence (Rolstad et al., 2002). The spatial scale at which indicators 1992; Selva, 1994, 2003; Kuusinen, 1996; Økland, 1996; Lindblad, are assumed to work is usually undefined, the spatial precision of 1998) when historical data are difficult to obtain. The EC concept the indicator species being dependent on dispersal capacity (Rols- tad et al., 2002; Sverdrup-Thygeson and Lindenmayer, 2003; Kal- * Corresponding author. Address: Museum of Evolution, Evolutionary Biology Centre, Uppsala University, Norbyvägen 16, SE-752 36 Uppsala, Sweden. Tel.: +46 wij et al., 2005). Finally, it cannot be uncritically assumed that 18 471 28 21; fax: +46 471 27 94. species richness or the number and abundance of rare species is E-mail addresses: [email protected] (H. Lättman), [email protected] strictly positively correlated with temporal continuity (Ohlson (L. Lindblom), [email protected] (J.-E. Mattsson), [email protected] et al., 1997; Fenton and Bergeron, 2008; Lõhmus and Lõhmus, (P. Milberg), [email protected] (M. Skage), stefan.ekman@evolmuseum. uu.se (S. Ekman). 2008). However, there is ample evidence that some species are

0006-3207/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocon.2009.03.026 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 1871 indeed confined to EC habitats and that red-listed species or cer- C. corrugatum, like all ascomycetes except some yeasts, is pre- tain taxonomic groups are represented by more species in older- sumed to have a dominantly haploid life cycle. Dikaryotic and dip- than-average and unmanaged forests compared to younger and loid stages appear only as very small amounts of hyphae confined managed ones (e.g., Gustafsson and Hallingbäck, 1988; Goward, to the apothecia. C. corrugatum is rare in northern Europe and 1994; Fritz and Larsson, 1996; Spence et al., 1996; Nilsson and Bar- red-listed in, e.g., Sweden (Near Threatened; Gärdenfors, 2005), anowski, 1997; Martikainen et al., 1999, 2000; Uliczka and Angel- Norway (Critically Endangered; Kålås et al., 2006), Denmark (‘Vul- stam, 1999; Hedenås and Ericson, 2000; Cameron, 2002; Penttilä nerable’ but not evaluated according to recent IUCN criteria; Stol- et al., 2004; Tikkanen et al., 2006; Rivas Plata et al., 2008; Fritz tze and Pihl, 1998), Finland (Near Threatened; Rassi et al., 2001), et al., 2008). Germany (‘Critically Endangered’ but not evaluated according to A central question is why some organisms are confined to sites recent IUCN criteria; Ludwig and Schnittler, 1996), and the United with long temporal continuity, or at least prefer them. There are Kingdom (Vulnerable; Woods and Coppins, 2003). The lichen is two main, not mutually exclusive, explanations for this phenome- small but its distinctive morphology and habitat makes it relatively non: (1) limitation by dispersal, dispersal primarily taking place easy to detect in the field. only at very short distances and being virtually absent at larger distances, and (2) limitation by habitat availability. If dispersal is the 2.2. Sample area and sites primary limiting factor, occurrence in EC habitats is likely to be of a relictual nature, whereas this is not necessarily the case when hab- Samples of C. corrugatum were collected on tree trunks of Q. ro- itat availability is the primary limiting factor. Limitation by habitat bur up to 2 m above the ground between January 5 and February 4 availability is indeed a realistic phenomenon, as the structural 2005 at five sites in central Östergötland, south-eastern Sweden complexity and consequently the number of microhabitats in a for- (Fig. 1; Table 1). Quercus colonized this area approximately est has been shown to be higher under old-growth conditions 7000 years BP (Brewer et al., 2002). Central Östergötland supports (Zenner, 2004). In both cases, preserving currently occupied one of the highest densities of old oaks in Sweden. Altitudinal dif- patches and creating new habitat may be necessary for the long- ferences in this region are small and the soil is fertile and consists term preservation of an organism restricted to EC habitats. How- of sedimentary silt and clay particles deposited during or after the ever, the distances that can be allowed in this network of currently end of the last glaciation. The ice retreated from this region and potentially occupied patches depend crucially on the dispersal 10,000 years ago (Lundqvist, 1998), but due to land depression, capabilities of the organism in question. Unfortunately, a scientifi- all sites were initially below sea level. Pairwise distances between cally based knowledge of effective dispersal (i.e. dispersal followed sites were evenly distributed and ranged from 6.5 to 83 km. The by establishment) at the landscape level is currently missing in smallest tree trunk inhabited by C. corrugatum, out of all investi- most species. This includes also species restricted to EC habitats, gated, was 0.65 m in diameter at breast height (dbh). Johansson many of which are also red-listed and in need of proper manage- et al. (in press) demonstrated a positive correlation between the ment for long-term persistence. probability of occurrence of C. corrugatum and tree trunk size of The objective of this study was to estimate the rate of dispersal Q. robur. Owing to difficulties determining the boundaries between at the landscape level in an organism restricted to forests with long adjacent lichen thalli, only one sample was taken per tree to avoid temporal continuity. We selected the lichen-forming ascomycete the risk of sampling the same individual twice. The minimum Cliostomum corrugatum (Ach.) Fr. as a study species. C. corrugatum number of samples per site was set to 15. is a rare lichen that is largely confined to very old stands of Quercus robur (Berg et al., 2002) and has been suggested to be an indicator 2.3. Laboratory methods and sequence editing of EC (Arup, 1997). We used DNA sequence data from a nuclear marker, combined with a population genetics approach, to address Methods for DNA extraction, PCR amplification, and sequencing the question at hand. We are not aware of any previous investiga- followed Lindblom and Ekman (2006), except that we mainly used tions of genetic variation at small spatial scales in a crustose (crust- apothecial tissue (or tissue from the thallus or pycnidia, when apo- forming) lichen. thecia were not available). We first targeted the internal tran- scribed spacer region (ITS) and the intergenic spacer (IGS) of the 2. Materials and methods nuclear ribosomal DNA. Because of low variability, we turned our attention to the group 1 intron situated between positions 1516 2.1. Study species and 1517 at the end of the small subunit (SSU) of the nuclear ribosomal RNA gene (Gargas et al., 1995). This region was amplified The epiphytic crustose lichen C. corrugatum (Ach.) Fr. (Ascomy- using the forward primer ITS1F (White et al., 1990), situated at cota, Lecanoromycetes, Lecanorales, Ramalinaceae) possesses a the end of the SSU but upstream of the intron site, and the newly greyish thallus containing a green alga as its photosynthesizing designed reverse primer ITS1-Cc1-R, situated in the first part of symbiont. The thallus bears conspicuous, light yellow to light ITS1. The new primer was designed because the widely used com- brown, 0.5–1.2 mm wide apothecia (fruiting bodies producing bination of ITS1F and ITS4 (Gardes and Bruns, 1993) to amplify the putatively meiotic ascospores) as well as black, 0.2–0.5 mm wide entire ITS region in many cases either failed or resulted in multiple pycnidia (producing mitotic conidia that may function as sperma- PCR products. The sequence of the new primer is 50-ATG GTA AGG tia, fungal diaspores, or both) (Thor and Arvidsson, 1999). In Swe- TAA TCA CAG GGT GTA-30. The amplification, PCR clean-up, and den, its geographical distribution largely follows that of Quercus. Its sequencing procedures were identical to the ones used by Lindb- primary habitat is coarse bark of old trunks of Q. robur in relatively lom and Ekman (2006) for the ITS region. Sequencing was per- dry and semi-open forests or parklands (Thor and Arvidsson, formed using both PCR primers. The technique used here, 1999), mainly on the flat terminal parts of the bark structure and particularly when using overlapping forward and reverse reads, not on either side of the cracks in the bark (pers. obs.). C. corruga- has by far the lowest error rate of any currently used sequencing tum has occasionally been encountered on the coarse bark of other procedure (Johnson and Slatkin, 2007). Only sequences from reads deciduous trees as well as on wood of decorticated stumps, old with a low level of noise relative to the signal were processed for wood structures, and twigs of Picea abies (Thor and Arvidsson, further analysis. Resulting sequences were manually edited and 1999). No vegetative diaspores (soredia or isidia) containing tissue aligned using BioEdit version 7.0.5.3 (Hall, 1999). The identity of from both symbionts are produced (Thor and Arvidsson, 1999). the sequences obtained with the new primer pair was confirmed 1872 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878

Fig. 1. The ﬁve sample sites of Cliostomum corrugatum in the province of Östergötland, southern Sweden, indicated with black dots.

Table 1 Geographical location and short description of the ﬁve sites where Cliostomum corrugatum was sampled in Östergötland, south-eastern Sweden.

Site Latitude/longitude Description Bjärka-Säby 58°16029.9}N, 15°440 24.0}E Parkland mixed with arable land Orräng 58°170 32.2}N, 15°510 19.9}E Grassland with patches of open oak stands Solberga 58°200 19.2}N, 15°110 49.8}E Forest adjacent to river Bråborg 58°360 56.5}N, 16°220 00.4}E Mixed forest in NE facing slope to Baltic sea Stegeborg 58°260 17.2}N, 16°350 58.0}E Small patchy tree stands in agricultural landscape

by comparing with sequences initially generated with ITS1F and linearized FST) and pair-wise values of the natural logarithms of ITS4. The ITS part of these sequences were in turn subjected to a geographic distance between populations (Rousset, 1997) was BLAST (megablast) search against the nr/nt database at the Na- used to check for the presence of isolation-by-distance (Wright, tional Center for Biotechnology Information (http:// 1943), i.e. a spatial aggregation of genetically similar individuals. www.ncbi.nlm.nih.gov) on 22 October 2008. The top 69 hits Significance was assessed by 10,000 permutations. Wherever against the amplified ITS were taxa in the Ramalinaceae (expecta- applicable, Tamura distances were used, as they most closely cor- tion scores 63 10137 and query coverage P72%). Each haplo- responded to HKY85. In all calculation involving Arlequin, indels type sequence was submitted to GenBank and given accession were given weight 1 (i.e., they were counted as the ‘fifth character numbers EU218541-EU218551. state’). The main part of our analysis, however, consisted of a direct 2.4. Statistical analyses estimation of population parameters, including dispersal, using Bayesian Markov chain Monte Carlo (MCMC), as implemented in In order to create a simple overview of haplotype relationships the software LAMARC 2.0.2 (Kuhner et al., 2005; Kuhner, 2006; and frequencies present in the sample, we constructed a haplotype Beerli, 2006; Kuhner and Smith, 2007). In its official terminology, network using the 95% statistical parsimony criterion as imple- LAMARC measures ‘migration’. This word is usually interpreted mented in the software TCS version 1.13 (Clement et al., 2000). A as one or more individuals contributing to Ne (the effective popu- likelihood model was fitted to the data using a hierarchical likeli- lation size) in one population leaving that population and entering hood ratio test as implemented in MODELTEST version 3.7 (Posada a new population. In our case, individuals are sessile and dispersal and Crandall, 1998). The best model reported was HKY85 without is expected to occur via ascospores or perhaps conidia, without any rate heterogeneity. Using Arlequin version 3.1 (Schneider et al., individuals contributing to Ne ever moving between populations. 2000), we first conducted a Ewens–Watterson–Slatkin exact test However, migration in the true sense is not a prerequisite for pop- of selective neutrality in the entire sample (Slatkin, 1994, 1996). ulation parameter estimates by LAMARC to be valid for dispersing The null distribution under neutrality was obtained by simulating sessile organisms (Peter Beerli and Lucian Smith pers. comm. the null hypothesis 10,000 times (Stewart, 1977). We conducted an 2007). This is because the model in LAMARC decouples migration analysis of molecular variance (AMOVA) (Excoffier et al., 1992), from size fluctuations, and anyway restores population size to its assessing significance by 10,000 permutations. A Mantel matrix original size when individuals migrate. Bayesian MCMC has the correlation test between pair-wise values of FST/(1 FST) (Slatkin’s advantage of explicitly handling uncertainty in parameter H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 1873 estimates, which is all the more important when the amount of 500,000 generations. Using this scheme, we arrived at a model data, as in this case, is small and uncertainty about estimates can treating size fluctuations as absent (set to zero), and h and migra- consequently be expected to be large. In the Bayesian MCMC, the tion as equal across all populations. This does not mean that the F84 model was used, because this is the model implemented in true scenario was this simple, only that the current data contained LAMARC that most closely corresponds to HKY85. The transition no information to support a more complex model. Final estimates to transversion ratio, which is treated as fixed by LAMARC, was cal- of h and migration were obtained by summing results across three culated under maximum likelihood using PAUP* 4.0b10 (Swofford, identical runs, each 8.5 million generations in length and discard- 2003) on a collapsed version of our dataset. For this estimate (ra- ing the initial 500,000 generations as burn-in. In the haploid case, tio = 7.01), empirical nucleotide frequencies were used under the h =2Nel, where l is the per site mutation rate per generation. F84 likelihood model. A void population was added to account Migration is measured as M = m/l, where m is the proportion of for ‘ghost populations’ (Slatkin, 2005) and unsampled populations, immigrants into a population per generation. Software was written following the recommendation by Kuhner (2003) and Beerli in RealBasic in order to extract the posterior distribution of

(2004). Prior distributions were set to uniform in linear space on Nem = hM/2, the absolute number of immigrants into a population the interval [103,104] for migration and uniform in logarithmic per generation, from the joint distribution of h and M. The uni- space on the interval [108, 10] for the population mutation rate modal posterior probability distributions were finally transformed h (corresponding to the extreme lower and upper boundaries al- into 95% equal-tail credible intervals by removing 2.5% of the total lowed by the software). These priors essentially meant assuming probability at each end of the posterior probability distribution. that small and large values of migration are equally likely a priori, Finally, we performed a simulation study using SimCoal 2.1.2 whereas small values of h are more likely than large values a priori. (Laval and Excoffier, 2004), with the purpose of evaluating the The prior distribution of population size fluctuations, when appli- temporal information contained in the migration rate estimates cable, was uniform on the interval [500, 1000] in linear space obtained by LAMARC under different demographic histories. LAM- (LAMARC allows only linear priors for size fluctuations). The pro- ARC assumes migration rates to be constant over time, from the posal rate for population parameters was set to ten times the pro- present to coalescence of the sample, but this is rarely the case posal rate for genealogy rearrangements. This was necessary to in real populations. Therefore, estimated migration rates might alleviate problems with poor effective sample sizes of population be averages over recent evolutionary time, with limited informa- parameter estimates, particularly h. Preliminary runs with adaptive tion about ongoing migration. We wanted to answer two specific heating indicated that Metropolis coupling, the use of heated questions: can we separate between a model with and a model chains, was unnecessary. However, we discovered that population without migration after the start of the dramatic decrease of oaks parameter estimates were more precisely repeatable when using a few hundred years ago? Similarly, can we separate between a heating. Consequently, all subsequent runs were conducted with model with and a model without migration after the immigration one heated chain at a temperature of 1.1, allowing information of oaks to Östergötland 6000 years ago? We assumed a generation from the heated chain to be swapped into the cold chain every time of 30 years for C. corrugatum, based on a combination of the 10 generations. LAMARC by default treats population parameters demography and phenology of the lichen as well as the growth rate as unconstrained. This means that when h, migration, and size fluc- of the inhabited oak trees (Lättman et al., unpublished results). The tuation are estimated jointly, there is one h and one size fluctuation demographic history was divided into three phases (in backward parameter for each population as well as one migration parameter time): (1) the first 13 generations, corresponding to the time dur- in each direction between each pair of populations (migration is ing which old oaks decreased dramatically in the region (Eliasson treated as asymmetric). With five sampled populations and one and Nilsson, 2002); (2) generations 14–200, corresponding to the void population, as in our case, this amounts to a large total num- period limited by the immigration of oak to the region; (3) gener- ber of parameters that may not be supported by the data. There- ations 201 until coalescence, corresponding to the history of the fore, we performed tests of model adequacy, which involved the sample during which oaks had not yet immigrated into the region. use of Bayes factor (Kass and Raftery, 1995) to compare models We assumed the most probable estimates of h and M obtained based on the harmonic mean estimator (Newton et al., 1994). from the LAMARC analysis and translated them into Ne and m Starting with simple models, we added parameters only if there using an estimate of l. The (short-term) pedigree rate of mutation, was ‘strong’ support (Kass and Raftery, 1995, p. 777) for a more which should not be confused with the (long-term) phylogenetic complex model, i.e. if twice the difference in harmonic mean ln substitution rate (Howell et al., 2003; Ho et al., 2005, 2007), has likelihood exceeded six. For h, we tested a model where all values been found to be approximately 1–2 108 for several organisms are identical against an unconstrained model. Size fluctuation was (Drake et al., 1998; Nachman and Crowell, 2000; Denver et al., either set to zero, treated as equal across all populations, or uncon- 2004). However, Lutzoni and Pagel (1997) reported up to 10-fold strained. Migration was set to zero, treated as equal across all pop- higher mutation rates in lichen-forming fungi compared to non- ulations, as different between population pairs but symmetric, or mutualistic relatives that could not be ascribed to significantly re- unconstrained (asymmetric between population pairs). In all cases, laxed negative selection. Therefore, we settled for l =107 per site the void population was treated as unconstrained, because we do per generation. The following scenarios were simulated (subscripts not know how many real-world populations it represents. Because of m refer to the three phases described above): (A)

LAMARC 2.0.2 only reports the data ln likelihood for the last sam- m1 = m2 = m3 = 0.001 (migration has remained constant and is ple of the MCMC chain, we created a workaround by splitting the ongoing), (B) m1 =0,m2 = m3 = 0.001 (migration ceased 13 genera- analysis into several consecutive ‘initial chains’, the likelihood tions ago in connection with a decrease in available oak habitat), being reported at the end of each such chain. We allowed 42,500 and (C) m1 = m2 =0,m3 = 0.002 (migration ceased once the postgla- initial chains, each 200 generations long, i.e. a total of 8.5 million cial expansion of oak reached the region). In scenario C, we dou- generations. Software was written in RealBasic to extract data ln bled the migration rate in phase 3 in order to maintain an likelihoods from the output (‘outsumfile’). Likelihoods were subse- approximate average migration rate of 0.001 over the entire time quently imported in a Microsoft Excel spreadsheet, likelihoods span (assuming that coalescence occurred during the bottleneck plotted, and the harmonic mean ln likelihood calculated across caused by the latest glaciation). In all three scenarios, we simulated the stationary phase of the run. Plotting ln likelihood against gen- a 20-fold increase in population size from phase 1 to phase 2 (cor- eration indicated that the true burn-in was in the order of responding to a 95% population reduction in forward time). Two 100,000–150,000 generations, but we anyway discarded the first hundred data sets were simulated per scenario. The transition to 1874 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 transversion rate used in LAMARC was maintained in SimCoal. Each data set simulated by SimCoal was analyzed using Arlequin 3.1, and three types of summary statistics were collected in order to compare them with observed values: (1) the proportion of within-population variation inferred by an AMOVA, (2) Tajima’s D (Taj- ima, 1989) averaged over populations, and (3) the average number of polymorphic (segregating) sites per population. Ninety five percentage ranges were constructed by removing the five most extreme values at each end of the distributions.

3. Results

We found three IGS haplotypes, two of which were represented by a single individual each and the third by 79 individuals. ITS proved to be difficult to amplify and sequence. Eight samples, which were successfully sequenced, displayed no variation at all. We found the variation in IGS and ITS to be insufficient and discarded this data in subsequent statistical analyses. The SSU intron, on the other hand, was represented by 11 haplotypes (Table 2). Out of the 96 samples, 85 were successfully extracted, the SSU intron amplified, sequenced, and consequently included in the statistical analyses. The SSU intron length varied from 612 to 613 nucleotides, and the resulting alignment was 614 positions including gaps. Ten positions were variable. A haplotype network is presented in Fig. 2. This shows that the 11 haplotypes are separated by single mutational steps and that the two common haplotypes (represented by 30 and 46 thalli, respectively) are internal and Fig. 2. Unrooted haplotype network for Cliostomum corrugatum with the two hence presumably older than the infrequent terminal haplotypes. common haplotypes 1 (n = 30) and 2 (n = 46) in the centre. Remaining terminal The neutrality test indicated no deviation from neutral conditions haplotypes are represented by one thallus each. One mutational step between (all populations p = 1.00, Bjärka-Säby p = 0.79, Orräng p = 0.68, Sol- haplotypes is represented by a line. berga p = 0.84, Bråborg p = 0.38, and Stegeborg p = 0.40). The AMO- VA indicated that 0.4% of the variance is between populations and 99.6% within populations. The reported fixation index (U = 0.004) Table 3 ST Population parameter estimates obtained using the coalescent in a Bayesian MCMC was not significant (p = 0.35). Consequently, the AMOVA provided framework, as implemented in LAMARC 2.0.2. h =2Nel, M = m/l, and Nem = hM/2, no evidence of significant neutral differentiation among popula- where Ne is the effective population size, l the per site mutation rate per generation, tions. The Mantel test revealed no indication of significant isola- and m the per generation proportion of immigrants into a population from another population. MPE = most probable estimate, corresponding to the mode of the tion-by-distance, the correlation between Slatkin’s linearized FST posterior probability distribution. 95% CI = 95% equal-tail credible interval. and the logarithm of geographic distance being non-significant

(p = 0.70). Estimates of h, M, and Nem obtained via LAMARC, includ- Parameter MPE 95% CI ing 95% equal-tail credible intervals, are reported in Table 3. The h 6.1 104 3.2 104–1.2 103 absolute number of successful establishments per generation per M 9541 4225–9988 population was estimated to be between 1 and 5, with the median Nem 1.9 1.1–4.8 of the posterior probability being two. With four other sampled populations, from which dispersal and establishment into a population can take place, the total number of successful establish- between 4 and 20 with the median at eight successful establishments from the sampled populations is four times higher, i.e. ments per generation from the other four populations. Dispersal

Table 2 Variable nucleotide sites in the alignment of 11 haplotypes of the position 1516 SSU intron. The alignment was 614 sites in length. GenBank accession numbers for each haplotype are indicated. Dots denote nucleotides that are identical to haplotype 1.

Haplotype Sequence and base number Origin 0011122455Bjärka-Säby Orräng Solberga Bråborg Stegeborg GenBank accession nos 2704569712 6931145163 1 GACCC–AAA–8 6 5 7 4 EU218541

2 –C– 7 9 10 7 13 EU218542

3 G – G C 1 EU218543

4 T –C– 1 EU218544

5 T – – 1 EU218545

6A T –C– 1 EU218546

7 T–C– 1 EU218547

8 C – 1 EU218548

9 G – C 1 EU218549

10 – G – 1 EU218550

11 G – – 1 EU218551 Total 18 17 18 15 17 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 1875 from unsampled populations, the number of which is unknown but gal cells have been shown to be viable after having passed through presumably rather large, comes in addition. Migration (M) may the gut of mites (Meier et al., 2002). Mites, in turn, could be carried have been underestimated, or at least truncated, because much over large distances with the help of mammals or birds. A number of the posterior probability accumulated right below the highest of insects have been shown to be faithful to the kind of oak trees upper limit allowed by LAMARC for that parameter. Table 4 ac- that C. corrugatum inhabits (Niklasson and Nilsson, 2005). counts for the simulation of three different demographic scenarios. An assumption of the Bayesian coalescent analysis was that dis- This simulation shows that the observed values of the proportion persal rates have remained approximately constant from the pres- of within-population variation, the average Tajima’s D across pop- ent to coalescence in backward time, otherwise inferred rates will ulations, and the average number of polymorphic sites across pop- reduce to averages over time. The simulation study, although a ulation is compatible with both a model of ongoing migration as simplistic picture of the real events, efficiently demonstrates that well as a model where migration ceased 13 generations ago. The we cannot separate between a model with constant and ongoing observed values are, however, incompatible with a model where dispersal from a model where migration ceased at the time when migration ceased 200 generations ago, because of the higher ex- oak habitat started to decrease dramatically around four centuries pected proportion of within-population variation. ago (Table 4, scenario A and B). In other words, we cannot know whether effective dispersal is ongoing or whether recent fragmen- 4. Discussion tation, owing to human influence on landscape characteristics, has caused connectivity between populations to decrease. On the other 4.1. Dispersal in C. corrugatum hand, a scenario where effective dispersal ceased already at the time when the postglacial reinvasion of oak had reached the region Knowledge about effective dispersal rates and dispersal dis- is implausible (Table 4, scenario C). tances are paramount to any scientifically-based conservation Could the high inferred dispersal rates be a consequence of high measure. Yet, such knowledge is unavailable for most organisms. connectivity in large and effectively continuous populations of C. We used DNA sequence data from an intron near the terminal corrugatum from the time of oak reinvasion until around 1600 AD? end of the nuclear small subunit ribosomal DNA to infer effective Current knowledge of the vegetation history of southern Sweden dispersal rates between populations of C. corrugatum. Other mark- allows some inferences about the dispersal capabilities of C. ers, ITS and IGS, failed to produce useful amounts of variation. The corrugatum, although we can say nothing about ongoing dispersal. limited amount of data available to us, a single gene, made it In the historic agricultural landscape of Sweden, we know that from imperative to apply analytical methods that reveal the uncertainty 1558 until 1830 oaks were considered state property to meet the in parameter estimates (in addition to methods that calculate point needs of timber for the navy. This royal decree was increasingly estimates). We handled uncertainty in our estimates by use of a being disregarded by peasants, and the 1825 reinventory of oaks Bayesian as well as a simulation approach. made some 30 years earlier disclosed an 80% reduction of timber Indirect estimates of dispersal rates and dispersal distances oaks during this short period of time (Eliasson and Nilsson, 2002). through a point estimate of population differentiation (AMOVA), Another inventory of oaks in Östergötland in 1813 demonstrated a point estimate of the correlation between interpopulational ge- that more than 80% of the oaks were found in the enclosed meadows netic and physical distances (Mantel test), as well as a Bayesian di- and fields surrounding the villages (Eliasson and Nilsson, 2002). The rect measure of dispersal (Table 3) all indicate that effective remainder of the oaks were found outside the village enclosures, dispersal between populations at this spatial scale has been sub- grazing intensity (and thereby the amount of sun-lit oaks) progres- stantial and without measurable restrictions. The Bayesian ap- sively decreasing with increasing distance from the villages. Before proach indicates that the most likely number of successful and after the 1558–1830 period, oaks were probably uncommon in- establishments per generation between the sampled populations side village enclosures. The Swedish system of a clear division be- is between 4 and 20. The number of immigrants needed to prevent tween areas inside and outside village enclosures has a tradition neutral divergence of populations has been suggested to be ca 5 that goes back at least 1000 years but probably as much as (Lacy, 1987), 1–10 (Mills and Allendorf, 1996), or more than 10 2000 years (Ekstam et al., 1988; Niklasson and Nilsson, 2005). Dur- (Vucetich and Waite, 2000) per generation. The very wide Bayesian ing this period, C. corrugatum habitat was probably patchily distrib- posterior distribution of dispersal rate also demonstrates that our uted, oaks occurring under semi-open and sun-lit conditions almost estimate is indeed associated with considerable uncertainty. How- exclusively being found in or near villages. Further back in time, prior ever, although it remains unknown exactly how high the rate of to the advent of agricultural landscape, human influence was pri- dispersal is at this spatial scale, the credible interval clearly ex- marily by cultivation of temporary clearings in the forest (Niklasson cludes low dispersal rates even though we used a uniform prior and Nilsson, 2005). Recent developments in paleoecology (Mitchell, distribution ranging from no dispersal at all to very high dispersal 2005; Birks, 2005) indicate that in the early and mid-Holocene, much rates. A possible explanation for the high rates of successful dis- of lowland Europe was covered by closed-canopy forests, contrary to persal is that dispersal is not as passive as one might think. Perhaps earlier suggestions involving wood-pastures kept open by megaher- dispersal in C. corrugatum is facilitated by winged insects carrying bivores (Vera, 2000). In closed-canopy forests, C. corrugatum would ascospores, conidia, or pieces of lichen thallus. Ascospores and al- have been restricted to steep, south- or west-facing slopes (Ek

Table 4 Median and 95% ranges of the proportion of within-population variation, average value of Tajima’s D per population, and the average number of polymorphic sites per population obtained when simulating three different demographic scenarios: A (constant and ongoing migration), B (migration ceased around the time when the oak habitat started to decrease dramatically around four centuries ago), and C (migration ceased once the postglacial reinvasion of oak had reached the region). Details of the simulation parameters are found in the text. The observed parameter values are included for comparison.

Proportion of within-population variation Tajima’s D (average per population) No. of polymorphic sites (average per population)

Observed 0.004 0.312 3.6 Scenario A 0.086 (0.010–0.230) 0.358 (0.690–1.717) 4.6 (1.0–12.2) Scenario B 0.086 (0.001–0.274) 0.363 (0.692–1.441) 4.7 (1.4–12.2) Scenario C 0.144 (0.014–0.307) 0.410 (0.615–1.306) 4.0 (1.0–12.6) 1876 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 et al., 1995) and lake and river edges. This type of habitat is likely to (Bhattacharya et al., 2002) and even interspecific horizontal transfer have been highly patchily distributed in the landscape. In conclu- (Martin et al., 2003; Simon et al., 2005) in the nuclear SSU rRNA gene sion, dispersal between patches suitable for C. corrugatum during has been implicated in a phylogenetic perspective. However our the last 6000 years in Östergötland must commonly have involved haplotype network (Fig. 2), which is typically star-shaped and sepa- crossing distances of at least several kilometres, even if oaks were rates the 11 haplotypes by single mutational steps, strongly indi- notably more common than in the present-day landscape of south- cates that horizontal transfer did not affect our study. Secondly, we ern Sweden. Indeed, inferred dispersal rates are high enough to sus- primarily used fruiting bodies (apothecia) for DNA extraction, be- pect that the postglacial occurrence of C. corrugatum was primarily cause extractions from the vegetative thallus were more likely to limited by the availability of habitat and not by dispersal. Limitation be troubled with contamination by other lichenized or non-liche- by habitat availability has been suggested also for lichens in stands of nized fungi present in the habitat. Apothecia contain very small aspen (Populus tremula), inferred from a combination of occupancy amounts of dikaryotic tissue as well as meiotic ascospores that could patterns, successional history, and stand characteristics (Hedenås potentially contain genetic material from another, presumably near- and Ericson, 2004). by, individual, the ‘father’. However, we did not experience problems with multiple mixed PCR products as evidenced by chromatograms 4.2. Dispersal in lichens – the current state of knowledge with superimposed base calls. We cannot say whether this means that C. corrugatum is homothallic (the haploid equivalent of self-fer- There is considerable disagreement in the literature concerning tilizing) or just that the amount of ‘father DNA’ was too small to be the ability of lichens to disperse and establish. Like in most other detected among the dominant ‘mother DNA’ in the vegetative hy- organisms, effective dispersal rates seem to be scale-dependent, phae making up the vast majority of the apothecial tissue. but this explains only part of the disagreement. Vegetative diaspore dispersal at short distances, up to a few hundred meters, has been Acknowledgements suggested to be effective, although several studies did not investigate the success rate of establishment (Armstrong, 1987, 1990; Tap- We thank Wala and Folke Danielssons stiftelse for financial sup- per, 1976; Heinken, 1999; Lorentsson and Mattsson, 1999). port, Nicklas Jansson at the County Administrative Board in Linkö- Dispersal limitation has been reported for tree-living lichens within ping for suggesting potential sample sites for C. corrugatum, and tree stands, between tree stands in close proximity, or up to a few members of the plant ecology group at Linköping University for kilometres apart (Dettki et al., 2000; Sillett et al., 2000; Hilmo and comments on the manuscript and David Lawrence for last spelling Såstad, 2001; Johansson and Ehrlen, 2003; Walser, 2004; Öckinger corrections. et al., 2005), as well as between populations of an asexual terrestrial lichen at a distance of up to a few kilometres (Cassie and Piercey- Normore, 2008). However, genetic studies of Xanthoria parietina References and Lobaria pulmonaria suggest otherwise: effective dispersal at this Armstrong, R.A., 1987. Dispersal in a population of the lichen Hypogymnia physodes. scale shows no sign of being restricted, although ascospores have Environmental and Experimental Botany 27, 357–363. been found to disperse, on average, at longer distances than heavier Armstrong, R.A., 1990. Dispersal, establishment and survival of soredia and vegetative diaspores (Lindblom and Ekman, 2006, 2007; Wagner fragments of the lichen, Hypogymnia physodes (L.) Nyl. New Phytologist 114, 239–245. et al., 2006; Werth et al., 2006a,b). At large spatial scales, populations Arup, U., 1997. Skoglig kontinuitet. In: Arup, U., Ekman, S., Kärnefelt, I., Mattsson, J.- being separated by hundreds of kilometres or more, genetic studies E. (Eds.), Skyddsvärda lavar i sydvästra Sverige, SBF-förlaget, Lund, pp. 92–95. of lichen populations revealed severe dispersal restrictions (Printzen Beerli, P., 2004. Effect of unsampled populations on the estimation of population sizes and migration rates between sampled populations. Molecular Ecology 13, et al., 2003; Palice and Printzen, 2004; Walser et al., 2005), whereas 827–836. studies relying on biogeographic patterns (Munoz et al., 2004), trap- Beerli, P., 2006. Comparison of Bayesian and maximum-likelihood inference of ping of lichen fragments in the atmosphere (Harmata and Olech, population genetic parameters. Bioinformatics 22, 341–345. Berg, A., Ehnström, B., Gustafsson, L., Hallingbäck, T., Jonsell, M., Weslien, J., 1994. 1991), or observations of lichen fragments on bird feet (Coppins Threatened plant, animal, and fungus species in Swedish forests – distribution and James, 1979) implicitly proposed effective dispersal to be fre- and habitat associations. Conservation Biology 8, 718–731. quent. Finally, the small size and weight of the ascospores has been Berg, A., Gärdenfors, U., Hallingbäck, T., Noren, M., 2002. Habitat preferences of red- taken as indirect evidence in favour of lichens being able to disperse listed fungi and bryophytes in woodland key habitats in southern Sweden – analyses of data from a national survey. Biodiversity and Conservation 11, ‘‘widely” (Nordén and Appelqvist, 2001). 1479–1503. What is the conservation message contained in our results? We Bhattacharya, D., Friedl, T., Helms, G., 2002. Vertical evolution and intragenic spread have inferred high rates of dispersal at landscape level in the his- of lichen-fungal group I introns. Journal of Molecular Evolution 55, 74–84. Birks, H.J.B., 2005. Mind the gap: how open were European primeval forests? Trends tory of a set of populations of a red-listed crustose lichen confined in Ecology and Evolution 20, 154–156. to EC habitats. The often-repeated claim that lichens confined to EC Brewer, S., Cheddadi, R., de Beaulieu, J.L., Reille, M., 2002. The spread of deciduous habitats are poor dispersers at more than very local scales may be a Quercus throughout Europe since the last glacial period. Forest Ecology and Management 156, 27–48. severe underestimate of their capabilities. As mentioned above, Cameron, R.P., 2002. Habitat associations of epiphytic lichens in managed and there are indications that some rare taxa restricted to EC forests unmanaged forest stands in Nova Scotia. Northeastern Naturalist 9, 27–46. are indeed poor dispersers at the landscape level, but that conclu- Cassie, D.M., Piercey-Normore, M.D., 2008. Dispersal in a sterile lichen-forming fungus, Thamnolia subuliformis (Ascomycotina: Icmadophilaceae). Botany 86, sion might not apply universally. Furthermore, there is a non-neg- 751–762. ligible risk that the species so far investigated are not Clement, M., Posada, D., Crandall, K.A., 2000. TCS: a computer program to estimate representative among the lichens, the majority of EC species being gene genealogies. Molecular Ecology 9, 1657–1659. Coppins, B.J., James, P.W., 1979. Birds and the dispersal of lichen propagules. crustose like C. corrugatum. Unfortunately, life-history traits might Lichenologist 11, 105–106. not help us to accurately predict dispersal ability (Johansson and Denver, D.R., Morris, K., Lynch, M., Thomas, W.K., 2004. High mutation rate and Ehrlen, 2003; Duminil et al., 2007). Our knowledge of the dispersal predominance of insertions in the Caenorhabditis elegans nuclear genome. capabilities of lichenized fungi therefore remains in its infancy. Nature 430, 679–682. Dettki, H., Klintberg, P., Esseen, P.A., 2000. Are epiphytic lichens in young forests limited by local dispersal? Ecoscience 7, 317–325. 4.3. Methodological issues Drake, J.W., Charlesworth, B., Charlesworth, D., Crow, J.F., 1998. Rates of spontaneous mutation. Genetics 148, 1667–1686. Duminil, J., Fineschi, S., Hampe, A., Jordano, P., Salvini, D., Vendramin, G.G., Petit, R.J., There are two methodological issues that need to be discussed 2007. Can population genetic structure be predicted from life-history traits? briefly. Firstly, lateral transfer of group I introns between positions American Naturalist 169, 662–672. H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878 1877

Ek, T., Wadstein, M., Johannesson, J., 1995. Varifrån kommer lavar knutna till gamla Laval, G., Excoffier, L., 2004. SIMCOAL 2.0: a program to simulate genomic diversity ekar? Svensk Botanisk Tidskrift 89, 335–343. over large recombining regions in a subdivided population with a complex Ekstam, U., Aronsson, M., Forshed, N., 1988. Ängar. LTs förlag, Stockholm. history. Bioinformatics 20, 2485–2487. Eliasson, P., Nilsson, S.G., 2002. ‘You should hate young oaks and young noblemen’. Lindblad, I., 1998. Wood-inhabiting fungi on fallen logs of Norway spruce: relations The environmental history of oaks in eighteenth- and nineteenth-century to forest management and substrate quality. Nordic Journal of Botany 18, 243– Sweden. Environmental History 7, 659–677. 255. Excoffier, L., Smouse, P.E., Quattro, J.M., 1992. Analysis of molecular variance Lindblom, L., Ekman, S., 2006. Genetic variation and population differentiation in inferred from metric distances among DNA haplotypes – application to human the lichen-forming ascomycete Xanthoria parietina on the island Storfosna, mitochondrial-DNA restriction data. Genetics 131, 479–491. central Norway. Molecular Ecology 15, 1545–1559. Fenton, N.J., Bergeron, Y., 2008. Does time or habitat make old-growth forests Lindblom, L., Ekman, S., 2007. New evidence corroborates population differentiation species rich? Bryophyte richness in boreal Picea mariana forests. Biological in Xanthoria parietina. Lichenologist 39, 259–271. Conservation 141, 1389–1399. Lõhmus, A., Lõhmus, P., 2008. First-generation forests are not necessary worse than Fritz, Ö., Gustafsson, L., Larsson, K., 2008. Does forest continuity matter in long-term managed forests for lichens and bryophytes. Restoration Ecology 16, conservation? – A study of epiphytic lichens and bryophytes in beech forests 231–239. of southern Sweden. Biological Conservation 141, 655–668. Lorentsson, S., Mattsson, J.-E., 1999. New reports of soredia dispersed by ants, Fritz, Ö., Larsson, K., 1996. Betydelsen av skoglig kontinuitet för rödlistade lavar. En Formica cunicularia. Lichenologist 31, 204–207. studie av halländsk bokskog. Svensk Botanisk Tidskrift 90, 241–262. Ludwig, G., Schnittler, M. (Eds.), 1996. Rote Liste gefährdeter pflanzen Deutschlands. Gärdenfors, U. (Ed.), 2005. Rödlistade arter i Sverige 2005. ArtDatabanken, SLU, Schriftenreihe für vegetationskunde 28, 1–744. Uppsala. Lundqvist, J., 1998. Isavsmältning och israndlinjer i Sverige. In: Andersen, S., Gardes, M., Bruns, T.D., 1993. ITS primers with enhanced specificity for Pedersen, S.S. (Eds.), Israndslinier i Norden. TemaNord, København, pp. 61–82. basidiomycetes – application to the identification of mycorrhizae and rusts. Lutzoni, F., Pagel, M., 1997. Accelerated evolution as a consequence of transitions to Molecular Ecology 2, 113–118. mutualism. Proceedings of the National Academy of Sciences of the USA 94, Gargas, A., Depriest, P.T., Grube, M., Tehler, A., 1995. Multiple origins of lichen 11422–11427. symbioses in fungi suggested by SSU rDNA phylogeny. Science 268, 1492–1495. Martikainen, P., Siitonen, J., Kaila, L., Punttila, P., Rauh, J., 1999. Bark beetles Gauslaa, Y., Ohlson, M., 1997. Et historisk perspektiv på kontinuitet og forekomst av (Coleoptera, Scolytidae) and associated beetle species in mature managed and epifyttiske laver i norske skoger. Blyttia 55, 15–27. old-growth boreal forests in southern Finland. Forest Ecology and Management Goward, T., 1994. Notes on oldgrowth-dependent epiphytic macrolichens in inland 116, 233–245. British Columbia, Canada. Acta Botanica Fennica 150, 31–38. Martikainen, P., Siitonen, J., Punttila, P., Kaila, L., Rauh, J., 2000. Species richness of Gustafsson, L., Hallingbäck, T., 1988. Bryophyte flora and vegetation of managed and Coleoptera in mature managed and old-growth boreal forests in southern virgin coniferous forests in south-west Sweden. Biological Conservation 44, Finland. Biological Conservation 94, 199–209. 283–300. Martin, M.P., Coucheron, D.H., Johansen, S., 2003. Structural features and Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and evolutionary considerations of group IB introns in SSU rDNA of the lichen analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41, fungus Teloschistes. Fungal Genetics and Biology 40, 252–260. 95–98. Meier, F.A., Scherrer, S., Honegger, R., 2002. Faecal pellets of lichenivorous mites Harmata, K., Olech, M., 1991. Transect for aerobiological studies from Antarctica to contain viable cells of the lichen-forming ascomycete Xanthoria parietina and its Poland. Grana 30, 458–463. green algal photobiont, Trebouxia arboricola. Biological Journal of the Linnean Hedenås, H., Ericson, L., 2000. Epiphytic macrolichens as conservation indicators: Society 76, 259–268. successional sequence in Populus tremula stands. Biological Conservation 93, Mills, L.S., Allendorf, F.W., 1996. The one-migrant-per-generation rule in 43–53. conservation and management. Conservation Biology 10, 1509–1518. Hedenås, H., Ericson, L., 2004. Aspen lichens in agricultural and forest landscapes: Mitchell, F.J.G., 2005. How open were European primeval forests? Hypothesis the importance of habitat quality. Ecography 27, 521–531. testing using palaeoecological data. Journal of Ecology 93, 168–177. Heinken, T., 1999. Dispersal patterns of terricolous lichens by thallus fragments. Munoz, J., Felicisimo, A.M., Cabezas, F., Burgaz, A.R., Martinez, I., 2004. Wind as a Lichenologist 31, 603–612. long-distance dispersal vehicle in the Southern Hemisphere. Science 304, 1144– Hilmo, O., Såstad, S.M., 2001. Colonization of old-forest lichens in a young and an 1147. old boreal Picea abies forest: an experimental approach. Biological Conservation Nachman, M.W., Crowell, S.L., 2000. Estimate of the mutation rate per nucleotide in 102, 251–259. humans. Genetics 156, 297–304. Ho, S.Y.W., Phillips, M.J., Cooper, A., Drummond, A.J., 2005. Time dependency of Newton, M.A., Raftery, A.E., Davison, A.C., Bacha, M., Celeux, G., Carlin, B.P., Clifford, molecular rate estimates and systematic overestimation of recent divergence P., Lu, C., Sherman, M., Tanner, M.A., Gelfand, A.E., Mallick, B.K., Gelman, A., times. Molecular Biology and Evolution 22, 1561–1568. Grieve, A.P., Kunsch, H.R., Leonard, T., Hsu, J.S.J., Liu, J.S., Rubin, D.B., Lo, A.Y., Ho, S.Y.W., Shapiro, B., Phillips, M., Cooper, A., Drummond, A.J., 2007. Evidence for Louis, T.A., Neal, R.M., Owen, A.B., Tu, D.S., Gilks, W.R., Roberts, G., Sweeting, T., time dependency of molecular rate estimates. Systematic Biology 56, 515–522. Bates, D., Ritter, G., Worton, B.J., Barnard, G.A., Gibbens, R., Silverman, B., 1994. Howell, N., Smejkal, C.B., Mackey, D.A., Chinnery, P.F., Turnbull, D.M., Herrnstadt, C., Approximate Bayesian-inference with the weighted likelihood bootstrap. 2003. The pedigree rate of sequence divergence in the human mitochondrial Journal of the Royal Statistical Society Series B – Methodological 56, 3–48. genome: there is a difference between phylogenetic and pedigree rates. Niklasson, M., Nilsson, S.G., 2005. Skogsdynamik och arters bevarande. American Journal of Human Genetics 72, 659–670. Studentlitteratur, Lund. Johansson, P., Ehrlen, J., 2003. Influence of habitat quantity, quality and isolation on Nilsson, S.G., Baranowski, R., 1997. Habitat predictability and the occurrence of the distribution and abundance of two epiphytic lichens. Journal of Ecology 91, wood beetles in old-growth beech forests. Ecography 20, 491–498. 213–221. Nordén, B., Appelqvist, T., 2001. Conceptual problems of ecological continuity and Johansson, V., Bergman, K.-O., Lättman, H., Milberg, P., in press. Tree and site quality its bioindicators. Biodiversity and Conservation 10, 779–791. preferences of six epiphytic lichens growing on oaks in south eastern Sweden. Öckinger, E., Niklasson, M., Nilsson, S.G., 2005. Is local distribution of the epiphytic Annales Botanici Fennici. lichen Lobaria pulmonaria limited by dispersal capacity or habitat quality? Johnson, P.L.F., Slatkin, M., 2007. Accounting for bias from sequencing error in Biodiversity and Conservation 14, 759–773. population genetic estimates. Molecular Biology and Evolution 25, 199–206. Ohlson, M., Söderström, L., Hörnberg, G., Zackrisson, O., Hermansson, J., 1997. Kålås, J.A., Viken, Å., Bakken, T. (Eds.), 2006. Norsk rødliste 2006. Artsdatabanken, Habitat qualities versus long-term continuity as determinants of biodiversity in Trondheim. boreal old-growth swamp forests. Biological Conservation 81, 221–231. Kalwij, J.M., Wagner, H.H., Scheidegger, C., 2005. Effects of stand-level disturbances Økland, B., 1996. Unlogged forests: important sites for preserving the diversity of on the spatial distribution of a lichen indicator. Ecological Applications 15, mycetophilids (Diptera: Sciaroidea). Biological Conservation 76, 297–310. 2015–2024. Palice, Z., Printzen, C., 2004. Genetic variability in tropical and temperate Kass, R.E., Raftery, A.E., 1995. Bayes factors. Journal of the American Statistical populations of Trapeliopsis glaucolepidea: evidence against long-range Association 90, 773–795. dispersal in a lichen with disjunct distribution. Mycotaxon 90, 43–54. Kuhner, M.K., 2003. LAMARC: estimating population genetic parameters from Penttilä, R., Siitonen, J., Kuusinen, M., 2004. Polypore diversity in managed and old- molecular data. In: Salemi, M., Vandamme, A. (Eds.), The Phylogenetic growth boreal Picea abies forests in southern Finland. Biological Conservation Handbook: A Practical Approach to DNA and Protein Phylogeny. Cambridge 117, 271–283. University Press, Cambridge, pp. 379–399. Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution. Kuhner, M.K., Yamato, J., Beerli, P., Smith, L.P., Rynes, E., Walkup, E., Li, C., Sloan, J., Bioinformatics 14, 817–818. Colacurcio, P., Felsenstein, J., 2005. LAMARC v 2.0. University of Washington. Printzen, C., Ekman, S., Tønsberg, T., 2003. Phylogeography of Cavernularia hultenii: (Cited 7 September 2006). evidence of slow genetic drift in a widely disjunct lichen. Molecular Ecology 12, Kuhner, M.K., 2006. LAMARC 2.0: maximum likelihood and Bayesian estimation of 1473–1486. population parameters. Bioinformatics 22, 768–770. Rassi, P., Alanen, A., Kanerva, T., Mannerkoski, I. (Eds.), 2001. The Red List of Finnish Kuhner, M.K., Smith, L.P., 2007. Comparing likelihood and Bayesian coalescent Species. Ministry of the Environment, Finnish Environment Institute, Helsinki. estimation of population parameters. Genetics 175, 155–165. Rivas Plata, E., Lücking, R., Lumbsch, H.T., 2008. When family matters: an analysis of Kuusinen, M., 1996. Cyanobacterial macrolichens on Populus tremula as indicators Thelotremataceae (lichenized Ascomycota: Ostropales) as bioindicators of of forest continuity in Finland. Biological Conservation 75, 43–49. ecological continuity in tropical forests. Biodiversity and Conservation 17, Lacy, R.C., 1987. Loss of genetic diversity from managed populations: interacting 1319–1351. effects of drift, mutation, immigration, selection, and population subdivision. Rolstad, J., Gjerde, I., Gundersen, V.S., Stersdal, M., 2002. Use of indicator species to Conservation Biology 1, 143–158. assess forest continuity: a critique. Conservation Biology 16, 253–257. 1878 H. Lättman et al. / Biological Conservation 142 (2009) 1870–1878

Rose, F., 1974. The epiphytes of oak. In: Morris, M.G., Perring, F.H. (Eds.), The British Tapper, R., 1976. Dispersal and changes in the local distributions of Evernia prunastri Oak: Its History and Natural History. Classey, Faringdon, pp. 250–273. and Ramalina farinacea. New Phytologist 77, 725–734. Rousset, F., 1997. Genetic differentiation and estimation of gene flow from F- Thor, G., Arvidsson, L. (Eds.), 1999. Rödlistade lavar i Sverige – artfakta. statistics under isolation by distance. Genetics 145, 1219–1228. ArtDatabanken, SLU, Uppsala. Schneider, S., Roessli, D., Excoffier, L., 2000. Arlequin: A Software for Population Tibell, L., 1992. Crustose lichens as indicators of forest continuity in boreal Genetics Data Analysis. Version 2.000. Genetics and Biometry Laboratory, coniferous forests. Nordic Journal of Botany 12, 427–450. Department of Anthropology, University of Geneva. Tikkanen, O.P., Martikainen, P., Hyvärinen, E., Junninen, K., Kouki, J., 2006. Red-listed Selva, S.B., 1994. Lichen diversity and stand continuity in the northern hardwoods boreal forest species of Finland: associations with forest structure, tree species, and spruce-fir forests of northern New England and western New Brunswick. and decaying wood. Annales Zoologici Fennici 43, 373–383. Bryologist 97, 424–429. Uliczka, H., Angelstam, P., 1999. Occurrence of epiphytic macrolichens in relation to Selva, S.B., 2003. Using calicioid lichens and fungi to assess ecological continuity in tree species and age in managed boreal forest. Ecography 22, 396–405. the Acadian Forest Ecoregion of the Canadian Maritimes. Forestry Chronicle 79, Vera, F.W.M., 2000. Grazing Ecology and Forest History. CABI International, 550–558. Wallingford, UK. Sillett, S.C., McCune, B., Peck, J.E., Rambo, T.R., Ruchty, A., 2000. Dispersal limitations Vucetich, J.A., Waite, T.A., 2000. Is one migrant per generation sufficient for the of epiphytic lichens result in species dependent on old-growth forests. genetic management of fluctuating populations? Animal Conservation 3, 261– Ecological Applications 10, 789–799. 266. Simon, D., Moline, J., Helms, G., Friedl, T., Bhattacharya, D., 2005. Divergent histories Wagner, H.H., Werth, S., Kalwij, J.M., Bolli, J.C., Scheidegger, C., 2006. Modelling of rDNA group I introns in the lichen family Physciaceae. Journal of Molecular forest recolonization by an epiphytic lichen using a landscape genetic approach. Evolution 60, 434–446. Landscape Ecology 21, 849–865. Slatkin, M., 1994. An exact test for neutrality based on the Ewens sampling Walser, J.C., 2004. Molecular evidence for limited dispersal of vegetative propagules distribution. Genetical Research 64, 71–74. in the epiphytic lichen Lobaria pulmonaria. American Journal of Botany 91, Slatkin, M., 1996. A correction to the exact test based on the Ewens sampling 1273–1276. distribution. Genetical Research 68, 259–260. Walser, J.C., Holderegger, R., Gugerli, F., Hoebee, S.E., Scheidegger, C., 2005. Slatkin, M., 2005. Seeing ghosts: the effect of unsampled populations on migration Microsatellites reveal regional population differentiation and isolation in rates estimated for sampled populations. Molecular Ecology 14, 67–73. Lobaria pulmonaria, an epiphytic lichen. Molecular Ecology 14, 457–467. Spence, J.R., Langor, D.W., Niemela, J., Carcamo, H.A., Currie, C.R., 1996. Northern Werth, S., Wagner, H.H., Gugerli, F., Holderegger, R., Csencsics, D., Kalwij, J.M., forestry and carabids: the case for concern about old-growth species. Annales Scheidegger, C., 2006a. Quantifying dispersal and establishment limitation in a Zoologici Fennici 33, 173–184. population of an epiphytic lichen. Ecology 87, 2037–2046. Stewart, F.M., 1977. Computer algorithm for obtaining a random set of allele Werth, S., Wagner, H.H., Holderegger, R., Kalwij, J.M., Scheidegger, C., 2006b. Effect frequencies for a locus in an equilibrium population. Genetics 86, 482–483. of disturbances on the genetic diversity of an old-forest associated lichen. Stoltze, M., Pihl, S. (Eds.), 1998. Rødliste 1997 over planter og dyr i Danmark. Miljø- Molecular Ecology 15, 911–921. og energiministeriet, Danmarks miljøundersøgelser, Skov- og naturstyrelsen, White, T.J., Bruns, T., Lee, S., Taylor, J., 1990. Amplification and direct sequencing of København. fungal ribosomal RNA genes for phylogenetics. In: Innis, M.A., Gelfand, D.H., Sverdrup-Thygeson, A., Lindenmayer, D.B., 2003. Ecological continuity and assumed Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide to Methods and indicator fungi in boreal forest: the importance of the landscape matrix. Forest Applications. Academic Press, San Diego, California, pp. 315–322. Ecology and Management 174, 353–363. Woods, R.G., Coppins, B.J., 2003. A Conservation Evaluation of British Lichens. Swofford, D.L., 2003. PAUP. Phylogenetic Analysis Using Parsimony ( and other British Lichen Society, London. Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Wright, S., 1943. Isolation by distance. Genetics 28, 114–138. Tajima, F., 1989. Statistical method for testing the neutral mutation hypothesis by Zenner, E.K., 2004. Does old-growth condition imply high live-tree structural DNA polymorphism. Genetics 123, 585–595. complexity? Forest Ecology and Management 195, 243–258.

PAPER III

Nordic Journal of Botany 27: 413Á418, 2009 doi: 10.1111/j.1756-1051.2009.00425.x, # The Authors. Journal compilation # Nordic Journal of Botany 2009 Subject Editor: Torbjo¨rn Tyler. Accepted 18 June 2009

Changes in the distributions of epiphytic lichens in southern Sweden using a new statistical method

Ha˚kan La¨ttman, Per Milberg, Michael W. Palmer and Jan-Eric Mattsson

H. La¨ttman ([email protected]) and J.-E. Mattsson, School of Life Sciences, So¨derto¨rn Univ., SEÁ181 89 Huddinge, Sweden. Á HL and P. Milberg, IFM Biology, Division of Ecology, Linko¨ping Univ., SEÁ581 83 Linko¨ping, Sweden. PM also at: Dept of Crop Production Ecology, SLU, PO Box 7043, SEÁ750 07 Uppsala, Sweden. Á M. W. Palmer, Dept of Botany, Oklahoma State Univ., Stillwater, Oklahoma 74078, USA.

Past studies on changes in species distribution have mainly been based on analysis of range boundaries. In contrast, the method used here evaluates shifts in species’ geographic centroids within a predefined area. We used presence/absence data on epiphytic lichens collected 1986 and 2003 from 64 sites in southern Sweden. A centroid was calculated each year, for each lichen species and substrate. The distance of centroid movement was evaluated in a permutation procedure. In total, 56 lichen species on 22 tree species were involved in the analyses, yielding 30 cases that had sufficient sample sizes both years to be evaluated. Out of these, three exhibited a significant movement of their centroid. The shift of lichen centroids of Hypogymnia physodes (L.) Nyl. and Vulpicida pinastri (Scop.) J.-E. Mattsson & M. J. Lai on the tree species Juniperus communis L. was 50 and 151km with the direction 278 and 488, respectively. For Hypogymnia physodes on Pinus sylvestris L., corresponding values were 41 km and 308. The northnortheast shifts of these species in Sweden could be a response to a warming climate.

Large-scale environmental changes, such as global warming, of 329 lichen species collected in 1979, 1984, 1989, 1995 are likely to affect several organisms in a similar way, e.g. by and 2001, they conclude that lichens can respond to global range shifts in similar directions, and also through a change warming over a time frame of only a few decades. Lichens in the density of individuals within populations. So far, with a northern distribution and a southern border in the most studies of distributional changes attributed to global Netherlands have declined. On the other hand, species with warming in the northern hemisphere have been examined as a southern distribution and a northern border are at present the northward expansion of the studied organism (Thomas invading the Netherlands. and Lennon 1999, Warren et al. 2001, Parmesan and Yohe In the present study, we illustrate an approach using 2003, Root et al. 2003, Hickling et al. 2005). The number presence/absence data to compare calculated centroids of of studies based on population densities within a species geographic location, within the area studied (also known as distribution (Thomas and Lennon 1999, Warren et al. ‘centre of occurrence’, La Sorte and Thompson 2007), from 2001, Hickling et al. 2005), or to the likelihood of encoun- investigations at two points in time (1986 and 2003). The tering a species (Bridle and Vines 2007) is much lower. changes of distance and direction of the centroid was easily As the appropriate spatial and temporal scales involved calculated and for distance, significance (null hypothesis of are large, relatively few data sets exist. Further, available no change in centroid) can be tested with a permutation data are often not straightforward and rarely detailed; for procedure. One advantage is that no assumption is needed example field notes from sites with long time gaps, data not about the direction (most analyses so far test for changes in collected under a uniform protocol, or data with low northern and southern borders). Another is that our focus is information content (e.g. presence/absence only). Data of on the general distribution of the species within any this type, however, still provide opportunities for interesting predefined area of any shape, not just on the boundaries. questions to be addressed. Thus, simple methods for The latter might be a poor representation of a large-scale analysis of uneven-quality data would be valuable. change in abundance, is sensitive to grid size, might be Epiphytic lichens on trees, and other transient substrates, highly influenced by single observations and be very are likely to respond relatively quickly to broad-scale sensitive to data deficiencies (Dormann 2007). Further- changes in the climate. van Herk et al. (2002) used more, data for most species are not available to determine checklists, data from field meetings, herbarium material clear distribution boundaries. We apply the method to a and long-term monitoring data to study large-scale changes data set on epiphytic lichens, a group likely to respond of lichens in the Netherlands. Based on the documentation relatively quickly, as both the lichen and its host might

413 respond to environmental changes on temporal and spatial or different land use. Different habitats common to scales relevant for the data at hand. The data come from 64 southern Sweden are represented at the sites, i.e. coniferous, sites in southern Sweden visited in 1986 and 2003. deciduous and mixed forests, wooded pastures, parks, and on occasion single trees in towns, villages and farms. Tree trunks, branches and twigs were examined to a Material and methods height of 2 m above the ground in search of all epiphytic macrolichens. And on each site, all epiphytic lichen species Sampling were collected on all substrates present. The number of tree species per site varied from one to eleven and the aim was to Epiphytic lichens on 64 sites in southern Sweden were investigate approximately 50 trees with lichens at each site. sampled in 1986 and 2003 (Fig. 1a). One of the authors Tree trunks without visual lichens were observed, but were Jan-Eric Mattsson, (JEM) originally selected and visited the not included in the study. Most of the lichen species sites for a study to which the extent Vulpicida juniperinus (L.) J.-E. Mattsson and M. J. Lai and Vulpicida pinastri recorded are widespread and abundant with some excep- (Scop.) J.-E. Mattsson and M. J. Lai were still present at tions, e.g. the redlisted Melanelia laciniatula (Flagey ex former localities (Mattsson 1988). The identification of H. Olivier) Essl. and Usnea barbata (L.) Weber ex F. H. species and selection of the sites, from the first study, were Wigg. (Thor and Arvidsson 1999). The species are, with a based on herbarium specimens of the genus Vulpicida few exceptions, easy to identify based on macro-characters collected over a period of about one hundred years. Hence, only. Most of the lichen species are foliose or fruticose. the sites are not a random selection but follow a previously Some of the species are habitat generalists and may occur on known occurrence of these target species. The sites span several of the tree species surveyed, as well as on rock. One over a wide range of ecological habitats with no apparent observer (JEM) investigated the sites in 1986 and two bias towards a particular tree diameter (successional stage). observers JEM and Ha˚kan La¨ttman (HL) in 2003. Total The sites were of different sizes ranging from a single tree to time, spent to record all species present, per site was similar one hectare and often delimited by natural boundaries such both times and the search method was identical. In 1986, as creeks or ridges, or by man-made borders such as fences, JEM used 40 minutes on each site and in 2003, JEM and

Figure 1. (a)Á(d) distribution of 64 sites in southern Sweden where collection of epiphytic lichen species were recorded 1986 and 2003, and the results of the three signiﬁcant cases of centroid movement (arrows) (Table 1). (a) the 64 investigated sites. The centroid shift of (b) Hypogymnia physodes on Pinus sylvstris, (c) Hypogymnia physodes on Juniperus communis and (d) Vulpicida pinastri on Juniperus communis. A cross indicates a site occupied by the lichen species 1986, a plus indicate a site occupied by the lichen species 2003 and the star indicate a site occupied by the lichen species both years. The arrows indicate length and direction of centroid movements.

414 HL used 20 minutes each per site. Tree and lichen species (Fig. 1). Furthermore, in these three cases there was a nomenclature follows Karlsson (1997) and Santesson et al. general decrease in occurrence. Hypogymnia physodes and (2004) respectively. Vulpicida pinastri on Juniperus communis decreased from 14 In total, the data included 64 sites, 22 tree species and to 6 and from 14 to 3 observations, respectively. For standing dead wood and 56 lichen species. Longitude and Hypogymnia physodes on Pinus sylvestris, the corresponding latitude were recorded in field and converted to WGS 84 values were 8 to 4 (Table 1). before analysis. If applying a control for false discovery rates (following Benjamini and Hochberg 1995), due to multiple testing (n30), one of the three cases remained significant (i.e. Analysis Vulpicida pinastri on Juniperus communis:p0.0002). It can be argued, however, whether the 30 tests conducted We developed a permutation procedure to assess whether actually belong to the same family of tests (cf. Perneger the distance moved by the distributional centroid (the 1998, Proschan and Waclawiw 2000). arithmetic mean of coordinates of sites) was greater than The cumulative distribution of all possible directions expected due to chance. Each combination of lichen and and of all permuted directions were similar, with two soft tree species was analyzed separately. For a site to be bumps (Fig. 2); a consequence of the elongated shape of included for a particular combination of lichen and tree the study area (Fig. 1a). The observed distribution of species analysis, the tree species had to be recorded at the the 30 analyses, that were possible to conduct (Fig. 2), site both years and the lichen species in at least one of the deviated from both of the above-mentioned in the KS tests years. From the information included per combination of (pB0.025, pB0.005). Overall, the direction of movement lichen and tree, we calculated the geographic centroid of the was, for the 30 evaluated cases, dominated by 16 in northeast lichen species in 1986 and 2003, and then the distance and 9 in southwest (Fig. 2). between these centroids. We then randomly permutated the status (present both times, present in 1986 only, and present in 2003 only) for all those sites in which the lichen occurred on that tree species at least once (Table 1). Thirty Discussion combination of trees and lichen species were possible to Movements and directions as judged by centroids analyze, involving eight and 17 tree and lichens species, respectively (Table 1). There are two main results in this study. First, it provides We calculated the centroids and associated distance for evidence that movement of centroids (representing the each of 5000 permutations, and compared these distances probability of finding a species, within an area), of epiphytic to our measured value to obtain a p-value. This analysis was lichens can be detected over a time frame of less than two performed using an Excel visual basic macro written by decades. The strength of the evidence is further discussed MWP. below. This conclusion corroborates the findings of van Although it may seem that a reasonable null hypothesis Herk et al. (2002); a study that was conducted on com- for direction of movement is that all compass directions are parable spatial and temporal scales which suggested that equally likely, irregularities in the distribution of samples epiphytic lichens can respond very quickly to climatic as well as the elongated shape of the study area in Sweden changes. It is easy to speculate on the cause for shifts in (Fig. 1) mean that this is not the case. Therefore, we stored distribution but more difficult to disentangle the possible the directions generated by the permuted data to compile an contributions of trends in, e.g. temperature, precipitation expected distribution of directions. In addition, we calcu- and pollution. Yearly average temperature and precipitation lated the directions between all possible pair-wise combina- at sites within the investigated area increased slightly over the tions of sites. The cumulative distribution of these two data study period (0.0568C year 1 (29 sites) and 5.0 mm year 1 sets were compared with the corresponding of the observed (28 sites), respectively; SMHI 1987Á2003). During the changes of directions of centroids (including those whose same period air pollution decreased strongly and. e.g. NO distance had scored NS), using the KolmogorovÁSmirnov 2 (14 sites) and SO (6 sites) in the air dropped by (KS) test. 2 approximately 50 and 90%, respectively (IVL 2009). Other studies of epiphytes in southern Sweden have documented a local ‘reinvasion/recolonisation’ of lichens attributed to the Results improved air quality (Hultengren et al. 2004). The three significant cases documented in the present study involved Centroid movements of lichen species two lichens (Hypogymnia physodes and Vulpicida pinastri) that both have mainly a northerly distribution in Sweden, We recorded a significant change (pB0.05) in distance and that were shown to be generally on the retreat between the centroids calculated for 1986 and 2003 for (Table 1), and specifically so in the southwest (Fig. 1). three of the 30 analysis possible to conduct (Table 1). This region was also the one with the poorest air quality at Centroid movements of the lichens Hypogymnia physodes the onset of our study, so unless air pollutant concentrations and Vulpicida pinastri on the tree species Juniperus communis recorded in the 1980s turn out to be beneficial to the lichens were 50 and 151 km, respectively, and corresponding in question, it is difficult to see how the documented movement of Hypogymnia physodes on Pinus sylvestris was patterns could be driven by a decrease in air pollution. 41 km (Table 1). The direction of centroid movements of Hence, our tentative interpretation is that the shifts seen are these three significant cases were in a northeast direction climate-driven.

415 416

Table 1. Presence/absence of epiphytic lichens were recorded on 64 sites in southern Sweden in 1986 and 2003. The shift in centroid (Distance) was calculated and its direction could be calculated for 30 cases (lichen species on tree species). The statistical signiﬁcance of the shift in centroid was evaluated in a permutation test. *p-valueB0.05, **p-value highly signiﬁcant also after adjusting for false discovery rate.

Lichen species Tree species Occurrences 1986 only/both Total occurrences Distance (km) Direction (0Á3608) p-value years/2003 only (max 128)

Hypogymnia physodes L. Betula spp. 9/35/15 94 23 44.3 0.2126 Hypogymnia physodes Pinus sylvestris L. 8/31/4 74 41 30.5 0.0066* Hypogymnia physodes Picea abies (L.) H. Karst 9/29/3 70 13 348.4 0.5594 Hypogymnia physodes Juniperus communis L. 14/23/6 66 50 26.9 0.0258* Hypocenomyce scalaris (Ach.) M. Choisy Pinus sylvestris 6/10/15 41 31 41.7 0.5262 Pseudevernia furfuracea (L.) Zopf Betula spp. 16/6/12 40 71 43.2 0.1028 Pseudevernia furfuracea Picea abies 9/9/13 40 70 93.2 0.0688 Platismatia glauca (L.) W. L. Culb. & C. F. Culb. Betula spp. 10/7/9 33 70 35.8 0.0808 Parmeliopsis ambigua (Wulfen) Nyl. Betula spp. 5/3/19 30 106 45.0 0.1160 Vulpicida pinastri (Scop.) J.-E. Mattsson & M. J. Lai Juniperus communis 14/5/3 27 151 47.7 0.0002** Parmelia sulcata Taylor Betula spp. 7/1/14 23 106 51.6 0.1132 Vulpicida pinastri Pinus sylvestris 8/3/2 16 43 27.2 0.4002 Tuckermanopsis chlorophylla (Willd.) Hale Picea abies 6/1/7 15 101 71.2 0.2192 Hypogymnia tubulosa (Schaer.) Hav. Juniperus communis 2/1/10 14 100 16.6 0.4590 Usnea hirta (L.) Weber ex F. H. Wigg. Pinus sylvestris 8/2/2 14 70 212.8 0.4262 Vulpicida pinastri Picea abies 13/0/1 14 106 44.6 0.6408 Xanthoria parietina (L.) Th. Fr. Populus tremula L. 1/2/8 13 35 94.7 0.6994 Ramalina farinacea (L.) Ach. Acer platanoides L. 0/4/4 12 92 222.9 0.0882 Evernia prunastri (L.) Ach. Populus tremula 5/1/4 11 111 197.9 0.0604 Bryoria fuscescens (Gyeln.) Brodo & D. Hawskw. Betula spp. 6/2/0 10 97 59.0 0.3570 Parmelia sulcata Taylor Picea abies 1/2/5 10 19 224.8 0.9146 Usnea hirta Betula spp. 7/1/1 10 48 195.5 0.5198 Evernia prunastri Picea abies 2/1/5 9 42 222.6 0.8922 Usnea hirta Juniperus communis 5/2/0 9 70 189.0 0.7186 Usnea subfloridana Stirt. Betula spp. 8/0/1 9 258 53.7 0.2196 Ramalina farinacea Fraxinus excelsior L. 1/0/6 7 322 16.3 0.2768 Calicium viride Pers. Quercus robur L. 1/1/2 5 211 169.3 0.3292 Ramalina fastgiata (Pers.) Ach. Fraxinus excelsior 1/0/4 5 149 302.1 0.8000 Tuckermanopsis chlorophylla Pinus sylvestris 0/1/3 5 115 223.4 0.7524 Usnea subfloridana Stirt. Picea abies 3/0/2 5 169 204.0 0.0958 The estimates of shift of the centroids presented, 3Á10 km whether the current analyses should be considered to year1, is in range with the weighted centroids of counts of belong to the same family of tests or not (i.e. a justified wintering shorebirds in western Europe (1.5Á6.0, MacLean need for correction or none), so we present both. But et al. 2008). Although not comparable, it might be noted we focus on the uncorrected p-values for two reasons. that published estimates of northward movement show an Firstly, as generally in monitoring, type-II errors (accepting average of 6.1 km per decade (Parmesan and Yohe 2003, a true null hypothesis) might be equally, or more, harm- several organisms), 9.5 km per decade (Thomas et al. 1999, ful than type-I errors (Legg and Nagy 2006). Secondly, birds) and 98.8 km per decade (Perry et al. 2005, marine we were interested in including all species to be able to fishes). Further, simulations on the range expansions needed consider the directions Á several cases, each with a weak for different organism groups to keep track with expected signal, might together indicate a trend. This was, in fact, climate warming, suggest less than 1 km in most cases, and the outcome Á also non-significant cases contributed to rarely longer than 10 km per annum (Malcolm et al. 2002). the evidence suggesting a prevailing northeast direction Second, our study also showed that the prevailing (Fig. 2, Table 1). direction of movements of lichens is likely to be in a In conclusion, despite the low power of the data our new northeast direction, rather than north, in southern Scan- method proved to work well, and there were significant dinavia. Most studies of species evaluating range shifts case(s) of centroid movements with a discernable prevailing assume a northward movement of organisms in the northeast direction. Northern Hemisphere (Parmesan 2006). It is not easy to justify this particular direction when considering the global air circulation; the prevailing northeast direction The versatile permutation procedure documented in the present study are more in line with The above-mentioned shortcomings are mainly related to the global atmospheric circulations (de Blij and Muller this particular data set and its small sample size, and the 1996) and the gradients in temperature and rainfall in the method of analysis has some general advantages and might study area (SMHI 1987Á2003). be useful in a broader range of analyses. First of all, it is possible to study changes within distribution areas without Methodological considerations any knowledge of distribution boundaries. These are often The current lichen data set and its low power difficult to determine and also probably have poor All data sets suffer from shortcomings. In our case, field statistical properties (Dormann 2007). Some species do sampling was designed to be quick, recording only not have their outermost localities within an area studied, presence/absence of lichens that, in most cases, did not severely reducing the data sets available for boundary require a specialist for identification. Consequently, we analysis. With the current method, many more types of were able to include more sites than if including, e.g. data set could be analyzed. assessment of abundance or demographic data. Future power analyses would be welcome to strike an appropriate 100 compromise between data quality and quantity when p<0.05 setting up monitoring studies in general. p 0.05-0.1 p 0.1-0.25 We lacked information about trees without lichens and 80 were therefore unable to discriminate between actual losses p 0.25-0.5 p>0.5 of lichens on specific substrates and the loss of the substrate on a site. An additional consequence of this is that it prevented us from evaluating a possible change in substrate 60 preference (e.g. a species might, over time, occur on a wider range of tree species). Another drawback of the simple field method is that the 40 age or size of the tree species, and where the lichen was found (trunk or branches), were not recorded. Although Cumulative percentage most of the epiphytic species sampled are not sensitive to tree species or tree size, some may, e.g. prefer thin branches 20 over trunks and thus suffer a bias due to tree size All pairwise distances composition. Distances from permutations Three out of 30 tests turned out to be significant, which 0 might be interpreted as negligible change. It must be 0 45 90 135180 225270 315 360 remembered, though, that many of the tests were based Angle on very few occurrences (Table 1). Thus, the statistical power was, in most cases, low or very low and only large Figure 2. The cumulative distribution of the calculated directions movements would be possible to detect. of centroid movement (Table 1), that were based on inventories of The relatively large number of tests conducted (n30) epiphytic lichens on trees at 64 sites in southern Sweden in 1986 and 2003. The size of the mark shows the probability of a might imply a need for adjusting for the family-wise movement of centroid between the two years. The dotted line type-I error (rejecting a true null hypothesis). Only one show the angles of all possible pair-wise combinations of sites and of three significant tests was found significant after apply- the unbroken line show the angles generated in the 30 permutation ing the adjustment for false discovery rate (Benjamini tests. The two soft bumps are due to the slightly elongated shape of and Hochberg 1995). It is mainly a matter of opinion the study area (Fig. 1a).

417 With presence/absence data, power decreases when La Sorte, F. A. and Thompson III, F. R. 2007. Poleward whifts in approaching both zero and 100% frequency of the winter ranges of North American birds. Á Ecology 88: 1803Á phenomenon under study. Hence, species with intermediate 1812. abundance, occurring in ca 50% of sample points, would Legg, C. J. and Nagy, L. 2006. Why most conservation have the strongest power in the present analysis. Basing monitoring is, but need not be, a waste of time. Á J. Environ. centroids on abundance data, would allow a strong analysis Manage. 78: 194Á199. also of very frequent species (cf. the weighted centroids used MacLean, I. L. M. D. et al. 2008. Climate change causes rapid changes in the distribution and site abundance of birds in by MacLean et al. 2008). winter. Á Global Change Biol. 14: 2489Á2500. Malcolm, J. A. et al. 2002. Estimated migration rates under Acknowledgements Á The study was funded by the Swedish Royal scenarios of global climate change. Á J. Biogeogr. 29: 835Á849. Academy of Science, Lunds Botaniska Fo¨rening (1986) and the Mattsson, J.-E. 1988. Na˚gra Cetraria-arters tillbakaga˚ng i so¨dra County of Stockholm. PM and MWP were partly funded through Sverige. Á Sv. Bot. Tidskr. 82: 27Á31. the research program ENGO sponsored by the Swedish Environ- Parmesan, C. 2006. Ecological and evolutionary responses mental Protection Agency. We thank Karl-Olof Bergman, Anders to recent climate change. Á Annu. Rev. Ecol. Evol. Syst. 37: Hargeby, Lars Westerberg and Mikael Lo¨nn for discussions, 637Á669. Christopher Zetterberg for drawing the maps and Jennifer Larsson Parmesan, C. and Yohe, G. 2003. A globally coherent fingerprint for linguistic revision. Thanks also Kristina Articus (Usnea) and of climate change impacts across natural systems. Á Nature Anders Nordin (Bryoria and Melanelia) for your help to determine 421: 37Á42. some rare lichen species. Perneger, T. V. 1998. What’s wrong with Bonferroni adjustments. Á Brit. Med. J. 316: 1236Á1238. Perry, A. L. et al. 2005. Climate change and distribution shifts in References marine fishes. Á Science 308: 1912Á1915. Proschan, M. A. and Waclawiw, M. A. 2000. Practical guidelines Benjamini, Y. and Hochberg, Y. 1995. Controlling the false for multiplicity adjustment in clinical trials. Á Control. Clin. discovery rate: a practical and powerful approach to multiple Trials 21: 527Á539. testing. Á J. R. Stat. Soc. Ser. B 57: 289Á300. Root, T. L. et al. 2003. Fingerprints of global warming on wild Bridle, J. R. and Vines, T. H. 2007. Limits to evolution at range animals and plants. Á Nature 421: 57Á60. margins: when and why does adaptation fail? Á Trends Ecol. Santesson, R. et al. 2004. Lichen-forming and lichenicolous fungi Evol. 22: 140Á147. of Fennoscandia. Á Uppsala Univ. de Blij, H. J. and Muller, P. O. 1996. Physical geography of the SMHI 1987Á2003. Va¨der och vatten. Á Swedish Meteorological global environment (2nd ed.). Á John Wiley and Sons. and Hydrological Institute, Norrko¨ping, Sweden. Dormann, C. F. 2007. Promising the future? Global change Thor, G. and Arvidsson, L. 1999. Ro¨dlistade lavar i Sverige, projections of species distributions. Á Basic Appl. Ecol. 8: 387Á artfakta. Á ArtDatabanken. 397. Thomas, C. D. and Lennon, J. J. 1999. Birds extend their ranges Hickling, R. et al. 2005. A northward shift of range margins in northwards. Á Nature 399: 213Á213. British Odonata. Á Global Change Biol. 11: 502Á506. van Herk, C. M. et al. 2002. Long-term monitoring in the Hultengren, S. et al. 2004. Recovery of the epiphytic lichen flora Netherlands suggests that lichens respond to global warming. following air quality improvement in southwest Sweden. Á Lichenologist 34: 141Á154. Á Water Air Soil Pollut. 154: 203Á211. Warren, M. S. et al. 2001. Rapid responses of British butterflies to IVL 2009. Swedish environmental research institute, Bwww. opposing forces of climate and habitat change. Á Nature 414: ivl.se, accessed May 2009. 65Á69. Karlsson, T. 1997. Fo¨rteckning o¨ver svenska ka¨rlva¨xter. Á Sv. Bot. Tidskr. 91: 241Á560.

418 PAPER IV

Biodiversity in the wake of urban sprawl: loss among epiphytic lichens on large oaks

Håkan Lättman, Karl-Olof Bergman, Malin Rapp, Malin Tälle, Lars Westerberg and Per Milberg

Håkan Lättman, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden, and School of Life Sciences, Södertörn University, SE-181 49 Huddinge, Sweden

Karl-Olof Bergman, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden

Malin Rapp, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden

Malin Tälle, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden

Lars Westerberg, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden

Per Milberg, IFM Biology, Linköping University, SE-581 83 Linköping, Sweden

Abstract

Biodiversity often suffers from urbanisation. In the present study, we focused on how the age of urbanisation affects the richness of 17 epiphytic lichens species and their cover on large oaks, with a minimum spacing of 250 m, in urban environments in the city of Linköping (100,000 inhabitants), SE Sweden. We also surveyed trees in adjacent rural areas, selected to have similar distributions of tree trunk circumference and oak density within 300 m. Lichen richness and cover were significantly lower on urban trees compared to rural trees. Furthermore, richness and cover decreased with the length of time that urban trees had been surrounded by houses. Roughly one species is lost every 30 years. Most of the species that were analysed demonstrated a drop in occurrence with respect to the duration of housing development. The reduction in the probability of occurrence varied from 60% (Calicium viride, Evernia prunastri), 80% (Chrysotrix candelaris) to 90% (Ramalina spp.) during the 160-year period of urbanisation considered. Therefore, even if valuable trees survive over the course of development, their lichen flora are likely to become depleted over time.

Key words: housing; landscape; Linköping; Quercus; spatial; urbanization

Introduction

Urbanisation alters the natural environment in a number of ways and causes local extinction of species (McKinney, 2006). In the USA, urbanisation is considered to be one of the worst threats to endangered species (Czech, Krausman, & Devers, 2000). Many studies have shown that when moving from rural areas to the central parts of a city, there is a gradual decrease in species richness in a number of organism groups, such as butterflies (Blair, 1999; Blair & Launer, 1997), beetles (Ishitani, Kotze, & Niemelä, 2003; Niemelä & Kotze, 2009; Su, Zhang, & Qiu, 2011), birds (Blair, 1999; Gagne & Fahrig, 2011), amphibians (Price, Dorcas, Gallant, Klaver & Willson, 2006; McKinney, 2008), reptiles (McKinney, 2008), mammals (McKinney, 2008), and plants (McKinney, 2008). However, there are a few studies that show that urbanisation has no effect on biodiversity. For instance, Carabidae assemblages showed no changes across urban gradients in three cities in Finland, Bulgaria and Canada with three different levels of urbanisation (Niemelä et al., 2002). A study in the lower Florida Keys concluded that the native ant fauna was not, at the moment, negatively affected by age of urbanisation (Forys & Allen, 2005). However, with a few exceptions, urbanisation generally seems to reduce biodiversity. Urbanisation affects various environmental factors that in turn affect survival and reproduction of species. The mechanisms involved in species loss vary, spanning from habitat loss and fragmentation (Robinson, Newell, & Marzluff, 2005) to assumed shifts in biotic interactions and possible adverse effects of pollution (Tarhanen, Poikolainen, Holopainen, & Oksanen, 2000). One group that is particularly sensitive to pollution is lichens. When pollution exceeds a critical level, some species become locally extinct (van Herk, 2001). In urban areas, sulphur and nitrogen oxides primarily cause the greatest damage to lichens (Giordani, 2007). Some lichens are less sensitive to air pollution, but if levels of sulphur oxide exceed 125 µg/m3, very few lichens survive (Johnson, 1979; LeBlanc, Rao, & Comeau, 1972; de Wit, 1976). Based on the fact that different lichen species have different tolerances for air pollutants, Hawksworth and Rose (1970) created a method to assess the level of sulphur dioxide in the air based on presence of epiphytic lichens. Lichens also have different tolerances and preferences for substrates. Some lichens are restricted to large trees (Washburn & Culley, 2006) that usually have a coarse bark structure. Such specialised lichens have declined as the density of that substrate has diminished due to agriculture and forestry (Hedenås & Ericson, 2008; Belinchón et al., 2009). However, trees are often an important feature of cities and are often allowed to grow old for a number of reasons, including human wellbeing, economic value, aesthetics, shade production, and in some circumstances, fire prevention (Lohr, Pearson-Mims, Tarnai, & Dillman, 2004; Donovan & Butry, 2010, 2011). An example is the large number of Quercus robur trees, which have been maintained throughout urbanisation, in urban Linköping, a city with 100,000 inhabitants in southeastern Sweden. Mobile organisms can escape urban habitats, while trees (and lichens) are inevitably incorporated into the urban environment. Lichens on the remaining trees are thus isolated, and their dispersal and establishment may be negatively affected. However, the future

3 prospects for epiphytic organisms are better than those of many other elements of urban biodiversity as trees will continue to be an important feature of cityscapes. In the present study, a selection of lichen species were surveyed on Quercus robur trees in urban and rural environments. The objectives of this study can be divided into four questions: Is there any difference with regard to (i) species richness and (ii) cover of epiphytic lichens on Q. robur in urban or rural environment? If there are any differences are these affected by (iii) species richness and (iv) cover of epiphytic lichens on Q. robur affected by the age of urbanization? To make these comparisons, we matched a population of urban oaks with a selection of rural oaks based on tree circumference and the density of surrounding oaks because oak lichens are sensitive to both the size of their host tree (Johansson, Bergman, Lättman, & Milberg, 2009) and the density of surrounding oaks (Muhammadi, 2011). The design of this study allowed us to separate the effect of habitat loss from the effect from the age of urbanization, a feature rarely considered.

Materials and Methods

Study area The study was conducted in Linköping city and surrounding rural areas south of the city, in the province of Östergötland, Sweden (Fig. 1). The area north of Linköping municipality was avoided due to differences in soil and vegetation characteristics from the urban areas. The study area is located at an altitude of 30 to 125 m above sea level and consists of the urban city centre of Linköping, urban district and surrounding rural environments. The city of Linköping is the seventh largest city in Sweden with a population of 104,232 in 2010 (Linköpings kommun, 2011). It was founded in the 11th century, but the majority of buildings have been built since the mid-1800s. The annual average temperature at the nearest meteorological station (urban district Malmslätt) is 6.1 ºC, with an average temperature of –3.2 ºC in January and 16.2 ºC in July; and the average precipitation is 516 mm per year (Statistics Sweden, 2011). Linköping has expanded concentrically (Fig. 2), and the city centre consists largely of three to five storey buildings with shops at the street level. Blocks composed of flats and one to two-storey houses make up the greater part of the residential areas in the urban district. Squares, parks and paved roads are both in the city centre and urban district. The surrounding rural areas consist mainly of a mixed landscape with deciduous forests, planted conifers, arable land and pastures.

Selection of Quercus robur and lichen species We studied the lichen flora on pedunculate oak (Quercus robur). Quercus robur is common in and around Linköping, and urban trees were selected for study from an existing tree database. The database contains the position and circumference of all large Q. robur, as well as several other tree species in the province of Östergötland, and was developed by the County Administration Board in Linköping (available at http://gis.lst.se/lstgis/). Using ArcMap 9.3 (ESRI, 2011), all Q. robur within the municipality of Linköping with a circumference >250 cm were chosen because these trees can be assumed to be old enough to host a high diversity of lichen flora (e.g., Johansson et al., 2009). The age of the sampled trees can be estimated from dendrological studies in the area (Berg, 2006) to be at least 180 to 240 years; therefore, the establishment of most trees predates urbanisation by at least 100 to 150 years. A grid was added over the urban area, and in each cell in the grid where Q. robur trees were present, the oak tree closest to the centre of the cell was chosen; this method ensured that study trees were at least 250 m apart. From the database, we selected 105 urban trees, which were matched with a population of rural trees (Fig. 1). The criteria for matching urban trees with rural trees were that (i) the distribution of trunk circumference and (ii) the density of trees within 302 m should be similar since Muhammadi (2011) has shown that species richness has a high degree of explanation at this distance. Some selected rural trees were excluded during the fieldwork because they had

5 been cut down, or were in private gardens; others were substituted with similarly sized trees for the same reasons. In the end, we used 105 and 109 trees in urban and rural environments. The seventeen lichen species selected for our survey are shown together with their red-list category (Gärdenfors, 2010) and, when appropriate, if the species are used as an indicator species (Nitare, 2010), as well as with substrate preference (Santesson, Moberg, Nordin, Tønsberg, & Vitikainen, 2004) (Table 1). Nine species were commonly found, and the majority occur more or less throughout all of Sweden. Some of them are considered resistant to air pollutants. Eight lichens are rare and red-listed by the threat categories near threatened (NT) or vulnerable (VU), and/or indicator species for high nature conservation and are thought to be sensitive to air pollution. Two additional lichen species were initially included but were excluded in the beginning of the fieldwork as they turned out too difficult to reliably identify in the field. After the fieldwork, it was concluded that three species (Calicium quercinum, Lecanographa amylacea and Schismatomma decolorans) had never been registered.

Field survey The fieldwork was conducted during April, May, October and November, 2011. For every oak, circumference at breast height was measured to the nearest centimetre. The depth of tree trunk bark crevices was measured in the north, east, south and west sides of the trunk to the nearest millimetre using a ruler. Based on these four measurements, a mean bark crevice depth was calculated for each tree. Sun exposure was estimated by assessing how much sun in per cent to the trunk was received, taking into account the shade cast by nearby buildings, shrubs and trees. On each tree trunk, lichens were searched for over a one-metre mantle area starting 50 cm above the ground to avoid the influence of environmental effects at the base as they vary greatly between individual trees. For each of the target lichen species, the area they covered was estimated to the nearest cm2 and was expressed as the percentage of the inspected trunk area.

Additional parameters Some complementary explanatory variables were also considered (Table 2). The number of Q. robur trees around each tree, including the tree being examined, was used as an explanatory variable. Six different radii were used: 150, 250, 350, 500, 700 and 1000 m. The same range of radii was also used to determine the area of buildings around each studied Q. robur. The data on buildings were drawn from the GSD-Topographic Map (Lantmäteriet, http://www.lantmateriet.se/) and combined four different classes of buildings (industrial buildings, one and two storey houses, apartment buildings and city blocks). Age of urbanisation was estimated from the average year of construction of the five building closest to the target tree, using a radius of approximately 200 m (data from the municipality of Linköping). Fewer buildings were used if the year of construction was not available for all five buildings or if fewer than five buildings were within 200 m of the tree. For the city centre (Fig. 2), where some older buildings have been demolished and replaced with newer ones, the above approach would underestimate the age. We therefore assigned the age 1900 to seven buildings that were erected by 1950 or later (hence replacing older buildings). One tree associated with buildings in the area dated to the eighteenth century (Fig. 2) was assigned the age 1850. In total, these age assignments affected the urban age estimates of eight trees in the city centre, and are most likely to be underestimates of the true ages.

Statistical analyses The number of urban and rural trees of Q. robur used in the statistical analyses was 105 and 109, respectively, and in total, 14 lichen taxa were observed: nine common and five rare. To evaluate the data, regression analyses (generalised linear model, GLM) were run in Statistica 10 (Statsoft, 2011). The first set of analyses involved the number of target species per tree on urban vs. rural trees (normal distribution; identity link), and the total cover of target lichen per tree for urban vs. rural trees (normal distribution; log link). Rural trees were not included in any further analysis. The second set of analyses evaluated the relationship between (i) the number of target species per tree (normal distribution; identity link), and (ii) the total cover of target lichen per tree (normal distribution; log link), as related to number of years in an urban setting (i.e., the average age of the five closest buildings). The third set of analyses involved the species-wise occurrence of the eight most frequent species (binomial distribution; logit link). For each radius (150–1000 m), a model was made including the following candidate explanatory variables: (i) age of urbanisation, (ii) tree circumference, (iii) bark crevice depth, (iv) sun exposure, (v) density of oaks, and (iv) cover of buildings. Using AIC, explanatory variables, including radii, were selected. Depth of bark crevices was chosen as it is often considered a better proxy for tree age than is tree circumference (Barkman, 1958; Pedersen, 1980; see Johansson et al., 2009). However, as the depth and circumference were poorly correlated, both were considered as candidate variables. Tree circumference and density of oaks were log transformed before the analysis. Finally, five attributes of the fourteen species were analysed in relation to their average cover (square-root transformed) on urban oaks: 1) spores (no, rarely, yes; ordinal multinomial; logit); 2) diaspore (whether soredia and isidia occur or not; binomial; logit); 3); spore area (a measure of spore size; normal distribution; identity link function); 4) pycnidium (no, rarely, yes; ordinal multinomial; logit); and 5) growth form (crustose, fruticose, foliose; multinomial; logit) (Foucard, 2001; Nash III, Ryan, Gries, & Bungartz, 2002; Nash III, Ryan, Diederich, Gries, & Bungartz, 2004; Nash III, Gries, & Bungartz, 2007).

Results

In total, 214 trees of Quercus robur were studied, of which 105 and 109 were in urban and rural environments, respectively. Out of the 17 lichen species that were searched for, three rare and red-listed species (Calicium quercinum, Lecanographa amylacea and Schismatomma decolorans) were not found (Table 1). The number of lichen species per oak varied between 0 and 9. The total number of observations of lichen species was 954, with 400 observations in urban areas and 554 in rural areas.

Species richness and cover: urban vs. rural There were clear differences in lichen species occurrence and cover of individual lichen species on oaks in the urban area compared to oaks in rural environments (Fig. 3). Four taxa were only found in the rural environment (Calicium adspersum, Cyphelium inquinans, Sclerophora coniophaea, Usnea spp.). All lichen taxa, except Lepraria incana, occurred more frequently on rural oaks than on urban oaks (Fig. 3a). All lichen taxa (including Lepraria incana) had a higher cover on rural than urban oaks, both when looking at averages over all trees (including zero cover observations) and when considering only trees on which the species were present (Fig. 3b and c). The number of target species per tree and the cover of target lichens were significantly higher on oaks in rural environments than on oaks in urban area (P < 0.0001 in both cases) (Fig. 4a and c). The mean species richness was 34% higher on rural oaks than urban oaks (5.1 and 3.8 lichen species, respectively) (Fig. 4a). When looking at the total cover of the 14 lichen species, the differences between urban oaks and rural oaks were even more pronounced, with a mean cover many times larger on rural oaks than on urban oaks (0.041% and 0.007%, respectively) (Fig. 4c).

Species richness and cover: effects from age of urbanisation The number of lichen species per tree decreased significantly with the increasing age of the surrounding buildings (P = 0.00031) (Fig. 4b). The average number of target species in the most recently urbanised areas was four times higher than in the oldest parts (Fig. 4b), and a similar decline was found in the cover of target lichens (Fig. 4d).

Species-wise responses to environmental factors in urban areas Out of the 14 lichen species, eight were sufficiently frequent to be analysed individually in consideration of all explanatory variables. The selected models proved significant for seven of the eight species (Table 3). The majority of the species were affected by urbanisation factors or oak size factors, while few species were affected by other biotic factors such as oak density. The age of urbanisation had a significant negative effect on five species and the area covered by buildings on two species. Chrysotrix candelaris and Hypogymnia physodes were positively

8 affected by the area covered by buildings (Table 3). The lichen species responded mostly to area covered by buildings at larger scales, ≥350 m. A higher tree circumference or deeper bark crevices had a significant positive effect on the occurrence of four species and a negative effect on one. Sun exposure had a significant negative impact on two species (Table 3). Oak density affected only one species significantly negatively and none positively. For the four species where the age of urbanisation significantly affected the occurrence, separate binomial (logit link) GLMs were conducted. All species showed a decrease in prevalence in the older parts of the city in comparison to the newer parts. The models predicted a reduction of 60% (Calicium viride, Evernia prunastri), 80% (Chrysotrix candelaris) and 90% (Ramalina spp.) probability of occurrence during the 160 years of urbanisation (Fig. 5). Out of the six tested species attributes (Table 4), only one had a significant association with cover on urban trees: „spores‟, i.e., to what extent a species rely on spore dispersal (P = 0.039).

Discussion

This study shows a clear reduction in species richness and abundance of lichens on oaks in urban areas compared to rural oaks, as well as with age of urbanisation in the urban areas, a finding that is in line with studies on other groups (Blair, 1999; Blair & Launer, 1997; Gagne & Fahrig, 2011; Ishitani et al., 2003; McKinney, 2008; Niemelä & Kotze, 2009; Price et al., 2006; Su et al., 2011). Our results indicate that the dispersal mode may be an important factor in predicting which species will be affected by urbanisation.

Lower richness and cover on urban oaks Species in urban areas are affected by a complex interaction between factors, such as pollution, temperature, moisture, disturbance and habitat configuration (McDonnell et al., 1997). Generally habitat loss and fragmentation are among the most important factors in urban areas (McKinney, 2006; Niemelä, 1999), which for most species is negative. A high density of large oak trees has proved to be important for species richness and the occurrence of some lichens (Paltto, Thomasson, & Nordén, 2010; Muhammadi, 2011), showing that habitat configuration (or connectivity) is important. The sample populations of urban and rural oaks were matched for both circumference and surrounding tree density, thereby concurrently allowing for comparison while controlling for habitat quality (circumference) and connectivity. Several species are generalist and not confined to oaks, but because previous studies emphasize the importance of oak (Paltto et al., 2010; Muhammadi, 2011), we argue that our sample design can control for differences in habitat connectivity. Thus, by controlling for habitat age and habitat composition (density of surrounding oaks), we were able to single out the effect of urbanisation, from that of a general decrease in habitat. In cases where it was possible to analyse the species, our study showed that urbanisation had a clear effect on species richness and cover (Table 3). The explanatory variable, i.e., age of urbanisation, showed that five species were negatively affected, while area covered by buildings showed that three species were negatively affected and one species was positively affected. A decrease of lichens in urban environments was reported very early during industrialisation in several places in Europe (Grindon, 1859; Nylander, 1866). The main factor for the decline is sulphur dioxides from industries and traffic (Gilbert, 1968). Sulphur dioxides peaked during the 1960s and 1970s, and lichen deserts in urban areas were reported at these elevated emission levels (Hawksworth & Rose, 1970). However, air quality has recently been shown to have improved in many urban areas (Lisowska, 2011). In Tampere, Finland, sulphur dioxide levels were reduced from 160 μg m–3 in 1973 to 2 μg m–3 in 1999, which coincided with an increase in lichen epiphytic richness (>10 times) and cover (>200 times) since 1980 (Ranta, 2001). A reduction has also occurred in the county of Östergötland, where sulphur dioxide levels have decreased from 5290 tonnes per year in 1990 to 1588 tonnes per year in 2009, and from 2786 to 347 tonnes per year for Linköping municipality (RUS, 2012). In our study area, there is no longer any ground for talking of a “lichen desert” in the city: There were no differences between

10 the number of oaks without target species in urban and rural areas (4 out of 105 urban oaks without target lichens and 3 of 109 rural oaks). However, there were still large differences in species diversity and the cover of individual species between urban and rural oaks. Our results show that species that disperse by spores were often absent or had low cover, while those that were not spore-dispersed were among those with the highest cover (Table 4). Species dependent on the process of re-lichenisation with a photobiont after successful dispersal seem to be less likely to succeed in urban areas. The species most sensitive to urbanisation in our study all belong to this group. Calicium adspersum, Cyphelium inquinans, Sclerophora coniophaea and Usnea spp. could not be found on oaks in the urban environment, but they were present in the rural environment. Calicium adspersum, C. inquinans and S. coniophaea are classified as indicators of forest areas with high conservation values (Nitare, 2010), while the taxon Usnea spp. is known to be sensitive to air pollution (Hawksworth & Rose, 1970). In addition to the dispersal mode, the photobiont may be important. Tarhanen et al. (2000) have demonstrated that the green algae Trebouxia is sensitive to air pollution. Their results showed that high concentrations of pollutants in the air increased plasmolysis and mitochondrial changes in cells. Furthermore, degenerated cells showed altered chloroplasts and electron-translucent pyrenoglobuli as far as 35 to 50 km from the pollutant source. The low cover on many urban oaks compared to that on rural oaks indicates that something affects lichens in urban areas, even though the air quality has improved. Armstrong and Bradwell (2010) reviewed 52 studies with reported growth rates, and of which 33 of the surveys had been performed in Europe. If we adopt an average European radial growth rate of 0.92 mm per year, which seems reasonable, the lichen thallus would need 6 years to reach a size of one centimetre in diameter and thus become clearly visible to the naked eye. For the five foliose and fruticose lichens, the growth rate is available for Hypogymnia physodes: approximately 3 to 5 mm in diameter per year (Gorbach & Kobzar, 1981; unpublished data), with a corresponding shorter time to reach visibility. Given that levels of sulphur dioxide have decreased over the past 30 years, there should have been ample time for regrowth for most lichens, given that they are not dispersal or photobiont limited and that current SO2 levels are sufficiently low. Lichens are also known to be sensitive to changes in temperature and to respond to global warming (van Herk, Aptroot, & van Dobben, 2002), and most of the species favour a moist environment. Therefore, shifts in temperature and moisture might be important for urban lichens. Several studies have shown that the temperature is higher in urban than in rural areas (Bulut, Toy, Irmak, Yilmaz, & Yilmas, 2008; George, Ziska, Bunce, & Quebedeaux, 2007; Liu, You, & Dou, 2009), a phenomenon known as “urban heat island”. Hughes (2006) investigated the “urban heat island effect” in four cities in the United Kingdom of varying population size. Norwich had on average urban temperature of 0.5°C higher than the rural surroundings. Because the population size of Norwich is similar to Linköping‟s, we can assume the latter has a similar temperature difference. Relative humidity (RH) is, to a large extent, a function of temperature, so we would expect the urban heat island effect to coincide with lower RH. Furthermore, hard surfaces, efficient canalisation of runoff water, and small volumes of vegetation (low evapotranspiration) in urban areas also suggest lower RH. Nevertheless, evidence from the measurements is conflicting (George et al., 2007 found no differences, while Liu et al., 2009 did).

Loss rate over the course of urbanisation Numerous studies have investigated how a spatial gradient from rural, suburban to urban areas affects the number of organisms and their quantities. Most studies have concluded that the number of species and their amount decreases with an increasing degree of urbanisation (Mercado Cárdenas & Buddle, 2009), but there are exceptions where the cause is often alien species (Dolan, Moore, & Stephens, 2011) and generalist species (Magura, Tóthmérész, & Molnár, 2008; Tóthmérész, Máthé, Balázs, & Magura, 2011). The same reduction of species and their amount toward the city centre can also be observed in temporal studies. For instance, Fattorini (2012) made a reconstruction of the extinction trends of four insect groups in urban Rome, Italy from 1885–1999. Her results showed a clear decline in species richness in each group of insects. Price et al. (2006) examined the incidence of species from the family Salamandridae near Davidson, North Carolina, USA, for a period of 30 years and found that the populations declined over time. Our results showed a clear effect on species number and cover with the length of time that urbanisation had been taking place. Of all 14 lichens, eight were analysed individually (Table 3). Age of urbanisation, or the area covered by buildings around the trees, affected seven species negatively. It is of interest to know the rate of species loss, or cover, during urbanisation. We estimated a decrease in species richness and the total cover of target lichen species per tree (Fig. 4b and 4d) and a lowering probability of occurrence by 3.7–5.6% per decade depending on lichen species (Figure 5). In summary, the relatively slow but steady loss of biodiversity confirms the assumption of Hahs et al. (2009) that modern cities potentially carry a large extinction debt; however, the interpretation is complicated by the return of lichens following recent improvements in air quality (see above).

Conclusion We have shown major differences in lichen species richness and cover between urban and rural environments, as well as a clear decrease with the age and degree of urbanisation on remnant large oaks in a city. Therefore, even if the prospects are good for a continuous supply of epiphytic substrate in cities – where trees are likely to exist – their value for epiphytic lichens seems limited. For the urban planner, this means two things. First, retaining individual valuable trees during urbanisation does not automatically preserve their lichen biodiversity. Second, the preservation of lichen biodiversity is most likely best achieved if urban development leaves valuable areas of trees with rich lichen populations intact and relatively large and plan for connecting fragments of old trees instead of embedding them.

Acknowledgements

Thanks to Lena Samuelsson and Linköping municipality for assistance with obtaining maps of Linköping‟s different urban district boundaries and the age of buildings.

References

Armstrong, R., & Bradwell T. (2010). Growth of crustose lichens: a review. Geografiska Annaler Series A Physical Geography, 92(1), 3-17. Barkman, J. J. (1958). Phytosociology and ecology of cryptogamic epiphytes: including a taxonomic survey and description of their vegetation units in Europe. Assen, Van Gorcum. Belinchón, R., Martínez, I., Otálora, M. A. G., Aragón, G., Dimas, J., & Escudero, A. (2009). Fragment quality and matrix affect epiphytic performance in a Mediterranean forest landscape. American Journal of Botany, 96, 1974-1982. Berg, N. (2006). Age and size of hollow oaks and their associated lichen flora and beetle fauna. Final thesis, Linköping University, LiTH-IFM-Ex--05/1605--S. Blair, R. B. (1999). Birds and butterflies along an urban gradient: surrogate taxa for assessing biodiversity? Ecological Applications, 9, 164-170. Blair, R. B., & Launer, A. E. (1997). Butterfly diversity and human land use: species assemblages along an urban gradient. Biological Conservation, 80, 113-125. Bulut, Y., Toy, S., Irmak, M. A., Yilmaz, H., & Yilmas, S. (2008). Urban-rural climatic differences over a 2-year period in the city of Erzurum, Turkey. Atmósfera, 21(2), 121-133. Czech, B., Krausman, P. R., & Devers, P. K. (2000). Economic associations among causes of species endangerment in the United States. BioScience, 50, 593-601. de Wit, T. (1976). Epiphytic lichens and air pollution in the Netherlands. Bibliotheca Lichenologica, 5, 1-227. Dolan, R. W., Moore, M. E., & Stephens, J. D. (2011). Documenting effects of urbanization on flora using herbarium records. Journal of Ecology, 99, 1055-1062. Donovan, G. H., & Butry, D. T. (2010). Trees in the city: valuing street trees in Portland, Oregon. Landscape and Urban Planning, 94, 77-83. Donovan, G. H., & Butry, D. T. (2011). The effect of urban trees on the rental price of single- family homes in Portland, Oregon. Urban Forestry and Urban Greening, 10, 163-168. ESRI. (2011). ArcGIS Desktop: Release 10. Redlands, CA: Environmental Systems Research Institute. Fattorini, S. (2012). Insect extinction by urbanization: a long term study in Rome. Biological Conservation, 144, 370-375. Forys, E. A., & Allen, C. R. (2005). The impacts of sprawl on biodiversity: the ant fauna of the lower florida keys. Ecology and Society, 10(1), 25. [online] URL: http://www.ecologyandsociety.org/vol10/iss1/art25/. Foucard, T. (2001). Svenska skorplavar och svampar som växer på dem. Interpublishing, Stockholm. [in Swedish] Gagne, S. A., & Fahrig, L. (2011). Do birds and beetles show similar responses to urbanization? Ecological Applications, 21, 2297-2312.

George, K., Ziska, L. H., Bunce, J. A., & Quebedeaux, B. (2007). Elevated atmospheric CO2 concentration and temperature across an urban–rural transect. Atmospheric Environment, 41, 7654-7665. Gilbert, O. L. (1968). Bryophytes as indicators of air pollution in the Tyne Valley. New Phytologist, 67(1), 15-30. Giordani, P. (2007). Is the diversity of epiphytic lichens a reliable indicator of air pollution? A case study from Italy. Environmental Pollution, 146, 317-323. Gorbach, N V., & Kobzar, N. N. (1981). Determination of the age of epiphyte lichen Hypogymnia physodes (L.) Nyl. Doklady Akademii Nauk BSSR [Minsk], 25, 848-851. Grindon, L. H. (1859). The Manchester flora. W. White, London. Gärdenfors, U. ed (2010). The 2010 red list of Swedish species. ArtDatabanken, SLU, Uppsala. Hahs, A. K., McDonnell, M. J., McCarthy, M. A., Vesk, P. A., Corlett, R. T., Norton, B. A., Clemants, S. E., Duncan, R. P., Thompson, K., Schwartz, M. W., & Williams, N. S. G. (2009). A global synthesis of plant extinction rates in urban areas. Ecology Letters, 12, 1165-1173. Hawksworth, D. L., & Rose, F. (1970). Qualitative scale for estimating sulphur dioxide pollution in England and Wales using epiphytic lichens. Nature, 227, 145-148. Hedenås, H., & Ericson, L. (2008). Species occurrences at stand level cannot be understood without considering the landscape context: cyanolichens on aspen in boreal Sweden. Biological Conservation, 141, 710-718. Hughes, K. (2006). The impact of urban areas on climate in the UK: a spatial and temporal analysis, with an emphasis on temperature and precipitation effects. Earth & E-nvironment, 2, 54-83. Ishitani, M., Kotze, D. J., & Niemelä, J. (2003). Changes in carabid beetle assemblages across an urban-rural gradient in Japan. Ecography, 26, 481-489. Johnson, D. W. (1979). Air pollution and the distribution of corticolous lichens in Seattle, Washington. Northwest Science, 53, 257-263. Johansson, V., Bergman, K.-O., Lättman, H., & Milberg, P. (2009). Tree and site quality preferences of six epiphytic lichens growing on oaks in southeastern Sweden. Annales Botanici Fennici, 46, 496-506. LeBlanc, F., Rao, D. N., & Comeau, G. (1972). The epiphytic vegetation of Populus balsamifera and its significance as an air pollution indicator in Sudbury, Ontario. Canadian Journal of Botany, 50, 519-528. Linköpings kommun. (2011). Linköpings kommun, Statistik & utredningar, Statistikinfo (Linköping kommun, Statistics & reports, statistical information) 2011:7. Lisowska, M. (2011). Lichen recolonisation in an urban-industrial area of southern Poland as a result of air quality improvement. Environmental Monitoring and Assessment, 179, 177-190. Liu, W., You, H., & Dou, J. (2009). Urban-rural humidity and temperature differences in the Beijing area. Theoretical and Applied Climatology, 96, 201-207. Lohr, V. I., Pearson-Mims, C. H., Tarnai, J., & Dillman, D. A. (2004). How urban residents rate and rank the benefits and problems associated with trees in cities. Journal of Arboriculture, 30, 28-35. Magura, T., Tóthmérész, B., & Molnár, T. (2008). A species-level comparison of occurrence patterns in carabids along an urbanisation gradient. Landscape and Urban Planning, 86, 134-140.

McDonnell, M. J., Pickett, S. T. A., Groffman, P., Bohlen, P., Pouyat, R. V., Zipperer, W. C., Parmelee, R. W., Carreiro, M. M., & Medley, K. (1997). Ecosystem processes along an urban-to-rural gradient. Urban Ecosystems, 1, 21-36. McKinney, M. L. (2006). Urbanization as a major cause of biotic homogenization. Biological Conservation, 127, 247-260. McKinney, M. L. (2008). Effects of urbanization on species richness: a review of plants and animals. Urban Ecosystems, 11, 161-176. doi:10.1007/s11252-007-0045-4 Mercado Cárdenas, A., & Buddle, C. M. (2009). Introduced and native ground beetle assemblages (Coleoptera: Carabidae) along a successional gradient in an urban landscape. Journal of Insect Conservation, 13, 151-163. Muhammadi, U. H. (2011). Spatial pattern of occurrence of eleven epiphytic lichen species in a heterogeneous landscape. Master thesis, Linköping University, LiTH-IFM-Ex--11/2441--SE. Nash III, T. H., Ryan, B. D., Gries, C., & Bungartz, F. (eds.) (2002). Lichen flora of the greater Sonoran Desert region. Vol. I. State University, Tepme, Arizona. Nash III, T. H., Ryan, B.D., Diederich, P., Gries, C., & Bungartz, F. (eds.) (2004). Lichen flora of the greater Sonoran Desert region. Vol. II. State University, Tepme, Arizona. Nash III, T. H., Gries, C., & Bungartz, F. (eds.) (2007). Lichen flora of the greater Sonoran Desert region. Vol. III. State University, Tepme, Arizona. Niemelä, J. (1999). Is there a need for a theory of urban ecology? Urban Ecosystems, 3, 57-65. Niemelä, J., & Kotze, D. J. (2009). Carabid beetle assemblages along urban to rural gradients: a review. Landscape and Urban Planning, 92, 65-71. Niemelä, J., Kotze, D. J., Venn, S., Penev, L., Stoyanov, I., Spence, J., Hartley, D., & de Oca, E. M. (2002). Carabid beetle assemblages (Coleoptera, Carabidae) across urban-rural gradients: an international comparison. Landscape Ecology, 17, 387-401. Nitare, J. (2010). Signalarter. Indikatorer på skyddsvärd skog. Flora över kryptogamer (4th ed.) (Indicators for valuable forest areas. A cryptogam flora). Jönköping: Skogsstyrelsen. Nylander, W. (1866). Les lichens du Jardin du Luxembourg (Lichens in Jardin du Luxembourg). Bulletin de la Société Botanique de France, 13, 364-372. Paltto, H., Thomasson, I., & Nordén, B. (2010). Multispecies and multiscale conservation planning: setting quantitative targets for red-listed lichens on ancient oaks. Conservation Biology, 24, 758-768. Pedersen, I. (1980). Epiphytic lichen vegetation in an old oak wood, Kaas skov. Botanisk tidsskrift, 75, 105-120. Price, S. J., Dorcas, M. E., Gallant, A. L., Klaver, R. W., & Willson, J. D. (2006). Three decades of urbanization: estimating the impact of land-cover change on stream salamander populations. Biological Conservation, 133, 436-441. Ranta, P. (2001) Changes in urban lichen diversity after a fall in sulphur dioxide levels in the city of Tampere, SW Finland. Annales Botanici Fennici, 38, 295-304. Robinson, L., Newell, J. P., & Marzluff, J. M. (2005). Twenty-five years of sprawl in the Seattle region: growth management responses and implications for conservation. Landscape and Urban Planning, 71, 51-72. RUS. (2012). Regional utveckling och samverkan i miljömålssystemet (Regional development and cooperation in the environmental system). Nationella emissionsdatabasen. http://www.rus.lst.se/nyheter.html.

Santesson, R., Moberg, R., Nordin, A., Tønsberg, T., & Vitikainen, O. (2004). Lichen-forming and lichenicolous fungi of Fennoscandia. Museum of Evolution, Uppsala University, Uppsala. Statistics Sweden. (2011). Statistisk årsbok för Sverige 2011 (Statistical yearbook of Sweden 2011). Official statistics of Sweden, Örebro. Statsoft, I. (2011). STATISTICA (Data Analysis Software System), Version 10. www.statsoft.com. Su, Z., Zhang, R., & Qiu, J. (2011). Decline in the diversity of willow trunk-dwelling weevils (Coleoptera: Curculionoidea) as a result of urban expansion in Beijing, China. Journal of Insect Conservation, 15, 367-377. doi:10.1007/s10841-010-9310-6 Tarhanen, S., Poikolainen, J., Holopainen, T., & Oksanen, J. (2000). Severe photobiont injuries of lichens are strongly associated with air pollution. New Phytologist, 147, 579-590. Tóthmérész, B., Máthé, I., Balázs, E., & Magura, T. (2011). Responses of carabid beetles to urbanization in Transylvania (Romania). Landscape and Urban Planning, 101, 330-337. van Herk, C. M. (2001). Bark pH and susceptibility to toxic air pollutants as independent causes of changes in epiphytic lichen composition in space and time. The Lichenologist, 33, 419- 441. van Herk, C. M., Aptroot, A., & van Dobben, H. F. (2002). Long-term monitoring in the Netherlands suggests that lichens respond to global warming. The Lichenologist, 34(2), 141- 154. Washburn, S. J., & Culley, T. M. (2006). Epiphytic macrolichens of the greater Cincinnati metropolitan area―Part II: distribution, diversity and urban mycology. The Bryologist, 109, 516-526.

Fig. 1. The study area in and around Linköping, SE Sweden. The map in the middle shows the position of the rural oaks (triangles) around Linköping (grey), while the bottom map shows urban (circles) and rural oaks in and near Linköping (grey).

Fig. 2. The concentric expansion of Linköping, SE Sweden, over the past 300 years. Reproduced with kind permission from the municipality of Linköping.

Fig. 3. (a)-(c) A comparison of the 14 investigated lichen species‟ occurrence and cover on oaks in urban and rural environments. (a) shows the proportion of trees with occurrence. (b) shows the average cover (%) on trees when present. (c) shows the average cover (%) based on all oaks. Error bars show 95% confidence intervals (binomial ditto in 3a).

Fig. 4. The median of 14 target lichen species and the per cent of the trunks of Quercus robur covered in rural and urban areas (a and c). In b and d, the cover of target lichen species on urban oaks is shown as a function of when urbanisation occurred (based on the average age of the five closest buildings).

Fig. 5. The predicted values based on results from a logistic regression of finding Calicium viride, Chrysotrix candelaris, Evernia prunastri and Ramalina spp. on trees in relation to the age of urbanisation.

Table 1The 17 target lichens searched for on Quercus robur. Nine of the lichens are common, and eight are red-listed and/or indicator species with varying requirements of the substrate. Status Lichen species Common Threat1 Indicator2 Substrate preference3 Calicium adspersum X On rough bark of Quercus, rarely on old wood. Calicium quercinum VU Corticolous on Quercus, etc., sometimes on old wood. Calicium viride X Corticolous (Alnus, Quercus, Betula, Pinus, etc.) and lignicolous. Chaenotheca phaeocephala X On lignum and bark of various trees (Quercus, etc.). Chrysothrix candelaris X Corticolous and lignicolous. Cliostomum corrugatum NT X On rough bark of Quercus, Ulmus, etc., also lignicolous. Cyphelium inquinans X Lignicolous, rarely corticolous (on Picea, Quercus and Betula) and saxicolous. Evernia prunastri X Corticolous or sometimes lignicolous, rarely saxicolous. Hypocenomyce scalaris X Corticolous (esp. on Pinus) and lignicolous. Toxitolerant. Hypogymnia physodes X Corticolous, lignicolous and less often saxicolous. Lecanographa amylacea VU Corticolous on Quercus. Lepraria incana X Corticolous and lignicolous, also saxicolous (under overhangs). Parmelia sulcata X Corticolous, lignicolous and saxicolous, often on rocks manured by birds. Ramalina spp. X - Schismatomma decolorans NT X Corticolous, esp. on old Quercus. Sclerophora coniophaea NT X Corticolous on Quercus, Ulmus, etc., in northern areas usually lignicolous on conifers. Usnea spp. X - 1 Gärdenfors (2010) 2 Nitare (2010) 3 Santesson Moberg, Nordin, Tønsberg, and Vitikainen (2004)

530 60.0 60.0 2.55 1.53 1.35 1.04 0.91 0.71 15.8 24.8 28.3 35.3 30.9 19.8 Max

were

es es

4.0 0.0 0.0 0.0 0.0 0.0 0.0 250 Min 12.6 0.14 0.05 0.03 0.01 0.01 0.00 Rural (n=109) Rural

0.2 0.5 0.7 0.9 1.1 1.4 351 29.2 32.6 0.57 0.36 0.29 0.22 0.18 0.15 Average

574 57.5 63.0 3.96 1.58 1.01 0.71 0.56 0.36 88.2 78.7 77.0 71.8 75.6 66.3 Max

0 8.8 0.0 0.0 0.0 5.1 7.1 230 Min 0.14 0.05 0.03 0.01 0.01 0.02 10.4

Urban (n=105) Urban

341 24.8 36.8 0.85 0.41 0.29 0.20 0.18 0.13 35.9 39.8 40.8 40.6 38.6 35.8 Average

in urban environments. The average depth of bark crevices is based on averages from each from tree iseach depthcrevices based on averages of bark environments.average The in urban

within:

north, east, south and west. The summarised number of oak trees (including the focal tree) and summarised density summarised the (including density tree) and focal of Thetrees number oak north, south summarised and west. east, ) within: 1 - Quercus robur

the ation of the urban and rural oak populations studied, and the explanatory variables used when modelling the occurrence of the occurrence when modelling variables used oak populations rural the urban and explanatory studied, of ation and from is

m m m m m m m

exposure (%) exposure lculated from a tree database and maps, and respectively. database a tree lculated from 150 250 350 500 700 1000 150 250 m 350 m 500 m 700 m 1000 m Table 2 Character on lichens target measured tree on sixdistanc radius encloseddifferent the focal high, low and surrounding industrial area, buildings buildings; of ca (cm) height breast at circumference Oaks (mm) crevices bark Depth of Sun (ha of oaks Density (%) buildings by Area covered 24

/0.022* +/0.16 incana – Lepraria Lepraria +/0.036* 0.0037** or or

value is

viride /0.017* +/0.069 /0.0033** – +/0.033* Calicium Calicium – 0.00007***

+/0.061 +/0.099 /0.0043** candelaris – +/0.0014** Chrysothrix Chrysothrix

<0.00001***

/0.046* /0.044* 0.021* Evernia Evernia – – prunastri

/0.039* 0.034* sulcata – Parmelia Parmelia

spp. /0.055 /0.013* – – Ramalina 0.00066*** value <0.05; <0.05; *** <0.001. value ** <0.01; -

were analysed in relation to the explanatory variables shown. The P in relation towere the explanatory analysed

+/0.013* * P indicates /0.00076*** 0.000043*** Chaenotheca Chaenotheca – phaeocephala

on urban trees

/0.13 /0.14 /0.12 0.078 – – – +/0.014* physodes Hypogymnia Hypogymnia

model

m m

m m m m m m m m m m

value for the selected selected the for value 150 250 350 500 700 1000 150 250 350 500 700 1000 - indicator lichen indicator

Table 3 of species lichens eight occurrence The variables includedmodel positive species (based onindicate each a for in the best minus Plus and signs for AIC). displayed variables. association response with the negative of urbanisation Age Tree circumference crevices Bark Sun exposure within of oaks Density within buildings by Area covered P a 25

0.27 mode of mode of Foliose Foliose Crustose Crustose Crustose Crustose Crustose Crustose Crustose Crustose Crustose Fruticose Fruticose Fruticose Growth form Growth

value <0.05; <0.05; value ** <0.01; - NA Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Trebouxia Photobiont Trentepohlia Chlorococcus Chlorococcus Chlorococcus

No No No No No No No Yes Yes Yes Yes 0.16 Rarely Rarely Rarely Pycnidium

spp. by isidium. * spp. by P indicates

No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes 0.19 Diaspores Usnea

) 2

six traits, i.e. if they mainly dispersespores or diaspores, six by the i.e. if they traits, 123 82.5 21.6 42.4 23.6 20.6 29.4 81.0 33.2 68.7 67.5 63.8 0.95 Spore area (µmarea Spore h an addition in addition taxah an

No No Yes Yes Yes Yes Yes Yes Yes Rarely Rarely Rarely Rarely Rarely Spores 0.039* (2004)

08 07 06 06 06 06 06 06 05 05 ------

0 0 0 0 on urban trees as affected by on urbanby trees affected as

6.66 E 5.62 E 1.60 E 2.02 E 2.22 E 4.75 E 6.14 E 8.41 E 1.19 2.89 E Cover (%) Cover

1,3

2 1,3 1,3

1,3 scalaris 1,3

inquinans 4 spp.

spp.

value - (2001) Foucard (2002) Bungartz and Gries, Ryan, III, Nash and Bungartz Gries, Diederich, Ryan, III, Nash (2007) Gries, and Bungartz III, Nash

Table 4 14 lichen cover species of cent Per to and the dispersal growth The photobiont or not, the present belong to which the form. if pycnidiumgroup area, are spores but soredium the included islichens for wit diaspores *** <0.001. Species adspersum Calicium Cyphelium coniophaea Sclerophora Usnea corrugatum Cliostomum Hypocenomyce physodes Hypogymnia Ramalina phaeocephala Chaenotheca sulcata Parmelia prunastri Evernia candelaris Chrysothrix viride Calicium incana Lepraria P 1 2 3 4 26