Urbanisation Impacts on Rare and Endangered Plant Species

Urbanisation impacts on rare and endangered plant species

vorgelegt von Dipl. Biol. Gregorio Planchuelo ORCID: 0000-0003-0760-5478

von der Fakultät VI - Planen | Bauen | Umwelt der Technischen Universität Berlin zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften - Dr. rer. nat. - genehmigte Dissertation

Promotionsausschuss:

Vorsitzender: Prof. Dr. Norbert Kühn

Gutachter: Prof. Dr. Ingo Kowarik

Gutachter: Dr. Moritz von der Lippe

Gutachter: Prof. Dr. Jonathan Jeschke

Tag der wissenschaftlichen Aussprache: 2. September 2019

Berlin 2019 2 Abstract

Given the rising urban land cover worldwide, the contribution of cities to the conservation of biodiversity becomes increasingly important. Previous research shows that urban environments can host high numbers of species, including endangered plant species. Yet, key questions on the urban contribution to plant conservation remain critically open, as little information is available on how populations of endangered plant species occur across different biotope types within cities and to what extent anthropogenically shaped vs. natural ecosystems provide habitats for endangered plants. This thesis is based on the analysis of a unique dataset with the exact geographical location of 1,742 populations of 213 endangered plant species in the city of Berlin, as well as information on the survival rate of 858 of these populations. Firstly, I assessed the relative importance of Berlin’s major biotope classes as habitats of endangered plant species and applied the novel ecosystem concept to quantify endangered plant populations for natural remnants vs. hybrid vs. novel ecosystems within Berlin. Secondly, I evaluated to which extent these populations of endangered plant species may occur in anthropogenically shaped hybrid or novel ecosystems spatially independently from natural remnants. Finally, I unravelled the most important drivers of population survival of endangered plant species in cities by linking population survival to various parameters (biotope class, patch size, habitat continuity, conservation status, proportion of impervious surfaces, floor space index, human population density, human population constancy, nearby street length, distance to the nearest street, distance to the nearest remnant ecosystem, proportion of nearby forests and grasslands, and plant traits). Results show that populations of endangered plant species were generally, although unevenly, associated with specific biotope classes; with forest, grassland, and ruderal biotopes as the most important habitats. Surprisingly, novel ecosystems harboured the highest numbers of total populations, of total species, and their populations were found to occur independently from natural remnants. Population success was highest in anthropogenically influenced biotopes such as green spaces, built up areas, fields and grasslands as well as in competitive species, while surprisingly it was lowest inside nature protection areas. Quantifying the relative importance of biotope classes and novel vs. (near-)natural ecosystems as habitats of endangered species demonstrates that the urban contribution to biodiversity conservation is best ensured by providing a wide range of ecosystems. Rather than prioritizing only natural remnants, I thus argue for broad approaches to urban biodiversity conservation that include novel ecosystems.

3 Zusammenfassung

Angesichts der weltweit steigenden Landnutzung durch städtische Gebiete wird der Beitrag, den Städte zum Erhalt der Biodiversität leisten können, immer wichtiger. Frühere Untersuchungen haben gezeigt, dass städtische Flächen eine hohe Artenzahl beherbergen können, einschließlich gefährdeter Pflanzenarten. Dennoch bleiben Schlüsselfragen zum Beitrag, den Städte zum Pflanzenartenschutz leisten können, offen, da wenig Informationen darüber vorliegen, wie Populationen gefährdeter Pflanzenarten in verschiedenen Biotoptypen innerhalb der Städte auftreten und inwieweit anthropogen geprägte vs. natürliche Ökosysteme Lebensräume für gefährdete Pflanzen bereitstellen. Die vorliegende Studie verwendet einen einzigartigen Datensatz mit der genauen geographischen Lage von 1.742 Populationen 213 gefährdeter Pflanzenarten in der Stadt Berlin sowie Informationen über die Überlebensrate von 858 dieser Populationen. Zunächst habe ich die relative Bedeutung der großen Berliner Biotopklassen als Lebensraum für bedrohte Pflanzenarten bewertet und das “Novel ecosystem concept” angewendet, um die Eignung von „natural remnants“ vs. „hybrid“ vs. „novel ecosystems“ in Berlin für gefährdete Pflanzenpopulationen zu bestimmen. Zweitens habe ich untersucht, inwieweit diese Populationen gefährdeter Pflanzenarten in anthropogen geprägten „hybrid“ oder „novel ecosystems“ räumlich unabhängig von „remnant ecosystems“ auftreten können. Schließlich ermöglichte die Verknüpfung der Wuchsorte der Populationen mit Parametern auf verschiedenen räumlichen Skalen, die wichtigsten Faktoren für den Populationserfolg zu ermitteln. Die Ergebnisse zeigen, dass Populationen gefährdeter Pflanzenarten im Allgemeinen, wenn auch ungleichmäßig, mit bestimmten Biotopklassen assoziiert waren. Dabei waren Wald-, Grünland- und Ruderalbiotope die wichtigsten Lebensräume. Überraschenderweise beherbergten „novel ecosystems“ die höchste Anzahl von Gesamtpopulationen und Gesamtarten, und ihre Pflanzenpopulationen traten räumlich unabhängig von „natural remants“ auf. Am größten war der Populationserfolg in anthropogen beeinflussten Biotopen wie Grünflächen, bebauten Flächen, Feldern und Grünland sowie bei konkurrierenden Arten, und erstaunlicherweise niedriger innerhalb von Shutzgebieten als außerhalb. Die Ergebnisse zeigen, dass der städtische Beitrag zum Erhalt der Biodiversität am besten durch die Bereitstellung einer vielseitigen Bandbreite an Ökosystemen gewährleistet ist. Anstatt nur Relikte von Naturlandschaften zu priorisieren, plädiere ich für weitreichende Ansätze zum Schutz der städtischen Biodiversität, die auch neuartige Ökosysteme beinhalten.

4 Acknowledgements

Even if just my name appears below the title of the thesis, this has been a journey coloured by the closest people of my academic and personal life. Their direct and indirect participation has been essential in making this possible.

First of all, I would like to thank my supervisors Ingo Kowarik and Moritz von der Lippe for having taught me so much - academically and personally. Ingo, for having given me the great opportunity to pursue my doctoral studies in such a beautiful topic in such a special city of the world, and for the inspiration that has transformed my view of cities forever. Moritz, for having opened my eyes to a new world of statistics and for being there not only as a supervisor, but also as a friend.

A special mention also to Juan Antonio for discovering me for the first time the field of ecology many years ago, and for opening the door to my first research. To Luis Balaguer for his immense inspiration, even in such a limited time frame.

I would also like to thank Vanda, whom I admire so much, for helping me grow as fast as a tree. To my parents Gregorio and Regina and to my sister Elisa for being my pillar in life.

Special thanks likewise to all the researchers at the Institute of Ecology for their encouragement and support throughout the doctoral research. Divya, for sharing a laugh whenever we can, and for breaking the ice and guiding me in the bureaucratic chaos that is the submission of a doctoral thesis. Andreas, for his perseverance and steady work. Leonie, for her support, especially during my participation at the SER conference. Birgit, for her wonderful insights on the Berlin flora. Robert, for his uplifting sense of humour. Heinke, for keeping me company in the empty faculty many times during late-night hours. Anne, for helping me navigate the enormous plant trait databases. Bela, for her uplifting spirit and the nice talks.

Many thanks to Justus Meißner and the Stiftung Naturschutz Berlin for the amazing datasets that have enabled such detailed analyses. Likewise I would like to thank Berlin’s Senate Department for the Environment, Transport and Climate Protection for providing GIS layers of many maps of Berlin.

Thanks to Jonas for being a best friend. Ever onward! To Yuri, for showing me his approach to life, and for so many incredible and crazy moments together. Best day ever! To Ricky for the babies. To Alex, for his drinks. To Flora, for her high vibes. To Javi, for being the best neighbour. Special thanks to Urska, for being the trigger of this journey. To my good old Gonzalo, Alfonso and Lucas, for letting me know that even if far away, they are always there. To Carlos for our late night scientific talks that soon divert into philosophy. To Ulf and Lara, for knowing that I have PhD buddy at the other side of the country and the world, but very close to my heart.

I am also extremely thankful to the Hans-Böckler Stiftung and the TU-Berlin + DAAD for awarding me scholarships. Without their financial support, this project would not have been possible.

And to Berlin, for being such a wonderful niche for growing, for learning, and for living.

5 Publications

The following publication is part of this thesis:

• G. Planchuelo, M. von der Lippe M, I. Kowarik, 2019, Untangling the role of urban ecosystems as habitats for endangered plant species. Landscape and Urban Planning. DOI: https://doi.org/10.1016/j.landurbplan.2019.05.007

- The pre-print version of the publication can be found in Chapter 2.

This dissertation also includes manuscripts that have been submitted for consideration for publication:

• G. Planchuelo, I. Kowarik, M. von der Lippe M, 2019, Endangered plants in novel urban ecosystems are filtered by strategy type and dispersal syndrome, not by spatial dependence on natural remnants

The pre-print version of the manuscript submitted to Frontiers in Ecology and Evolution can be found in Chapter 3.

• G. Planchuelo, I. Kowarik, M. von der Lippe M, 2019, Survival of endangered plants in the city: the role of plant traits, biotopes, protected areas and urbanization

- The pre-print version of the manuscript submitted to Journal of Applied Ecology can be found in Chapter 4.

6 Contents Abstract ...... 3 Acknowledgements ...... 5 Publications ...... 6 Chapter 1: Introduction ...... 8 Importance of cities as habitats for endangered plant species ...... 9 The novel ecosystem concept ...... 11 Research area ...... 12 Gaps in knowledge and research questions ...... 13 Data sources and methodological approach ...... 14 Outline of the thesis ...... 19 Chapter 2: Untangling the role of urban ecosystems as habitats for endangered plant species ...... 25 Abstract ...... 26 Introduction ...... 27 Methods ...... 29 Results ...... 34 Discussion ...... 40 Chapter 3: Plant strategy and dispersal syndrome drive endangered species’ colonization into the urban matrix ...... 51 Abstract ...... 52 Introduction ...... 53 Methods ...... 55 Results ...... 61 Discussion ...... 66 Chapter 4: Survival in the city? Population persistence of endangered plant species across different urban ecosystems ...... 74 Abstract ...... 75 Introduction ...... 76 Methods ...... 80 Results ...... 85 Discussion ...... 89 Chapter 5: Synthesis ...... 101 Implications for conservation in cities ...... 105 Future research ...... 108 Appendix ...... 115

7 Chapter 1

Chapter 1: Introduction

Chapter 1

“One of the first conditions of happiness is that the link between man and nature shall not be broken.”

L. Tolstoy

Importance of cities as habitats for endangered plant species

Cities cover a rather small proportion of the land surface (Schneider et al. 2009) but host more than half of the world population (UN 2015). It is projected that in the next 30 years there will be an increase of 2.5 billion urban dwellers worldwide (UN 2015), while urban land cover will grow at a rate twice as fast (Angel et al. 2011; Seto et al. 2011). Only by 2030, there will be an increase of 1.2 million km2 to urban land cover (Seto et al. 2012), an area roughly the size of Mongolia.

With these prospects of accelerating urbanisation, the importance of cities to biodiversity conservation becomes increasingly important. Urban regions have been shown to host high numbers of species due to their high habitat heterogeneity, human agency, and the location chosen for their establishment, which is often naturally high in geomorphological and biological richness (Kühn et al. 2004; McKinney 2008). For instance, a wide range of sites with high disturbance levels (Kowarik 1990), as well as wastelands (Bonthoux et al. 2014), heavily urbanised areas (Jokimaki et al. 2018), gardens (Goddard et al. 2010), remnant patches (Aronson et al. 2017), novel ecosystems (Kowarik & von der Lippe 2018) or green roofs (Dvorak & Volder 2010) can all serve as suitable habitats for species from the regional species pool (Aronson et al. 2014) as well as for endangered species (Schwartz et al. 2002; Lawson et al. 2008; Lenzen et al. 2009; Shwartz et al. 2014; Ives et al. 2016).

Indeed, 30% of Australia’s endangered species occur in cities, which have been shown to support more endangered species per unit area than rural landscapes, with some of them being highly restricted to urban environments (Ives et al. 2016). In the USA numbers are similar, and 22% of endangered species are found in the 40 biggest cities (Schwartz et al. 2002). This is consistent with the findings of Lenzen et al. (2009), which reveal that built-up areas seem to have a conserving influence on endangered species when assessing land use patterns at a country scale, even in small, isolated populations such as the ones found in the fragmented landscapes of urban areas (Lawson et al. 2008). A special focus on threatened urban grasslands through a floristic survey of differently sized conservation reserves in Melbourne concluded that small patches have similar yet complementing conservation potential when compared to bigger areas, highlighting the importance of a broad range of patch sizes for urban biodiversity

Chapter 1 conservation (Kendal et al. 2017). In consequence, the last years have seen a rise in conservation policies applied to urban ecosystems, with arguments that a great variety of ecosystems offer habitat functions for native species and could conserve biodiversity (Goddard et al. 2010; Kowarik 2011; Lepczyk et al. 2017; Nilon et al. 2017). This topic gains special relevance in light of the growing evidence suggesting positive links between biodiversity and ecosystem services (Schmitt-Harsh et al. 2013; Briguiche & Zidane 2016; Capotorti et al. 2016), which show that biodiversity conservation could have positive benefits on the well-being and mental health of urban dwellers (Dean et al. 2011; Engemann et al. 2019).

While the significance of urban remnant biotopes for biodiversity conservation is well understood (e.g. Godefroid and Koedam 2003), there is increasing evidence that some anthropogenic biotopes can also serve for biodiversity conservation in cities - such as parks (Cornelis & Hermy 2004; Nielsen et al. 2014), vacant land (Bonthoux et al. 2014; Anderson & Minor 2017) or urban grasslands (Smith et al. 2006; Fischer et al. 2013a). To date, however, the extent to which urban ecosystems that differ in ecological novelty (i.e., along the spectrum from natural remnants to novel ecosystems; Hobbs et al. 2009) function as habitats of large sets of endangered plant species has been only quantified once (Kowarik & von der Lippe 2018), revealing that endangered species were unevenly distributed across the different ecosystems, with a higher proportion in natural remnants (Kowarik & von der Lippe 2018). Yet, a key question on the urban contribution to biodiversity conservation remains critically open, as little information is available on how populations of endangered plant species occur across different biotope types within cities and to what extent anthropogenically shaped vs. natural ecosystems provide habitats for populations of endangered plants (Shwartz et al. 2014).

Indeed, varying human impacts and a range of environmental barriers filter the composition of urban species assemblages (Williams et al. 2009; Aronson et al. 2016; Knapp et al. 2017; Kowarik & von der Lippe 2018), so whether endangered plant species can colonize into different urban ecosystems depends on how they negotiate with a set of barriers related to dispersal (Kowarik & von der Lippe 2018). A yet unexplored question for understanding the urban contribution to biodiversity conservation is whether populations of endangered plant species may colonize anthropogenically-shaped urban habitats independently from source populations located in natural remnants - or whether populations currently located outside these remnants are still spatially confined to them.

However, the mere occurrence of endangered plant species across a wide range of urban ecosystems does not necessarily indicate their successful establishment (Shwartz et al. 2014), as harsh environmental settings in some areas can act as a reproduction barrier that prevents

Chapter 1 the long-term establishment of populations (Lundholm et al. 2010; Kowarik & von der Lippe 2018). Thus, the occurrence of endangered plant populations in urban ecosystems may possibly mask future extinction debts (Hahs & McDonnell 2014). Unfortunately, studies on population dynamics in cities are scarce and do not distinguish between different urban habitats (i.e. Schwartz et al. 2013) nor do they account for specific environmental and socioeconomic factors to explain the survival of species of conservation concern in urban settings.

The novel ecosystem concept

As urban land use types with apparently similar conditions can host different assemblages of species (Godefroid & Ricotta 2018), new approaches that divide the urban landscape into different categories can be important to better understand species distribution patterns in cities (Godefroid & Koedam 2007; Godefroid & Ricotta 2018). A recent study from Berlin (Kowarik & von der Lippe 2018) has made a first step by assessing the role of highly aggregated novelty types (i.e., natural remnants, hybrid and novel ecosystems) for alien, native and endangered plant species. They have adapted the classification of ecological novelty from Hobbs et al. (2009) and specified it for urban settings by considering the origin of the site (natural vs. anthropogenic) and whether it is dominated by natural ecosystem processes or by human interference (Table 1). This has led to three novelty categories:

• Natural remnant ecosystems are relicts of natural ecosystems that remain within their historical range of variability, although they are often slightly affected by urban impacts. Examples range from near-natural forests, mires and wetlands to moderately used dry or wet grasslands. • Hybrid ecosystems are human-mediated ecosystems that have been modified from their historical state but still retain the potential to approach historical conditions. These include many young tree plantings in forests and parks, managed grasslands, extensively managed urban green spaces, or low-intensity pastures. • Novel ecosystems are human-modified ecosystems that have likely been irreversibly changed by profound impacts on abiotic conditions or biotic composition. Novel ecosystems include built-up areas, vacant lots, rooftops, abandoned industrial areas, or high-intensity agricultural land.

Chapter 1

Natural remnant Hybrid ecosystems Novel ecosystems ecosystems Origin Natural Natural or Anthropogenic anthropogenic Level of natural High Medium Low or high ecosystem processes Level of human Low Medium Low or high interference Examples Near-natural forests, Young tree plantings Built-up areas, mires and wetlands to in forests and parks, vacant lots, rooftops, moderately used dry managed grasslands, wastelands, or wet grasslands urban green spaces, abandoned or low-intensity industrial areas, or pastures high-intensity agricultural land Table 1 Differentiation of three major novelty categories in cities according to different levels of ecosystem novelty (i.e. natural remnant, hybrid, and novel ecosystems). Types have been specified according to the origin, the level of natural ecosystem processes and the level of human interference (source: chapter 2).

Research area

This study was carried out in the city of Berlin, the capital and most populous city of Germany with 3.6 million inhabitants as of 2017. It has a total area of 891km2, from which 59% is covered by built-up areas and streets, while green and blue spaces occupy 41% of the area - including forests (17.7%), lakes and rivers (6.1%), parks (5.6%), allotment gardens (5.3%), fields (5%) and meadows (1.3%) (SenStadtUm 2014). The climate is temperate, with forests and wetlands as natural ecosystems. The total area of Berlin encompasses remnants of natural and agrarian landscapes, urban greenspaces with different land-use histories, and a range of novel ecosystems on vacant urban land and within the built-up areas (Sukopp 1990).

Berlin was founded in the Medieval (1237) and remained embedded in the rural countryside for centuries. Berlin was a small city for centuries until its population rapidly grew from ca. 424,000 inhabitants in 1849 to four million inhabitants in 1925. Due to bombings during World War II and a stagnated urban development until the German reunification in 1989, there was a considerable amount of vacant land within the city that was recolonized by natural succession into a range of novel ecosystems. While some of these areas were integrated into urban parks, others were re-built as Berlin is growing again (Sukopp 1990; Lachmund 2013).

Chapter 1

Gaps in knowledge and research questions

Although the number of related urban conservation studies is sharply increasing, the contributions of urban areas to nature conservation are not yet convincingly supported by empirical research and important gaps of knowledge still remain (Shwartz et al. 2014): a) Recent advances in understanding the processes that shape urban biodiversity are nearly exclusively derived from analyses at larger spatial scales (e.g. Lenzen et al. 2009 or Ives et al. 2016) or from mapping the plant assemblages of grid cells with highly heterogeneous land use types (e.g. Kühn et al. 2004). While these datasets portray the overall consequences of habitat function, little information is available on how populations of endangered plant species occur across different biotope types within cities and to what extent anthropogenically shaped vs. natural ecosystems provide habitats for endangered plants. b) Research on the urban contribution to biodiversity conservation often resorts to biodiversity measures in terms of total species, failing to identify individual species and their relevance for nature conservation (e.g., common species versus endangered species), as high numbers in common or introduced species may mask the decline of rare, endangered or species of special conservation interest (Chocholouskova & Pysek 2003; Knapp et al. 2010). c) Added to this, assessing biodiversity exclusively in terms of species numbers often overlooks the population level (e.g. Ives et al. 2016), yet considering population survival would unravel more precisely related extinction risks and the drivers and future perspectives of species survival (Shwartz et al. 2014).

The present research makes further steps forward towards understanding the potential of cities to nature conservation, first, by working with extremely detailed maps at the biotope level that reflect the varying degrees of novelty of the different ecosystems of Berlin; second, by focusing on spatially referenced populations plant species that had been identified as endangered species; and third, by assessing the status of these populations throughout a period of various years in order to understand the drivers of population survival. In detail, I address the following research questions:

1. What is the relative contribution of different biotope classes and of natural remnant vs hybrid vs novel ecosystems within the city of Berlin in harbouring populations of endangered plant species?

Chapter 1

2. To which extent endangered plant species in an urban landscape may occur in anthropogenically shaped hybrid or novel ecosystems spatially independently from natural remnants?

3. What are the most important urban drivers in determining population survival of endangered plant species within the metropolitan region of Berlin?

I hypothesise significant differences between biotope classes and novelty types in their role as habitats for endangered plant species in cities. I foresee an important role of natural remnants in hosting endangered plant species, though anthropogenic biotopes and novel ecosystems will also offer great opportunities for biodiversity conservation in cities. I also believe that populations of endangered plant species in highly anthropogenic urban sites such as novel ecosystems can occur independently from source populations in natural remnants, and that these will tend to have ruderal strategies. Finally, I expect that population survival will vary considerably amongst biotope classes and will tend to increase in species with competitive strategies.

Data sources and methodological approach

Populations of endangered plant species

A set of 230 plant species of highest conservation priority in Berlin had been identified within the framework of Berlin’s Flora Protection Program, based on a systematic assessment approach (Seitz 2007; Senatsverwaltung für Stadtentwicklung 2010). All of these priority species are being endangered in Berlin and/or at higher spatial scales (i.e. in the surrounding federal state of Brandenburg, in Germany, Europe, or at a global scale). In this study, I used data on the spatial location of populations of priority species that were available for 213 out of 230 species (Appendix A). These data covered a total of 1,742 populations (Fig. 1) and resulted from an initial survey that was carried out between 1990 and 2014 by a range of experts on behalf of the Stiftung Naturschutz Berlin (2015). These mapped occurrences were revisited again by the experts after a mean time span of 7.6 years as part of the monitoring process, giving us precise data on the individual survival of 858 of these populations, comprising a total of 179 species.

Chapter 1

Water bodies Forests Built up areas Other green spaces

Fig. 1 Location of all populations of rare and endangered plant species (n=1742, red dots) across some of the biotope types in Berlin.

Biotope map

Between 2001 and 2003, a combined work of 64 different mapping projects generated a detailed biotope map of Berlin (SenStadtUm 2014). Currently, the city of Berlin is divided into 79,268 patches, with each of these patches being assigned to one of the twelve major biotope classes: forests, grasslands, ruderal sites, standing bodies of water, built-up areas, bogs and marshes, groves and hedges, green spaces, fields, moving bodies of water, heaths, and “other types”. For the purpose of this study, I merged biotope classes with a low number of endangered plant populations with a similar biotope class. In particular I merged standing and moving bodies of water to water bodies, heaths were incorporated into grasslands, and “other types” (which was a plant nursery) was incorporated into fields. This resulted in a total of nine biotope classes that are further subdivided in up to 7 hierarchical levels that sum up to a total of 7,483 biotope types (Fig. 2).

15 Chapter 1

Fig. 2 Cropped section of the Berlin Tempelhof Airport showing the high detail of the biotope mapping (left, SenStadtUm 2014). Each patch is coloured differently according to its biotope type. The same part of the city can be seen in aerial photography (right, colour photo from SenStadtUm 2014).

The methodological approach of this thesis includes the idea to generate an ecosystem novelty map of Berlin by spatially merging the aforementioned biotope data with land use data. This map will be used to test for spatial relationships between the populations of endangered plant species and the different novelty categories. Some examples of how this ecological novelty classification could result for different locations in Berlin can be seen in Fig. 3.

Fig. 3 Examples of possible outcomes of the ecological novelty classification of different sites in Berlin.

Chapter 1

Drivers of population survival

In order to unravel the most important drivers of population survival, I used the following variables from the Berlin Senate Department of Urban Development and Environment / Berlin Senate Department of Urban Development and Housing:

• Habitat features: biotope class, patch size, habitat continuity, conservation status.

• Degree of urbanization in surrounding matrix: proportion of impervious surfaces, floor space index, human population density, human population constancy (percentage of residents that remain at the same address for at least 5 years), nearby street length, distance to the nearest street.

• Proximity of near natural ecosystems: distance to the nearest remnant ecosystem and proportion of nearby forests and grasslands.

Additionally, I also used data on the following plant traits from the BIOFLOR database (Klotz et al. 2002): Morphological traits, CSR strategy (according to Grime 1977) and realized niche of the species (according to Ellenberg et al. 1991).

The different habitat features were used in this thesis as variables to understand population survival because their relative importance for endangered plant species is critically understudied (Shwartz et al. 2014). Variables related to the degree of urbanisation were chosen because previous research has shown that urbanisation parameters are relevant in understanding urban biodiversity patterns (e.g., Beninde et al. 2015; Matthies et al. 2015; Anderson & Minor 2019). I also used variables related to plant traits for their great importance in explaining the success of specific sets of plants in urban ecosystems (Knapp et al. 2009; Fischer et al. 2013b; Williams et al. 2015).

Methodological approach

In order to address the first research question (chapter 2) I first linked with a GIS the geographical position of each of the 1742 populations of the 213 endangered plant species to the biotope class and the ecological novelty it is located at by means of spatial queries. The data was then statistically analysed to unravel the relative importance of different biotopes classes and ecological novelty types as habitats of endangered plant species (Fig. 4).

Chapter 1

Addressing the second research question (chapter 3) required using the same dataset. The geographical point data of all the populations was spatially analysed in order to unravel their clustering pattern as well as their spatial correlation towards natural remnant ecosystems. Some dispersal traits were used in a conditional inference tree to assess the reasons for this spatial patterning (Fig. 4).

Finally, in order to address the third research question (chapter 4) I spatially linked the geographical location of a subset of 858 populations of endangered plant species (with information on population survival) to the aforementioned urban parameters (biotope class, patch size, habitat continuity, legal conservation status, proportion of impervious surfaces, floor space index human population density, human population constancy, nearby street length, distance to the nearest street, distance to the nearest remnant ecosystem, proportion of nearby forests and grasslands, and plant traits). The data was then statistically analysed (GLMM and conditional inference tree) to find the most important drivers of population survival (Fig. 4).

Chapter 2 Statistical Analyses Relative importance of Proportion of populations Spatial queries biotope type and ecological in each biotope type and novelty for endangered plant ecological novelty species

Precise geographical location of Chapter 3 Statistical Analyses Dependance of endangered - Clustering patterns - 1742 populations plant species towards natural - Spatial dependance - 213 endangered species remnants and drivers of in - Cond. inference tree urban ecosystems

Map data Dispersal traits - Dispersal syndrome subset - Biotope classes - Seed morphology - Ecological novelty - CSR strategy

Monitoring population Statistical Analyses survival Chapter 4 - Descriptive statistics Spatial queries Drivers of population survival - GLMM - 858 populations of endangered plant species - Cond. inference tree - 179 endangered species

Map data Plant traits - Habitat continuity - Morphological traits - Conservation status - CSR strategy - Impervious surfaces - Realized niche - Floor space index - Human pop. density - Human pop. constancy - Street Network - Remnant ecosystems - Forests and grasslands Fig. 4 Visual summary of the methodological process of this study. Green boxes represent the data sources, blue boxes represent the methodological procedures, and white boxes represent the knowledge generated from each of the methodological procedures. Lines indicate which data is used in each procedure.

Chapter 1

Outline of the thesis

The research questions highlighted above are tackled throughout the three next chapters in the form of scientific articles. The main focus of each chapter along with the general structure of this thesis is indicated in Fig. 5. A concluding chapter synthesising the findings is presented after the three scientific papers.

Fig. 5 General structure of this thesis.

Chapter 2 explores the relative importance of different biotope types and of natural remnant vs hybrid vs novel ecosystems within the city of Berlin in harbouring populations of endangered plant species.

Chapter 3 is a focused research that deals with the findings discovered in the previous chapter and centres on the spatial distribution of these populations in order to unravel to which extent endangered plant species in an urban landscape may occur in anthropogenically shaped hybrid or novel ecosystems spatially independently from natural remnants.

Chapter 4 attempts to explain the mechanisms that underlie population survival in urban environments by relating a range of variables at different spatial scales to population survival.

Finally, chapter 5 gives a synthesis of the main results and provides some insights for conservation practices in cities, while indicating the perspectives of future research on this topic.

Chapter 1

References

Anderson E.C. and Minor E.S. 2017. Vacant lots: An underexplored resource for ecological and social benefits in cities. Urban Forestry & Urban Greening 21: 146-152. Anderson EC, Minor ES. 2019. Assessing social and biophysical drivers of spontaneous plant diversity and structure in urban vacant lots. Sci Total Environ 653:1272-1281. Angel S., Parent J., Civco D.L., Blei A. and Potere D. 2011. The dimensions of global urban expansion: Estimates and projections for all countries, 2000-2050. Progress in Planning 75: 53-107. Aronson M.F., La Sorte F.A., Nilon C.H., Katti M., Goddard M.A., Lepczyk C.A., Warren P.S., Williams N.S., Cilliers S. and Clarkson B. 2014. A global analysis of the impacts of urbanization on bird and plant diversity reveals key anthropogenic drivers. Proc. R. Soc. B. The Royal Society, p. 20133330. Aronson M.F., Nilon C.H., Lepczyk C.A., Parker T.S., Warren P.S., Cilliers S.S., Goddard M.A., Hahs A.K., Herzog C., Katti M., La Sorte F.A., Williams N.S. and Zipperer W. 2016. Hierarchical filters determine community assembly of urban species pools. Ecology 97: 2952-2963. Aronson M.F.J., Lepczyk C.A., Evans K.L., Goddard M.A., Lerman S.B., MacIvor J.S., Nilon C.H. and Vargo T. 2017. Biodiversity in the city: key challenges for urban green space management. Frontiers in Ecology and the Environment 15: 189-196. Beninde J, Veith M, Hochkirch A. 2015. Biodiversity in cities needs space: a meta-analysis of factors determining intra-urban biodiversity variation. Ecol Lett 18:581-592. Bonthoux S., Brun M., Di Pietro F., Greulich S. and Bouché-Pillon S. 2014. How can wastelands promote biodiversity in cities? A review. Landscape and Urban Planning 132: 79-88. Briguiche H, Zidane L. 2016. Ethnobotanical study of medicinal plants from El-Jadida City (Morocco). Lazaroa 37:145-151. Capotorti G, Del Vico E, Anzellotti I, Celesti-Grapow L. 2016. Combining the conservation of biodiversity with the provision of ecosystem services in urban green infrastructure planning: critical features arising from a case study in the metropolitan area of Rome. Sustainability 9:10. Chocholouskova Z. and Pysek P. 2003. Changes in composition and structure of urban flora over 120 years: a case study of the city of Plzen. Flora 198: 366-376. Colchester M. 1994. Salvaging nature: indigenous peoples, protected areas and biodiversity conservation. Diane Publishing. Cornelis J. and Hermy M. 2004. Biodiversity relationships in urban and suburban parks in Flanders. Landscape and Urban Planning 69: 385-401.

Chapter 1

Dean J, van Dooren K, Weinstein P. 2011. Does biodiversity improve mental health in urban settings? Medical hypotheses 76:877-880. Dvorak B. and Volder A. 2010. Green roof vegetation for North American ecoregions: a literature review. Landscape and Urban Planning 96: 197-213. Ellenberg H, Weber H, Dull R, Wirth V, Werner W, Paulissen D. 1991. Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobotanica 18:248 Engemann K, Pedersen CB, Arge L, Tsirogiannis C, Mortensen PB, Svenning J-C. 2019. Residential green space in childhood is associated with lower risk of psychiatric disorders from adolescence into adulthood. Proceedings of the National Academy of Sciences:201807504. Fischer L.K., von der Lippe M. and Kowarik I. 2013a. Urban land use types contribute to grassland conservation: The example of Berlin. Urban Forestry & Urban Greening 12: 263- 272. Fischer LK, von der Lippe M, Kowarik I. 2013b. Urban grassland restoration: which plant traits make desired species successful colonizers? Applied Vegetation Science 16:272-285. Goddard M.A., Dougill A.J. and Benton T.G. 2010. Scaling up from gardens: biodiversity conservation in urban environments. Trends Ecol Evol 25: 90-98. Godefroid S. and Koedam N. 2003. How important are large vs. small forest remnants for the conservation of the woodland flora in an urban context? Global ecology and biogeography 12: 287-298. Godefroid S. and Koedam N. 2007. Urban plant species patterns are highly driven by density and function of built-up areas. Landscape Ecology 22: 1227-1239. Godefroid S. and Ricotta C. 2018. Alien plant species do have a clear preference for different land uses within urban environments. Urban Ecosystems: 1-10. Grime JP. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. The American Naturalist 111:1169-1194. Hahs A.K. and McDonnell M.J. 2014. Extinction debt of cities and ways to minimise their realisation: a focus on Melbourne. Ecological management & restoration 15: 102-110. Hobbs R.J., Higgs E. and Harris J.A. 2009. Novel ecosystems: implications for conservation and restoration. Trends Ecol Evol 24: 599-605. Ives C.D., Lentini P.E., Threlfall C.G., Ikin K., Shanahan D.F., Garrard G.E., Bekessy S.A., Fuller R.A., Mumaw L., Rayner L., Rowe R., Valentine L.E. and Kendal D. 2016. Cities are hotspots for threatened species. Global ecology and biogeography 25: 117-126. Jokimaki J., Suhonen J. and Kaisanlahti-Jokimaki M.L. 2018. Urban core areas are important for species conservation: A European-level analysis of breeding bird species. Landscape and Urban Planning 178: 73-81.

Chapter 1

Kendal D., Zeeman B.J., Ikin K., Lunt I.D., McDonnell M.J., Farrar A., Pearce L.M. and Morgan J.W. 2017. The importance of small urban reserves for plant conservation. Biological conservation 213: 146-153. Klotz S, Kühn I, Durka W, Briemle G. 2002. BIOLFLOR: Eine Datenbank mit biologisch- ökologischen Merkmalen zur Flora von Deutschland. Knapp S., Kühn I., Mosbrugger V. and Klotz S. 2008. Do protected areas in urban and rural landscapes differ in species diversity? Biodiversity and Conservation 17: 1595-1612. Knapp S, Kühn I, Bakker JP, Kleyer M, Klotz S, Ozinga WA, Poschlod P, Thompson K, Thuiller W, Römermann C. 2009. How species traits and affinity to urban land use control large-scale species frequency. Diversity and Distributions 15:533-546. Knapp S., Kühn I., Stolle J. and Klotz S. 2010. Changes in the functional composition of a Central European urban flora over three centuries. Perspectives in Plant Ecology, Evolution and Systematics 12: 235-244. Knapp S., Winter M. and Klotz S. 2017. Increasing species richness but decreasing phylogenetic richness and divergence over a 320-year period of urbanization. Journal of Applied Ecology 54: 1152-1160. Kowarik I. 1990. Some response of flora and vegetation to urbanization in central Europe. Urban ecology: 45-74. Kowarik I. 2011. Novel urban ecosystems, biodiversity, and conservation. Environ Pollut 159: 1974-1983. Kowarik I. and von der Lippe M. 2018. Plant population success across urban ecosystems: A framework to inform biodiversity conservation in cities. Journal of Applied Ecology 55: 2354-2361. Kühn I., Brandl R. and Klotz S. 2004. The flora of German cities is naturally species rich. Evolutionary Ecology Research 6: 749-764. Lachmund J. 2013. Greening Berlin: The Co-production of Science, Politics, and Urban Nature. MIT Press. Lawson D.M., Lamar C.K. and Schwartz M.W. 2008. Quantifying plant population persistence in human-dominated landscapes. Conserv Biol 22: 922-928. Lenzen M., Lane A., Widmer-Cooper A. and Williams M. 2009. Effects of land use on threatened species. Conserv Biol 23: 294-306. Lepczyk C.A., Aronson M.F.J., Evans K.L., Goddard M.A., Lerman S.B. and Macivor J.S. 2017. Biodiversity in the City: Fundamental Questions for Understanding the Ecology of Urban Green Spaces for Biodiversity Conservation. BioScience 67: 799-807. Louf R. and Barthelemy M. 2014. A typology of street patterns. Journal of The Royal Society Interface 11: 20140924.

Chapter 1

Lundholm J., Macivor J.S., Macdougall Z. and Ranalli M. 2010. Plant species and functional group combinations affect green roof ecosystem functions. PLoS One 5: e9677. Matthies SA, Rüter S, Prasse R, Schaarschmidt F. 2015. Factors driving the vascular plant species richness in urban green spaces: using a multivariable approach. Landscape and Urban Planning 134:177-187. Maurer U., Peschel T. and Schmitz S. 2000. The flora of selected urban land-use types in Berlin and Potsdam with regard to nature conservation in cities. Landscape and Urban Planning 46: 209-215. McKinney M.L. 2008. Effects of urbanization on species richness: A review of plants and animals. Urban Ecosystems 11: 161-176. Naughton-Treves L., Holland M.B. and Brandon K. 2005. The role of protected areas in conserving biodiversity and sustaining local livelihoods. Annu. Rev. Environ. Resour. 30: 219-252. Nielsen A.B., van den Bosch M., Maruthaveeran S. and van den Bosch C.K. 2014. Species richness in urban parks and its drivers: A review of empirical evidence. Urban Ecosystems 17: 305- 327. Nilon C.H., Aronson M.F.J., Cilliers S.S., Dobbs C., Frazee L.J., Goddard M.A., O'Neill K.M., Roberts D., Stander E.K., Werner P., Winter M. and Yocom K.P. 2017. Planning for the Future of Urban Biodiversity: A Global Review of City-Scale Initiatives. BioScience 67: 331-341. Noss R.F. 1997. Last stand: Protected areas and the defense of tropical diversity: Edited by R. Kramer, C. van Schaik and J. Johnson Oxford University Press, 1997.£ 29.50 hbk (xiv+ 242 pages) ISBN0 19 509554 5. Trends Ecol Evol 12: 450-451. Pimbert M.P. and Pretty J.N. 2013. Parks, people and professionals: putting ‘participation’into protected area management. Social change and conservation 16: 297-330. Ramalho C.E. and Hobbs R.J. 2012. Time for a change: dynamic urban ecology. Trends Ecol Evol 27: 179-188. Ramalho C.E., Laliberte E., Poot P. and Hobbs R.J. 2014. Complex effects of fragmentation on remnant woodland plant communities of a rapidly urbanizing biodiversity hotspot. Ecology 95: 2466-2478. Schmitt-Harsh M, Mincey SK, Patterson M, Fischer BC, Evans TP. 2013. Private residential urban forest structure and carbon storage in a moderate-sized urban area in the Midwest, United States. Urban Forestry & Urban Greening 12:454-463. Schneider A., Friedl M.A. and Potere D. 2009. A new map of global urban extent from MODIS satellite data. Environmental Research Letters 4: 044003. Schwartz M.W., Jurjavcic N.L. and O'Brien J.M. 2002. Conservation's disenfranchised urban poor. BioScience 52: 601-606.

Chapter 1

Schwartz M.W., Smith L.M. and Steel Z.L. 2013. Conservation investment for rare plants in urban environments. PLoS One 8: e83809. Seitz B. 2007. Konzeption zum Florenschutz im Land Berlin. Gutachten im Auftrag des Landesbeauftragten für Naturschutz und Landschaftspflege Berlin. Senatsverwaltung für Stadtentwicklung. 2010. Das Berliner Florenschutzkonzept. Natur und Landschaft 85: 16. SenStadtUm. 2014. Berlin Environmental Atlas, Biotope Types. SenStadtUm. 2015. Actual Use of Built-up Areas, Inventory of Green and Open Spaces, Actual Use and Vegetation Cover. Seto K.C., Fragkias M., Guneralp B. and Reilly M.K. 2011. A meta-analysis of global urban land expansion. PLoS One 6: e23777. Seto K.C., Güneralp B. and Hutyra L.R. 2012. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proceedings of the National Academy of Sciences 109: 16083-16088. Shwartz A., Turbe A., Julliard R., Simon L. and Prevot A.C. 2014. Outstanding challenges for urban conservation research and action. Global Environmental Change-Human and Policy Dimensions 28: 39-49. Smith R.M., Thompson K., Hodgson J.G., Warren P.H. and Gaston K.J. 2006. Urban domestic gardens (IX): Composition and richness of the vascular plant flora, and implications for native biodiversity. Biological conservation 129: 312-322. Stiftung Naturschutz Berlin. 2015. Koordinierungsstelle Florenschutz – ein Projekt zur Umsetzung des Florenschutzkonzeptes Berlin. Stoll-Kleemann S. 2001. Barriers to nature conservation in Germany: A model explaining opposition to protected areas. Journal of Environmental Psychology 21: 369-385. Sukopp H. 1990. Stadtökologie. Das Beispiel Berlin. Reimer Verlag.–1990.–455 s. UN. 2015. World Urbanization Prospects: The 2014 Revision. (ST/ESA/SER.A/366). Wells M. and Bradon K. 1992. People and parks: linking protected area management with local communities. World Bank. Williams N.S.G., Schwartz M.W., Vesk P.A., McCarthy M.A., Hahs A.K., Clemants S.E., Corlett R.T., Duncan R.P., Norton B.A., Thompson K. and McDonnell M.J. 2009. A conceptual framework for predicting the effects of urban environments on floras. Journal of Ecology 97: 4-9. Williams NSG, Hahs AK, Vesk PA. 2015. Urbanisation, plant traits and the composition of urban floras. Perspectives in Plant Ecology Evolution and Systematics 17:78-86.

Chapter 2

Chapter 2: Untangling the role of urban ecosystems as habitats for endangered plant species

Chapter 2

This chapter is the preprint version of the journal article published as: G. Planchuelo, M. von der Lippe M, I. Kowarik, 2019, Untangling the role of urban ecosystems as habitats for endangered plant species. Landscape and Urban Planning. DOI: https://doi.org/10.1016/j.landurbplan.2019.05.007

Abstract

As urbanization accelerates globally, a better understanding of how cities contribute to biodiversity conservation is increasingly pressing. Previous studies reveal that cities can harbour a considerable biological richness, including endangered plant species. Yet, a key question on the urban contribution to plant conservation remains critically open, as little information is available on how populations of endangered plant species occur across different biotope types within cities and to what extent anthropogenically shaped vs. natural ecosystems provide habitats for endangered plants. We analysed a unique dataset on the exact geographical position of 1,742 populations of 213 highly endangered plant species in the city of Berlin. We first assessed the relative importance of Berlin’s nine major biotope classes as habitats of endangered plant species. Second, we applied the novel ecosystem concept to quantify endangered plant populations for natural remnants vs. hybrid vs. novel ecosystems within Berlin. Populations of endangered plant species were generally, although unevenly, associated with specific biotope classes, with forest, grassland, and ruderal biotopes as the most important habitats. Surprisingly, novel ecosystems harboured the highest numbers of total populations, of total species, and of species that were exclusively confined to one type of ecosystem novelty. Quantifying the relative importance of biotope classes and novel vs. (near-) natural ecosystems as habitats of endangered species demonstrates that the urban contribution to biodiversity conservation is best ensured by providing a range of ecosystems. Rather than prioritizing only natural remnants, we thus argue for broad approaches to urban biodiversity conservation that include novel ecosystems.

Chapter 2

Introduction

The potential contribution of cities to biodiversity conservation is becoming increasingly important in the Anthropocene period, in which urbanization and biodiversity loss are coinciding and accelerating global trends (Ellis 2015). While urban sprawl can threaten biodiversity outside cities (Mcdonald et al. 2008; Guneralp & Seto 2013), urban regions have been reported to harbour high numbers of species due to habitat heterogeneity, human agency and non-random location of many cities in landscape with a naturally high geomorphological and biological richness (Kühn et al. 2004; McKinney 2008). Moreover, cities reflect their regional species pool (Aronson et al. 2014) including rare and endangered species (Schwartz et al. 2002; Lawson et al. 2008; Lenzen et al. 2009; Shwartz et al. 2014; Ives et al. 2016). Of Australia’s threatened plant and animal species, for example, 30% occur in cities, and some are exclusively found in them (Ives et al. 2016). Even heavily urbanized areas (Jokimaki et al. 2018) and novel urban ecosystems offer suitable habitats for some endangered species as shown for European cities (Jokimaki et al. 2018; Kowarik & von der Lippe 2018). In consequence, conservation policies have increasingly focused on urban areas in recent years, with claims that a vast variety of urban ecosystems offer habitat functions for native species (Goddard et al. 2010; Kowarik 2011; Lepczyk et al. 2017; Nilon et al. 2017)

Yet, in their meta-analysis on urban biodiversity conservation, Shwartz et al. (2014) highlight gaps in knowledge that are vital for confirming the role of cities for biodiversity conservation. One major point is that studies of endangered species in urban regions largely do not account for the conspicuous heterogeneity in time and space that exists among urban ecosystems (Ramalho & Hobbs 2012). Indeed, varying human impacts and a range of environmental barriers filter the composition of urban species assemblages (Williams et al. 2009; Aronson et al. 2016; Knapp et al. 2017; Kowarik & von der Lippe 2018), since species are differently pre- adapted or able to adapt to urban conditions (McDonnell & Hahs 2015). In consequence, species loss is seen in urban regions as well (Hahs et al. 2009). While the “human threat hypothesis” posits a negative relationship between increasing human dominance and the persistence of species (Lawson et al. 2008), a comparison of endangered plant species occurring in differently urbanized regions of California found no evidence for negative urbanization effects (Lawson et al. 2008). However, this may be due to the large scale of the study, which did not allow for the habitat functions of natural remnants to be untangled from other, anthropogenically shaped urban biotopes.

While there is growing evidence of the factors underlying species richness in urban habitats, (Beninde et al. 2015; Lepczyk et al. 2017) these factors are not necessarily the same that

Chapter 2

promote the existence of endangered species in cities. For example, Beninde et al. (2015) revealed patch size as the most important driver of urban species richness while a recent study in urban regions of Australia indicated the importance of both big and small habitat patches for plant species of conservation interest (Kendal et al. 2017).

The importance of remnant biotopes for urban biodiversity conservation has been well established (e.g. Godefroid and Koedam 2003). There is increasing evidence that anthropogenic biotopes such as urban grassland (Fischer et al. 2013; Chollet et al. 2018), domestic gardens (Smith et al. 2006; Loram et al. 2008) vacant land (Bonthoux et al. 2014; Anderson & Minor 2017), parks (Cornelis & Hermy 2004; Nielsen et al. 2014), or cemeteries (Kowarik et al. 2016; Yılmaz et al. 2018) can harbour a broad range of native species as well. Yet there are open questions on the role of such biotopes for endangered species (e.g. Williams et al. 2014 on green roofs). However, comparative analyses of the relevance of different types of (near-)natural and anthropogenic biotopes as habitats of endangered species in cities are missing thus far.

While natural remnants and conservation areas are the traditional conservation focus in cities (e.g. Godefroid & Koedam 2003; Knapp et al. 2008; Diamond & Heinen 2016; Kendal et al. 2017; Zeeman & Morgan 2018), a key question remains as to how novel urban ecosystems should be addressed from a conservation perspective (Kowarik 2011; Hobbs et al. 2014). Novel ecosystems, which are characterized by profound and likely irreversible changes to ecosystem features and/or species assemblages (Hobbs et al. 2009; 2013), are important components of urban regions (Kowarik 2011) and have been shown to harbour endangered species (Goddard et al. 2010; Bonthoux et al. 2014; Kowarik & von der Lippe 2018; Maclagan et al. 2018). To date, however, the extent to which urban ecosystems that differ in ecological novelty (i.e., along the spectrum from natural remnants to novel ecosystems; Hobbs et al. 2013) function as habitats of large sets of endangered plant species has not been quantified in comparative studies.

A recent study from Berlin (Kowarik & von der Lippe 2018) made a first step by assessing the role of highly aggregated novelty types (i.e., natural remnants, hybrid and novel ecosystems according to Hobbs et al. 2013) for alien, native and endangered plant species. Based on expert assessments of the general occurrence and population establishment of species in each novelty type, this study revealed that endangered species were unevenly distributed across the novelty categories, with highest richness in natural remnants. Our present study makes further steps forward towards understanding the role of different types of biotopes within novel, hybrid and remnant ecosystems for plant species of conservation concern: first, by focusing on 1,742 spatially referenced populations of 213 plant species that had been identified as priority species for biodiversity conservation by the responsible authority (Stiftung Naturschutz 2015); second,

Chapter 2

by relating the exact location of these populations to biotope types that reflect the variety of ecological conditions and land uses in the metropolitan area of Berlin; and, third, by assigning each biotope patch to one of the major novelty types (i.e., natural remnant, hybrid, novel ecosystems).

Based on previous research we expect that all novelty categories harbour plant species of major conservation concern (i.e. priority species for biodiversity conservation in Berlin), but with a prominent role for natural remnants in terms of numbers of species, populations and of species being exclusively confined to one novelty type. We further hypothesize significant differences between biotope types within each novelty type in their role as habitats of these priority species, as well as patch size effects on some biotope types. In detail, we address the following research questions: What is the relative contribution of (1) different biotope classes and (2) of natural remnant vs hybrid vs novel ecosystems within the city of Berlin in harbouring populations of endangered plant species? (3) To what extent do ecosystems that differ in ecological novelty share or exclusively host populations of endangered plant species?

Methods

Study area and study system

The study was conducted in the city of Berlin, Germany’s capital and largest city, with 3.6 million inhabitants in 2017 and a total area of 891 km2. About 59% of Berlin’s surface is covered by built-up areas and streets, while green and blue spaces occupy 41% of the area— including forests (17.7%), lakes and rivers (6.1%), parks (5.6%), allotment gardens (5.3%), fields (5%) and meadows (1.3%) (SenStadtUm 2016). The climate is temperate, with forests and wetlands as natural ecosystems. The total area of Berlin encompasses remnants of natural and agrarian landscapes, urban greenspaces with different land-use histories, and a range of novel ecosystems on vacant urban land and within the built-up areas (Sukopp 1990).

Berlin was founded in the Medieval (1237), but remained for centuries a rather small city, embedded in the rural countryside. Berlin’s population rapidly grew from ca. 424,000 inhabitants in 1849 to four million inhabitants in 1925. Today, the total area of Berlin encompasses remnants of natural landscapes (forests, wetlands) and agrarian landscapes (fields, grassland) as well as urban greenspaces (parks, cemeteries, community gardens), including some historical parks older than 200 years. Due to bombings during World War II, and a slow urban development until the German reunification in 1989, a considerable amount of

Chapter 2

vacant land with a range of novel ecosystems remained within the city and was recolonized by natural succession. While some of these areas were integrated into urban parks, others were re- built as Berlin is growing again (Sukopp 1990; Lachmund 2013).

Berlin’s flora has been well studied since the 18th century (Sukopp 1987). Starting in the 1970s, information about species extinctions and declines within the area of today’s Berlin has been synthesized and updated several times in Red Lists of endangered species (Seitz et al. 2018). Berlin’s flora comprises 1,527 previously present taxa; 17% of these taxa have gone extinct and 29% are currently being endangered in Berlin, referring to the well-known flora of the mid-19th century as a baseline (Seitz et al. 2018). A set of 230 plant species of highest conservation priority in Berlin had been identified within the framework of Berlin’s Flora Protection Program, based on a systematic assessment approach (Berliner Florenschutzkonzept; Seitz 2008). All of these priority species are being endangered in Berlin and/or at higher spatial scales (i.e. in the surrounding federal state of Brandenburg, in Germany, Europe, or at a global scale). In this study, we used data about the spatial location of populations of priority species that were available for 213 out of 230 species. These data covered a total of 1,742 populations and resulted from a monitoring that was carried out by Stiftung Naturschutz (2015) since 2009. Populations were separated from each other during the monitoring based on a standardized protocol. Experts discerned populations when individuals from one population were clearly separated from individuals of another population by a distance of more than 30 m or by paved pathways or roads. Appendix A includes the species list, with numbers of populations per species and further information.

Methodological approach

We first linked the geographical position of each of the populations of the endangered plant species with a GIS to the biotope it is located in to unravel the relative importance of different biotopes as habitats of endangered plant species (biotope scale). We then assigned all biotopes to a level of ecological novelty according to the novel ecosystem concept (Hobbs et al. 2013) to elucidate the relative importance of natural remnant vs hybrid vs novel ecosystems for the populations of the target species (novelty scale).

Chapter 2

Biotope types

In the 1980s, a methodological approach was developed to assign the entire area of Berlin to biotope types that integrate various combinations of environmental conditions and land uses (Sukopp & Weiler 1988). The biotope mapping was based on aerial images, combined with ground proofing, and has been updated over time. Biotopes with multiple and complex vegetation structures (such as forests) or with a high conservation value were covered by field biotope mapping, while others (e.g. lakes, agricultural fields, residential areas) were differentiated by using colour infrared aerial photography (SenStadtUm 2014). Today, the Berlin Environmental Atlas provides a comprehensive biotope map (SenStadtUm 2014) that divides Berlin into twelve major biotope classes: forests, grasslands, ruderal sites, standing bodies of water, built-up areas, bogs and marshes, groves and hedges, green spaces, fields, moving bodies of water, heaths, and “other types”. For the purpose of this study, we merged biotope classes with a rather low number of endangered plant populations with a similar biotope class. In particular we merged standing and moving bodies of water to water bodies, heaths were incorporated into grasslands, and “other types” (which was a plant nursery) was incorporated into fields. This resulted in a total of nine biotope classes that are further subdivided in up to 7 hierarchical levels that sum up to a total of 7,483 biotope types. In the mapping dataset of the Berlin Environmental Atlas, Berlin is differentiated into 79,268 patches, and each of these patches is assigned to one of the 7,483 possible biotope types. With the help of spatial queries in GIS software, we linked the geographical positions of each of the 1,742 populations of the target species with the corresponding fine-grained biotope type and then aggregated the results at the level of the nine major biotope classes.

Ecological novelty

To relate the locations of populations of endangered species to ecological novelty, we generated an ecosystem novelty map of Berlin (Fig. 1) by spatially combining data on biotope types with land use data as follows: We first assigned a preliminary novelty category to each of the 7,483 biotope types of the aforementioned biotope map. Following the approach of Kowarik and von der Lippe (2018), we classified biotope types by considering the origin of the type (natural vs. anthropogenic) and whether it is dominated by natural ecosystem processes or by human interference (Table 1 Chapter 1). This led to three novelty categories (adapted from Hobbs et al. 2009 and specified for urban settings by Kowarik and von der Lippe 2018):

Chapter 2

• Hybrid ecosystems are human-mediated ecosystems that have been modified from their historical state but still retain the potential to approach historical conditions. These include many young tree plantings in forests and parks, managed grasslands, extensively managed urban green spaces, or low-intensity pastures.

• Novel ecosystems are human-modified ecosystems that have likely been irreversibly changed by profound impacts on abiotic conditions or biotic composition. Novel ecosystems include built-up areas, vacant lots, rooftops, abandoned industrial areas, or high-intensity agricultural land.

Fig. 1 Novelty map of Berlin with locations of populations (n=1742; red dots) of endangered plant species (n=213) that are priority species for biodiversity conservation in Berlin. The pie chart displays the different surface area occupied by natural remnant, hybrid, and novel ecosystems.

Chapter 2

While most biotopes could be easily classified into one of the ecosystem novelty types, we further refined the novelty classification in other cases by considering land use data (Appendix B). We identified the locations of wastelands, rooftops, or former sewage farms from the Berlin Environmental Atlas’s land use map (SenStadtUm 2015), and classified all biotope patches within these locations as novel ecosystems due to the generally profound changes to environmental features in these land-use types. Grasslands, for example, were usually classified as hybrid ecosystems—but as novel ecosystems when located in a wasteland, a rooftop, or a former sewage farm.

Spatial and statistical analyses

Merging the biotope map and the land use map to generate the novelty map was done by spatial intersections and later unions in Quantum-GIS (QGIS). To determine the biotope class and the ecosystem novelty type for the locations of the 1,742 populations, spatial intersections were performed in QGIS. To reveal how novelty categories differ in their relative contribution in hosting populations of endangered plant species we calculated two measures of population density for endangered plant species. First we calculated population density per patch by dividing the area of all habitat patches by the number of populations they host, including zero- values for habitat patches which are not occupied by endangered plant species. Second, we calculated population density per patch as described above but only within occupied patches.

We tested for overall patch size effects on the occurrence of endangered species by comparing the mean patch size of all habitat patches occupied by endangered species to the size all unoccupied patches. As the numbers of occupied and unoccupied patches differed at about two orders of magnitude, we compared both groups by means of a permutation test (R-function independence_test in package coin).

To reveal the relative importance of patch size for the occurrence of endangered species within the different biotope classes and novelty categories, we built a conditional inference tree (R package “party”, Hothorn et al. 2006) with biotope class, ecological novelty and patch size as predictors of patch occupancy by populations of endangered plant species in Berlin. This decision tree technique repeatedly partitions the dataset into two groups at a threshold value of an environmental predictor that shows the most significant effect on the response, i.e. the ratio of occupied patches in the resulting subgroups. The terminal nodes of the resulting tree diagram represent the share of habitat patches occupied by endangered plant species in a subgroup of all

Chapter 2

habitat patches with habitat characteristics determined by all preceding nodes of the respective branch in the tree diagram. Tree depth was restricted to three hierarchy levels.

To test for possible significant differences in mean densities of populations of endangered species between natural remnant, hybrid, and novel ecosystem patches, we performed a one- way permutation test based on a maxT statistic. This is similar to a one-way ANOVA but avoiding biased estimates due to non-normal distribution and zero inflation of the response. This procedure was followed by paired permutation tests with Holm correction to reveal significant differences between the three novelty categories. To reveal if habitat novelty has an effect on the density of endangered species in occupied habitat patches, we performed a linear mixed effect model with population density as a response, novelty type as fixed effect and genus as a random effect to correct for phylogenetic dependencies. This was followed by a post-hoc Tukey test to discriminate the population density between the three novelty categories.

To examine how populations of endangered species from the nine main biotope classes are distributed across the three novelty categories, we created a Sankey diagram by using the online tool SankeyMATIC (www.sankeymatic.com). To analyze if populations within the main biotope categories contribute disproportionally to the novelty categories, we run a log-linear model using the R-function loglin and computed standardized residuals for all possible combinations of factor levels between main biotope classes and novelty categories to reveal any deviations of the observed frequencies from the frequencies estimated by the model. These deviations were than incorporated as symbols in the Sankey diagram, displaying disproportional contributions of populations within the main biotope classes to the novelty categories.

To elucidate whether natural remnants, hybrid, and novel ecosystems hold similar or distinct sets of endangered species, we generated an Euler diagram showing overlaps in the species pools of novelty types (eulerAPE: Micallef and Rodgers 2014).

All statistical analyses were performed in R version 3.5.0 (R Core Team 2018).

Results

Biotope types

Results showed that each of the nine major biotope classes harbour populations of endangered plant species—but with conspicuous differences among classes (Fig. 2, left). Most of the 1,742 populations (70%) of the 213 high priority species were located within three biotope classes:

Chapter 2

forests (34%), grasslands (26%), and ruderal sites (10%), while the remaining 30% were spread throughout the other six classes.

Fig. 2 Number of populations of 213 endangered plant species across biotope classes and types of ecological novelty in Berlin. The locations of 1,742 populations were assigned to nine biotope classes (left) and to natural remnant, hybrid, or novel ecosystems (right). The width of the lines

Chapter 2

is proportional to the number of populations located in each combination of biotope class and novelty type. Positive and negative deviations of expected frequencies from a log-linear model are displayed by symbols on the left side of the respective lines in the diagram (++: > 4 standardized residuals; +: 2 to 4 standardized residuals; -: -2 to -4 standardized residuals; --: < - 4 standardized residuals; ○: -2 to 2 standardized residuals).

Patches with populations of endangered species were on average significantly bigger in size than patches without populations (Z = -11.653, p<0.001, Fig. 3). Populated patches showed a lower variability in patch size compared to unpopulated patches and had a markedly higher minimum patch size of 118m2 compared to 4m2 in unpopulated patches.

Fig. 3 Boxplots of patch size in unpopulated habitat patches and habitat patches populated by endangered plant species in Berlin. Please note the log scale of the y axis. Significant differences in average patch size between unpopulated (n=78320) and populated (n=948) patches were determined by a permutation test for the two groups (Z = -11.653, p<0.001).

Chapter 2

Ecological novelty

Combining information about the fine-grained types of biotope with land use made it possible to assign the 79,268 patches of Berlin’s biotope map to one of the three categories of ecological novelty. The resulting novelty map of Berlin (Fig. 1) revealed that natural remnants make up 7% of the area of Berlin, hybrid ecosystems 16%, and novel ecosystems 78%; the latter largely includes built-up areas and transportation corridors but also open spaces such as vacant urban- industrial land and sewage farms.

Each novelty type harboured more than 400 populations of endangered species (Fig. 2, right). Surprisingly, novel ecosystems supported the highest number of populations (39.4%). These were located across a wide range of biotope classes, but most prominently within grassland and ruderal biotopes. Hybrid ecosystems had a slightly smaller share of populations (36%), and these were found mostly in forest and grassland biotopes. About one-fourth of all populations (24%) were located in natural remnant ecosystems, with forest and wetland biotopes as major habitats.

The log-linear model revealed clear disproportional distributions of populations growing in the main biotope classes to the three novelty categories (χ²= 1257, p>0.001). From the three biotope classes hosting the majority of populations, populations in forests were overrepresented in remnant ecosystems, populations in grassland biotopes were overrepresented in hybrid ecosystems and ruderal sites contributed more populations than expected to novel ecosystems (Fig. 2).

Population density, however, was significantly higher in natural remnants (0.12 pop./patch) compared to hybrid (0.04 pop./patch) and novel ecosystems (0.01 pop./patch) (Fig. 4A, maxT = 17.108, p<0.001). Moreover, there is a similar effect of habitat novelty on population density of endangered species in occupied habitat patches as revealed by a linear mixed effect model corrected for phylogenetic dependencies (Fig. 4B, χ²= 37.92, p<0.001).

Chapter 2

Fig. 4 A) Mean density of populations of endangered plant species in patches of natural remnant (n=3,426), hybrid (n=15,681), and novel ecosystems (n=58,102), including unoccupied patches. Significant differences between ecosystem types were determined by a one-way permutation test (maxT = 17.108, p<0.001). Different lower case letters above the bars indicate significant differences in mean population density between novelty categories according to pairwise permutation tests with Holm correction. B) Density of populations of endangered plant species within occupied patches of natural remnant (n=173), hybrid (n=320), and novel ecosystems (n=456). Significant differences between ecosystem types were calculated by a linear mixed effect model and corrected for phylogenetic dependencies by incorporating genus as a random factor. Different lower case letters above the bars indicate significant differences in mean population density between novelty categories according to a Post-hoc Tukey-test.

Total species numbers were similar in the two anthropogenic ecosystem categories (141 in hybrid, 142 in novel ecosystems), while only 102 species had populations in natural remnants. Importantly, each novelty type harboured a considerable number of exclusive species (species confined solely to one novelty type; Fig. 5). Novel ecosystems had the highest share of exclusive species (20%), followed by hybrid ecosystems (15%), and natural remnant ecosystems (9%). We also found important overlaps in species that were present in all novelty types (26%) or two of the three novelty types (30%; Fig. 5).

Chapter 2

Fig. 5 Endangered plant species with shared or exclusive occurrences in natural remnant, hybrid, and novel ecosystems in Berlin (100% = 213 species), illustrated by an area- proportional Euler diagram.

A conditional inference tree analysis revealed the relative importance of biotope class, ecological novelty and patch size in explaining patch occupancy by populations of endangered plant species (Fig. 6). Biotope class is the most important factor affecting patch occupancy, followed by ecological novelty, and finally patch size. The tree diagram shows the highest patch occupancy for remnant bogs and marshes followed by remnant forests, grasslands and waterbodies (pie charts in Fig. 6). Patch size matters for discriminating patch occupancy in the biotope classes built up area, green spaces and groves and hedges, where patches bigger than 9 ha show a significantly higher share of occupied patches than smaller ones. For all other biotope classes patch size was only relevant in hybrid and novel habitats with a higher occupancy of endangered species in patches bigger than 0.4 ha.

Chapter 2

Fig. 6 Conditional inference tree showing the relative importance of biotope class, ecological novelty and patch size in explaining patch occupancy by populations of endangered plant species in Berlin. Pie charts at the terminal nodes of the diagram indicate the proportion of patches occupied by endangered species in the final partition. All splits of the dataset are indicated by the predictor that causes the split and have p values < 0.001. The final tree was pruned to three hierarchy levels to reduce complexity. The different biotope classes have been abbreviated as follows: Fr forests, Gr grasslands, Rs ruderal sites, WB water bodies, BA built up areas, BM bogs and marshes, GH groves and hedges, GS green spaces, Fi fields. The different ecological novelty types have been abbreviated as follows: R remnant ecosystems, H hybrid ecosystems, N novel ecosystems.

Discussion

Urban regions have important potential for the conservation of plant and animal species (Shwartz et al. 2014), but ambiguous trends for species of conservation concern have been reported for cites, including population decline, extinctions (Hahs et al. 2009; Knapp et al.

Chapter 2

2017), and biotic homogenization (Zeeman et al. 2017) as well as population persistence and colonization of new habitats (Lawson et al. 2008; Lundholm & Richardson 2010; Kowarik & von der Lippe 2018). A key question for biodiversity conservation thus is not simply whether cities support biodiversity conservation, but which environmental settings within urban regions support species of conservation interest (Lepczyk et al. 2017).

Previous urban studies have revealed that different land-use types harbour different sets of species (e.g., Celesti-Grapow et al. 2006; Godefroid & Koedam 2007; Lososova et al. 2012; Godefroid & Ricotta 2018). As this can be true also for land-use types with similar environmental conditions, e.g. due to legacy effects of previous land use, new approaches towards determining the role of urban environments as habitats for plants and animals are necessary (Ramalho & Hobbs 2012; Godefroid & Ricotta 2018). This study is likely the first that quantifies the relative contribution of different biotope classes within a large metropolitan region for harbouring populations of a large set of endangered plant species of the highest conservation priority at the city scale. Moreover, applying the novel ecosystem concept (Hobbs et al. 2013) to a fine-scaled map of biotope types revealed the importance that different levels of ecological novelty in urban ecosystems may have for endangered plant species. Results of this study are relevant for other large historical cities, at least in Europe, with a similar structure and land-use history. Yet results from Berlin cannot be generalized for cities around the globe as these can represent different typologies in regard to the history of landscape configuration, the proportion of native vegetation (Hahs et al. 2009), land use legacies (Ramalho & Hobbs 2012), and the configuration of urban elements (Louf & Barthelemy 2014).

Biotope types

Our analysis revealed conspicuous differences in habitat functions of biotope classes for populations of endangered plant species (Fig. 2). Built-up biotopes cover the majority of Berlin (54%), but only 92 populations of endangered species were found there, while bogs and marshes hold almost as many populations (85) in a fraction of that area (>1%). A very high density of populations of endangered species was also found in grassland and ruderal biotope classes, which hosted 26% and 10% of populations in 5% and 2% of the area respectively. These results demonstrate the importance of anthropogenically influenced biotopes for urban biodiversity conservation as grassland communities are absent in Berlin’s natural vegetation (with a few exceptions for wetlands; Sukopp 1990). Previous studies have highlighted the conservation potential of urban grasslands, considering grasslands either as remnants of natural vegetation in other biogeographical regions (e.g., Cilliers et al. 2004; Zeeman et al. 2017)

Chapter 2

or as culturally shaped grassland types such as the ones in urban parks or airfields (e.g., Fischer et al. 2013a; Klaus 2013; Nielsen et al. 2014). This study confirms the high importance of urban grassland biotopes by quantifying grassland populations of endangered plant species.

Highly disturbed ruderal sites such as the ones found on vacant land or in transportation corridors usually fall outside the conservation focus. Recent studies have illustrated the potential of such sites for biodiversity or restoration efforts (Gardiner et al. 2013; Bonthoux et al. 2014; Anderson & Minor 2017) and have documented the appreciation of biodiverse ruderal vegetation by urban people (Fischer et al. 2018). This study substantiates the relative importance ruderal biotopes may have for endangered plant species compared to other biotopes.

Forests biotopes, which represent the largest vegetation-dominated biotope class in Berlin (Fig. 2), provided habitats for the highest number of populations of endangered species. This unsurprising result corroborates previous studies about the significance of forest patches for urban biodiversity conservation (e.g. Godefroid and Koedam 2003; Diamond and Heinen 2016). Species that are strongly limited in their dispersal capacity, such as ancient forest species (Hermy et al. 1999; Dyderski et al. 2017), may be unable to colonize other urban biotopes and thus remain confined to forest biotopes with their long habitat continuity.

While we were able to demonstrate a higher overall patch size of habitats occupied by endangered plant species compared to unoccupied patches, patch size was less relevant than biotope type or ecological novelty in explaining the presence of endangered plant species (Fig. 6). Patch size has a major effect in some more urban biotope classes (built up areas, groves and hedges, and green spaces). In those biotopes, large patches (>9 ha) have the highest occupancy rates (Fig. 6). This is likely due to the fact that small sized patches of highly urbanized biotopes are less capable of holding populations of endangered species (Fig. 3) as the probability of incorporating suitable microhabitats increases with patch size. In the other biotope classes, patch size was only relevant in hybrid and novel habitats, with a much lower threshold of >0.4 ha delimitating patches with a higher occupancy (Fig. 6). The fact that there was no threshold in patch size detectable for predicting patch occupancy in remnant ecosystems could result from a generally small patch size in this novelty class due to constrictions by urban pressures. In contrast, the meta-analysis by Beninde et al. (2015) revealed patch size to be the most important predictor for urban biodiversity (i.e., species richness). Yet, this meta-analysis aggregated data from a large range of taxa and had a clear focus on species richness. We here show that spatial requirements of endangered plant species clearly differ between biotopes and novelty classes. Correspondingly, a recent Australian study highlighted the importance of both

Chapter 2

big and small urban conservation areas for plant species of conservation interest (Kendal et al. 2017).

Ecological novelty

Urbanization implies the step-by-step transformation of natural to anthropogenically disturbed ecosystems. Among the latter, a range of novel ecosystems have emerged following the destruction or profound alterations of former ecosystems (Kowarik 2011). Considering the ecological features, current land use, and the history of urban sites allows deeper insights into the response of species to urbanization (Ramalho & Hobbs 2012). We approached this challenge by overlaying biotope maps with land use maps. This allowed us to differentiate among novelty categories sometimes even within the same type of biotope (Appendix D).

The prominence of novel ecosystems in harbouring the largest number of populations of endangered plant species in Berlin is a major and unexpected result of this study. Additionally, novel ecosystems include populations from a wider range of biotope classes than remnant or hybrid ecosystems with ruderal sites, grasslands and built up areas contributing the majority of populations to this class. Former studies have shown the potential of novel urban ecosystems for native and some endangered species (e.g., Lundholm and Richardson 2010; Bonthoux et al. 2014)—but without quantifying novel habitat functions in relation to others (but see case study from Maclagan et al. 2018). As an exception, a previous study from Berlin revealed a considerable number of native and endangered species in novel ecosystems, but with clearly more in natural remnants (Kowarik & von der Lippe 2018). That the present study was able to reveal an even higher importance of novel ecosystems for endangered species is likely due to methodological reasons. The previous study relied on a dataset with expert ratings on the occurrence of about 1200 species across novelty categories. In contrast, here we used a current data set on a smaller subset of species of highest conservation concern (n=213), with information about 1,742 spatially referenced populations. Relating these data to the level of ecological novelty of each biotope patch, and aggregating fine-grained novelty assessments ultimately to the three major novelty categories likely yielded more precise insights into the relevance of ecological novelty for endangered plant species in cities.

However, the data analysed also revealed a twelve times higher density of endangered species in natural remnant compared to novel ecosystems and a more than three times higher density in hybrid compared to novel ecosystems (Fig. 4A). A similar pattern, but with less pronounced differences, was found for population density only within occupied patches of the respective

Chapter 2

habitat class (Fig. 4B). The rather low density of endangered species in novel ecosystems mainly resulted from the inclusion of a large amount of built-up areas with sealed surfaces. Although excluding these areas would have possibly modified this relationship, it was not possible due to an unknown extent of the sealed surface.

In terms of cost efficiency of practical conservation measures, it may still be more efficient to preserve the same amount of habitat in natural remnant than in novel ecosystems due to the higher density of populations in the former. A further argument for allocating conservation efforts to natural remnants is the fact that their area cannot be increased and that smaller patches of hybrid or novel ecosystems might be more sensitive to urban pressures and isolation (Ramalho et al. 2014).

However, all novelty types are important for urban biodiversity conservation. Analysing the overlap of species pools across the three novelty categories revealed a considerable number of species that occur exclusively in one category, with 9-20% of species confined to one novelty type, and surprisingly, of these most were confined to novel ecosystems (Fig. 5). These results provide further evidence on the importance of biotopes across all novelty categories for urban biodiversity conservation. One novelty type can thus not substitute for the habitat functions of another category.

The high share of species that were only found in novel ecosystems is likely due to the urbanization-mediated habitat transformation of natural remnants or historical agrarian ecosystems. Even some dispersal-limited species may occur in novel ecosystems owing to human-mediated dispersal pathways that can counteract spatial isolation. For example, non- generalist grassland species are often dispersal limited (Deák et al. 2018), but they can nonetheless be found on ca. 100-year-old rooftops of waterworks in Berlin (data not shown), suggesting historical practices of transferring seeds, sods, or natural soil from near-natural settings for rooftop greening (Brenneisen 2006; Jim 2017).

It is important to note that, due to a range of human-mediated environmental barriers (Williams et al. 2009; Kowarik & von der Lippe 2018), not all urban populations of a given plant species are ultimately successful at establishment, and some may be at risk of extirpation (Hahs & McDonnell 2014). A study on the performance of rare plant species along California’s coastline found no correlation between species’ performance and human population density (Schwartz et al. 2013). However, of all species that were documented for different types of anthropogenic, hybrid, or novel ecosystems in Berlin, 21-57% had only casual populations, i.e., these species were not able to establish self-sustaining populations in cultural ecosystems (Kowarik & von der Lippe 2018). The urban success of endangered species indicated by the present study may

Chapter 2

thus be overestimated until casual or decreasing populations are accounted for. Untangling the long-term population performance of endangered species across urban biotope classes is thus an important future direction.

Conclusions

While the urban contribution to biodiversity conservation is increasingly highlighted in conservation policies and management approaches (e.g., Aronson et al. 2017, Garrard et al. 2018), important barriers to investing in urban conservation efforts still remain. Soanes et al. (2018) recently complained of “a pervasive narrative in policy, practice and the public psyche” that urban environments offer limited conservation value; the authors argued for making use of small spaces and unconventional habitats. By quantifying the relative importance of novel urban ecosystems for plant species of high conservation priority in a metropolitan region, this study provides strong support for making use of the opportunities that unconventional habitats offer for biodiversity conservation in urban regions. At the same time, results also demonstrate that neither novel nor remnant ecosystems alone were able to harbour all plant species of high conservation concern in Berlin. This study thus further supports a diversified urban conservation approach to cover the potential of all types of urban ecosystems for biodiversity conservation.

Literature

Anderson EC, Minor ES. 2017. Vacant lots: An underexplored resource for ecological and social benefits in cities. Urban Forestry & Urban Greening 21:146-152. Aronson MF, et al. 2014. A global analysis of the impacts of urbanization on bird and plant diversity reveals key anthropogenic drivers. Page 20133330. Proc. R. Soc. B. The Royal Society. Aronson MF, et al. 2016. Hierarchical filters determine community assembly of urban species pools. Ecology 97:2952-2963. Aronson MFJ, Lepczyk CA, Evans KL, Goddard MA, Lerman SB, MacIvor JS, Nilon CH, Vargo T. 2017. Biodiversity in the city: key challenges for urban green space management. Frontiers in Ecology and the Environment 15:189-196. Beninde J, Veith M, Hochkirch A. 2015. Biodiversity in cities needs space: a meta-analysis of factors determining intra-urban biodiversity variation. Ecol Lett 18:581-592.

Chapter 2

Bonthoux S, Brun M, Di Pietro F, Greulich S, Bouché-Pillon S. 2014. How can wastelands promote biodiversity in cities? A review. Landscape and urban planning 132:79-88. Brenneisen S. 2006. Space for urban wildlife: designing green roofs as habitats in Switzerland. Urban habitats 4. Celesti-Grapow L, Pyšek P, Jarosik V, Blasi C. 2006. Determinants of native and alien species richness in the urban flora of Rome. Diversity and Distributions 12:490-501. Chollet S, Brabant C, Tessier S, Jung V 2018. From urban lawns to urban meadows: Reduction of mowing frequency increases plant taxonomic, functional and phylogenetic diversity. Landscape and Urban Planning, 180, 121-124. Cornelis J, Hermy M. 2004. Biodiversity relationships in urban and suburban parks in Flanders. Landscape and Urban Planning 69:385-401. Cilliers SS, Müller N, Drewes E. 2004. Overview on urban nature conservation: situation in the western-grassland biome of South Africa. Urban Forestry & Urban Greening 3:49-62. Deák B, Valka O, Török P, Kelemen A, Bede A, Csathó AI, Tóthmérész B. 2018. Landscape and habitat filters jointly drive richness and abundance of specialist plants in terrestrial habitat islands. Landscape Ecology 33:1117-1132. Diamond JM, Heinen JT. 2016. Conserving rare plants in locally-protected urban forest fragments: A case study from Miami-Dade County, Florida. Urban Forestry & Urban Greening 20:1-11. Dyderski MK, Tyborski J, Jagodzinski AM. 2017. The utility of ancient forest indicator species in urban environments: A case study from Poznan, Poland. Urban Forestry & Urban Greening 27:76-83. Ellis EC. 2015. Ecology in an anthropogenic biosphere. Ecological Monographs 85:287-331. Fischer LK, et al. 2018. Beyond green: Broad support for biodiversity in multicultural European cities. Global Environmental Change-Human and Policy Dimensions 49:35-45. Fischer LK, von der Lippe M, Kowarik I. 2013a. Urban grassland restoration: which plant traits make desired species successful colonizers? Applied Vegetation Science 16:272-285. Fischer LK, von der Lippe M, Kowarik I. 2013b. Urban land use types contribute to grassland conservation: The example of Berlin. Urban Forestry & Urban Greening 12:263-272. Gardiner MM, Burkman CE, Prajzner SP. 2013. The value of urban vacant land to support arthropod biodiversity and ecosystem services. Environ Entomol 42:1123-1136. Garrard GE, Williams NSG, Mata L, Thomas J, Bekessy SA. 2018. Biodiversity Sensitive Urban Design. Conservation Letters 11:e12411. Goddard MA, Dougill AJ, Benton TG. 2010. Scaling up from gardens: biodiversity conservation in urban environments. Trends Ecol Evol 25:90-98.

Chapter 2

Godefroid S, Koedam N. 2003. How important are large vs. small forest remnants for the conservation of the woodland flora in an urban context? Global Ecology and Biogeography 12:287-298. Godefroid S, Koedam N. 2007. Urban plant species patterns are highly driven by density and function of built-up areas. Landscape Ecology 22:1227-1239. Godefroid S, Ricotta C. 2018. Alien plant species do have a clear preference for different land uses within urban environments. Urban Ecosystems:1-10. Guneralp B, Seto KC. 2013. Futures of global urban expansion: uncertainties and implications for biodiversity conservation. Environmental Research Letters 8:014025. Hahs AK, McDonnell MJ. 2014. Extinction debt of cities and ways to minimise their realisation: a focus on Melbourne. Ecological Management & Restoration 15:102-110. Hahs AK, et al. 2009. A global synthesis of plant extinction rates in urban areas. Ecol Lett 12:1165-1173. Hermy M, Honnay O, Firbank L, Grashof-Bokdam C, Lawesson JE. 1999. An ecological comparison between ancient and other forest plant species of Europe, and the implications for forest conservation. Biological Conservation 91:9-22. Hobbs RJ, et al. 2014. Managing the whole landscape: historical, hybrid, and novel ecosystems. Frontiers in Ecology and the Environment 12:557-564. Hobbs RJ, Higgs E, Harris JA. 2009. Novel ecosystems: implications for conservation and restoration. Trends Ecol Evol 24:599-605. Hobbs RJ, Higgs ES, Hall C 2013. Novel ecosystems: intervening in the new ecological world order. John Wiley & Sons. Hothorn, T., Hornik, K., Zeileis, A. 2006. Unbiased recursive partitioning: A conditional inference framework. Journal of Computational and Graphical Statistics, 15:651-674. Ives CD, et al. 2016. Cities are hotspots for threatened species. Global Ecology and Biogeography 25:117-126. Jim CY. 2017. An archaeological and historical exploration of the origins of green roofs. Urban Forestry & Urban Greening 27:32-42. Jokimaki J, Suhonen J, Kaisanlahti-Jokimaki ML. 2018. Urban core areas are important for species conservation: A European-level analysis of breeding bird species. Landscape and Urban Planning 178:73-81. Kendal D, Zeeman BJ, Ikin K, Lunt ID, McDonnell MJ, Farrar A, Pearce LM, Morgan JW. 2017. The importance of small urban reserves for plant conservation. Biological Conservation 213:146-153. Klaus VH. 2013. Urban Grassland Restoration: A Neglected opportunity for biodiversity conservation. Restoration Ecology 21:665-669.

Chapter 2

Knapp S, Kühn I, Mosbrugger V, Klotz S. 2008. Do protected areas in urban and rural landscapes differ in species diversity? Biodiversity and Conservation 17:1595-1612. Knapp S, Winter M, Klotz S. 2017. Increasing species richness but decreasing phylogenetic richness and divergence over a 320-year period of urbanization. Journal of Applied Ecology 54:1152-1160. Kowarik I. 2011. Novel urban ecosystems, biodiversity, and conservation. Environ Pollut 159:1974-1983. Kowarik I, von der Lippe M. 2018. Plant population success across urban ecosystems: A framework to inform biodiversity conservation in cities. Journal of Applied Ecology 55:2354-2361. Kowarik I, Buchholz S, von der Lippe M, Seitz B 2016. Biodiversity functions of urban cemeteries: Evidence from one of the largest Jewish cemeteries in Europe. Urban Forestry & Urban Greening 19:68-78. Kühn I, Brandl R, Klotz S. 2004. The flora of German cities is naturally species rich. Evolutionary Ecology Research 6:749-764. Lachmund J 2013. Greening Berlin: The Co-production of Science, Politics, and Urban Nature. MIT Press. Lawson DM, Lamar CK, Schwartz MW. 2008. Quantifying plant population persistence in human-dominated landscapes. Conserv Biol 22:922-928. Lenzen M, Lane A, Widmer-Cooper A, Williams M. 2009. Effects of land use on threatened species. Conserv Biol 23:294-306. Lepczyk CA, Aronson MFJ, Evans KL, Goddard MA, Lerman SB, Macivor JS. 2017. Biodiversity in the City: Fundamental Questions for Understanding the Ecology of Urban Green Spaces for Biodiversity Conservation. Bioscience 67:799-807. Loram A, Thompson K, Warren PH, Gaston KJ 2008. Urban domestic gardens (XII): the richness and composition of the flora in five UK cities. Journal of Vegetation Science 19:321-330. Lososova Z, et al. 2012. Native and alien floras in urban habitats: a comparison across 32 cities of central Europe. Global Ecology and Biogeography 21:545-555. Louf R, Barthelemy M. 2014. A typology of street patterns. Journal of The Royal Society Interface 11:20140924. Lundholm JT, Richardson PJ. 2010. MINI-REVIEW: Habitat analogues for reconciliation ecology in urban and industrial environments. Journal of Applied Ecology 47:966-975. Maclagan SJ, Coates T, Ritchie EG. 2018. Don't judge habitat on its novelty: Assessing the value of novel habitats for an endangered mammal in a peri-urban landscape. Biological Conservation 223:11-18.

Chapter 2

Mcdonald RI, Kareiva P, Formana RTT. 2008. The implications of current and future urbanization for global protected areas and biodiversity conservation. Biological Conservation 141:1695-1703. McDonnell MJ, Hahs AK. 2015. Adaptation and Adaptedness of Organisms to Urban Environments. Annual Review of Ecology, Evolution, and Systematics, Vol 46 46:261-280. McKinney ML. 2008. Effects of urbanization on species richness: A review of plants and animals. Urban Ecosystems 11:161-176. Micallef L, Rodgers P. 2014. eulerAPE: drawing area-proportional 3-Venn diagrams using ellipses. PLoS One 9:e101717. Nielsen AB, van den Bosch M, Maruthaveeran S, van den Bosch CK. 2014. Species richness in urban parks and its drivers: A review of empirical evidence. Urban Ecosystems 17:305-327. Nilon CH, et al. 2017. Planning for the Future of Urban Biodiversity: A Global Review of City- Scale Initiatives. Bioscience 67:331-341. R Core Team. 2018. R: A Language and Environment for Statistical Computing. Ramalho CE, Hobbs RJ. 2012. Time for a change: dynamic urban ecology. Trends Ecol Evol 27:179-188. Ramalho CE, Laliberte E, Poot P, Hobbs RJ. 2014. Complex effects of fragmentation on remnant woodland plant communities of a rapidly urbanizing biodiversity hotspot. Ecology 95:2466- 2478. Schwartz MW, Jurjavcic NL, O'Brien JM. 2002. Conservation's disenfranchised urban poor. Bioscience 52:601-606. Schwartz MW, Smith LM, Steel ZL. 2013. Conservation investment for rare plants in urban environments. PLoS One 8:e83809. Seitz B, Ristow M, Meißner J, Machatzi B, Sukopp H 2018. Rote Liste und Gesamtartenliste der etablierten Farn-und Blütenpflanzen von Berlin. Universitätsverlag der TU Berlin. SenStadtUm. 2014. Berlin Environmental Atlas, Biotope Types. www.stadtentwicklung.berlin.de/umwelt/umweltatlas/ek508.htm SenStadtUm. 2015. Actual Use of Built-up Areas, Inventory of Green and Open Spaces, Actual Use and Vegetation Cover. SenStadtUm. 2016. Berlin Environmental Atlas, Map Actual Use and Vegetation. Shwartz A, Turbe A, Julliard R, Simon L, Prevot AC. 2014. Outstanding challenges for urban conservation research and action. Global Environmental Change-Human and Policy Dimensions 28:39-49. Smith RM, Thompson K, Hodgson JG, Warren PH, Gaston KJ. 2006. Urban domestic gardens (IX): Composition and richness of the vascular plant flora, and implications for native biodiversity. Biological Conservation 129:312-322.

Chapter 2

Soanes K, Sievers M, Chee YE, Williams NSG, Bhardwaj M, Marshall AJ, Parris KM. 2018. Correcting common misconceptions to inspire conservation action in urban environments. Conserv Biol. https://doi.org/10.1111/cobi.13193. Stiftung Naturschutz Berlin. 2015. Koordinierungsstelle Florenschutz – ein Projekt zur Umsetzung des Florenschutzkonzeptes Berlin. Sukopp H. 1987. On the history of plant geography and plant ecology in Berlin. Englera 7:85- 103. Sukopp H. (ed) 1990. Stadtökologie. Das Beispiel Berlin.–Berlin: D. Reimer Verlag.–1990.–455 s. Sukopp H, Weiler S. 1988. Biotope Mapping and Nature Conservation Strategies in Urban Areas of the Federal-Republic-of-Germany. Landscape and Urban Planning 15:39-58. Williams NSG, et al. 2009. A conceptual framework for predicting the effects of urban environments on floras. Journal of Ecology 97:4-9. Williams NSG, Lundholm J, Scott MacIvor J. 2014. Do green roofs help urban biodiversity conservation? Journal of Applied Ecology 51:1643-1649. Yılmaz H, Kuşak B, Akkemik Ü 2018. The role of Aşiyan Cemetery (İstanbul) as a green urban space from an ecological perspective and its importance in urban plant diversity. Urban Forestry & Urban Greening 33: 92-98. Zeeman BJ, McDonnell MJ, Kendal D, Morgan JW. 2017. Biotic homogenization in an increasingly urbanized temperate grassland ecosystem. Journal of Vegetation Science 28:550-561. Zeeman BJ, Morgan JW. 2018. Increasing and declining native species in urban remnant grasslands respond differently to nitrogen addition and disturbance. Ann Bot 121:691-697.

Chapter 3

Chapter 3: Plant strategy and dispersal syndrome drive endangered species’ colonization into the urban matrix

Chapter 3

This chapter is the preprint version of the article submitted to Frontiers in Ecology and Evolution for consideration for publication as: G. Planchuelo, I. Kowarik, M. von der Lippe M, 2019, Endangered plants in novel urban ecosystems are filtered by strategy type and dispersal syndrome, not by spatial dependence on natural remnants

Abstract

The contribution of cities to nature conservation is gaining increasing importance with a globally accelerating urbanisation. Previous research shows that urban environments can host high numbers of species, including endangered species, and that these can occur throughout a wide range of urban ecosystems that differ in ecological novelty (i.e. from natural remnants to novel ecosystems). However, as natural remnants host the highest diversity and population density of endangered plant species, there is still one fundamental open question for urban nature conservation: are there dispersal limitations that prevent these species from successfully colonising from their source populations in natural remnants into the rest of the urban matrix? This study uses a unique dataset of precisely mapped point data of 1,742 populations of 213 endangered plant species in the city of Berlin and applies the most recent techniques in point pattern analysis to unravel their spatial dependence towards natural remnant ecosystems. Results show that populations located in novel ecosystems, which comprise more than a third of the total, are not spatially dependant to natural remnant ecosystems. Additionally, these populations tend to have significantly higher proportions of ruderal species with anemochory and stomatochory (e.g. in the beaks of birds) as dispersal syndromes. Our research thus demonstrates that populations of endangered plant species in novel habitats have overcome the urban dispersal barriers as a consequence of their predominantly ruderal qualities and from benefiting from dispersal syndromes that facilitate seed transport in cities.

Chapter 3

Introduction

The contribution of cities to biodiversity conservation is increasingly important with a globally accelerating urbanization (Angel et al. 2011; Seto et al. 2012), as they are environments that can potentially aid in preserving nature not only locally, but also worldwide (McKinney 2008; Kowarik 2011). Urban floras have been found to be rich in species due to human agency and habitat heterogeneity (Kühn et al. 2004; McKinney 2008), and can host a surprisingly high share of endangered species (Schwartz et al. 2002; Lawson et al. 2008; Lenzen et al. 2009; Shwartz et al. 2014; Ives et al. 2016; Kowarik & von der Lippe 2018; Planchuelo et al. in revision). For instance, 22% of the endangered plant species in the United States are found in the 40 biggest cities (Schwartz et al. 2002), and urban areas host 30% of the endangered plant species in Australia, supporting more endangered species per unit area than rural landscapes (Ives et al. 2016). Indeed, urban ecosystems are acting as substitute habitats for some species that have already gone extinct in (near-)natural landscapes, as some endangered species are exclusively confined to urban areas (Ives et al. 2016).

Cities hold a wide mosaic of small patches with a wide range of environmental conditions that generate a very heterogeneous landscape (Grimm et al. 2000; Cadenasso et al. 2007; Kowarik 2011).Correspondingly, urban biodiversity is not randomly distributed in the space, but rather tends to follow specific spatial patterns (Kowarik 1990; McKinney 2002; Shochat et al. 2006; McKinney 2008; Swan et al. 2011; Schmidt et al. 2014). For instance, different types of urban ecosystems in of Berlin (Germany) - ranging from natural remnants to novel ecosystems - have been shown to harbour endangered plant species (Kowarik & von der Lippe 2018; Planchuelo et al. in revision). Interestingly, the highest number of endangered species (Kowarik & von der Lippe 2018) and the highest density of populations of endangered plant species (Planchuelo et al. in revision) was found to occur in urban natural remnants ecosystems, with numbers decreasing the further away from them (Jarošík et al. 2011), suggesting possible dispersal limitations that prevent these species from successfully colonising other urban ecosystems. Indeed, the composition of urban species assemblages is filtered by multiple human impacts and a wide range of environmental barriers (Williams et al. 2009; Aronson et al. 2016; Knapp et al. 2017; Kowarik & von der Lippe 2018), that also determine to which extent endangered species colonize different urban ecosystems from natural remnants as source habitats. Thus, one critical question for understanding the urban contribution to biodiversity conservation is to which extent endangered species are confined to natural remnants and their vicinity (source sink dynamic; Pulliam 1988), or whether they can spread beyond their limits into anthropogenic urban habitats and establish independent populations there.

Chapter 3

As dispersal is an ecological process that causes spatial autocorrelation in species distributions (Epperson 2005; Bahn et al. 2008), spatial statistics can be used to assess whether or not populations located in anthropogenic urban ecosystems are still dependent to populations in natural remnants. Point pattern analysis has already been extensively applied on endangered plant species in non-urban areas to assess their dependence on habitat patches of other species (Wehenkel et al. 2015). Examples involve work on the effects of an animal vector on the dispersal of an endangered shrub (Rodríguez-Pérez et al. 2012), or on the minimum areas for the conservation of an endangered plant (Lim et al. 2008). Even if there is growing evidence of the importance of spatial patterning for biodiversity conservation in cities (Rastandeh et al. 2017), insights to these spatial patterns and processes in urban landscapes are scarce as exact spatial information on the locations of endangered species is often not available.

Whether endangered plant species colonize into anthropogenic urban ecosystems depends on how they negotiate with a set of barriers related to dispersal (Kowarik & von der Lippe 2018). Many of the plants that are able to pass dispersal filters in cities tend to be ruderal species, as they can rapidly colonise new urban habitats while tolerating a wide range of urban disturbances (Prach et al. 2001; McKinney 2008). Additionally, some species might possess dispersal traits that are pre-adapted or adapting to urbanity (McDonnell & Hahs 2015). For instance, it has been shown that some dispersal syndromes increase long-distance dispersal in cities as a consequence of human activity (Kowarik & Von der Lippe 2011), including native species (von der Lippe & Kowarik 2008). Indeed, species that have human vectors as an additional dispersal pathway were found to be increasingly abundant over time in densely urbanized neighbourhoods (Johnson et al. 2017) and to be less affected by urban dispersal limitations (Chytrý et al. 2012).

Here, we tested to which extent and why endangered plant species in an urban landscape can occur in anthropogenic ecosystems (i.e. hybrid ecosystems or novel ecosystems according to the novel ecosystem concept; Hobbs et al. 2009) spatially independently from natural remnants. We used a unique dataset with the exact geographical position of 1742 populations of 213 highly endangered plant species in Berlin to test, first whether their locations are randomly distributed throughout the urban matrix or spatially clustered. In a subsequent step we performed spatial analyses to unravel the spatial correlation of the populations towards remnant ecosystems patches. We repeated the same spatial analyses with two subsets of the populations; those located in hybrid ecosystems and those located in novel ecosystems, in order to separately evaluate their spatial relation to remnant ecosystems. Finally, we linked the population occurrence across remnant, hybrid and novel ecosystems by means of a classification tree to plant traits that have been shown in previous urban studies to be relevant for population

Chapter 3

dispersal and are related to the CSR strategy (Grime 1977), the seed traits, and the dispersal syndrome of the different endangered plant species.

We hypothesised that populations of endangered plant species located in hybrid ecosystems are spatially related to seed sources in natural remnants, differently to populations in located in novel ecosystems, which are occurring independently from source populations in natural remnants. Populations in novel ecosystems were expected to have the highest proportion of ruderal species and have dispersal syndromes that facilitate seed movement in cities (e.g. as a consequence of human-mediated dispersal).

Methods

Study area and population data

This study was carried out in the city of Berlin, the capital and most populous city of Germany with 3.6 million inhabitants as of 2017. Berlin has a total area of 891km2, from which 59% is covered by built-up areas and streets, while green and blue spaces occupy 41% of the area - including forests (17.7%), lakes and rivers (6.1%), parks (5.6%), allotment gardens (5.3%), fields (5%) and meadows (1.3%) (SenStadtUm 2016). Berlin is an example of European historical city (Sukopp 1990; Lachmund 2013) that shares with a big majority of other European cities a similar configuration of the urban matrix (Louf & Barthelemy 2014), proportion of native vegetation, plant extinctions and history of landscape arrangement (Hahs et al. 2009). The city offers great opportunities in urban ecology exploration, as it is an extremely heterogeneous environment that comprises remnants of natural and agrarian landscapes, urban greenspaces with varying land use histories as well as a diverse array of novel ecosystems. Within the built up areas a large amount of vacant land has still remained – as a result of the bombings during World War II and a subsequent slow urban development (Sukopp 1990; Lachmund 2013).

Since the 19th century, the flora of Berlin has been thoroughly studied (Sukopp 1987). Information on populations dynamics has been collected and updated numerous times in Red Lists of endangered species (Seitz et al. 2018). 17% of Berlin’s flora has gone extinct since the mid-19th century and 29% is currently being endangered. In this study, we used a detailed dataset with the precise geographical location of 1742 populations of 213 highly endangered plant species (Fig. 1, Appendix 1) from Berlin’s Flora Protection Program (Berliner Florenschutzkonzept). These species have been selected for the Flora Protection Program

Chapter 3

according to their: 1) red list status 2) long term population trends (50-150 years), 3) short term population trends (10-15 years) and 4) species specific or anthropogenic risk factors (Seitz 2007). The populations of these species are being monitored since 2009 by several experts on behalf of the Stiftung Naturschutz Berlin (2015).

Fig. 1 Location of all populations of endangered plant species (1742 populations, 213 species red dots) inside the city limits of Berlin. Remnant ecosystems are displayed in green.

Ecological novelty

A methodological approach was developed in the 1980s to assign different biotope types to the whole area of Berlin by incorporating various combinations of land uses and environmental conditions (Sukopp & Weiler 1988). The mapping was done as part of a combined work of 64 different projects between 2001 and 2013, where areas with high conservation value or low visibility such as forests were covered by field biotope mapping while built up areas and transportation corridors were primarily covered by data of the urban planning authorities

Chapter 3

(SenStadtUm 2014). Today, the Berlin Environmental Atlas includes a very detailed area wide biotope mapping that is updated regularly (SenStadtUm 2014). Currently, the city of Berlin is divided into 79,268 patches, with each of these patches being assigned to one of 12 biotope classes. These are forests, grasslands, ruderal sites, standing bodies of waters, built-up areas, bogs and marshes, hedges, green spaces, fields, flowing moving bodies of waters, heaths, and other types. These biotope classes are in turn further subdivided in 7,483 biotope types at several hierarchical levels.

In order to relate the geographical position of each of the 1,742 populations of endangered plant species to remnant, hybrid and novel ecosystems and assess their spatial correlation towards natural remnants, we used the ecological novelty map of Berlin (Planchuelo et al. in revision). It spatially merges land use data with the aforementioned biotope information to classify the 79,268 biotope patches of Berlin into three novelty categories, which were adapted from Hobbs et al. (2009), specified for urban settings by Kowarik and von der Lippe (2018) and used by Planchuelo et al. (in revision):

• Natural remnant ecosystems are relicts of natural ecosystems that are often slightly affected by urban impacts but still remain within their historical range of modifications. These include biotopes from mires and wetlands to near-natural forests or moderately used dry or wet grasslands (Fig. 1).

• Hybrid ecosystems are human mediated ecosystems that have been modified from their historical counterparts but still have the potential to return to historical conditions. Examples range from managed grasslands to urban greenspaces, young tree plantations in forests and parks or intermediate succession stages following abandonment.

• Novel ecosystems are human shaped ecosystems that have potentially been irreversibly changed by large modifications to their abiotic conditions or biotic composition. They include built up areas, rooftops, vacant lots, abandoned industrial areas or intensively managed agricultural fields.

The classification of biotope types into novelty categories can be challenging in some cases. For instance, grassland biotopes include a wide range of near-natural and anthropogenically shaped ecosystems in Berlin (Fischer et al. 2013), and should be assigned to different novelty categories as portrayed in Fig. 2. By overlaying the biotope map with the land use map from Berlin’s Environmental Atlas (SenStadtWo 2018), Planchuelo et al. (in revision) were able to distinguish between cases where the same grassland biotope type was associated with far reaching anthropogenic impacts such as agricultural fields, rooftops or sewage farms. The grassland

Chapter 3

patches in the first case were assigned to hybrid ecosystems, while the two latter were assigned to novel ecosystems (Fig. 2). This process allowed to further refine the preliminary novelty categorisation when the same biotope type was found in different land uses.

Fig 2 Decision tree illustrating the process of novelty categorization of biotope types by using the example of grassland biotopes. These can be assigned to different novelty categories depending on the land use type.

Spatial analyses

Merging the biotope map and the land use map to generate the novelty map was done by spatial intersections and later unions in Quantum-GIS (QGIS). To enable spatial analyses of point patterns between patches of ecosystem types and populations of endangered species, locations of patches were represented by a point positioned at the centroid of each patch (Baddeley 2008). Using the “polygon centroid” function in QGIS, we converted all ecosystem patches into point data. From those, we only selected for further analysis the 3426 points belonging to natural remnant ecosystems.

Spatial analyses consisted of an exploration of the spatial clustering of the populations, followed by an analysis of the spatial correlation between the populations and natural remnant ecosystems. Analyses were performed for the whole set of 1742 populations, as well as on two

Chapter 3

subgroups: The 632 populations located in hybrid ecosystems and the 687 populations located in novel ecosystems.

The spatial clustering patterns of populations located in the three different novelty categories were tested by means of Ripley’s L function (Dixon 2002), which allows to verify if, and at which spatial scales, populations are clustered, regularly or randomly distributed. This test compares the spatial distribution of a point pattern against a generated set of points with a random distribution (which is named “Poisson distribution”, see Stoyan and Stoyan 1994). If the curve of the L function of a point pattern lies above the Poisson line, the populations are clustered; if it lies below, the populations are regularly distributed; if it overlaps it, the populations are randomly distributed (Dixon 2002). L plots of Ripley’s K functions were used (Dixon 2002), as they stabilise the variance and are more powerful than simple plots of the K function (Baddeley et al. 2015).

Statistically significant spatial correlations between the populations and remnant ecosystems patches were assessed with cross L functions, which compare the observed association between the two point patterns of the empirical data against the association of the locations of remnant ecosystems with a simulated point pattern, named null model (Baddeley et al. 2015). Global envelopes were generated from 19 simulations of the null model to achieve a 95% confidence level (Baddeley et al. 2014). Global envelopes were chosen because they avoid the problem of data snooping of pointwise envelopes and give an accurate significance level (Baddeley et al. 2014). If the curve of the cross L function for the empirical point pattern transgresses outside of the global envelope of the null model at any moment, there is a statistically significant spatial correlation between the populations and natural remnant ecosystems (Baddeley et al. 2014; Baddeley et al. 2015). Alternatively, if the curve does not transgress outside of the global envelope, we consider that there is no spatial correlation between the populations and natural remnant ecosystems. L plots of cross K functions were used (Dixon 2002), as they stabilise the variance and are more powerful than the K plots (Baddeley et al. 2015).

Because the cross L function only reveals a spatial correlation between the populations and natural remnant ecosystems, but not if it is positive or negative, we used a cross pair correlation function in a successive step to assess the type of correlation at any given distance between the populations and the natural remnant ecosystems (Penttinen et al. 1992; Baddeley 2008). There is a spatial dependence between the populations and an ecosystem type when the values of the cross pair correlation function lay over 1, with values below 1 indicating spatial repulsion (Penttinen et al. 1992; Baddeley et al. 2015). The higher or lower the values, the stronger the dependence or the repulsion.

Chapter 3

All spatial statistics were performed in the R package “spatstat” (Baddeley et al. 2015). Ripley’s isotropic edge corrections (Ripley 1991) were applied to all calculations. As our point data proved to be non-stationary through a test for spatial heterogeneity, we conducted only spatial analyses that were adapted for inhomogeneous data.

Dispersal traits and CSR strategy

Information on the different plant dispersal traits (dispersal syndrome, seed mass, seed length, seed width, seed number, seed terminal velocity) and the CSR strategy (Grime 1977) of the species was gathered from the BIOFLOR database (Klotz et al. 2002) (Table 1).

Variable Units Scale Range

Dispersal traits

Dispersal syndrome / Categorical 5 categories

Seed mass mg Continuous 0.00009-166.3

Seed length mm Continuous 0.33 - 9

Seed width mm Continuous 0.1 - 4.5

Seed number n Continuous 0 - 4720600

Seed terminal velocity m/s Continuous 0.17 - 4.94

CSR Strategy / Categorical 7 categories

Table 1 List of predictor variables of population survival used in the conditional inference tree. Their units, scale and range in the model are shown.

In order to unravel the importance of the plant dispersal traits and the CSR strategy in determining the spatial distribution of populations of endangered plant species across the different types of ecological novelty, we performed a conditional inference tree analysis with the predictors in Table 1. A conditional inference tree is a non-parametric type of decision tree where the dataset is recursively split into dichotomous subsets which are discriminated by the most significant predictor (Hothorn et al. 2006). To reveal any positive or negative deviations from the expected frequencies of population distribution across the different types of ecological novelty, we additionally performed a log linear model with the “loglin” R-Function and

Chapter 3

computed the standardized residuals, which were included in the conditional inference tree. All analyses were performed with the statistical and programming software R version 3.5.2 (R Core Team 2018).

Results

Spatial clustering of populations of endangered plant species in the urban matrix

An analysis of the distribution pattern of the endangered plant populations by means of Ripley’s L function proved that they are grouped in clusters over the entire range of distances that occur between two single points. Furthermore, populations that are located in natural remnant ecosystems have a higher level of clustering than those located in hybrid or novel ecosystems (Fig. 3).

Fig. 3 Populations of endangered plant species in Berlin are located in clusters. The levels of clustering is higher for populations located in natural remnant ecosystems (line R, n=423) than for populations located in hybrid (line H, n=632) or novel (line N, n=687) ecosystems, as the Ripley’s L function of the first group lies far above that of the two latter (Dixon 2002). A random Poisson distribution is marked as a red dashed line.

Chapter 3

Spatial relation between populations of endangered plant species and remnant ecosystems

Results from the cross L function show that the whole set of populations of endangered plant species is spatially correlated to natural remnant ecosystems (Fig. 4a), as the observed curve of the point process crosses outside of the boundaries of the global envelope of the null model (Baddeley et al. 2014). An additional cross pair correlation function indicates that this correlation is strongly positive (dependence), especially at closer distances from 0 to 500m (Fig. 4b), where values lie very far above 1 (Penttinen et al. 1992; Baddeley et al. 2015).

a b

Fig. 4 Spatial correlation between populations of endangered plant species (n=1742) and patches of natural remnant ecosystems (n=3426). The cross L function (a) shows that there is a significant (p<0.05) spatial correlation between the populations and natural remnant ecosystems, as the observed curve of the point process (black line) crosses outside of the boundaries of the global envelope of the null model (dashed red line with grey envelope generated from 19 simulations) (Baddeley et al. 2014). The cross pair correlation function (b) shows that this correlation is strongly positive (dependence), as the correlation curve (black line) lies very far above 1 at most distances (Penttinen et al. 1992; Baddeley et al. 2015).

The cross L function for populations of endangered plant species located outside of remnant ecosystems but inside hybrid ecosystems show that they are spatially correlated to natural remnant ecosystems (Fig. 5a). A cross pair correlation function shows positive spatial interactions, especially at distances from 0m to 700m (Fig. 5b), though this interaction is not as strong as with the whole set of populations.

Chapter 3

a b Fig. 5 Spatial correlation between populations of endangered plant species located in hybrid ecosystems (n=632) and natural remnant ecosystem patches (n=3426). The cross L function (a) shows that there is a significant (p<0.05) spatial correlation between the populations and natural remnant ecosystems, as the observed curve of the point process (black line) crosses outside of the boundaries of the global envelope of the null model (dashed red line with grey envelope generated from 19 simulations) (Baddeley et al. 2014). The cross pair correlation function (b) shows that this correlation is positive (dependence), as the correlation curve (black line) lies above 1 at most distances (Penttinen et al. 1992; Baddeley et al. 2015).

In contrast to populations in hybrid systems, populations that are currently located in novel ecosystems are not spatially correlated (neither dependence nor repulsion) to natural remnant ecosystems (Fig. 6), as the observed curve of the point process does not cross outside of the boundaries of the global envelope of the null model (Baddeley et al. 2014).

Fig. 6 Spatial correlation between populations of endangered plant species located in novel ecosystems (n=687) and natural remnant ecosystem patches (n=3426). The cross L function

Chapter 3

shows that there is no spatial correlation between the populations and natural remnant ecosystems with a significance level of 0.05, as the observed curve of the point process (black line) does not cross outside of the boundaries of the global envelope of the null model (dashed red line with grey envelope generated from 19 simulations) (Baddeley et al. 2014).

Spatial relation between hybrid ecosystems and remnant ecosystems

To unravel if the spatial correlation of populations in hybrid ecosystems towards patches of natural remnant ecosystems was merely fruit of a spatial correlation between remnant and hybrid ecosystem patches, we performed a cross L function for the latter. Results show that patches of hybrid ecosystems are not spatially correlated (neither dependence nor repulsion) to patches of remnant ecosystems (Fig. 7), as the observed curve of the point process does not cross outside of the boundaries of the global envelope of the null model (Baddeley et al. 2014). Cross L Function Envelope Remnant and Hybrid Ecosystems Cross L function L Cross 500 1000 1500 2000 2500 0

0 500 1000 1500 2000 2500

Distance (m)

Fig. 7 Spatial correlation between hybrid ecosystem patches (n=15681) and natural remnant ecosystem patches (n=3426). The cross L function shows that there is no spatial correlation between the populations and natural remnant ecosystems with a significance level of 0.05, as the observed curve of the point process (black line) does not cross outside of the boundaries of the global envelope of the null model (dashed red line with grey envelope generated from 19 simulations) (Baddeley et al. 2014).

Chapter 3

Dispersal traits and CSR strategy

From all the predictor variables, the conditional inference tree retained CSR strategy (p<0.001) and dispersal syndrome (p=0.001) as significant predictors of population occurrence across the different types of ecological novelty (Fig. 8). Ruderal species with anemochory and stomatochory (e.g. in the beaks of birds) as dispersal syndromes were exceedingly represented in novel ecosystems (Node 3, Fig. 8), with non-ruderal species being much more common in remnant ecosystems than elsewhere (Node 5, Fig.8). Finally, ruderal populations with other dispersal syndromes were represented equally amongst natural remnants, hybrid and novel ecosystems (Node 4, Fig. 8).

CSR strategy p < 0.001

Ruderal species Non-ruderal species

Dispersal syndrome p = 0.001

anemochory, autochory, hydrochory, stomatochory endo/epizoochory

Node 3 (n=504) Node 4 (n=585) Node 5 (n=653)

o _ o o + +

o o Population occurrence occurrence Population

Fig. 8 Conditional inference tree showing the partitioning effects of plant dispersal traits and CSR strategy on the occurrence of populations of endangered plant species across natural remnant (green), hybrid (yellow) and novel (blue) ecosystems in Berlin. Positive and negative deviations of expected frequencies from a log-linear model are displayed by symbols in the pie charts (+, more than 4 standardised residuals; –, less than 4 standarised residuals; o, -2 to 2 standarised residuals).

Chapter 3

Discussion

Previous research has shown that urban biodiversity is not randomly distributed in the space, but rather tends to follow specific spatial patterns (Kowarik 1990; McKinney 2002; Shochat et al. 2006; McKinney 2008; Swan et al. 2011; Schmidt et al. 2014). Gaining an understanding on the spatial distribution of urban biodiversity could offer great advantages to develop better management strategies (Rastandeh et al. 2017), but this information is often missing (Shwartz et al. 2014). However, recent studies from Berlin (i.e. Kowarik and von der Lippe 2018; Planchuelo et al. in revision) have shown that natural remnant ecosystems host the highest numbers of endangered plant species and the highest density of their populations. These results point to an important of species’ ability to colonize anthopogenic urban habitats independently from source populations in natural remnants.

By analysing the geographical arrangement of the populations of endangered plant species and assessing the spatial correlation of populations towards natural remnant ecosystems, this study makes a step forward in unravelling the distribution pattern of endangered species throughout the urban matrix. This study shows for the first time with an extensive dataset that endangered plant species are not randomly distributed throughout the city, but are overall clustered towards natural remnants as primary seed sources of endangered species in cities. This can be expected taking into account that natural remnant ecosystems have species composition and environmental characteristics close to their historical counterparts where many of these species still persist (Hobbs et al. 2009; Kowarik & von der Lippe 2018). Additionally, the populations of endangered plant species located in hybrid ecosystems also exhibited a strong spatial dependence towards natural remnants - even when there was a lack of spatial correlation between remnant patches and hybrid patches. This indicates that dispersal processes are taking place (Epperson 2005; Bahn et al. 2008), and that remnant ecosystem patches act as seed sources for the spread of endangered plants into hybrid ecosystems through a source-sink dynamic (Schreiber 2010). Alternatively, some populations in hybrid ecosystems might be relicts of former remnants that have experienced land use changes and increasing levels of urbanisation (Hobbs et al. 2009; Kowarik & von der Lippe 2018).

As novel ecosystems have experienced much deeper changes to the abiotic conditions and species assemblages (Hobbs et al. 2009; Kowarik 2011), populations of endangered plant species in novel ecosystems can only originate from dispersal processes. Results clearly support this hypothesis, as populations located in novel ecosystems did not show any spatial correlation to natural remnant ecosystems (nor dependence nor repulsion). This holds for 687 populations of 142 species that have obviously overcome dispersal barriers from natural remnants and

Chapter 3

novel urban ecosystems and successfully established independent populations. In our study, ruderal species had a much higher frequency in novel ecosystems than it was expected. Ruderal species produce many seeds that can easily colonize new urban habitats while coping with their high disturbance levels, and previous urban research has found higher frequency of these plants at increasing urbanisation (Prach et al. 2001; McKinney 2008).

Surprisingly, our results showed that those ruderal species that use wind dispersal were exceedingly represented in novel ecosystems, whose populations were shown in our study to have overcome urban dispersal barriers. This differs from previous research that shows that anemochorous species often have reduced dispersal distances in fragmented landscapes (Soons et al. 2005), and are more likely to disappear from urban ecosystems (Williams et al. 2015). An explanation for our results lies on the fact that many species in cities benefit from alternative dispersal vectors such as human mediated dispersal (Johnson et al. 2017). In detail, it has been shown that seeds from native species that are pre-adapted to anemochory increase their dispersal distances from vehicle airflow as an additional dispersal vector (von der Lippe & Kowarik 2008). This process has been shown to enhance long-distance dispersal in urban anemochorous species by facilitating seed movement throughout the road network (Kowarik & Von der Lippe 2011).

Our study also revealed a higher than expected representation of stomatochorous species (e.g. dispersal in the beaks of birds) in novel ecosystems. Luna et al. (2018) have shown that birds can disperse native plant species effectively in cities when they pick up a fruit from a plant and then consume it at a distant perching tree. Indeed, dispersal distances were found to reach up to 155 meters even in the urban fragmented landscape (Luna et al. 2018).

Implications for conservation

Our results emphasize the significance of natural remnant ecosystems not only as habitats of endangered plant species but also as seed sources for the colonization of adjacent urban ecosystems - which combined add up to 60,6% of the total populations of endangered plant species in Berlin (Planchuelo et al. in revision). Indeed, in Rome remnant ecosystems have already been shown to host large numbers of plants of high conservation interest (Celesti- Grapow et al. 2013), while in Berlin natural remnants hold the highest numbers of native and endangered species (Kowarik & von der Lippe 2018) as well as of the highest density of populations of endangered plant species (Planchuelo et al. in revision). However, conservation

Chapter 3

practices should not only be reduced to natural remnant patches, but also towards their surrounding areas in order to include the endangered populations that are dependent on them.

On the other hand, novel ecosystems offer new opportunities for biodiversity conservation, as the largest number of the populations of endangered plant species is deeply embedded in them and no longer dependent on natural remnants. Indeed, many novel urban spaces such as wastelands (Bonthoux et al. 2014) have been shown to act as suitable habitat analogues for plant species (Lundholm & Richardson 2010), and even if they can be highly fragmented, small urban patches have been shown to have a comparable conservation potential to bigger areas (Kendal et al. 2017), especially for plants, which in general need smaller habitat requirements (Gaston et al. 1998).

As human mediated dispersal by the airflow of vehicles might be responsible for the dispersal of some endangered plant species throughout a city, appropriate management of road verges might be most profitable for biodiversity conservation. Because dispersal of invasive species can also occur by the same means (von der Lippe & Kowarik 2007), promoting native species vs. invasive species in urban road corridors might offer double benefits for urban conservation goals – more so considering that these locations have a great potential to offer many additional ecosystem services in cities (Säumel et al. 2015).

Finally, as dispersal in the beaks of birds was found to be an important vector aiding to overcome the dispersal barriers of endangered plant species in cities, we recommend management practices that promote the ample distribution of birds throughout the urban matrix. These can include preserving the heterogeneity of native vegetation and lowering the management of some urban parks to promote bird species diversity in them (Shwartz et al. 2008).

Future research should address the survival of individual populations in order to reveal if endangered plant species in novel ecosystems are self-sustaining and can persist in those environments, and if natural remnant ecosystems can continue acting as seed sources for a significant number of urban populations of endangered plant species. The mere occurrence of these populations is not indicative of their survival because there are often different time lags in species responses according to distinct histories of urbanisation (Du Toit et al. 2016). Indeed, the current numbers of populations in novel or remnant ecosystems could actually be masking future extinction debts (Hahs & McDonnell 2014), as harsh environmental settings in some urban environments have been shown to act as a barrier that hinders reproduction (Lundholm & Richardson 2010; Kowarik & von der Lippe 2018) and decreases plant establishment (Kowarik & von der Lippe 2018). Extinction processes need not only be assessed by studying

Chapter 3

population survival, but also by considering the different traits of the plants in order to unravel which ones might increase plant fitness throughout different urban ecosystems (Duncan et al. 2011).Considering the findings of this study, it additionally would be most interesting to assess if anemochorous endangered species in novel urban ecosystems tend to be located closer to roads or on road verges. This would unravel if they are indeed experiencing long-distance dispersal by the airflow of vehicles through the road network, or if other dispersal vectors might be responsible of their overabundance in novel ecosystems.

Literature

Angel S, Parent J, Civco DL, Blei A, Potere D. 2011. The dimensions of global urban expansion: Estimates and projections for all countries, 2000-2050. Progress in Planning 75:53-107. Aronson MF, et al. 2016. Hierarchical filters determine community assembly of urban species pools. Ecology 97:2952-2963. Baddeley A. 2008. Analysing spatial point patterns in R. Citeseer. Baddeley A, Diggle PJ, Hardegen A, Lawrence T, Milne RK, Nair G. 2014. On tests of spatial pattern based on simulation envelopes. Ecological Monographs 84:477-489. Baddeley A, Rubak E, Turner R 2015. Spatial point patterns: methodology and applications with R. Chapman and Hall/CRC. Bahn V, Krohn WB, O’Connor RJ. 2008. Dispersal leads to spatial autocorrelation in species distributions: A simulation model. Ecological Modelling 213:285-292. Bonthoux S, Brun M, Di Pietro F, Greulich S, Bouché-Pillon S. 2014. How can wastelands promote biodiversity in cities? A review. Landscape and urban planning 132:79-88. Cadenasso ML, Pickett STA, Schwarz K. 2007. Spatial heterogeneity in urban ecosystems: reconceptualizing land cover and a framework for classification. Frontiers in Ecology and the Environment 5:80-88. Celesti-Grapow L, Capotorti G, Del Vico E, Lattanzi E, Tilia A, Blasi C. 2013. The vascular flora of Rome. Plant Biosystems 147:1059-1087. Chytrý M, Lososová Z, Horsák M, Uher B, Čejka T, Danihelka J, Fajmon K, Hájek O, Juřičková L, Kintrová K. 2012. Dispersal limitation is stronger in communities of microorganisms than macroorganisms across Central European cities. Journal of biogeography 39:1101-1111. Dixon PM. 2002. Ripley's K function. Encyclopedia of environmetrics. Du Toit MJ, Kotze DJ, Cilliers SS. 2016. Landscape history, time lags and drivers of change: urban natural grassland remnants in Potchefstroom, South Africa. Landscape ecology 31:2133- 2150.

Chapter 3

Duncan RP, Clemants SE, Corlett RT, Hahs AK, McCarthy MA, McDonnell MJ, Schwartz MW, Thompson K, Vesk PA, Williams NS. 2011. Plant traits and extinction in urban areas: a meta- analysis of 11 cities. Global Ecology and Biogeography 20:509-519. Epperson BK. 2005. Estimating dispersal from short distance spatial autocorrelation. Heredity 95:7. Fischer LK, von der Lippe M, Kowarik I. 2013. Urban land use types contribute to grassland conservation: The example of Berlin. Urban Forestry & Urban Greening 12:263-272. Gaston KJ, Quinn RM, Blackburn TM, Eversham BC. 1998. Species-range size distributions in Britain. Ecography 21:361-370. Grime JP. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. The American Naturalist 111:1169-1194. Grimm NB, Grove JG, Pickett ST, Redman CL. 2000. Integrated approaches to long-term studies of urban ecological systems: Urban ecological systems present multiple challenges to ecologists—Pervasive human impact and extreme heterogeneity of cities, and the need to integrate social and ecological approaches, concepts, and theory. BioScience 50:571-584. Hahs AK, McDonnell MJ. 2014. Extinction debt of cities and ways to minimise their realisation: a focus on Melbourne. Ecological Management & Restoration 15:102-110. Hahs AK, McDonnell MJ, McCarthy MA, Vesk PA, Corlett RT, Norton BA, Clemants SE, Duncan RP, Thompson K, Schwartz MW. 2009. A global synthesis of plant extinction rates in urban areas. Ecology Letters 12:1165-1173. Hobbs RJ, Higgs E, Harris JA. 2009. Novel ecosystems: implications for conservation and restoration. Trends Ecol Evol 24:599-605. Hothorn T, Hornik K, Zeileis A. 2006. Unbiased recursive partitioning: A conditional inference framework. Journal of Computational and Graphical statistics 15:651-674. Ives CD, et al. 2016. Cities are hotspots for threatened species. Global Ecology and Biogeography 25:117-126. Jarošík V, Konvička M, Pyšek P, Kadlec T, Beneš J. 2011. Conservation in a city: Do the same principles apply to different taxa? Biological Conservation 144:490-499. Johnson AL, Borowy D, Swan CM. 2017. Land use history and seed dispersal drive divergent plant community assembly patterns in urban vacant lots. Journal of Applied Ecology. Kendal D, Zeeman BJ, Ikin K, Lunt ID, McDonnell MJ, Farrar A, Pearce LM, Morgan JW. 2017. The importance of small urban reserves for plant conservation. Biological Conservation 213:146-153. Klotz S, Kühn I, Durka W, Briemle G. 2002. BIOLFLOR: Eine Datenbank mit biologisch- ökologischen Merkmalen zur Flora von Deutschland.

Chapter 3

Knapp S, Winter M, Klotz S. 2017. Increasing species richness but decreasing phylogenetic richness and divergence over a 320-year period of urbanization. Journal of Applied Ecology 54:1152-1160. Kowarik I. 1990. Some response of flora and vegetation to urbanization in central Europe. Urban ecology:45-74. Kowarik I. 2011. Novel urban ecosystems, biodiversity, and conservation. Environ Pollut 159:1974-1983. Kowarik I, Von der Lippe M. 2011. Secondary wind dispersal enhances long-distance dispersal of an invasive species in urban road corridors. NeoBiota 9:49-70. Kowarik I, von der Lippe M. 2018. Plant population success across urban ecosystems: A framework to inform biodiversity conservation in cities. Journal of Applied Ecology 55:2354-2361. Kühn I, Brandl R, Klotz S. 2004. The flora of German cities is naturally species rich. Evolutionary Ecology Research 6:749-764. Lachmund J 2013. Greening Berlin: The Co-production of Science, Politics, and Urban Nature. MIT Press. Lawson DM, Lamar CK, Schwartz MW. 2008. Quantifying plant population persistence in human-dominated landscapes. Conserv Biol 22:922-928. Lenzen M, Lane A, Widmer-Cooper A, Williams M. 2009. Effects of land use on threatened species. Conserv Biol 23:294-306. Lim Y, Na S-T, Lee S-J, Cho KH, Shin H. 2008. Spatial distribution patterns and implications for conservation ofScrophularia takesimensis (Scrophulariaceae), an endangered endemic species on Ulleung Island, Korea. Journal of Plant Biology 51:213. Louf R, Barthelemy M. 2014. A typology of street patterns. Journal of The Royal Society Interface 11:20140924. Luna Á, Romero-Vidal P, Hiraldo F, Tella JL. 2018. Cities may save some threatened species but not their ecological functions. PeerJ 6:e4908. Lundholm JT, Richardson PJ. 2010. MINI-REVIEW: Habitat analogues for reconciliation ecology in urban and industrial environments. Journal of Applied Ecology 47:966-975. McDonnell MJ, Hahs AK. 2015. Adaptation and Adaptedness of Organisms to Urban Environments. Annual Review of Ecology, Evolution, and Systematics, Vol 46 46:261-280. McKinney ML. 2002. Urbanization, Biodiversity, and ConservationThe impacts of urbanization on native species are poorly studied, but educating a highly urbanized human population about these impacts can greatly improve species conservation in all ecosystems. Bioscience 52:883-890.

Chapter 3

McKinney ML. 2008. Effects of urbanization on species richness: A review of plants and animals. Urban Ecosystems 11:161-176. Penttinen A, Stoyan D, Henttonen HM. 1992. Marked point processes in forest statistics. Forest science 38:806-824. Planchuelo G, von der Lippe M, Kowarik I. In revision. Untangling the role of urban ecosystems as habitats for endangered plant species. Landscape and Urban Planning. Prach K, Pyšek P, Bastl M. 2001. Spontaneous vegetation succession in human-disturbed habitats: a pattern across seres. Applied Vegetation Science 4:83-88. Pulliam HR. 1988. Sources, Sinks, and Population Regulation. American Naturalist 132:652-661. R Core Team. 2018. R: A Language and Environment for Statistical Computing. Rastandeh A, Brown DK, Pedersen Zari M. 2017. Biodiversity conservation in urban environments: a review on the importance of spatial patterning of landscapes. Eco-city World Summit:12-14. Ripley BD 1991. Statistical inference for spatial processes. Cambridge university press. Rodríguez-Pérez J, Wiegand T, Traveset A. 2012. Adult proximity and frugivore's activity structure the spatial pattern in an endangered plant. Functional Ecology 26:1221-1229. Säumel I, Weber F, Kowarik I. 2015. Toward livable and healthy urban streets: Roadside vegetation provides ecosystem services where people live and move. Environmental Science & Policy. Schmidt KJ, Poppendieck HH, Jensen K. 2014. Effects of urban structure on plant species richness in a large European city. Urban Ecosystems 17:427-444. Schreiber SJ. 2010. Interactive effects of temporal correlations, spatial heterogeneity and dispersal on population persistence. Proceedings of the Royal Society B: Biological Sciences 277:1907-1914. Schwartz MW, Jurjavcic NL, O'Brien JM. 2002. Conservation's disenfranchised urban poor. Bioscience 52:601-606. Seitz B. 2007. Konzeption zum Florenschutz im Land Berlin. Gutachten im Auftrag des Landesbeauftragten für Naturschutz und Landschaftspflege Berlin. Seitz B, Ristow M, Meißner J, Machatzi B, Sukopp H 2018. Rote Liste und Gesamtartenliste der etablierten Farn-und Blütenpflanzen von Berlin. Universitätsverlag der TU Berlin. SenStadtUm. 2014. Berlin Environmental Atlas, Biotope Types, www.stadtentwicklung.berlin.de/umwelt/umweltatlas/ek508.htm SenStadtUm. 2016. Berlin Environmental Atlas, Map Actual Use and Vegetation. SenStadtWo. 2018. Berlin Environmental Atlas.

Chapter 3

Seto KC, Güneralp B, Hutyra LR. 2012. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proceedings of the National Academy of Sciences 109:16083-16088. Shochat E, Warren PS, Faeth SH, McIntyre NE, Hope D. 2006. From patterns to emerging processes in mechanistic urban ecology. Trends Ecol Evol 21:186-191. Shwartz A, Shirley S, Kark S. 2008. How do habitat variability and management regime shape the spatial heterogeneity of birds within a large Mediterranean urban park? Landscape and Urban Planning 84:219-229. Shwartz A, Turbe A, Julliard R, Simon L, Prevot AC. 2014. Outstanding challenges for urban conservation research and action. Global Environmental Change-Human and Policy Dimensions 28:39-49. Soons M, Messelink J, Jongejans E, Heil G. 2005. Habitat fragmentation reduces grassland connectivity for both short-distance and long-distance wind-dispersed forbs. Journal of Ecology 93:1214-1225. Stiftung Naturschutz Berlin. 2015. Koordinierungsstelle Florenschutz – ein Projekt zur Umsetzung des Florenschutzkonzeptes Berlin. Stoyan D, Stoyan H 1994. Fractals, random shapes and point fields: methods of geometrical statistics. Sukopp H. 1987. On the history of plant geography and plant ecology in Berlin. Englera 7:85- 103. Sukopp H. 1990. Stadtökologie. Das Beispiel Berlin. Reimer Verlag.–1990.–455 s. Sukopp H, Weiler S. 1988. Biotope Mapping and Nature Conservation Strategies in Urban Areas of the Federal-Republic-of-Germany. Landscape and Urban Planning 15:39-58. Swan CM, Pickett ST, Szlavecz K, Warren P, Willey KT. 2011. Biodiversity and community composition in urban ecosystems: coupled human, spatial, and metacommunity processes. Handbook of urban ecology. Oxford University Press, New York:179-186. von der Lippe M, Kowarik I. 2007. Long-distance dispersal of plants by vehicles as a driver of plant invasions. Conserv Biol 21:986-996. von der Lippe M, Kowarik I. 2008. Do cities export biodiversity? Traffic as dispersal vector across urban-rural gradients. Diversity and Distributions 14:18-25. Wehenkel C, Brazão-Protázio JM, Carrillo-Parra A, Martínez-Guerrero JH, Crecente-Campo F. 2015. Spatial distribution patterns in the very rare and species-rich Picea chihuahuana tree community (Mexico). PloS one 10:e0140442. Williams NSG, Hahs AK, Vesk PA. 2015. Urbanisation, plant traits and the composition of urban floras. Perspectives in Plant Ecology Evolution and Systematics 17:78-86. Williams NSG, et al. 2009. A conceptual framework for predicting the effects of urban environments on floras. Journal of Ecology 97:4-9.

Chapter 4

Chapter 4: Survival in the city? Population persistence of endangered plant species across different urban ecosystems

Chapter 4

This chapter is the preprint version of the article submitted to Journal of Applied Ecology for consideration for publication as: G. Planchuelo, I. Kowarik, M. von der Lippe M, 2019, Survival of endangered plants in the city: the role of plant traits, biotopes, protected areas and urbanization

Abstract

Given the rising urban land cover worldwide, the contribution of cities to the conservation of biodiversity becomes increasingly important. Previous research shows that urban environments can host high numbers of species, including endangered plant species, and that their populations occur across different biotope types with varying degrees of ecological novelty. However, one fundamental question for urban nature conservation remains ambiguous: to which extent and where can endangered plant species persist in urbanized landscapes? In this paper we evaluate the survival of 858 precisely monitored populations of 179 endangered plant species by assessing population survival throughout different urban ecosystems during a period of on average 7.6 years. By linking population survival to various urban landscape variables throughout the city (biotope class, patch size, habitat continuity, legal conservation status, proportion of impervious surfaces, floor space index, human population density, human population constancy, nearby street length, distance to the nearest street, distance to the nearest remnant ecosystem and proportion of nearby forests and grasslands) and plant traits (morphological traits, CSR strategy and realized niche of the species) we unravel the drivers of long-term population survival in urban endangered plant species. Results show that on the landscape scale, population survival is highly dependent on the biotope type where the population is located at. Surprisingly, populations in near natural habitats like forests and bogs or populations inside conservation areas were more prone to local extinction than populations in anthropogenically modified habitats or outside conservation areas. Additionally, the species soil humidity requirements as well as their CSR strategy were significant predictors of population survival, with competitive species with preference for low and medium soil humidity having the highest rates of survival. Our results demonstrate the need to broaden conservation approaches in cities to include anthropogenically modified ecosystems but also show the difficulties in maintaining populations of endangered species in natural ecosystems within the urban matrix.

Chapter 4

Introduction

As urbanization accelerates at a global scale, the question of the urban contribution to biodiversity conservation is becoming increasingly important (McKinney 2002; Nilon et al. 2017; Kowarik & von der Lippe 2018; Parris et al. 2018). The number of urban conservation studies has risen sharply in recent years (Shwartz et al. 2014) and indicate considerable opportunities for biodiversity conservation. Indeed, cities can be very rich in native species (McKinney 2002; Kühn et al. 2004; Aronson et al. 2014), including a considerable amount of rare and endangered plant species. Endangered plant species have been reported from cities across all continents, as from Africa (Rebelo et al. 2011), Asia (Wang et al. 2007), Australia (Ives et al. 2016), Europe (Jarošík et al. 2011; Kowarik & von der Lippe 2018), and North America (Schwartz et al. 2002; Lawson et al. 2008). Moreover, there is growing evidence that urban biodiversity is often (though not always) positively correlated with ecosystem services (Schwarz et al. 2017) and appreciated by urban residents (Fuller et al. 2007; Carrus et al. 2015; Fischer et al. 2018). It is expected that the experiences of city dwellers with urban nature will also support their commitment to nature conservation outside cities (Miller 2005; Zhang et al. 2014; Soga & Gaston 2016).

From these insights, the narrative of cities' contribution to biodiversity conservation has been developed, and thus a bundle of strategies to promote biodiversity in cities. Classical approaches such as the designation of protected areas (Rebelo et al. 2011; Kendal et al. 2017) are increasingly complemented by integrative approaches arguing for a biodiversity-friendly management of green spaces (Shwartz et al. 2013; Aronson et al. 2017; Chollet et al. 2018), a reconciliation of urban land use and biodiversity conservation (Francis & Lorimer 2011; Elmqvist et al. 2013), the integration of biodiversity in design (Müller et al. 2010; Garrard et al. 2018), or the promotion of urban wilderness (McKinney et al. 2018; Hwang et al. 2019).

Yet strategies to promote urban biodiversity in general do not necessarily support endangered plant species in particular. While target species conservation is a major motive for urban biodiversity conservation (Dearborn & Kark 2010), the review by Shwartz et al. (2014) revealed that the empirical basis for its success is still limited. There is therefore a risk that conservation policies and the allocation of scarce resources will be misled by insufficient evidence-based prospects for the urban contribution to biodiversity conservation (Shwartz et al. 2014). An assessment of the sustainable survival of endangered plant species in cities is currently hampered by knowledge deficits in three key areas.

(a) Population persistence. The occurrence of species of conservation concern in cities does not necessarily indicate their effective protection (Shwartz et al. 2014). Environmental constraints

Chapter 4

due to urban land use, fragmentation or novel environmental stressors not only filter species’ colonization of urban habitats, but represent significant recruitment limitations for plants at all life stages, from seed dispersal to the establishment of individuals and populations (Williams et al. 2009; McDonnell & Hahs 2015; Kowarik & von der Lippe 2018). For example, studies in Berlin have shown that 18-57% of plant species occurring across different ecosystem types have not been able to establish sustainable populations (Kowarik & von der Lippe 2018). Local extinction of plant species is common in cities, due to habitat loss or change (Knapp et al. 2010; Duncan et al. 2011). Remaining small populations of a species can thus mask future extinction debts (Hahs & McDonnell 2014). Indeed, there is evidence for decreasing population sizes in urban plant species over time (Chocholouskova & Pysek 2003; Knapp et al. 2010). Population persistence is thus a key issue for sustainable target species conservation in cities – but has rarely been studied (Shwartz et al. 2014). Most urban biodiversity studies are snapshot studies that report the occurrence of species at a certain point in time, but do not allow conclusions to be drawn about the persistence of populations in urban settings (but see Lawson et al. 2008 and Schwartz et al. 2013).

(b) Type of habitat. While the occurrence of target species of conservation concern in cities has often been described, the relative importance of different urban habitats for such species is critically understudied (Shwartz et al. 2014). Species of conservation concern have been reported from very different habitat types within cities, e.g. forests (Godefroid & Koedam 2003; Croci et al. 2008), brownfields (Bonthoux et al. 2014), parks (Cornelis & Hermy 2004; Kümmerling & Müller 2012) cemeteries (Löki et al. 2015; Kowarik et al. 2016). However, an important bias is that not all urban habitat types have been studied equally. Rather, most urban conservation studies address larger habitats (roughly >2 ha), with natural remnants and large green spaces being overrepresented (Shwartz et al. 2014). Moreover, species inventories of nature reserves within cities are likely better known than those of unprotected land use types. However, a Californian study in which about half of the rare plant species were found on privately owned land shows how important areas outside nature reserves can be (Schwartz et al. 2002).

(c) Environmental and socioeconomic predictors. Which species occur in urban habitats not only depends on local habitat features such as patch size but also on features of the surrounding urban matrix as indicated by urbanization parameters (e.g. proportion of imperious surface, population density) or the vicinity to (near-)natural ecosystems. Previous urban studies have revealed the relative importance of a large set of biodiversity predictors (e.g., Kinzig et al. 2005; Williams et al. 2006; Westermann et al. 2011; Beninde et al. 2015; Matthies et al. 2015; Anderson & Minor 2019). However, results based on total species richness cannot necessarily

Chapter 4

be generalized to endangered plant species as shown for urban grassland reserves in Prague (Jarošík et al. 2011). Specific analyses for endangered plant species are rare though.

Moreover, urbanization effects on endangered plant species are largely scale-dependent. Studies at a large scale, for example, revealed positive relationships between urbanization parameters and species of conservation concern (Kühn et al. 2004; Lenzen et al. 2009; Shwartz et al. 2013). At a more fined grained city scale though, the degree of urbanization can be negatively related to endangered plant species (Schmidt et al. 2014). Furthermore, the landscape history beyond current land uses may modulate biodiversity patterns (e.g. habitat continuity; Du Toit et al. 2016; Johnson et al. 2017) as well as the connectivity to near-natural ecosystems (Bierwagen 2007; LaPoint et al. 2015). Yet many studies often analyze the importance of biodiversity predictors for heterogeneous urban landscapes (e.g. grids) without specifying for the habitat scale, i.e. the specific locations of endangered plant populations.

(d) Species traits. Besides characteristics of a population’s environment, species identity or more generally, species traits are held responsible for the vulnerability to local extinction in urban landscapes (Duncan et al. 2011). Although several studies have indicated shifts in specific traits due to urbanization (Knapp et al. 2008a; Williams et al. 2015), there was only a marginal impact of species traits on plant extinctions in urban areas in North America (Duncan et al. 2011). However, population survival or higher frequency in urban habitats has successfully been linked to competitive traits (e.g. plant height or specific leaf area; Fischer et al. 2013a), to competitive strategies (Chocholouskova & Pysek 2003; Knapp et al. 2008c), and to preference for nutrient rich sites (Knapp et al. 2009). While lower frequencies in cities are reported for species with high water requirements (Knapp et al. 2010).

Here we tried to better understand the persistence of endangered plant species in cities – and underlying mechanisms – by evaluating a unique data set from Berlin. These data contained monitoring results of a large number of precisely mapped plant populations (n=858) belonging to 179 endangered plant species. These plant species were red listed in Berlin or at higher spatial scales and have been identified as priority species of conservation concern in Berlin (Seitz et al. 2018). For this reason, the populations of these target species have been subjected to monitoring by experts on behalf of the Stiftung Naturschutz Berlin, which began with a compilation of all known precise mappings of these species after 1990 and a subsequent remapping of these known locations between 2009 and 2014. Therefore, information was available on which populations have survived or have gone locally extinct at the respective sites since the first survey.

Chapter 4

To unravel mechanisms that might underpin population survival, we assigned each of the originally mapped populations (i) to a specific biotope type and (ii) characterized each population site by a range of urban landscape predictors whose fundamental relevance has been shown in previous urban biodiversity studies and that were related to habitat features, the urbanization of the surroundings and the vicinity of (near)-natural ecosystems. This allowed revealing which landscape parameters are correlated with the survival of populations of endangered species in different types of urban habitats in Berlin. Additionally, we linked population survival to plant traits that have been shown to be relevant for population survival in previous urban studies and that are related to the plant morphology, the CSR strategy (Grime 1977), and the realized niche of the species (i.e. Ellenberg indicators; Ellenberg et al. 1991).

Based on results of previous urban biodiversity studies, we expected that population survival rates differ significantly among biotope types and decrease with increasing urbanization of their surroundings, indicated by parameters related either to the urban form (e.g. proportion of impervious surface, or proximity to roads) or human population density, or dynamics. On the other hand, we assumed a positive correlation between population survival and an increasing patch size, and with increasing proximity to natural remnants or large forest and grassland areas. We also suspected that the continuity of habitats over a period of more than 100 years and the conservation status of an area are positively related to population survival in endangered plant species. Furthermore we hypothesized a positive relationship between plant traits related to competitive ability like plant height or specific leaf area and population survival.

In detail, we addressed the following research questions: How is the survival rate of populations of endangered plant species in Berlin related to (1) features of the respective habitats (i.e., type of biotope, patch size, habitat continuity, legal conservation status); (2) the degree of urbanization of the surrounding urban matrix (i.e., percentage of impervious surface, floor space index, human population density and human population constancy, distance to nearest street, street length); (3) to the vicinity of (near)natural ecosystems (i.e. distance to natural remnants, proportion of forests, grasslands in buffer around the sites); and (4) plant functional traits of the respective endangered plant species.

Chapter 4

Methods

Study area

This study was carried out in the city of Berlin, the capital and largest city of Germany, with 3.6 million inhabitants in 2017. Berlin is a typical example of a large European city, including a wide range of ecosystems from remnants of natural and agrarian landscapes, urban greenspaces with different land use histories and novel ecosystems as on vacant urban-industrial land (Sukopp 1990; Lachmund 2013). Green and blue spaces such as forests, lakes and rivers, parks, allotment gardens, fields, or grassland areas make up 41% of the total area of Berlin, which is 891 km2 ; the remaining 59% is covered by built up areas (SenStadtUm 2016). The flora of Berlin is well studied since the 18th century (Sukopp 1987). Species extinctions and declines were originally analyzed in the 1970s and further updated in red lists of endangered plant species in 1982, 2001 and 2018 (Seitz et al. 2018). Since the mid-19th century, 17% of Berlin’s flora has gone extinct and 29% is currently being endangered.

Data on endangered plant populations

We used an extensive monitoring dataset from Berlin’s Flora Protection Program (Berliner Florenschutzkonzept) on the precise geographical location of 858 populations of 179 endangered plant species in Berlin (a subset of the species used in Chapters 2 and 3). These species have the highest conservation priority because they are threatened locally and/or at regional, national or global scales, and their populations are being monitored by the Stiftung Naturschutz Berlin (2015). The first inventory of populations of these endangered species was compiled from nature conservation expert surveys after the year 1990 that contained precise information on their location. On the basis of the georeferenced locations of this initial dataset, a re-mapping project took place between 2009 and 2014, aiming to relocate all previously mapped populations. The average period of time between the first and second mapping of a population was 7.6 (±0.2 years) years. This allowed differentiating persistent from locally extinct populations out of the initial dataset. For our analyses we used only data on populations of terrestrial biotopes as aquatic populations are more difficult to relocate which could result in methodological bias to the determination of locally extinct populations. We also excluded populations from the initial compilation with imprecise information on the spatial location, i.e. when no precise point map, coordinates or exact description of the spatial location was provided.

Chapter 4

Habitat features of endangered plant populations

For all remaining populations, we determined the habitat features of the biotope patch where it was located. To disclose the role of different biotope types for population survival, we intersected each mapped location of a population of endangered plant species with the respective biotope class from the Berlin biotope mapping. Between 2001 and 2003, 64 different mapping projects led to a detailed biotope map of Berlin (SenStadtUm 2014). Terrestrial populations of endangered plant species were present in eight major biotope classes (Fig. 1): forests, grasslands, ruderal sites, built-up areas, bogs and marshes, groves and hedges, green spaces and agricultural fields. We also calculated the area of the respective biotope patches to elucidate the role of patch size for population survival.

Water bodies Forests Built up areas Other green spaces

Fig. 1 Survival of populations (n=858) of endangered plant species in Berlin across the biotope classes. Blue circles represent populations that survived during the monitoring period; red circles indicate populations that have gone locally extinct during the monitoring period. Please note that “other green spaces” corresponds to grasslands, ruderal sites, bogs and marshes, groves and hedges, green spaces, fields, heaths and “other types”.

81 Chapter 4

To test whether habitat continuity is related to population survival, we analysed the temporal continuity of three main land use types (forests, grasslands, and built up areas) throughout the area of today’s Berlin. We used information from digitized historical maps (von der Lippe et al., unpublished) covering three points of time: 1774-75 ("Schmettausches Kartenwerk"), 1831-71 ("Preußische Uraufnahme"), and 1927-40 ("Preußische Neuaufnahme").

As conservation measures may largely shape the management an endangered species is exposed to, we attributed each site of a population to one of the following categories for legal protection status: nature reserve (Naturschutzgebiet), Natura 2000 site, protected landscape (Landschaftsschutzgebiet), and unprotected sites. The first three categories represent areas protected by law, with decreasingly strict protection status and management measures (SenStadtUm 2003).

Urban landscape predictors

We determined several characteristics of the urban matrix surrounding the populations of endangered plant that were used in previous studies to indicate effects of urbanization on plant populations. For a 500 m radius buffer around each population of endangered plant species we calculated the proportion of impervious surface and the length of the road network, based on data from the Berlin Environmental Atlas (SenStadtWo 2018). Moreover, we calculated data on the mean floor space index, human population density, and mean human population constancy (i.e. percentage of residents living at least 5 years in the same address) from The Berlin Social Urban Development Monitoring (SenStadtWo 2017). The scale of the map is 1:5,000 and values were calculated for each of the 447 districts (Planungsräume) in Berlin (SenStadtWo 2017). To account for effects of spatial isolation from near natural ecosystems we calculated the proportion of forests and grassland in the same buffer around each population of endangered plant species, based on information from the biotope map of Berlin (SenStadtUm 2014).

To calculate the average values of each parameter, we spatially intersected the corresponding map with each individual buffer in QGIS (2019) and then used the plugin “Dissolve with Stats” to obtain the average values for each buffer. The distance of each population to the nearest road was calculated in QGIS (2019) with the “NN Join” plugin.

Moreover, we calculated the distances of all populations to the nearest natural remnant patch with the “NN Join” plugin in QGIS (2019). We addressed those biotopes as natural remnants if they corresponded to historical natural ecosystems, being only slightly affected by human impacts, following the classification by (Kowarik & von der Lippe 2018). Such remnants include

82 Chapter 4 mires, wetlands, near-natural forests and some near-natural grassland. Spatial information on natural remnants was derived from an unpublished map (Planchuelo et al., in revision).

An overview of the methodological steps in this study can be seen in figure 2.

Fig. 2 Visual summary of the methodological process of this study. Green-coloured boxes represent the data sources, blue coloured boxes represent the methodological procedures, and white boxes represent the data generated from each of the methodological procedures.

Plant traits and niche indicators

Information on the different plant morphological traits (specific leaf area, plant height) and the CSR strategy (Grime 1977) of the species was gathered from the BIOFLOR database (Klotz et al. 2002). Plant traits (Diaz et al. 2004; Knapp et al. 2008a; Knapp et al. 2008b; Knapp et al. 2009; Soudzilovskaia et al. 2013) and CSR strategies (Chocholouskova & Pysek 2003; Fischer et al. 2013a; Williams et al. 2015) have been extensively used to describe trait variations in urban floras.

For information on the realized niche of a species, we used Ellenberg indicator values for the soil moisture, soil nitrogen and soil acidity requirements, as well as the light and temperature requirements of each species (Ellenberg et al. 1991). Ellenberg indicators have been used successfully in urban habitats as indicators of the realized niches of ample sets of species (Pywell et al. 2003; Knapp et al. 2009; Williams et al. 2015).

83 Chapter 4

Statistical analyses

To relate population survival to habitat characteristics and urban landscape variables (Table 1) we used a generalized linear mixed model (GLMM; R function glmer (Bates 2010)). We coded population survival as a binomial response and corrected for phylogenetic and spatial dependence of the data points by including the plant genus and the spatial coordinates of their locations as random factors. Because of the varying period between the first and the second monitoring of the different populations, this time span was also included as a random factor. All habitat and urban landscape predictors were tested for intercorrelation and included as fixed effects in an initial model when │r│<0.7. A minimal adequate model was chosen by backward- stepwise selection of the predictors based on minimal AIC.

Urban landscape variables Units Scale Range Effect Habitat Features Biotope class / Categorical 8 classes Fixed Patch size sq. meters Continuous 97 - 689256m2 Fixed Habitat continuity years Categorical 4 time frames Fixed Conservation status / Categorical 4 categories Fixed Degree of urbanization Proportion of impervious percentage Continuous 0 - 84% Fixed surface (500 buffer) Floor space index (500 index Continuous 0 - 1.8 Fixed buffer) Human population density pop/ha Continuous 0 - 186 Fixed (500 buffer) Human population percentage Continuous 0 - 77% Fixed constancy (500 buffer ) Street length (500m buffer) meters Continuous 0 - 845m Fixed Distance to nearest street meters Continuous 11 - 2160m Fixed Proximity of near natural ecosystems Distance to nearest remnant meters Continuous 0 - 1866m Fixed Proportion of forests (500m percentage Continuous 0 - 100% Fixed buffer) Proportion of grasslands percentage Continuous 0 – 71% Fixed (500m buffer) Variables for correction Spatial location of coordinates Continuous / Random populations Genus / Categorical 129 genera Random Period between first and years Categorical 22 years Random second monitoring

Table 1 List of urban landscape predictor variables of population survival used in the GLMM model. Their units, scale, range and effect in the model are shown.

84 Chapter 4

We evaluated the prediction accuracy of the model by calculating the area under the receiver operated curve (AUC) with the package “pROC” in R (Robin et al. 2011).

To estimate the effect of plant traits, plant strategies and niche indicators on population survival and to account for possible non-linear effects, we complemented the binomial GLMM with a conditional inference tree analysis with the same response (population survival) and a set of predictors related to the morphological plant traits of the species, their CSR strategy, and their realized niche (Table 2). A conditional inference tree is a non-parametric type of decision tree where the dataset is recursively split into dichotomous subsets which are discriminated by the most significant predictor (Hothorn et al. 2006).

Plant traits Units Scale Range

Morphological traits Specific leaf area mm2.mg-1 Continuous 3.25 - 123.42 Plant height m Continuous 0.045 - 5.278 CSR Strategy / Categorical 7 categories Realized niche Soil moisture requirements / Categorical 12 categories Soil nitrogen requirements / Categorical 9 categories Soil acidity requirements / Categorical 9 categories Light requirements / Categorical 9 categories Temperature requirements / Categorical 9 categories

Table 2 List of species-related predictor variables of population survival used in the conditional inference tree. Their units, scale and range in the model are shown.

All analyses were performed with the statistical and programming software R version 3.5.2 (R Core Team 2018).

Results

Survival rate

Almost two thirds of the populations of endangered plant species (64%) mapped in the first monitoring were confirmed in the second monitoring (551 populations). Thus, more than one third of all populations became locally extinct during the period between the two monitoring

85 Chapter 4

dates, which averaged 7.6 years. During that time, from a total of 179 endangered species, 49 (27%) went extinct for the area of Berlin.

Landscape predictors of population survival

In the binomial GLMM, biotope class had a strong significant effect on population survival. Additionally, the minimal model with the best fit based on AIC selection also included distance to the nearest natural remnant as a negative effect and conservation status as predictors, though they were not significant (Table 3). While random effects of genus and monitoring duration were retained in the minimal model, neither patch size, habitat continuity nor any of the urbanization variables were. Degrees Variable Chi Sq. of P value Freedom Biotope class 14.39 7 0.004

Conservation status 3.23 3 0.071 Distance to natural 3.89 1 0.274 remnants

Table 3 Analysis of deviance table (Type II Wald Chisquare Test) of the effects of predictor variables on population survival of endangered plant species from the minimal GLMM model after backward selection (AUC: 0.871). Random effects of genus and monitoring duration of the endangered plant populations were retained in the minimal model.

Population survival rates were highest in green spaces (0.90) and built up areas (0.78) and lowest in forests (0.58) and ruderal sites (0.61). While the first two biotope types only housed a few populations, the last two biotope types had significantly more populations. Forests that hosted the majority of populations of endangered plants had the lowest survival rate (Fig. 3a). Grasslands, which hosted the second highest number of populations (261), showed an intermediate survival rate of 0.69.

The nature conservation status of the area was also included in the minimal GLMM model as a predictor of population survival (Fig. 3b). Legally protected areas in total harbored more populations of endangered plant species (484) than unprotected areas (374). Yet population survival was highest outside protected areas (0.72). Across protected areas survival rate decreased from 0.64 in nature reserves to 0.47 in protected landscapes.

Chapter 4

a b 0.95 n=21 0.95 0.90 0.90 0.85 0.85 0.80 n=27 0.80 0.75 0.75 n=374 n=10 n=261 0.70 n=37 0.70 n=31 n=114 0.65 n=67 0.65 0.60 n=404 0.60 n=306 0.55 0.55 Popula(on survival rate Popula(on surivival rate 0.50 0.50 n=64 0.45 0.45

Biotope class Conserva(on Status

c d 0.95 0.95 0.90 n=14 n=23 n=7 0.90 0.85 0.85 0.80 0.80 n=96 n=49 0.75 0.75 n=101 n=99 0.70 0.70 n=97 0.65 n=21 0.65 Species survival rate Species survival rate 0.60 0.60 n=22 n=28 n=34 0.55 0.55 0.50 0.50 0.45 0.45

Biotope class Conserva(on Status

Fig. 3 Survival rate of populations (n=858) of endangered plant species and species survival rate (n=179) across (a and c) different biotope classes and (b and d) the different status of conservation in Berlin. N above the bars indicates the total number of populations (a and b) or the total number of species (c and d) across biotope classes and areas with different conservation status.

Chapter 4

The survival rates of species among the different biotope classes and sites of different nature conservation status showed a slightly different picture than the survival rates of populations (Fig. 3 c and d). Here, green spaces, built-up areas and agricultural fields showed the highest survival rates at 0.86. Ruderal sites as well as bogs and marshes had the lowest survival rates and lost nearly 50% of endangered species during the monitoring period. The survival rates of species showed similar differences among nature conservation areas as for population survival. However, the survival of species outside of conservation areas was clearly lower here compared to nature reserves and comparable to Natura 2000 sites.

The distance to a natural remnant was also a variable selected by the GLMM model as predictor of population survival. Based on our data, population survival decreased moderately at increasing distances from a natural remnant. Survival rates slightly decreased on average from 0.63 at 0m of a remnant patch to 0.59 at 1500m.

Plant traits as predictors of population survival

From all plant traits and ecological niche indicators, the conditional inference tree only retained soil humidity requirements (p<0.001) and CSR strategy (p=0.029) as significant predictors of population survival of endangered plant species (Fig. 4). Soil humidity requirements caused the first split in the data, with an overall lower survival in populations with the highest soil humidity requirements. Those with low to medium soil humidity requirements had overall higher survival rates and were further split according to their CSR strategy, with populations of competitive species having higher survival rates than those with ruderal strategies.

Chapter 4

Soil humidity requirements p < 0.001

Low-medium soil High soil humidity requirements humidity requirements

CSR strategy p = 0.029

Competitive species Ruderal species

Node 3 (n=645) Node 4 (n=181) Node 5 (n=32) 1 1 1

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

Survival rate Survival 0.2 0.2 0.2

0 0 0

Fig. 4 Conditional inference tree showing partitioning effects of plant traits on population survival of endangered plant species in Berlin.

Discussion

While cities are known to harbour many endangered plant species, the importance of urban areas for biodiversity conservation has been challenged (Shwartz et al. 2014). This study addressed one critically understudied key question: the extent to which endangered species not only occur in cities, but can also survive under urban conditions. To our knowledge, this is the first long-term urban study to investigate population survival for a large set of endangered plant species (179 species) at their respective growth sites in a large city. Relating each growth site of 858 populations to a range of parameters on different spatial scales allowed us to shed light on mechanisms of population survival over an average period of time from 7.6 years.

Chapter 4

Results are relevant beyond Berlin because this city is a typical European metropolis, similar to other historical cities in terms of history of development and habitat transformations, current land use and proportion of native vegetation within the city borders (Hahs et al. 2009). The systematic long-term monitoring started in the 1990s and thus falls at a time when Berlin was growing and becoming denser - like many cities worldwide (Haaland & van den Bosch 2015). Our results thus indicate how populations of endangered plant species can survive under the challenges of growing historical cities.

A key finding is that populations of endangered plant species can be highly dynamic in urban settings, with a considerable risk of local extinction within a period of less than ten years. At the second monitoring date, only about two thirds of the populations were confirmed, whereas one third disappeared. Yet population survival as reported in this study cannot be equated to the sustainable establishment of populations over longer periods of times, which is generally challenging in urban environments as recently reported from Berlin (Kowarik & von der Lippe 2018). Therefore, our results on population survival may be burdened with extinction debts - which should be verified by continued monitoring. However, it should also be borne in mind that this study has essentially illuminated the fate of populations between first and second monitoring. It cannot be ruled out that some species colonized a new site for the first time during this period - for which, however, no information was available.

Beyond the large picture, our results illustrate considerable differences in population survival with regard to the different biotopes, the protection status of the sites and their proximity to semi-natural areas.

Survival across biotope types

As assumed, biotope classes had the most pronounced effect on the survival of endangered plant populations (Table 1). Of all biotope types, forests and grasslands hosted by far the largest populations of endangered species and the highest number of endangered species represented. This emphasizes the importance of these semi-natural biotopes for biodiversity conservation, which has already been identified in previous studies on forests (Godefroid & Koedam 2003; Diamond & Heinen 2016) or on grassland in cities (Fischer et al. 2013b; Rudolph et al. 2017).

Surprisingly, however, the survival rate was lowest in forests (58%) (Fig. 4). The large number of endangered species of forest biotopes includes apparently many species that respond sensitively to changes in their habitat. The fact that 42% of all populations in forests could no longer be confirmed with the second monitoring is alarming, as forests in Berlin are generally

Chapter 4

protected, i.e. there is no significant loss of forest area. While habitat continuity is generally important for dispersal limited forest specialists (Hermy et al. 1999; Dyderski et al. 2017), the variable was not retained in the models. Instead, the low survival rates in forests could relate to changes to the forest structure. Many of Berlin's forests had been grazed until the 19th century (Sukopp 1990) and slowly developed a denser tree canopy after the abandoning of this traditional land use. In addition, tree plantings had been established, including native and alien tree species, which often establish dense ground and shrub layers by natural regeneration (Acer species, Prunus serotina; Sachse et al. 1990; Kowarik et al. 2013). Another driver of change is eutrophication from urban sources that had been related to losses of plants adapted to nutrient- poor sites in another urban study (Knapp et al. 2010). In forests of Berlin, nutrient inputs from recreational activities, including dogs, may have enhanced denser vegetation structures on previously open sites, e.g. along paths (personal observation). These drivers of change could have impaired many populations of endangered plants before the first monitoring date and led to plenty of local extinction afterwards. Indeed, the species with the lowest survival rates in forests were Ranunculus lingua, Festuca psammophila and Botrychium lunaria, all of which are not typical forest species that occur in open, nutrient poor habitats and adjacent woodlands.

Conversely, the survival rate was highest in the most anthropogenic biotopes (i.e. green spaces: 0.90; built-up areas: 0.78). However, only few endangered species have benefited from this. This pattern suggests that few highly endangered species are pre-adapted (sensu McDonnell and Hahs 2015) to intensively designed and used urban biotopes and thus have a relatively good chance of surviving here. Examples involve Chenopodium murale, Filago minima and Sagina apetala, rare species of highly disturbed sites.

Among the strongly anthropogenic biotope classes, ruderal sites hosted the largest number of populations of endangered species. However, we found the second lowest survival rate of the populations here (0.61). While novel urban ecosystems of Berlin have been shown to harbor many endangered plant species (Kowarik & von der Lippe 2018), population survival seems to be challenged in ruderal sites that are most prominent on vacant land. This may be related to the rising construction activities in Berlin during the last decades, which affect vacant land directly and indirectly. Alternative, progressing succession towards forest stages may impair the habitats of rare ruderal species that often rely on open habitats (Kattwinkel et al. 2009).

Grasslands, which only cover 5% of the area of Berlin (SenStadtUm 2014), hosted the second highest number of total populations and showed an intermediate survival rate of 69%. These results confirm, on a broad basis at the population scale, previous studies on the importance of

Chapter 4

urban grassland for biodiversity conservation (Fischer et al. 2013a; Klaus 2013; Nielsen et al. 2014; Kendal et al. 2017).

Patch size and urbanization

Interestingly, neither patch size nor any urbanization variables were retained in the models. The lack of a significant relationship between urbanization variables (and patch size) and population survival was a surprising result as many urban studies had shown the importance of these variables for biodiversity patterns (e.g. Beninde et al. 2015; Kendal et al. 2017). This finding points to the particular importance of the biotope class for endangered plant species and, indirectly, to local parameters that might determine habitat quality (e.g. management effects on vegetation structure; Shwartz et al. 2013; Chollet et al. 2018) but need to be untangled in further studies.

Survival within and outside conservation areas

Unprotected areas in Berlin harboured a considerable species richness and amount of populations of endangered plants (Fig. 3). This clearly illustrates opportunities for biodiversity conservation in urban settings outside conservation areas as shown in a Californian study (Schwartz et al. 2002). However, conservation areas hosted more populations, and also more endangered plant species. This was as an expected result as areas with abundant occurrences of endangered species usually have a higher probability of being protected than other sites.

Conservation status was also related to population survival in the binomial GLMM (albeit not significant; Table 2). Yet surprisingly, populations outside conservation areas had a higher survival rate than within conservations areas. In the same vein, survival rates decreased with the strictness of protection, i.e. from nature conservation areas to Natura 2000 areas to landscape conservation areas (Fig. 3). To our knowledge, comparative studies on the effectiveness of protected vs. unprotected areas are missing for cities. Studies on forests show that a protection status is not a mandatory prerequisite for higher protection effectiveness (e.g. Hayes 2006). While a protected status of an area in growing cities is likely to reduce the risk of its conversion to building land, our research also shows that a conservation status does not necessarily guarantee a high survival rate of endangered plant populations. This is a strong argument for allocating resources for target species-oriented management in protected areas.

Chapter 4

Proximity to (near-) natural ecosystems

The importance of remnant biotopes for urban biodiversity conservation is well established (e.g. Godefroid and Koedam 2003). In urban conservation areas in Prague, neighbourhood effects with other conservation areas had a positive effect on the occurrence of endangered plant species (Jarošík et al. 2011). Our study adds the insights that the distance of populations of endangered plant species to the nearest natural remnant also matters for population survival, as indicated by both models (Table 1, Fig. 4). Our results revealed an increased survival rate of populations the closer these were to remnant ecosystems. Some mutually not exclusive factors can contribute to this finding. Assuming that remnant sites within protected areas harbour many endangered species, the probability of successful dispersal increases with increasing proximity. This could support populations outside of remnants and reduce local extinctions. On the other hand, with increasing proximity to remnants, the probability of similar site factors (e.g. soils, hydrology), which can promote the survival of endangered populations, increases. Moreover, it can be assumed that populations within protected areas benefit significantly more from target species-related management measures than those outside protected areas. Our findings thus support conservation strategies to include larger areas around natural remnants in formally protected areas - not only as buffer areas but also as potential habitats for endangered species.

Species functional traits and ecological niche

While none of the metric functional traits was retained in the tree model as predictor of population survival, indicator values for soil moisture requirement and the strategy types of the CSR scheme differentiated three groups of species with distinct population survival. The first split of the model that delimits a group of species with the highest requirements on soil moisture with a rather low population survival is in accordance with previous findings on low performance of wetland species in urban areas (e.g. Knapp et al. 2010). In Berlin, as in other urban regions of the northern hemisphere, wetland species are particularly prone to extinction due to large-scale groundwater subsidence caused by extensive dealing and groundwater withdrawal.

The second predictor in the model, CSR strategy, split the remaining species into one group with medium survival rates that are characterized by ruderal strategies, and a second group with the highest survival rates and competitive strategies. Although this outcome could be expected as competitive species usually have higher prospects of long-term survival (Chocholouskova &

Chapter 4

Pysek 2003; Knapp et al. 2008c), it is interesting that ruderal strategies are not promoting survival in an urban – and largely ruderal – environment. This is probably linked to the elevated risk of local extinction in ruderal habitats due to high construction pressure on these sites as described above.

Conclusions

The survival of species of conservation concern in cities varies amongst biotope classes and the varying ecological strategies and niche requirements of the different species. Populations in habitats of continuous human impact like built up areas and green spaces perform rather well in the long run - most likely because the endangered species in these sites are well adapted to their conditions. However, survival rates are more alarming in those near-natural habitats like forests and grasslands that host the majority of populations of endangered plant species. Considering that population survival decreases in protected areas with the lowest degrees of maintenance, our study reveals the great importance of proper management in natural and semi-natural habitats of urban regions. Appropriate management can potentially counteract some urban impacts like nitrification or spread of invasive species. However, this strategy of preserving nature conservation reaches its limits when it comes to far reaching urban impacts on essential habitat requirements. The case of bog- and wetland-species, which are mainly threatened by urban groundwater subsidence, illustrates the dilemma in conserving relict species in urban natural remnants.

Literature

Anderson EC, Minor ES. 2019. Assessing social and biophysical drivers of spontaneous plant diversity and structure in urban vacant lots. Sci Total Environ 653:1272-1281. Aronson MF, La Sorte FA, Nilon CH, Katti M, Goddard MA, Lepczyk CA, Warren PS, Williams NS, Cilliers S, Clarkson B. 2014. A global analysis of the impacts of urbanization on bird and plant diversity reveals key anthropogenic drivers. Page 20133330. Proc. R. Soc. B. The Royal Society. Aronson MFJ, Lepczyk CA, Evans KL, Goddard MA, Lerman SB, MacIvor JS, Nilon CH, Vargo T. 2017. Biodiversity in the city: key challenges for urban green space management. Frontiers in Ecology and the Environment 15:189-196. Bates DM. 2010. lme4: Mixed-effects modeling with R. Berlin.

Chapter 4

Beninde J, Veith M, Hochkirch A. 2015. Biodiversity in cities needs space: a meta-analysis of factors determining intra-urban biodiversity variation. Ecol Lett 18:581-592. Bierwagen BG. 2007. Connectivity in urbanizing landscapes: The importance of habitat configuration, urban area size, and dispersal. Urban Ecosystems 10:29-42. Bonthoux S, Brun M, Di Pietro F, Greulich S, Bouché-Pillon S. 2014. How can wastelands promote biodiversity in cities? A review. Landscape and urban planning 132:79-88. Carrus G, Scopelliti M, Lafortezza R, Colangelo G, Ferrini F, Salbitano F, Agrimi M, Portoghesi L, Semenzato P, Sanesi G. 2015. Go greener, feel better? The positive effects of biodiversity on the well-being of individuals visiting urban and peri-urban green areas. Landscape and Urban Planning 134:221-228. Chocholouskova Z, Pysek P. 2003. Changes in composition and structure of urban flora over 120 years: a case study of the city of Plzen. Flora 198:366-376. Chollet S, Brabant C, Tessier S, Jung V. 2018. From urban lawns to urban meadows: Reduction of mowing frequency increases plant taxonomic, functional and phylogenetic diversity. Landscape and Urban Planning 180:121-124. Cornelis J, Hermy M. 2004. Biodiversity relationships in urban and suburban parks in Flanders. Landscape and Urban Planning 69:385-401. Croci S, Butet A, Georges A, Aguejdad R, Clergeau P. 2008. Small urban woodlands as biodiversity conservation hot-spot: a multi-taxon approach. Landscape Ecology 23:1171- 1186. Dearborn DC, Kark S. 2010. Motivations for conserving urban biodiversity. Conserv Biol 24:432- 440. Diamond JM, Heinen JT. 2016. Conserving rare plants in locally-protected urban forest fragments: A case study from Miami-Dade County, Florida. Urban Forestry & Urban Greening 20:1-11. Diaz S, et al. 2004. The plant traits that drive ecosystems: Evidence from three continents. Journal of Vegetation Science 15:295-304. Du Toit MJ, Kotze DJ, Cilliers SS. 2016. Landscape history, time lags and drivers of change: urban natural grassland remnants in Potchefstroom, South Africa. Landscape ecology 31:2133- 2150. Duncan RP, Clemants SE, Corlett RT, Hahs AK, McCarthy MA, McDonnell MJ, Schwartz MW, Thompson K, Vesk PA, Williams NS. 2011. Plant traits and extinction in urban areas: a meta- analysis of 11 cities. Global Ecology and Biogeography 20:509-519. Dyderski MK, Tyborski J, Jagodzinski AM. 2017. The utility of ancient forest indicator species in urban environments: A case study from Poznan, Poland. Urban Forestry & Urban Greening 27:76-83.

Chapter 4

Ellenberg H, Weber H, Dull R, Wirth V, Werner W, Paulissen D. 1991. Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobotanica 18:248. Elmqvist T, Fragkias M, Goodness J, Güneralp B, Marcotullio PJ, McDonald RI, Parnell S, Schewenius M, Sendstad M, Seto KC 2013. Urbanization, biodiversity and ecosystem services: challenges and opportunities: a global assessment. Springer. Fischer LK, et al. 2018. Beyond green: Broad support for biodiversity in multicultural European cities. Global Environmental Change-Human and Policy Dimensions 49:35-45. Fischer LK, von der Lippe M, Kowarik I. 2013a. Urban grassland restoration: which plant traits make desired species successful colonizers? Applied Vegetation Science 16:272-285. Fischer LK, von der Lippe M, Kowarik I. 2013b. Urban land use types contribute to grassland conservation: The example of Berlin. Urban Forestry & Urban Greening 12:263-272. Francis RA, Lorimer J. 2011. Urban reconciliation ecology: the potential of living roofs and walls. J Environ Manage 92:1429-1437. Fuller RA, Irvine KN, Devine-Wright P, Warren PH, Gaston KJ. 2007. Psychological benefits of greenspace increase with biodiversity. Biol Lett 3:390-394. Garrard GE, Williams NSG, Mata L, Thomas J, Bekessy SA. 2018. Biodiversity Sensitive Urban Design. Conservation Letters 11:e12411. Godefroid S, Koedam N. 2003. How important are large vs. small forest remnants for the conservation of the woodland flora in an urban context? Global Ecology and Biogeography 12:287-298. Grime JP. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. The American Naturalist 111:1169-1194. Haaland C, van den Bosch CK. 2015. Challenges and strategies for urban green-space planning in cities undergoing densification: A review. Urban Forestry & Urban Greening 14:760-771. Hahs AK, McDonnell MJ. 2014. Extinction debt of cities and ways to minimise their realisation: a focus on Melbourne. Ecological Management & Restoration 15:102-110. Hahs AK, et al. 2009. A global synthesis of plant extinction rates in urban areas. Ecol Lett 12:1165-1173. Hermy M, Honnay O, Firbank L, Grashof-Bokdam C, Lawesson JE. 1999. An ecological comparison between ancient and other forest plant species of Europe, and the implications for forest conservation. Biological Conservation 91:9-22. Hothorn T, Hornik K, Zeileis A. 2006. Unbiased recursive partitioning: A conditional inference framework. Journal of Computational and Graphical statistics 15:651-674. Hwang YH, Yue ZEJ, Ling SK, Tan HHV. 2019. It’s ok to be wilder: Preference for natural growth in urban green spaces in a tropical city. Urban Forestry & Urban Greening 38:165-176.

Chapter 4

Ives CD, et al. 2016. Cities are hotspots for threatened species. Global Ecology and Biogeography 25:117-126. Jarošík V, Konvička M, Pyšek P, Kadlec T, Beneš J. 2011. Conservation in a city: Do the same principles apply to different taxa? Biological Conservation 144:490-499. Johnson AL, Borowy D, Swan CM. 2017. Land use history and seed dispersal drive divergent plant community assembly patterns in urban vacant lots. Journal of Applied Ecology. Kattwinkel M, Strauss B, Biedermann R, Kleyer M. 2009. Modelling multi-species response to landscape dynamics: mosaic cycles support urban biodiversity. Landscape Ecology 24:929- 941. Kendal D, Zeeman BJ, Ikin K, Lunt ID, McDonnell MJ, Farrar A, Pearce LM, Morgan JW. 2017. The importance of small urban reserves for plant conservation. Biological Conservation 213:146-153. Kinzig AP, Warren P, Martin C, Hope D, Katti M. 2005. The effects of human socioeconomic status and cultural characteristics on urban patterns of biodiversity. Ecology and Society 10. Klaus VH. 2013. Urban Grassland Restoration: A Neglected Opportunity for Biodiversity Conservation. Restoration Ecology 21:665-669. Klotz S, Kühn I, Durka W, Briemle G. 2002. BIOLFLOR: Eine Datenbank mit biologisch- ökologischen Merkmalen zur Flora von Deutschland. Knapp S, Kuehn I, Wittig R, Ozinga WA, Poschlod P, Klotz S. 2008a. Urbanization causes shifts of species' trait state frequencies. Knapp S, Kühn I, Bakker JP, Kleyer M, Klotz S, Ozinga WA, Poschlod P, Thompson K, Thuiller W, Römermann C. 2009. How species traits and affinity to urban land use control large-scale species frequency. Diversity and Distributions 15:533-546. Knapp S, Kühn I, Mosbrugger V, Klotz S. 2008b. Do protected areas in urban and rural landscapes differ in species diversity? Biodiversity and Conservation 17:1595-1612. Knapp S, Kuhn I, Schweiger O, Klotz S. 2008c. Challenging urban species diversity: contrasting phylogenetic patterns across plant functional groups in Germany. Ecol Lett 11:1054-1064. Knapp S, Kühn I, Stolle J, Klotz S. 2010. Changes in the functional composition of a Central European urban flora over three centuries. Perspectives in Plant Ecology, Evolution and Systematics 12:235-244. Kowarik I, Buchholz S, von der Lippe M, Seitz B. 2016. Biodiversity functions of urban cemeteries: Evidence from one of the largest Jewish cemeteries in Europe. Urban Forestry & Urban Greening 19:68-78. Kowarik I, Lippe M, Cierjacks A. 2013. Prevalence of alien versus native species of woody plants in Berlin differs between habitats and at different scales. Preslia 85:113-132.

Chapter 4

Kowarik I, von der Lippe M. 2018. Plant population success across urban ecosystems: A framework to inform biodiversity conservation in cities. Journal of Applied Ecology 55:2354-2361. Kühn I, Brandl R, Klotz S. 2004. The flora of German cities is naturally species rich. Evolutionary Ecology Research 6:749-764. Kümmerling M, Müller N. 2012. The relationship between landscape design style and the conservation value of parks: A case study of a historical park in Weimar, Germany. Landscape and Urban Planning 107:111-117. Lachmund J 2013. Greening Berlin: The Co-production of Science, Politics, and Urban Nature. MIT Press. LaPoint S, Balkenhol N, Hale J, Sadler J, Ree R. 2015. Ecological connectivity research in urban areas. Functional Ecology 29:868-878. Lawson DM, Lamar CK, Schwartz MW. 2008. Quantifying plant population persistence in human-dominated landscapes. Conserv Biol 22:922-928. Lenzen M, Lane A, Widmer-Cooper A, Williams M. 2009. Effects of land use on threatened species. Conserv Biol 23:294-306. Löki V, Tökölyi J, Süveges K, Lovas-Kiss Á, Hürkan K, Sramkó G, Molnár A. 2015. The orchid flora of Turkish graveyards: a comprehensive field survey. Willdenowia 45:231-244. Matthies SA, Rüter S, Prasse R, Schaarschmidt F. 2015. Factors driving the vascular plant species richness in urban green spaces: using a multivariable approach. Landscape and Urban Planning 134:177-187. McDonnell MJ, Hahs AK. 2015. Adaptation and Adaptedness of Organisms to Urban Environments. Annual Review of Ecology, Evolution, and Systematics, Vol 46 46:261-280. McKinney ML. 2002. Urbanization, Biodiversity, and ConservationThe impacts of urbanization on native species are poorly studied, but educating a highly urbanized human population about these impacts can greatly improve species conservation in all ecosystems. Bioscience 52:883-890. McKinney ML, Kowarik I, Kendal D. 2018. The contribution of wild urban ecosystems to liveable cities. Urban Forestry & Urban Greening 29:334-335. Miller JR. 2005. Biodiversity conservation and the extinction of experience. Trends Ecol Evol 20:430-434. Müller N, Werner P, Kelcey JG 2010. Urban biodiversity and design. John Wiley & Sons. Nielsen AB, van den Bosch M, Maruthaveeran S, van den Bosch CK. 2014. Species richness in urban parks and its drivers: A review of empirical evidence. Urban Ecosystems 17:305-327. Nilon CH, et al. 2017. Planning for the Future of Urban Biodiversity: A Global Review of City- Scale Initiatives. Bioscience 67:331-341.

Chapter 4

Parris KM, Amati M, Bekessy SA, Dagenais D, Fryd O, Hahs AK, Hes D, Imberger SJ, Livesley SJ, Marshall AJ. 2018. The seven lamps of planning for biodiversity in the city. Cities 83:44-53. Pywell RF, Bullock JM, Roy DB, Warman LIZ, Walker KJ, Rothery P. 2003. Plant traits as predictors of performance in ecological restoration. Journal of Applied Ecology 40:65-77. QGIS. 2019. QGIS Geographic Information System. Open Source Geospatial Foundation Project. R Core Team. 2018. R: A Language and Environment for Statistical Computing. Rebelo AG, Holmes PM, Dorse C, Wood J. 2011. Impacts of urbanization in a biodiversity hotspot: Conservation challenges in Metropolitan Cape Town. South African Journal of Botany 77:20- 35. Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, Muller M. 2011. pROC: an open- source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 12:77. Rudolph M, Velbert F, Schwenzfeier S, Kleinebecker T, Klaus VH. 2017. Patterns and potentials of plant species richness in high-and low-maintenance urban grasslands. Applied Vegetation Science 20:18-27. Sachse U, Starfinger U, Kowarik I. 1990. Synanthropic woody species in the urban area of Berlin (West). Urban ecology. SPB Academic Publishing, The Hague:233-243. Schmidt KJ, Poppendieck HH, Jensen K. 2014. Effects of urban structure on plant species richness in a large European city. Urban Ecosystems 17:427-444. Schwartz MW, Jurjavcic NL, O'Brien JM. 2002. Conservation's disenfranchised urban poor. Bioscience 52:601-606. Schwartz MW, Smith LM, Steel ZL. 2013. Conservation investment for rare plants in urban environments. PLoS One 8:e83809. Schwarz N, Moretti M, Bugalho MN, Davies ZG, Haase D, Hack J, Hof A, Melero Y, Pett TJ, Knapp S. 2017. Understanding biodiversity-ecosystem service relationships in urban areas: A comprehensive literature review. Ecosystem Services 27:161-171. Seitz B, Ristow M, Meißner J, Machatzi B, Sukopp H 2018. Rote Liste und Gesamtartenliste der etablierten Farn-und Blütenpflanzen von Berlin. Universitätsverlag der TU Berlin. SenStadtUm. 2003. Protected Areas by Nature Conservation Legislation (incl. Natura 2000). SenStadtUm. 2014. Berlin Environmental Atlas, Biotope Types, www.stadtentwicklung.berlin.de/umwelt/umweltatlas/ek508.htm SenStadtUm. 2016. Berlin Environmental Atlas, Map Actual Use and Vegetation. SenStadtWo. 2017. Monitoring Soziale Stadtentwicklung Berlin. SenStadtWo. 2018. Berlin Environmental Atlas. Shwartz A, Muratet A, Simon L, Julliard R. 2013. Local and management variables outweigh landscape effects in enhancing the diversity of different taxa in a big metropolis. Biological Conservation 157:285-292.

Chapter 4

Shwartz A, Turbe A, Julliard R, Simon L, Prevot AC. 2014. Outstanding challenges for urban conservation research and action. Global Environmental Change-Human and Policy Dimensions 28:39-49. Soga M, Gaston KJ. 2016. Extinction of experience: the loss of human–nature interactions. Frontiers in Ecology and the Environment 14:94-101. Soudzilovskaia NA, Elumeeva TG, Onipchenko VG, Shidakov II, Salpagarova FS, Khubiev AB, Tekeev DK, Cornelissen JHC. 2013. Functional traits predict relationship between plant abundance dynamic and long-term climate warming. Proceedings of the National Academy of Sciences of the United States of America 110:18180-18184. Stiftung Naturschutz Berlin. 2015. Koordinierungsstelle Florenschutz – ein Projekt zur Umsetzung des Florenschutzkonzeptes Berlin. Sukopp H. 1987. On the history of plant geography and plant ecology in Berlin. Englera 7:85- 103. Sukopp H. 1990. Stadtökologie. Das Beispiel Berlin. Reimer Verlag.–1990.–455 s. Wang G, Jiang G, Zhou Y, Liu Q, Ji Y, Wang S, Chen S, Liu H. 2007. Biodiversity conservation in a fast-growing metropolitan area in China: a case study of plant diversity in Beijing. Biodiversity and conservation 16:4025-4038. Westermann JR, von der Lippe M, Kowarik I. 2011. Seed traits, landscape and environmental parameters as predictors of species occurrence in fragmented urban railway habitats. Basic and Applied Ecology 12:29-37. Williams NS, Morgan JW, McCarthy MA, McDonnell MJ. 2006. Local extinction of grassland plants: the landscape matrix is more important than patch attributes. Ecology 87:3000- 3006. Williams NSG, Hahs AK, Vesk PA. 2015. Urbanisation, plant traits and the composition of urban floras. Perspectives in Plant Ecology Evolution and Systematics 17:78-86. Williams NSG, et al. 2009. A conceptual framework for predicting the effects of urban environments on floras. Journal of Ecology 97:4-9. Zhang W, Goodale E, Chen J. 2014. How contact with nature affects children’s biophilia, biophobia and conservation attitude in China. Biological Conservation 177:109-116.

100

Chapter 5

Chapter 5: Synthesis

101

Chapter 5

“The tendency to overlook the conservation value of urban environments stems from misconceptions about the ability of native species to persist within cities and towns, and that this, in turn, hinders effective conservation action.”

Soanes et al. 2018

The importance of cities for the conservation of plant and animal species is not yet soundly demonstrated (Shwartz et al. 2014). Uncertain trends have been reported for species of conservation concern that range from population decline and extinctions (Hahs et al. 2009; Knapp et al. 2017) and biotic homogenization (Zeeman et al. 2017) to population survival and expansion into new habitats (Lawson et al. 2008; Lundholm & Richardson 2010; Kowarik & von der Lippe 2018). For this reason, the key question is not whether cities support biodiversity conservation, but where in cities and under which environmental conditions species of conservation interest can be supported (Lepczyk et al. 2017).

This study is likely the first to use such detailed datasets at the biotope and population level, shedding light for the first time on how populations of endangered plant species occur across different habitats and ecological novelty types within cities. Assessing population survival has allowed us to unravel endangered plant species’ extinction risks and the drivers and future perspectives of species survival in cities (Shwartz et al. 2014).

The first research paper (Chapter 2) titled “Untangling the role of urban ecosystems as habitats for endangered plant species”, tackles the first research question: What is the relative contribution of different biotope classes and of natural remnant vs hybrid vs novel ecosystems within the city of Berlin in harbouring populations of endangered plant species? In this study, I used an extensive dataset with the precise spatial location of 1742 populations of 213 endangered plant species in Berlin (Appendix A) and asses the relative importance of 9 biotope classes and 3 types of ecological novelty in hosting them (Fig. 2 Chapter 2).

A very high density of populations of endangered plant species was found in grassland and ruderal biotope classes, which together hosted 36% of populations in only 7% of the area of Berlin, demonstrating the great importance of anthropogenically-influenced biotopes for hosting endangered plant species. Unsurprisingly, forest biotopes provided habitats for the highest number of populations of endangered species, while built up areas, covering the majority of Berlin (54%), only hosted 92 populations.

102

Chapter 5

The importance of novel ecosystems in holding the largest number of populations of endangered plant species in Berlin is a major and unexpected result of this study - more so if we consider that novel ecosystems also harboured the highest number of species. However, remnant ecosystems also hosted a considerable amount of populations in a fraction of that size, conferring them with 12 times higher density of populations of endangered plant species than novel ecosystems. This disparity in density of populations points to an unresolved key question regarding whether or not populations of endangered plant species can successfully colonize into the urban matrix independently from source populations in remnant ecosystems. Indeed, varying human impacts and a variety of environmental barriers filter urban species compositions (Williams et al. 2009; Aronson et al. 2016; Knapp et al. 2017; Kowarik & von der Lippe 2018), so whether endangered plant species can colonize into different urban ecosystems depends on how they negotiate with a range of dispersal barriers.

This has been tackled in the second research paper (Chapter 3), which addresses the following research question: To which extent endangered plant species in an urban landscape may occur in anthropogenically shaped hybrid or novel ecosystems spatially independently from natural remnants? For this study I use the same dataset of endangered plant species as in the previous chapter and I divide them in two groups – those located inside hybrid ecosystems, and those located in novel ecosystems. I then assess the spatial relationship of these two groups of populations towards natural remnant ecosystems and evaluate if it is related to the dispersal traits of the different species. Findings show that populations located in hybrid ecosystems are in fact still dependent to source populations in natural remnants, while those located in novel ecosystems are not - indicating that they have overcome the dispersal barriers and colonised into the urban matrix. These species are predominantly ruderal plants with anemochory and stomatochory (e.g. in the beaks of birds) as dispersal syndromes (Fig. 8 Chapter 3). However, considering that many of these novel sites often have high levels of disturbances that affect population reproduction and long-term establishment (Lundholm et al. 2010; Kowarik & von der Lippe 2018), the mere occurrence of these populations is not indicative of their survival because time lags in species responses (Du Toit et al. 2016) could be masking future extinction debts (Hahs & McDonnell 2014). Thus, to better understand the role of cities to nature conservation it is essential to additionally consider the long-term establishment of these populations in order to unravel some of the drivers that might be involved in the survival of endangered species in cities.

The third research paper (Chapter 4) deals with this question about population survival: What are the most important urban drivers in determining population survival of endangered plant species within the metropolitan region of Berlin? This research uses a subset of 858 populations

103

Chapter 5

with information on population survival in order to find urban drivers of population persistence. Findings show that population survival is highly dependent on the biotope class where the populations are located at, with anthropogenically influenced biotopes such as green spaces or built up areas having highest survival rates, populations in forests having the lowest survival rates, and populations in grasslands having intermediate survival rates (Fig. 3a Chapter 4). Species numbers followed a similar survival trend, with the highest survival in built-up areas and green spaces (Fig. 3c Chapter 4). The conservation status and the distance to natural remnants also play a major role in the survival of endangered plant species in cities (Fig. 1). Surprisingly, there was an overall lower survival in populations inside conservation areas, and these performed worse if they were at a distance of more than 106 metres from a natural remnant patch. Finally, those populations with higher success rates were competitive species with medium to low soil humidity requirements (Fig. 4 Chapter 4).

Conservation status p < 0.001

Inside protected areas Outside protected areas

Distance to natural remnants p = 0.004

> 106 < 106

Node 3 (n=178) Node 4 (n=302) Node 5 (n=378) 1 1 1

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

Survival rate Survival 0.2 0.2 0.2

0 0 0 Fig. 1 Conditional inference tree showing partitioning effects of conservation status and distance to nearest remnant on the survival rates of the populations of endangered plant species in Berlin (n=858). Please note that the conservation status was only partitioned according to whether populations were or not inside a protected area. Biotope class was also included in the model, but not retained.

104

Chapter 5

Implications for conservation in cities

Biotopes

Anthropogenically influenced urban biotope types such as urban grasslands or green spaces have great potential for conservation in cities. Populations in grasslands have a considerable survival ratio of 69% (Fig. 3a Chapter 4), while they are the biotope class that holds the second highest total number of populations (Fig. 2 Chapter 2) and total species (Fig. 3c Chapter 4) of endangered plants in only 5% of the area of Berlin (SenStadtUm 2014). Previous studies have already highlighted the conservation potential of urban grasslands (Cilliers et al. 2004; Fischer et al. 2013b; Klaus 2013; Nielsen et al. 2014; Zeeman et al. 2017), and as native vegetation is almost not present in Berlin´s grassland communities (Sukopp 1990), they are again a great example of the great value that anthropogenic ecosystems can have in urban biodiversity conservation.

Interestingly, urban green spaces have highest population survival and the second species survival rate of all biotopes types (90% population survival, 86% species survival, Fig. 3a and 3c, Chapter 4). These spaces not only have high conservation value (Goddard et al. 2010), but promoting them could improve the well-being in citizens (Lim et al. 2018), improve cognitive development (Dadvand et al. 2015), mental health (Engemann et al. 2019), and allow the appreciation of biodiverse urban vegetation (Fischer et al. 2018). However, the low number of total species in this biotope (14, Fig. 3c Chapter 4) implies that only a reduced number of species can benefit from it. A similar phenomenon is observed with the other two biotope classes with higher survival rates, built up areas and fields, which hold respectively 23 and 7 species each (Fig. 3c Chapter 4). This pattern suggests than only a limited number of endangered plant species are pre-adapted (McDonnell and Hahs 2015) to succeed in highly anthropogenic biotopes.

Within the highly anthropogenic biotope classes, ruderal sites hosted the highest number of populations of endangered plant species (Fig. 2 Chapter 2), but population survival was the lowest (Fig. 3a Chapter 4). This could be due to the progressing levels of succession towards forest stages, which might detriment endangered species that rely on open habitats. Additionally, the low survival levels in ruderal sites could be related to the high levels of construction that ruderal vacant land is suffering in Berlin during these last decades.

Previous studies show the significance of forest patches for urban biodiversity conservation (e.g. Godefroid and Koedam 2003; Diamond and Heinen 2016), and while in this thesis forest biotopes provided habitats for the highest number of population of endangered plant species

105

Chapter 5

(Fig. 2 Chapter 2) and for the highest number of species (Fig. 3c Chapter 4), they also hold the lowest values of population survival (Fig. 3a Chapter 4). As forests often have a long habitat continuity, species with strong dispersal limitations such as ancient forest species might be constrained in them (Hermy et al. 1999; Dyderski et al. 2017) - unable to disperse elsewhere, they might be more vulnerable to urban changes. Additionally, the decline of traditional forest uses is affecting forest species that rely on light woodlands as vegetation structure is getting denser due in part to urban eutrophication.

Ecological novelty

Former studies have shown the potential of novel urban ecosystems for native and some endangered species (e.g., Lundholm & Richardson 2010; Bonthoux et al. 2014), and this thesis further supports the importance of novel ecosystems for endangered plant species, as they harboured the largest number of their populations in Berlin. Very importantly, these populations are occurring independently from remnant ecosystems, indicating that they have overcome the dispersal barriers and colonised into the urban matrix.

In terms of cost efficiency of practical conservation measures, it may still be more efficient to preserve the same amount of habitat in natural remnants than in novel ecosystems due to the higher density of populations in them. Considering that populations in hybrid ecosystems are clustered around remnant ecosystems and are still dependant on them through a source-sink dynamic, conservation efforts in remnants might benefit a wide range of populations located outside of them. More so, there is an argument to extend the management practices beyond the borders of remnant ecosystems as to englobe those populations clustered around them (i.e. those in hybrid ecosystems), especially considering that remnant ecosystems patches cannot increase their area and are particularly vulnerable to nearby urban pressures and isolation (Ramalho et al. 2014).

Conservation status

Surprisingly, populations inside nature conservation areas had as a whole lower survival rates than populations outside of them (Fig. 3b Chapter 4), illustrating the big opportunities for biodiversity conservation in urban settings outside protected areas. Previous research has already shown no significant differences in the conservation outcomes between protected and unprotected forests, with the latter having even higher vegetation density (Hayes 2006). The results in this thesis could be explained by the fact that protected areas are often located in

106

Chapter 5

already highly challenging spots for species survival (Riecken 2006). Additionally, the varying levels of management of the different conservation statuses might explain the different levels of survival within conservation areas. In this study, there were big decreases in population survival among the decreasing levels of management of the different statuses of conservation (Fig. 3b Chapter 4): “Nature reserves” are legally established areas with a strict protection status and the highest levels of management, while the approach of “Natura 2000” is less strict and keeps a lower management, and finally “Protected landscapes” have the lowest protection and management level of the three (SenStadtUm 2003).

Despite populations inside conservation areas as a whole had lower survival rates, these seemed to improve inside or on the proximity of remnant ecosystem patches (Fig. 1), suggesting that conservation investment in protected areas might be more effective when they are directed towards remnant ecosystem patches and their surroundings. This supports the idea of including management practices that go beyond the physical boundaries of remnant patches, as this thesis has shown that a big number of populations (i.e. those in hybrid ecosystems) are clustered around them and depend on their populations through a source-sink dynamic.

Plant traits

This thesis shows that the plant traits related to population colonisation into the urban matrix (Fig 8 Chapter 3) and population success within the urban matrix (Figure 4 Chapter 4) are not the same for endangered plant species. Plant traits related to species colonisation are connected to ruderal strategies and dispersal syndromes that can increase seed transport in cities (i.e. anemochory; von der Lippe et al. 2013, and stomatochory; Luna et al. 2018), while plant traits related to population survival are linked to competitive strategies and moderate to low humidity requirements in the soil. This opens the door to diverse conservation strategies that focus on both phenomena independently.

In this thesis, ruderal anemochorous species have been shown to overcome dispersal barriers and establish in novel ecosystems. Indeed, wind-dispersed species have been shown to benefit from human mediated dispersal in cities by vehicle airflow (von der Lippe et al. 2013), transporting the seeds of native species across the traffic network and nearby roadside verges (von der Lippe & Kowarik 2008). As dispersal of invasive species can occur by similar means (von der Lippe & Kowarik 2007), management of roadside verges in cities offers dual opportunities for biodiversity conservation. Supporting native species vs. alien species in road

107

Chapter 5

verges could help the former disperse across different urban ecosystems while limiting the unpredictable expansion of the latter (Säumel et al. 2015).

Along wind-dispersed species, stomatochorous species (i.e. dispersal in the beaks of birds) were also shown to overcome dispersal barriers in cities. Luna et al. 20018 have already shown an effective transport in cities by this means, with dispersal distances reaching up to 155 metres when birds pick up a fruit from a plant and then consume it at a distant perching tree. Management practices that support the wide distribution of birds in urban habitats (i.e. lowering management practice of some parks; Shwartz et al. 2008) could benefit endangered species that profit from this type of dispersal.

Our study shows that those species with very high soil humidity requirements have the lowest chance of survival. Similar results were found by Knapp et al. (2010), and evidence the problems these species might have in an urban context. However, this was only the case for 32 populations (Fig 4 Chapter 4). Management practices that support endangered plant species in bogs and marshes could improve their survival, and as these biotopes only occupy 0.2% of the area of Berlin (SenStadtUm 2014), conservation practices could be quite cost efficient.

Concluding remark

This study provides strong support for making use of the opportunities that unconventional habitats offer for biodiversity conservation in urban regions and supports a diversified urban conservation approach to cover the potential of all types of urban ecosystems for biodiversity conservation.

Future research

Berlin is an example of old European historical city (Sukopp 1990; Lachmund 2013) which shares with other European cities similar land use legacies (Ramalho & Hobbs 2012) and configuration of urban elements (Louf & Barthelemy 2014). While it is true that the findings of this thesis could apply to many other European metropoles, results from Berlin cannot be generalized for cities around the globe as these can represent different typologies in regard to the history of landscape configuration and the proportions of native vegetation (Hahs et al. 2009). As this is the first research of its kind, comparable meta-analyses around the globe are not yet possible. It would be most interesting if future research in urban ecology could adopt a

108

Chapter 5

common framework such as the one used in this thesis in order to make worldwide studies comparable and assess to which extent trends in cities vary across a decreasing gradient of similarity in age, land use legacies, proportion of native vegetation, culture, or socio-economic background.

This thesis shows that various plant traits of endangered species are related to their colonisation and survival in cities. Previous research has shown rapid changes and evolutionary adaptations in urban floras (Dubois & Cheptou 2017; McDonnel & Hahs 2015), and that trait frequencies can shift as a consecuence of urbanisation (Knapp et al. 2008) or urban stressors (Williams et al. 2015). However, studies including detailed information on the biotope and novelty level are missing thus far, yet it would be most interesting to assess whether a selected plant trait varies for the same species across populations located in different biotope types or different degrees of ecological novelty.

While current research still needs to progress on understanding the effects of urban ecosystems on biodiversity, it would be most interesting to widen the scope of this thesis and additionally study the effects of biodiversity on the city dwellers. With more than half of the world population living in cities (UN 2015) and a worldwide accelerating urbanization (Angel et al. 2011; Seto et al. 2011), urban environments are the quintessential human habitat. While the relationship between biodiversity and urban ecosystem services is complex (Schwarz et al. 2017), there is a growing evidence of positive links between them (Schmitt-Harsh et al. 2013; Briguiche & Zidane 2016; Capotorti et al. 2016). In particular, there are links showing that people in European cities largely prefer high species richness across different land use types (Botzat et al. 2016; Fischer 2018), and that a higher exposure to urban green spaces during childhood is linked to mental well-being later in the future (Engemann et al. 2019). Thus, further exploring the relationship between biodiversity and urban ecosystem services has the potential to have a positive impact on millions of city dwellers worldwide.

109

Chapter 5

References

Anderson E.C. and Minor E.S. 2017. Vacant lots: An underexplored resource for ecological and social benefits in cities. Urban Forestry & Urban Greening 21: 146-152.

Angel S, Parent J, Civco DL, Blei A, Potere D. 2011. The dimensions of global urban expansion: Estimates and projections for all countries, 2000-2050. Progress in Planning 75:53-107.

Aronson M.F., Nilon C.H., Lepczyk C.A., Parker T.S., Warren P.S., Cilliers S.S., Goddard M.A., Hahs A.K., Herzog C., Katti M., La Sorte F.A., Williams N.S. and Zipperer W. 2016. Hierarchical filters determine community assembly of urban species pools. Ecology 97: 2952-2963.

Auffret A.G. 2011. Can seed dispersal by human activity play a useful role for the conservation of European grasslands? Applied Vegetation Science 14: 291-303.

Bonthoux S., Brun M., Di Pietro F., Greulich S. and Bouché-Pillon S. 2014. How can wastelands promote biodiversity in cities? A review. Landscape and Urban Planning 132: 79-88.

Botzat A, Fischer LK, Kowarik I. 2016. Unexploited opportunities in understanding liveable and biodiverse cities. A review on urban biodiversity perception and valuation. Global Environmental Change-Human and Policy Dimensions 39:220-233.

Brenneisen S. 2006. Space for urban wildlife: designing green roofs as habitats in Switzerland. Urban habitats 4.

Briguiche H, Zidane L. 2016. Ethnobotanical study of medicinal plants from El-Jadida City (Morocco). Lazaroa 37:145-151.

Capotorti G, Del Vico E, Anzellotti I, Celesti-Grapow L. 2016. Combining the conservation of biodiversity with the provision of ecosystem services in urban green infrastructure planning: critical features arising from a case study in the metropolitan area of Rome. Sustainability 9:10.

Cilliers S.S., Müller N. and Drewes E. 2004. Overview on urban nature conservation: situation in the western-grassland biome of South Africa. Urban Forestry & Urban Greening 3: 49-62.

Dadvand P., Nieuwenhuijsen M.J., Esnaola M., Forns J., Basagana X., Alvarez-Pedrerol M., Rivas I., Lopez-Vicente M., De Castro Pascual M., Su J., Jerrett M., Querol X. and Sunyer J. 2015. Green spaces and cognitive development in primary schoolchildren. Proc Natl Acad Sci U S A 112: 7937-7942.

Diamond J.M. and Heinen J.T. 2016. Conserving rare plants in locally-protected urban forest fragments: A case study from Miami-Dade County, Florida. Urban Forestry & Urban Greening 20: 1-11.

Du Toit MJ, Kotze DJ, Cilliers SS. 2016. Landscape history, time lags and drivers of change: urban natural grassland remnants in Potchefstroom, South Africa. Landscape ecology 31:2133- 2150.

Dubois J. and Cheptou P.-O. 2017. Effects of fragmentation on plant adaptation to urban environments. Phil. Trans. R. Soc. B 372: 20160038.

110

Chapter 5

Duncan R.P., Clemants S.E., Corlett R.T., Hahs A.K., McCarthy M.A., McDonnell M.J., Schwartz M.W., Thompson K., Vesk P.A. and Williams N.S. 2011. Plant traits and extinction in urban areas: a meta-analysis of 11 cities. Global ecology and biogeography 20: 509-519.

Dyderski M.K., Tyborski J. and Jagodzinski A.M. 2017. The utility of ancient forest indicator species in urban environments: A case study from Poznan, Poland. Urban Forestry & Urban Greening 27: 76-83.

Engemann K, Pedersen CB, Arge L, Tsirogiannis C, Mortensen PB, Svenning J-C. 2019. Residential green space in childhood is associated with lower risk of psychiatric disorders from adolescence into adulthood. Proceedings of the National Academy of Sciences:201807504.

Fischer L.K., Honold J., Cvejic R., Delshammar T., Hilbert S., Lafortezza R., Nastran M., Nielsen A.B., Pintar M., van der Jagt A.P.N. and Kowarik I. 2018. Beyond green: Broad support for biodiversity in multicultural European cities. Global Environmental Change-Human and Policy Dimensions 49: 35-45.

Fischer L.K., von der Lippe M. and Kowarik I. 2013a. Urban grassland restoration: which plant traits make desired species successful colonizers? Applied Vegetation Science 16: 272-285.

Fischer L.K., von der Lippe M. and Kowarik I. 2013b. Urban land use types contribute to grassland conservation: The example of Berlin. Urban Forestry & Urban Greening 12: 263- 272.

Gardiner M.M., Burkman C.E. and Prajzner S.P. 2013. The value of urban vacant land to support arthropod biodiversity and ecosystem services. Environ Entomol 42: 1123-1136.

Goddard M.A., Dougill A.J. and Benton T.G. 2010. Scaling up from gardens: biodiversity conservation in urban environments. Trends Ecol Evol 25: 90-98.

Godefroid S. and Koedam N. 2003. How important are large vs. small forest remnants for the conservation of the woodland flora in an urban context? Global ecology and biogeography 12: 287-298.

Hahs A.K. and McDonnell M.J. 2014. Extinction debt of cities and ways to minimise their realisation: a focus on Melbourne. Ecological management & restoration 15: 102-110.

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. and Williams N.S. 2009. A global synthesis of plant extinction rates in urban areas. Ecol Lett 12: 1165-1173.

Hayes TM. 2006. Parks, People, and Forest Protection: An Institutional Assessment of the Effectiveness of Protected Areas. World Development 34:2064-2075.

Hermy M., Honnay O., Firbank L., Grashof-Bokdam C. and Lawesson J.E. 1999. An ecological comparison between ancient and other forest plant species of Europe, and the implications for forest conservation. Biological conservation 91: 9-22.

Jim C.Y. 2017. An archaeological and historical exploration of the origins of green roofs. Urban Forestry & Urban Greening 27: 32-42.

111

Chapter 5

Klaus V.H. 2013. Urban Grassland Restoration: A Neglected Opportunity for Biodiversity Conservation. Restoration Ecology 21: 665-669.

Knapp S., Kuehn I., Wittig R., Ozinga W.A., Poschlod P. and Klotz S. 2008. Urbanization causes shifts of species' trait state frequencies.

Knapp S, Kühn I, Stolle J, Klotz S. 2010. Changes in the functional composition of a Central European urban flora over three centuries. Perspectives in Plant Ecology, Evolution and Systematics 12:235-244.

Knapp S., Winter M. and Klotz S. 2017. Increasing species richness but decreasing phylogenetic richness and divergence over a 320-year period of urbanization. Journal of Applied Ecology 54: 1152-1160.

Kowarik I. and von der Lippe M. 2018. Plant population success across urban ecosystems: A framework to inform biodiversity conservation in cities. Journal of Applied Ecology 55: 2354-2361.

Kühn I, Klotz S. 2006. Urbanization and homogenization - Comparing the floras of urban and rural areas in Germany. Biological Conservation 127:292-300.

Lachmund J. 2013. Greening Berlin: The Co-production of Science, Politics, and Urban Nature. MIT Press.

Lawson D.M., Lamar C.K. and Schwartz M.W. 2008. Quantifying plant population persistence in human-dominated landscapes. Conserv Biol 22: 922-928.

Lepczyk C.A., Aronson M.F.J., Evans K.L., Goddard M.A., Lerman S.B. and Macivor J.S. 2017. Biodiversity in the City: Fundamental Questions for Understanding the Ecology of Urban Green Spaces for Biodiversity Conservation. BioScience 67: 799-807.

Lim K.H., Lee K.E., Kendal D., Rashidi L., Naghizade E., Winter S. and Vasardani M. 2018. The Grass is greener on the other Side: Understanding the effects of green spaces on Twitter user sentiments. Companion of the The Web Conference 2018 on The Web Conference 2018. International World Wide Web Conferences Steering Committee, pp. 275-282.

Louf R. and Barthelemy M. 2014. A typology of street patterns. Journal of The Royal Society Interface 11: 20140924.

Luna Á, Romero-Vidal P, Hiraldo F, Tella JL. 2018. Cities may save some threatened species but not their ecological functions. PeerJ 6:e4908.

Lundholm J., Macivor J.S., Macdougall Z. and Ranalli M. 2010. Plant species and functional group combinations affect green roof ecosystem functions. PLoS One 5: e9677.

Lundholm J.T. and Richardson P.J. 2010. MINI-REVIEW: Habitat analogues for reconciliation ecology in urban and industrial environments. Journal of Applied Ecology 47: 966-975.

McDonnell MJ, Hahs AK. 2015. Adaptation and Adaptedness of Organisms to Urban Environments. Annual Review of Ecology, Evolution, and Systematics, Vol 46 46:261-280.

112

Chapter 5

McKinney ML. 2006. Urbanization as a major cause of biotic homogenization. Biological Conservation 127:247-260.

Nielsen A.B., van den Bosch M., Maruthaveeran S. and van den Bosch C.K. 2014. Species richness in urban parks and its drivers: A review of empirical evidence. Urban Ecosystems 17: 305- 327.

Pimbert MP, Pretty JN. 2013. Parks, people and professionals: putting ‘participation’into protected area management. Social change and conservation 16:297-330.

Ramalho C.E. and Hobbs R.J. 2012. Time for a change: dynamic urban ecology. Trends Ecol Evol 27: 179-188.

Ramalho C.E., Laliberte E., Poot P. and Hobbs R.J. 2014. Complex effects of fragmentation on remnant woodland plant communities of a rapidly urbanizing biodiversity hotspot. Ecology 95: 2466-2478.

Riecken U 2006. Rote Liste der gefährdeten Biotoptypen Deutschlands: Zweite fortgeschriebene Fassung 2006. Bundesamt für Naturschutz.

Säumel I, Weber F, Kowarik I. 2015. Toward livable and healthy urban streets: Roadside vegetation provides ecosystem services where people live and move. Environmental Science & Policy.

Schmitt-Harsh M, Mincey SK, Patterson M, Fischer BC, Evans TP. 2013. Private residential urban forest structure and carbon storage in a moderate-sized urban area in the Midwest, United States. Urban Forestry & Urban Greening 12:454-463.

Schwarz N, Moretti M, Bugalho MN, Davies ZG, Haase D, Hack J, Hof A, Melero Y, Pett TJ, Knapp S. 2017. Understanding biodiversity-ecosystem service relationships in urban areas: A comprehensive literature review. Ecosystem Services 27:161-171.

SenStadtUm. 2003. Protected Areas by Nature Conservation Legislation (incl. Natura 2000).

SenStadtUm. 2014. Berlin Environmental Atlas, Biotope Types.

Seto KC, Fragkias M, Guneralp B, Reilly MK. 2011. A meta-analysis of global urban land expansion. PLoS One 6:e23777.

Shwartz A, Shirley S, Kark S. 2008. How do habitat variability and management regime shape the spatial heterogeneity of birds within a large Mediterranean urban park? Landscape and Urban Planning 84:219-229.

Shwartz A., Turbe A., Julliard R., Simon L. and Prevot A.C. 2014. Outstanding challenges for urban conservation research and action. Global Environmental Change-Human and Policy Dimensions 28: 39-49.

Sukopp H. 1990. Stadtökologie. Das Beispiel Berlin. Reimer Verlag.–1990.–455 s.

UN. 2015. World Urbanization Prospects: The 2014 Revision. (ST/ESA/SER.A/366).

113

Chapter 5

von der Lippe M, Kowarik I. 2007. Long-distance dispersal of plants by vehicles as a driver of plant invasions. Conserv Biol 21:986-996. von der Lippe M, Kowarik I. 2008. Do cities export biodiversity? Traffic as dispersal vector across urban-rural gradients. Diversity and Distributions 14:18-25. von der Lippe M, Bullock JM, Kowarik I, Knopp T, Wichmann MC. 2013. Human-mediated dispersal of seeds by the airflow of vehicles. PLoS One 8:e52733.

Wells M, Bradon K 1992. People and parks: linking protected area management with local communities. World Bank.

Williams N.S.G., Hahs A.K. and Vesk P.A. 2015. Urbanisation, plant traits and the composition of urban floras. Perspectives in Plant Ecology Evolution and Systematics 17: 78-86.

Williams N.S.G., Schwartz M.W., Vesk P.A., McCarthy M.A., Hahs A.K., Clemants S.E., Corlett R.T., Duncan R.P., Norton B.A., Thompson K. and McDonnell M.J. 2009. A conceptual framework for predicting the effects of urban environments on floras. Journal of Ecology 97: 4-9.

Zeeman B.J., McDonnell M.J., Kendal D. and Morgan J.W. 2017. Biotic homogenization in an increasingly urbanized temperate grassland ecosystem. Journal of Vegetation Science 28: 550-561.

114

Appendix

115

Appendix

Appendix A. Target species of Berlin’s Flora Protection Program and number of populations that had been monitored since 2009. All species are endangered in Berlin or at higher spatial scales, i.e. in the surrounding federal state of Brandenburg, in Germany, Europe, or at a global scale). As additional information, the current Red List status in Berlin is given: 1 threatened with extinction; 2 highly endangered; 3 compromised; 0 extinct or disappeared; G danger of unknown extent; R extremely rare; V early warning list; D data insufficient; * safe; Species that have been classified as ‘extinct’ in 2018 were present at the start of the Flora Protection Program in 2008, but have disappeared since the last monitoring. Moreover, information about the life form of species is given: Hp herbaceous perennial; Ha herbaceous annual; Tr tree; Sh shrub; DS dwarf shrub; Ge geophyte; Hy hydrophyte; Ch chamaephyte; Pp pseudophanerophyte.

Red Red Number of Life Number of Life Species List Species List Populations Form Populations Form Status Status Agrimonia procera 1 14 Hp Cannabis sativa s.l. 1 1 Ha Aira caryophyllea Carex 2 21 Hp subsp. 1 5 Hp appropinquata caryophyllea Carex cespitosa 1 9 Hp Ajuga reptans 1 5 Hp Carex demissa 1 1 Hp Alchemilla 1 2 Hp Carex diandra 1 3 Hp monticola Alchemilla plicata 0 3 Hp Carex hartmanii 1 37 Ge Alchemilla Carex lepidocarpa 1 5 Hp 1 1 Hp subcrenata Carex ligerica V 14 Ge Alisma Hy / 1 2 Carex limosa 1 13 Hp lanceolatum Ge Carex otrubae 1 2 Hp Allium angulosum 1 1 Ge Carex 1 3 Ge Alyssum alyssoides 1 5 Hp pseudobrizoides Andromeda 1 8 DS Carex supina 1 10 Ge polifolia Carex viridula Anemone 1 10 Hp 2 8 Ge subsp. viridula ranunculoides Carlina vulgaris 1 18 Hp Antennaria dioica 0 1 Ch agg. Anthericum Catabrosa 2 1 Hp 1 3 Hp ramosum aquatica Anthyllis 1 17 Hp Centaurea diffusa 1 3 Ha vulneraria s. l. Centaurium 2 1 Hp Arnoseris minima 1 8 Hp erythraea Asperula tinctoria 1 4 Hp Centaurium 1 9 Ha Asplenium pulchellum 2 1 Hp Chenopodium trichomanes 0 1 Hp Astragalus bonus-henricus 1 3 Hp Chenopodium arenarius 1 7 Ha Astragalus danicus 1 5 Hp murale Blysmus Chimaphila 1 2 Ge 0 2 DS compressus umbellata Chrysosplenium Botrychium lunaria 1 26 Ge 1 6 Hp alternifolium Botrychium 1 20 Ge Cicuta virosa 2 15 Hp matricariifolium Botrychium Colchicum 0 2 Ge 1 3 Ge multifidum autumnale Consolida regalis 1 15 Hp

116

Appendix

Red Red Number of Life Number of Life Species List Species List Populations Form Populations Form Status Status Corydalis Galeobdolon 1 3 Ge G 16 Ch intermedia luteum Crataegus 1 6 Sh / Tr Galeopsis ladanum 0 1 Ha macrocarpa Galium pumilum s. 0 1 Hp Crataegus media 1 1 Sh / Tr str. Crataegus 1 1 Tr / Sh Genista germanica 1 4 DS rhipidophylla s. str. Genista tinctoria 1 28 DS Crataegus 1 10 Sh / Tr Gentiana subsphaericea 1 8 Hp pneumonanthe Cuscuta epithymum 1 6 Ha Geranium Cuscuta 1 1 Hp 1 1 Ha columbinum lupuliformis Geranium Cystopteris fragilis 1 3 Hp 1 14 Hp sanguineum s. str. Gypsophila muralis 0 2 Ha Dactylis polygama G 19 Hp Helianthemum Dactylorhiza 1 15 Ge nummularium 1 1 DS incarnata subsp. obscurum Dactylorhiza Helictotrichon 1 10 Ge 1 2 Hp maculata agg. pratense Dactylorhiza 2 8 Ge Hepatica nobilis 1 4 Hp majalis s. str. Dactylorhiza x Hieracium bauhini 1 6 Ge 1 10 Hp aschersoniana subsp. heothinum Dianthus Hieracium 1 63 Hp 1 12 Hp carthusianorum caespitosum Hieracium fallax Dianthus superbus 1 16 Hp 1 3 Hp subsp. durisetum Drosera intermedia 1 2 Hp Hieracium Drosera maculatum subsp. 1 3 Hp 1 1 Hp rotundifolia fictum Dryopteris cristata 1 6 Hp Hieracium maculatum subsp. 1 6 Hp Elatine alsinastrum 1 2 Ha tinctum Epilobium 1 1 Hp Hieracium obscurum prussicum subsp. 0 1 Hp Epipactis palustris 1 1 Ge trichotum Equisetum 1 6 Ge Hierochloe hirta sylvaticum subsp. 1 5 Hp Equisetum 0 1 DS praetermissa variegatum Hierochloe odorata Erigeron 1 3 Hp 0 1 Hp subsp. odorata droebachiensis Hippuris vulgaris 0 8 Hp Euphorbia Hydrocharis 1 5 Hp 2 15 Hy palustris morsus-ranae Euphrasia 0 1 Ha Hypericum nemorosa s. l. desetangsii 1 2 Hp Euphrasia stricta 1 16 Ha nothosubsp. Festuca polesica 1 1 Hp carinthiacum Festuca Hypericum 1 31 Hp 1 5 Hp psammophila maculatum s. str. Hypochaeris Filago minima 2 15 Hp 0 1 Hp glabra Filago vulgaris 1 4 Hp Hypopitys Filipendula 1 4 Hp 2 23 Hp monotropa s. str. vulgaris Impatiens noli- 1 1 Ha Fragaria viridis 1 11 Hp tangere

117

Appendix

Red Red Number of Life Number of Life Species List Species List Populations Form Populations Form Status Status Potamogeton Inula salicina 1 3 Hp 1 5 Ha obtusifolius Iris sibirica 1 36 Hp Potamogeton 1 2 Hy Isolepis setacea 1 4 Hp perfoliatus Juncus Potamogeton 1 2 Hp G 7 Ha alpinoarticulatus pusillus Juncus capitatus 0 1 Ha Potentilla alba 1 17 Hp Juncus filiformis 1 5 Ge Potentilla 1 2 Hp Juncus heptaphylla 2 3 Ge subnodulosus Primula veris 1 6 Hp Juncus tenageia 1 2 Ha Pulicaria 0 1 Hp Juniperus dysenterica Pulsatilla pratensis communis subsp. 1 32 Tr / Sh 1 5 Hp communis subsp. nigricans Koeleria glauca 1 24 Hp Pyrola chlorantha 1 2 Hp Lathraea Pyrola minor 1 1 Hp 1 14 Ge squamaria Ranunculus 1 17 Ha Leersia oryzoides 1 5 Hp aquatilis Ranunculus Lotus tenuis 1 5 Hp 1 1 Ha circinatus Luzula pallescens 1 4 Hp Ranunculus lingua 1 15 Hp Lychnis viscaria 1 7 Hp Ranunculus Lycopodium 1 2 Ch peltatus subsp. 1 1 Ha annotinum peltatus Lythrum Ranunculus 0 2 Ha 1 22 Hp hyssopifolia sardous Medicago minima 2 9 Hp Ranunculus 1 7 Ha Myosotis discolor 0 1 Hp trichophyllus s. l. Myosotis Rhinanthus minor 1 32 Ha 2 7 Hp Rhododendron sparsiflora 1 5 DS Myosurus minimus 1 13 Ha tomentosum Najas marina Rhynchospora alba 1 7 Hp G 3 Hy subsp. intermedia Rosa caesia s. str. 1 3 Sh Najas marina 1 10 Ha Rosa dumalis 1 35 Sh subsp. marina Noccaea Rosa elliptica 1 4 Sh 0 2 Hp caerulescens Rosa marginata 0 1 Sh Oenothera Rosa 1 2 Hp 1 1 Sh parviflora s. str. pseudoscabriuscula Orchis militaris 1 10 Ge Rubus 1 1 Pp Osmunda regalis 1 28 Ge fasciculatiformis Rumex aquaticus 0 2 Hp Parnassia palustris 1 3 Hp Rumex sanguineus 1 20 Hp Platanthera bifolia 1 8 Ge Sagina apetala D 2 Ha Populus nigra G 103 Tr agg. Potamogeton 1 7 Ha Sagina nodosa 1 8 Hp acutifolius Sanguisorba minor 1 14 Hp Potamogeton friesii 1 2 Hy subsp. minor Potamogeton 0 4 Hy Scabiosa canescens 1 4 Hp gramineus Potamogeton Scilla amoena R 4 Ge 1 6 Hy Scolochloa lucens 0 2 Hp Potamogeton festucacea 1 3 Hy nodosus Scorzonera humilis 2 55 Hp

118

Appendix

Red Red Number of Life Number of Life Species List Species List Populations Form Populations Form Status Status Scorzonera Tragopogon 1 1 Hp 0 1 Hp purpurea orientalis Selinum dubium 1 15 Hp Trifolium alpestre 2 1 Hp Trifolium Senecio paludosus 1 10 Hp 0 2 Hp Serratula tinctoria montanum 1 23 Hp subsp. tinctoria Tulipa sylvestris * 6 Ge Silene chlorantha 2 2 Hp Urtica kioviensis 1 11 Hp Utricularia Silene conica 1 21 Ha 1 6 Hy australis Silene noctiflora 1 2 Hp Utricularia minor 1 1 Hy Silene otites 1 24 Hp s. str. Utricularia Silene tatarica 3 23 Hp 1 6 Hy Sparganium natans 1 3 Hy vulgaris Verbena officinalis 1 16 Hp Stipa capillata 1 1 Hp Veronica polita 2 8 Hp Stipa pennata s. str. 1 2 Hp Veronica praecox 1 1 Hp Stratiotes aloides 2 10 Hy Viola hirta 1 3 Hp Swertia perennis 0 1 Hp Taraxacum Viola rupestris 1 15 Hp 0 2 Hp nordstedtii Viola stagnina 0 2 Hp Tephroseris 1 3 Hp Vulpia myuros 3 4 Hp palustris Wolffia arrhiza 2 1 Hy Teucrium scordium 1 1 Hp Zannichellia subsp. scordium 1 3 Hy Thalictrum minus palustris 1 13 Hp subsp. minus Thelypteris 1 2 Ge limbosperma

119

Appendix

Appendix B. Assignment of biotope classes in Berlin to categories of ecological novelty. For each biotope class, the first and second levels of hierarchical subdivisions and the number of further subdivisions are shown. Moreover, it is indicated whether biotope classed were assigned directly to a novelty category or whether information about land use was used to modify the assignment of patches of a given biotope class to a novelty category. Most grassland biotopes on agricultural land, for example, were assigned to hybrid ecosystems while grassland patches located on rooftops, in wastelands, or former sewage farms were classified as novel ecosystems.

Modified Biotope # further by land Category of class Subdivision subdiv. use type ecological novelty

Forests Native forests 44 no Remnant Bog forests 3 no Remnant Swamp forests 7 no Remnant Native forest plantations 86 yes Hybrid or novel Forest clearings 1 no Hybrid Native pioneer forests 28 yes Hybrid or novel Alien forest plantations 43 no Novel Alien pioneer forests 5 no Novel Grasslands Flood grasslands 5 yes Remnant or novel Pastures 4 yes Hybrid or novel Predominantly native 20 yes Hybrid or novel vegetation Meadows 37 yes Hybrid or novel Heaths 6 yes Hybrid or novel Ruderal grasslands 3 no Novel Intensive grasslands 9 no Novel Ornamental lawns 9 no Novel Ruderal sites Exclusively native vegetation 2 yes Hybrid or novel With vegetation cover 49 no Novel Without vegetation cover 8 no Novel Water bodies Natural streams, rivers and 38 no Remnant lakes Reed beds 37 no Remnant Canalized water without banks 9 no Hybrid Built up artificial rivers or lakes 34 no Novel Channels 22 no Novel Built up areas

Residential core buildings 15 no Novel Courtyards 5 no Novel Single houses 8 no Novel Industrial, commercial, and 15 no Novel service buildings Traffic areas 41 no Novel Airports 3 no Novel Bogs and marshes

120

Appendix

Modified Biotope # further by land Category of class Subdivision subdiv. use type ecological novelty All types 79 no Remnant

Groves and hedges

Mostly native vegetation 5 yes Remnant or novel Some native vegetation 178 yes Hybrid or novel No native vegetation 4 no Novel Green spaces Inland dunes 2 no Remnant Small parks 1 25 no Novel Cemeteries 10 no Novel Gardens 6 no Novel Allotment gardens 6 Novel Sport facilities 10 Novel Golf courses 1 Novel Fields Arable land 10 yes Hybrid or novel Fallow land 5 yes Hybrid or novel Plant nurseries 3 no Hybrid Wild fields 3 yes Hybrid or novel

1 Large parks were not addressed as biotope type ‘park’, but differentiated according to present biotope types (e.g. biotope types of forests, grasslands, etc.)

121