HOLOCENE AND PLEISTOCENE GASTROPOD BIODIVERSITY IN EUROPEAN FRESHWATER ECOSYSTEMS

Dissertation

zur Erlangung des Doktorgrades der Naturwissenschaften

an der Karl-Franzens-Universität Graz

Institut für Erdwissenschaften

vorgelegt von

MSc. Elisavet Georgopoulou

Graz, July 2016

PREFACE

This thesis is part of the FWF project no. P25365-B25: “Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients” financed by the Austrian Science Fund. The project was held in the Natural History Museum of Vienna under the supervision of Priv. Doz. Dr. Mathias Harzhauser and in cooperation with Dr. Andreas Kroh, Dr. Oleg Mandic and Dr. Thomas A. Neubauer. The dissertation consists of six chapters and summarizes the results of the study of Quaternary gastropod biodiversity in freshwater ecosystems of Europe. Chapter 1 is a general introduction to the topic, including information on database setup and main objectives. Chapters 2 to 5 are the four scientific papers published or submitted for publication to peer- reviewed journals related to the project. They address the main scientific objectives set in Chapter 1. In Chapter 6, the collective results of these papers are summarized and discussed.

Chapter 2: Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2015. An outline of the European Quaternary localities with freshwater gastropods: Data on geography and updated stratigraphy. Palaeontologia Electronica, 18.3.48A, 1–9. Chapter 3: Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016. Distribution patterns of European lacustrine gastropods – a result of environmental factors and deglaciation history. Hydrobiologia, 775, 69–82. Chapter 4: Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. Beginning of a new age: How did freshwater gastropods respond to the Quaternary climate change in Europe? (Quaternary Science Reviews, in revision). Chapter 5: Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Kroh, A., Mandic, O. 2015. Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots. Proceedings of the National Academy of Sciences of the United States of America, 112 (37), 11478–11483.

Elisavet Georgopoulou Vienna, 2016

ii

Table of Contents

Abstract ...... 1 Zusammenfassung...... 3 Acknowlegedements ...... 5 CHAPTER 1: Introduction ...... 6 1.1. Long-lived lakes in Europe and gastropods ...... 6 1.2. The FreshGEN database setup ...... 9 1.3. Research objectives ...... 12 1.4. Rerefences ...... 13 CHAPTER 2: An outline of the European Quaternary localities with freshwater gastropods: Data on geography and updated stratigraphy ...... 20 2.1. Abstract ...... 20 2.2. Introduction ...... 20 2.3. Methods ...... 22 2.4. Data design ...... 25 2.5. Acknowledgments ...... 27 2.6. References ...... 28 2.7. Supplementary material...... 32 2.7.1. Supplementary material 2.1 ...... 32 2.7.2. Supplementary material 2.2 ...... 32 CHAPTER 3: Distribution patterns of European lacustrine gastropods: a result of environmental factors and deglaciation history ...... 33 3.1. Abstract ...... 33 3.2. Introduction ...... 33 3.3. Methods ...... 36 3.3.1. Dataset ...... 36 3.3.2. Reconstruction of the ice sheets ...... 36 3.3.3. Predictor variables ...... 38 3.3.4. Data analysis ...... 39 3.4. Results ...... 41 3.4.1. Gastropod community composition...... 41 3.4.2. Patterns of species richness and composition ...... 41 3.5. Discussion ...... 44

iii

3.5.1. Species richness ...... 44 3.5.2. Species composition ...... 46 3.5.3. Synthesis ...... 48 3.6. Acknowledgments ...... 50 3.7. References ...... 50 3.8. Supplementary material...... 56 3.8.1. Appendix S3.1 ...... 56 3.8.2. Supplementary figures and tables ...... 63 CHAPTER 4: Beginning of a new age: How did freshwater gastropods respond to the Quaternary climate change in Europe? ...... 72 4.1. Abstract ...... 72 4.2. Introduction ...... 73 4.3. Methods ...... 74 4.3.1. Data evaluation and temporal subdivision...... 74 4.3.2. Species richness ...... 75 4.3.3. Species temporal turnover and range-through assumption...... 75 4.3.4. Range sizes and nestedness ...... 76 4.4. Results ...... 77 4.5. Discussion ...... 81 4.5.1. Patterns of species richness ...... 81 4.5.2. Temporal turnover ...... 83 4.5.3. Nested patterns of species ranges ...... 84 4.5.4. Conclusions ...... 86 4.6. Acknowlegedements ...... 87 4.7. References ...... 88 4.8. Supplementary material...... 94 CHAPTER 5: Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots ...... 96 5.1. Abstract ...... 96 5.2. Significance ...... 96 5.3. Introduction ...... 97 5.4. Materials and methods ...... 98 5.5. Results ...... 99 5.5.1. Shifting hotspots through time ...... 99

iv

5.6. Discussion ...... 111 5.6.1. What drives species richness trends and patterns? ...... 111 5.6.2. Geodynamics as a primary driver of hotspot development ...... 113 5.6.3. The cradle of modern faunas ...... 113 5.7. Acknowledgments ...... 114 5.8. References ...... 115 5.9. Supplementary material...... 119 5.9.1. Data setup ...... 119 5.9.2. Assessment of area and temporal duration ...... 120 5.9.3. Climate data and regression analyses ...... 121 5.9.4. Potential effect of sampling bias ...... 121 5.9.5. References ...... 122 5.9.6. Supplementary figures and tables ...... 124 CHAPTER 6: Conclusions ...... 140 6.1. References ...... 143 Curriculum vitae ...... 146

v

Abstract

This thesis provides insights into the evolution of freshwater gastropod diversity of European continental aquatic systems during the Neogene and Quaternary. Three main aspects are explored: 1) the distribution of gastropods in modern lakes of Europe, 2) the spatial and temporal patterns of species during the Quaternary and 3) the evolution of richness hotspots during the last 23 myr. In total, 244 species of freshwater gastropods have been recorded for 898 European lakes. The effect of seven predictor variables, i.e., surface area, longitude and latitude of lake centroid, lake altitude, lake isolation, annual precipitation and annual mean temperature, and the influence of Late Pleistocene deglaciation on species richness were evaluated using multiple linear regression models. Beta diversity patterns between lake subsets with different deglaciation history, as well as the influence of varied species dispersal abilities and environment factors on species composition within these subsets, were also investigated. The results indicated that both species richness and beta diversity are partly explained by environmental factors and deglaciation history. Within the lake subsets, species composition is also controlled by dispersal limitations of species. The distribution of gastropods in modern European lakes is a young pattern. Since the end of the last Ice Age gastropods have recolonized European lakes – a process that is still ongoing. For the Quaternary, 398 species were recorded from the lacustrine and fluvio- lacustrine deposits of 1129 localities. Species and localities were separated into six distinct time intervals of the Quaternary, i.e., Gelasian, Calabrian, Ionian, Last Interglacial, Last Glacial and Holocene; for evaluation of the Plio-Pleistocene turnover the late Pliocene was included as well. Analyses of species richness, composition and geographical ranges revealed significant differences between the time intervals. Gelasian and Calabrian localities are exceptionally rich; the temporal turnover is high especially between the late Pliocene and the Gelasian. As for the Pliocene, early Pleistocene species still mostly derive from long-lived lakes, a result that is also reflected in the low biogeographical affinities of the localities. In comparison, few long-lived lakes are known for the Middle Pleistocene to Holocene and species derive mostly from short-lived fluvial and/or lacustrine systems. Temporal turnover between those time intervals is lower and the number of generalist species with wide geographical ranges is increased, resulting in the loss of the provincial character typical of the Neogene and Early Quaternary freshwater gastropod faunas.

1

The study of Miocene to Recent freshwater gastropod distributions revealed that the evolution of continental aquatic hotspots of Europe is related to its geodynamic history that facilitated the formation of several long-lived lakes with diverse faunas. Regression analyses further showed that species richness in long-lived lakes increases with warmer climate and larger surface area. The most prominent hotspots of the Neogene existed in the Late Miocene and Pliocene, including the biggest of all, Lake Pannon. The demise of most long-lived lakes at the end of the Pliocene and Quaternary glaciations contributed to the disappearance of such hotspots. In present-day Europe three hotspots of gastropod species richness are recognized. Two of them, the Caspian Sea and Lake Ohrid, evolved in basins that were present already during or before the Pleistocene. The third hotspot is located around the Baltic Sea and consists of a series of geologically young lakes and freshwater lagoons and likely results from constant species accumulation by immigration since deglaciation.

2

Zusammenfassung

Diese Arbeit gibt einen Einblick in die Entwicklung der Süßwassergastropoden-Vielfalt der europäischen kontinentalen Wassersysteme während des Neogens und Quartärs. Drei Hauptaspekte werden untersucht: 1) die Verteilung der Gastropoden in modernen Seen Europas, 2) die räumlichen und zeitlichen Muster der Arten im Quartär und 3) die Entwicklung von Artenreichtum-Hotspots während der letzten 23 Mill. Jahre. Insgesamt wurden 244 Arten von Süßwasserschnecken aus 898 Seen erfasst. Die Auswirkung von sieben Prädiktor-Variablen, nämlich Oberfläche, Längengrad und Breitengrad des Zentroids, Höhenlage und Isolation der Seen, jährlicher Niederschlag und Jahresmitteltemperatur, sowie der Einfluss der spätpleistozänen Deglaziation auf Artenreichtum wurden unter Verwendung von multiplen linearen Regressionsmodellen ausgewertet. Die Struktur der Beta-Diversität zwischen den Seen-Subsets mit verschiedener Deglaziationsgeschichte wurde ebenfalls untersucht. Die Ergebnisse zeigten, dass sowohl Artenreichtum als auch Beta-Diversität zum Teil durch Umweltfaktoren und Deglaziationsgeschichte erklärt werden. Innerhalb der Seen-Subsets wird die Artenzusammensetzung auch durch die beschränkte Ausbreitungsfähigkeit der Arten gesteuert. Die Verteilung der Gastropoden in den modernen europäischen Seen ist ein junges Muster. Seit dem Ende der letzten Eiszeit haben Gastropoden europäische Seen wiederbesiedelt – ein Prozess, der noch immer im Gange ist. Für das Quartär wurden 398 Arten aus 1129 Lokalitäten erfasst. Arten und Lokalitäten wurden in sechs verschiedene Zeitintervalle unterteilt, nämlich Gelasium, Calabrium, Ionium, letztes Interglazial, letztes Glazial und Holozän; für die Auswertung des plio-pleistozänen Turnovers wurde auch das späte Pliozän miteinbezogen. Die Analysen von Artenreichtum und die Zusammensetzung und geographischen Reichweiten der Arten zeigten signifikante Unterschiede zwischen den Zeitintervallen. Lokalitäten des Gelasiums und Calabriums sind außerordentlich reich; der zeitliche Turnover ist besonders hoch zwischen Spät-Pliozän und Gelasium. Wie im Pliozän stammten auch früh-pleistozäne Arten noch Großteils aus langlebigen Seen, ein Resultat, das sich auch in den geringen biogeographischen Affinitäten zwischen den Lokalitäten widerspiegelt. Im Vergleich dazu sind wenige langlebige Seen aus dem Mittelpleistozän bis Holozän bekannt und die Arten sind meist aus kurzlebigen fluvialen und/oder lakustrine Systemen. Der zeitliche Turnover zwischen diesen Zeitintervallen ist niedriger und die Anzahl von Generalisten mit großen geografischen Reichweiten ist erhöht,

3 was zu einem Verlust des für die neogenen und frühquartären Süßwassergastropoden-Faunen typischen provinziellen Charakters führte. Die Studie der miozänen bis heutigen Verteilungen von Süßwassergastropoden ergab, dass die Entwicklung von kontinentalen aquatischen Hotspots in Europa an seine geodynamische Geschichte geknüpft ist, die die Entstehung mehrerer langlebiger Seen mit diversen Faunen begünstigte. Regressionsanalysen zeigten ferner, dass Artenreichtum in langlebigen Seen mit wärmerem Klima und größerer Oberfläche ansteigt. Die prominentesten Hotspots des Neogens existierten im späten Miozän und Pliozän, inklusive des größten von allen, dem Pannon-See. Der Niedergang der meisten langlebigen Seen am Ende des Pliozäns und die quartären Vereisungenen trugen zum Verschwinden solcher Hotspots bei. Im heutigen Europa sind drei Gastropoden-Artenreichtums-Hotspots bekannt. Zwei von ihnen, das Kaspische Meer und der Ohrid-See, entwickelten sich in Becken, die während oder schon vor dem Pleistozän vorhanden waren. Der dritte Hotspot ist um die Ostsee gelegen und besteht aus geologisch jungen Seen und Süßwasserlagunen und resultiert wahrscheinlich aus konstanter Akkumulation von Arten durch Immigration seit der Deglaziation.

4

Acknowlegedements

Here, I would like to express my gratitude to all the people who supported me during my stay in Vienna and contributed to the accomplishment of my thesis. I would like to thank my two supervisors Mathias Harzhauser and Werner Piller who assisted and encouraged me through the three years of my work. Especially to Mathias, I would like to say thank you for giving me the opportunity and trust to come to Vienna! My gratitude extends to Thomas Neubauer, Andreas Kroh and Oleg Mandic for their excellent cooperation in our project. Thomas and Andreas, in the next room, were always ready for all possible questions and answers! Thanks to all the past and present members of the Geological-Paleontological Department of the Natural History Museum in Vienna (Uschi, Alex, Thomas, Herbert, Franz, Tony, Tony, Ulrike, Dörte, Matús, Paloma and Adriana) and special thanks to the librarians of the Museum (Sonja Herzog-Gutsch, Eva Hamberger and Wolfgang Brunnbauer) for providing me access to the library and helping me to find most valuable literature. Special thanks to Helmut and Anja from the Fish Collection for opening their library to me. To Anja, thank you for all the nice times we spent together! The seminars in Graz were always a nice break! Thanks to all the people there, not only for that, but also for their support when dealing with official University business. I would like to thank a lot of scientists around Europe for assisting me in my search of scientific literature, and for proving crucial advice on my work. They are all individually mentioned and thanked in the following chapters. Finally, I would like to thank my family and friends in Greece and Vienna for their moral support. And last but not least, I wish to thank Stelios Simaiakis for his love, support, advice and companionship in all these years.

5

CHAPTER 1: Introduction

1.1. Long-lived lakes in Europe and gastropods

European continental aquatic systems comprise ca. 1,000 gastropod species (estimated after de Jong, 2012; Strong et al., 2008), with several hundreds of them living in lakes (Økland, 1990; Strong et al., 2008). Overall freshwater gastropod richness is estimated to have been even higher for the European Neogene and Quaternary with over 2,500 species recorded so far (estimated after Kennard & Woodward, 1917, 1922; Mania, 1973; Ložek, 1964; Settepassi & Verdel, 1965; Esu & Girotti, 1975; Alexandrowicz, 1999; Sanko, 2007; Neubauer et al., 2014a, b and references therein). The majority of these fossil species is usually associated with long-lived lacustrine basins (lakes); lake or lowland deposits are better preserved compared to other freshwater environments such as springs, streams or groundwaters (Strong et al., 2008). Lakes are characterised as ancient or long-lived when they have existed continuously for a period longer than 100,000 years (Gorthner, 1994). Characteristics shared among long- lived lakes are the tectonic origin, relative stability (e.g., unaffected by the glaciations) and isolation (Martens, 1997; Wilke et al., 2007). In turn, taxa in long-lived lakes are characterised by high endemic rates e.g., Lake Ohrid (see Albrecht & Wilke, 2008) and Lake Pannon (see Neubauer et al., 2016a); the presence of species flock i.e., a group of speciose, endemic and monophyletic species (for a definition see Greenwood, 1984); and similarities to marine biota e.g., the thalassoid shell shape of Ohridopyrgula macedonica (Brusina, 1896) and Falsipyrgula pfeifferi (Weber, 1927) in Lakes Ohrid and Eğirdir, respectively (Wilke et al., 2007). However, it should be noted that the abovementioned features do not necessarily apply in all cases of known long-lived lakes. For example, the low rates of endemism in the putative ancient Lake Eğirdir (Wilke et al., 2007) or the absence of endemic gastropod species from fossil Lake Gacko (Neubauer et al., 2013a, 2016b), are not strong indicators of their ancient character. On the opposite, short-lived lakes can be rich in endemic species e.g., Middle Miocene Lake Groisenbach (Harzhauser et al., 2012). During the Neogene several long-lived lakes, which existed for several thousands to million years, developed across Europe (see Harzhauser and Mandic, 2008). Paradigms of such (now extinct) lakes include the Middle Miocene Lake Sinj of the Dinaride Lake System in the Balkans (see Krstić 2012; Mandic et al., 2009; Neubauer et al., 2011), Middle Miocene Lake Steinheim in southwestern Germany (see Gorthner, 1992; Rasser, 2014) and Late

6

Miocene Lake Pannon shared among Austria, Czech Republic, Slovakia, Hungary, Romania, Croatia, Slovenia, Bosnia and Serbia (see Neubauer et al., 2016b and references therein) (Fig. 1.1). Lake Pannon is the biggest recognised hotspot of freshwater gastropod biodiversity in the late Cenozoic of Europe, counting 579 species for its entire duration and several autochthonous intralacustrine radiations especially among the families Melanopsidae (Geary, 1990, 1992; Bandel, 2000; Geary et al., 2002, 2012; Neubauer et al., 2013b), Hydrobiidae (Geary et al., 2000; Harzhauser & Mandic, 2008) and Lymnaeidae (Gorjanović-Kramberger, 1901, 1923; Moos, 1944). By the end of the Pliocene almost all of the Neogene long-lived lakes disappeared (e.g., Krstić, 2003; Jipa & Olariu, 2009; Elezaj et al., 2010; Magyar et al., 2013; Andreescu et al., 2013). So far, few studied lakes of the early Quaternary present characteristics of long- lived lakes. Among them are Lake Tiberino (Esu & Girotti, 1975), Lake Bresse (Schlickum & Puisségur, 1977, 1978) and Lake Kos (Willmann, 1981). These lakes are rich in species and sometimes carry few characteristic endemic lineages e.g., the Melanopsidae of Lake Kos (Willmann, 1981). On the contrary, the majority of freshwater Quaternary deposits are of a short-lived lacustrine or fluviatile origin (e.g., Ložek, 1964; Szymanek, 2011; Alexandrowicz, 2013; Limondin-Lozouet et al., 2013).

7

Figure 1.1. Palinspastic maps for the Early Miocene, Middle Miocene, Late Miocene and Pliocene with indication of biogeographic regions (dashed lines). Light blue and green areas represent the main continental aquatic systems. Palaeogeographic reconstructions and abbreviations follow Neubauer et al. (2015 and references therein).

Contemporarily, the oldest lake of Europe is Lake Ohrid with an estimated duration of ca. 1.2 million years (Wagner et al., 2014), and several highly endemic taxa (Albrecht &

8

Wilke, 2008). Other putative long-lived lakes are Lake Eğirdir in Turkey (see Wilke et al., 2007) and Lakes Megali Prespa, Mikri Prespa, Pamvotis and Trichonis on the Balkan Peninsula (see Albrecht et al., 2006, 2009, 2012; Frogley & Preece, 2007). However, the majority of European lakes are geologically young, confined to the Late Pleistocene and Holocene (Cohen, 2003). Most northern European and Alpine lakes were formed after the Last Glacial Maximum (ca. 20,000 yr BP) by the glaciers’ retreat, or even after the end of the Younger Dryas (ca. 11,000 yr BP) (Fig. 1.2).

Figure 1.2. Modern map of Europe and the extent of the Last Glacial maximum (light grey) and Younger Dryas (dark grey) ice sheets. Reconstructions follow Ehlers et al. (2011). The geographic coordinate system used is WGS 1984.

1.2. The FreshGEN database setup

The current work is based on an extensive evaluation and documentation of scientific literature. In order to facilitate the accumulated information, an Access database (FreshGEN database – Freshwater Gastropods of the European Neogene database) with eight main tables and several interconnected supplementary tables was designed (Fig. 1.3). The FreshGEN database allows the user to design queries through a series of logical operators, and export information related to his/her specific requirements (e.g., the distribution of the

9

genus Gyraulus from the Early Miocene). Part of the FreshGEN database is available at http://www.marinespecies.org/freshgen/ (Neubauer et al., 2014). Below the eight main tables and the information they include are shortly described.

Taxa & Distributions (simple & with Figure 1.3. The initial interface of the FreshGEN synonyms) (Fig. 1.4): In these two tables the database in Access. name of the taxon and its authority are entered exactly as mentioned in the literature. The parent, the rank and the accepted taxon name are then added. Information on the status of the scientific name (e.g., accepted, temporary name, alternative representation) is provided along with justification and taxonomic references. Main sources used to validate and Figure 1.4. An example from the main table Taxa & Distributions. taxonomically update taxa names include Glöer (2002), Kantor et al. (2011), Fauna Europaea (de Jong et al., 2012), Welter-Schultes (2012) and Neubauer et al. (2014a, b). Records of distribution of the accepted taxon and its synonyms are presented in the distribution table. If available, an Aphia ID which connects the taxon records with World Register of Marine Species database (WoRMS – http://www.marinespecies.org/), is also provided. Other fields are “maximum height” and “maximum width” (in mm, available only for accepted species and subspecies), “Nominal subspecies” (if existing), environment (“brackish”, “freshwater”, “terrestrial”) and if the taxon is introduced or native to Europe (“alien in EU”).

10

Samples & Distribution (Fig. 1.5): Here the sample name and its unique ID are provided. The sample name indicates where the taxon (or taxa) was recorded from the literature, including information on the parent administrative unit Figure 1.5. An example from the main table Samples & Distributions. (“Locality”, “Province”, “Country”), coordinates (“Latitude”, “Longitude”) and their geographic precision (“Precision”: 1 to 6, most accurate to least accurate). Stratigraphic age of the sample (“Stratigraphic Unit”) and its bibliographic source, lake name (“Lake”, available only for recent lakes and selected fossil basins), and taxa and their bibliographic sources are inputted for each sample. When the distribution records are verified and no misidentification is reported from the literature, then the taxa records are marked as valid and certain. If the sample is the type locality of a taxon, then this field is also marked. Another field is the “Alternative names” which denotes misspelling, different notations and secondary names according to different countries (e.g., Çiftlikköy versus Ciftlikköy, Hegymagas versus Hegymagyas and Hegymagos).

Localities (Fig. 1.6): This small table includes locality, province and country name of the sample.

Stratigraphic unit (Fig. 1.7): Different stratigraphic ages as mentioned in the literature are recorded. The lower and upper boundary of the stratigraphic age is denoted; they indicate maximum and minimum possible boundaries of a time interval (e.g., Holocene, 0.0117-0 Ma) unless the exact age is known (e.g., 10.4 Ma). The fields “Unit type” (e.g., Biozone, climatic phase) and “Unit rank” (e.g., Epoch, Age) provide further information for the stratigraphic unit.

Lakes (Fig. 1.8): The most important fields of this table are “Latitude” and Figure 1.6. An example from the main table Localities. Longitude” (the coordinates of the lakes’

11

centroid), “Surface” (surface area in km2), “Perimeter” (in km) and “Elevation” (in m). The spatial information of this main table was extracted from the lake polygons designed in GoogleTM Earth (for fossil lakes see CHAPTER 2). It should be noted that this was possible for Figure 1.7. An example from the main table Stratigraphic unit. all recent lakes and a small number of fossil ones. Other fields are the “Notes”, “Volume” and “max. Depth”.

References (Sources): All references are stored in this main table. A full citation, year of publication and URL address (if available) is provided. The references mainly refer to scientific papers, reports, books and online databases.

Habitat: This field is relevant only for recent lakes; it is intended for recording habitat types or a lake’s trophic status. Currently no Figure 1.8. An example from the main table information is available under this table. Lakes.

The database includes more than 7,000 sample records and more than 30,000 individual distribution records of taxa. All taxa names, including synonyms, alternative representations and unaccepted names, number more than 12,000 entries. The FreshGEN database can be spatially and temporally expanded to incorporate study areas outside of Europe (e.g., Asia) and other time intervals older than the Neogene (e.g., the Paleogene).

1.3. Research objectives

12

In the last decades much research effort has documented broad-scale faunal patterns in aquatic ecosystems (e.g., Illies, 1978; Banarescu, 1990; Ribera et al., 2003; Reyjol et al., 2007; Hof et al., 2008). Invertebrates and particularly freshwater gastropods have received considerably less attention than other taxa, although several hundred species are described for Europe (de Jong, 2012), and the gastropod fauna of many lakes is relatively well-studied (e.g., Favre, 1927; Hubendick, 1947; Økland, 1990; Frogley & Preece, 2007; Hauffe et al., 2011; Albrecht et al., 2012). Notable works address regional biogeographic (e.g., Radoman, 1985; Økland, 1990) and phylogeographic relationships in lakes and springs (e.g., Wilke et al., 2007; Benke et al., 2011). However, research often focuses on the rich and highly endemic faunas of the ancient lakes of the Balkan Peninsula and Asia Minor (e.g., Albrecht et al. 2006, Wilke et al., 2007), while the biogeographic relationships among the majority of European lake faunas remain unreported. Recently, distinct Neogene biogeographic provinces based on gastropod composition of the continental aquatic system of the Early Miocene, Middle Miocene, Late Miocene and Pliocene, were recognised (Neubauer et al., 2015) (Fig. 1.1). Unfortunately, comparable pan- European biogeographic schemes for the Quaternary and contemporary lakes are missing. Similarly, no holistic approach on the evolution of freshwater faunas with emphasis on lacustrine systems throughout the Neogene and Quaternary is available. Overall, this work aims to explore past and present patterns and processes in the distribution and richness of freshwater gastropods in lacustrine systems of the European Neogene and Quaternary. Specifically, the following objectives will be addressed: a) Collection, record and stratigraphic and taxonomic update of all available published information on freshwater gastropods of the European Quaternary; b) Study of patterns of gastropod species richness, composition and spatial distribution in Quaternary continental aquatic systems; c) Assessment of the effect of the Last Ice Ages and geo-climatic predictors on contemporary gastropod fauna of European lakes; d) Examination of the mechanisms underlying the long-term patterns, and distribution of hotspots of lacustrine gastropod diversity during the European Neogene and Quaternary.

1.4. Rerefences

13

Albrecht, C., Trajanovski, S., Kuhn, K., Streit, B., Wilke, T. 2006. Rapid evolution of an ancient lake species flock: freshwater limpets (: Ancylidae) in the Balkan Lake Ohrid. Organisms, Diversity & Evolution 6, 294–307. Albrecht, C., Hauffe, T., Schreiber, K., Trajanovski, S., Wilke, T. 2009. Mollusc biodiversity and endemism in the putative ancient Lake Trichonis (Greece). Malacologia, 51, 357– 375. Albrecht, C., Wilke, T. 2008. Lake Ohrid: biodiversity and evolution. Hydrobiologia, 615, 103–40. Albrecht, C., Hauffe, T., Schreiber, K., Wilke, T. 2012. Mollusc biodiversity in a European ancient lake system: lakes Prespa and Mikri Prespa in the Balkans. Hydrobiologia, 682, 47–59. Alexandrowicz, W.P. 1999. Evolution of the malacological assemblages in north Poland during the Late Glacial and Early Holocene. Folia Quaternaria, 70, 39–69. Alexandrowicz, W.P. 2013. Late Glacial and Holocene molluscan assemblages in deposits filling palaeolakes in Northern Poland. Studia Quaternaria, 30, 5–17. Andreescu, I, Codrea, V., Lubenescu, V., Munteanu, T., Petculescu, A., Ştiucǎ, E., Terzea, E. 2013. New developments in the Upper Pliocene-Pleistocene stratigraphic units of the Dacian Basin (Eastern Paratethys), Romania. Quaternary International, 284, 15–29. Banarescu, P. (1990) Zoogeography of fresh waters. General distribution and dispersal of freshwater . AULA Verlag, Wiesbaden. Bandel, K. 2000. Speciation among the Melanopsidae (). Special emphasis to the Melanopsidae of the Pannonian Lake at Pontian time (Late Miocene) and the Pleistocene and Recent of Jordan. Mitteilungen des Geologisch-Palaontologischen Institutes Universität Hamburg, 84, 131–208. Benke, M., Brändle, M., Albrecht, C., Wilke, T. 2011. Patterns of freshwater biodiversity in Europe: lessons from the spring snail genus Bythinella. Journal of Biogeography, 38, 2021–2032. Cohen, A.S. 2003. Paleolimnology: the history and evolution of lake systems. Oxford University Press. de Jong, Y.S.D.M., 2012. Fauna Europaea version 2.5. Available online at http://www.faunaeur.org. Elezaj, Z., Tmava, A., Pashko, P., Vaso, P. 2010. The Neogene of the Dukagjini Basin. Oltenia. Studii şi comunicări. Ştiinţele Naturii, 26 (1), 291–295.

14

Esu, D., Girotti, O. 1975. La malacofauna continentale del Plio-Pleistocene dell’Italia centrale. I. Paleontologia. Geologica Romana, 13, 203–294. Favre, J. 1927. Les mollusques post-glaciaires et actuels du bassin de Genève. Mémoires de la Société de Physique et d'Histoire Naturelle de Genève, 40 (3), 171–434. Frogley, M.R., Preece, R.C. 2007. A review of the aquatic from lake Pamvotis, Ioannina, an ancient lake in NW Greece. Journal of Conchology, 39 (3), 271–295. Geary, D.H. 1990. Patterns of evolutionary tempo and mode in the radiation of Melanopsis (Gastropoda: Melanopsidae). Paleobiology, 16, 492–511. Geary, D.H. 1992. An unusual pattern of divergence between two fossil gastropods: ecophenotypy, dimorphism, or hybridization. Paleobiology, 18, 93–109. Geary, D.H., Magyar, I., Müller, P. 2000. Ancient Lake Pannon and its endemic molluscan fauna (Central Europe; Mio-Pliocene). Advances in Ecological Research, 31, 463– 482. Geary, D.H., Staley, A.W., Müller, P., Magyar, I. 2002. Iterative changes in Lake Pannon Melanopsis reflect a recurrent theme in gastropod morphological evolution. Paleobiology, 28, 208–221. Geary, D.H., Hoffmann, E., Magyar, I., Freiheit, J., Padilla, D. 2012. Body size, longevity, and growth rate in lake Pannon melanopsid gastropods and their predecessors. Paleobiology, 38, 554–568. Glöer, P. 2002. Die Tierwelt Deutschlands, 73. Teil: Die Süßwassergastropoden Nord- und Mitteleuropas. Bestimmungsschlüssel, Lebensweise, Verbreitung. ConchBooks, Hackenheim. Gorjanović-Kramberger, K. 1901. Über die Gattung Valenciennesia und einige unterpontische Limnaeen. Ein Beitrag zur Entwicklungsgeschichte der Gattung Valenciennesia und ihr Verhältnis zur Gattung Limnaea. Beiträge zur Paläontologie und Geologie Österreich-Ungarns und des Orients, 13, 121–140. Gorjanović-Kramberger, C. 1923. Die Valenciennesiden und einige anderen Limnaeiden der pontischen Stufe des unteren Pliocaens in ihrer stratigraphischen und genetischen Bedeutung. Glasnik Hrvatskoga Prirodoslovnoga Društva, 35, 87–114. Gorthner, A. 1992. Bau, Funktion und Evolution komplexer Gastropodenschalen in Langzeit- Seen. Mit einem Beitrag von Gyraulus ‘‘multiformis’’ im Steinheimer Becken. Stuttgarter Beiträge zur Naturkunde, 190, 1–173. Gorthner, A. 1994. What is an ancient lake? Archiv für Hydrobiologie, 44, 97–100.

15

Greenwood, P.H. 1984. What is a species flock? In: Echelle, A.A., Kornfield, I. (Eds.), Evolution of Fish Species Flocks. Orono Press, University of Maine, pp. 13–19. Harzhauser, M., Mandic, O. 2008. Neogene lake systems of Central and South-Eastern Europe: Faunal diversity, gradients and interrelations. Palaeogeography, Palaeoclimatology, Palaeoecology, 260, 417–434. Harzhauser, M., Neubauer, T.A., Mandic, O., Zuschin, M., Ćorić, S. 2012. A Middle Miocene endemic freshwater mollusc assemblage from an intramontane Alpine lake (Aflenz Basin, Eastern Alps, Austria). Paläontologische Zeitschrift, 86 (1), 23–41. Hauffe, T., Albrecht, C., Schreiber, K., Birkhofer, K., Trajanovski, S., Wilke, T. 2011. Spatially explicit analysis of gastropod biodiversity in ancient Lake Ohrid. Biogeosciences, 8, 175–188. Hof, C., Brändle, M., Brandl, R. 2008. Latitudinal variation of diversity in European freshwater animals is not concordant across habitat types. Global Ecology and Biogeography, 17, 539–546. Hubendick, B. 1947. Die Verbreitungsverhältnisse der limnischen Gastropoden in Südschweden. Zoologiska Bidrag från Upsala, 24, 419–559. Illies, J. 1978. Limnofauna Europaea, 2nd edition. Gustav Fischer Verlag, Stuttgart, New York. Jipa D.C., Olariu, C. 2009. Dacian Basin: Depositional Architecture and Sedimentary History of a Paratethys Sea (GeoEcoMar, Bucharest, Hungary). Kantor, Y.I., Vinarski, M.V., Schileyko, A.A., Sysoev, A.V. 2010. Catalogue of the continental mollusks of Russia and adjacent territories. Version 2.3.1. http://www.ruthenica.com/documents/Continental_Russian_molluscs_ver2-3-1.pdf. Kennard, A.S., Woodward, B.B. 1917. The Post-Pliocene non-marine Mollusca of Ireland. Proceedings of the Geologists’ Association, 28, 109–190. Kennard, A.S., Woodward, B.B. 1922. The Post-Pliocene non-marine Mollusca of the East of England. Proceedings of the Geologists’ Association, 33, 104–142. Krstić, N. 2003. Paludinian beds of Voivodina. Chronostratigraphie und Neostratotypen. Neogen der Zentrale Paratethys, Bd. X, Pl2. Romanien, eds Papaianopol, I., Marinescu, F., Krstić, N., Macaleţ, R. (Editura Academiei Române, Bucuresti, Hungary), pp 77–81. Krstić, N., Savić, L., Jovanović, G. 2012. The Neogene Lakes of the Balkan Land. Annales Géologiques de la Péninsule Balkanique, 73, 37–60.

16

Limondin-Lozouet, N., Preece, R.C., and Antoine, P. 2013. The Holocene tufa at Daours (Somme Valley, northern France): Malacological succession and palaeohydrological implications. Boreas, 42, 650–663. Ložek, V. 1964. Quartärmollusken der Tschechoslovakei. Rozpravy Ústředního Ústavu Geologického 31, Praha. Mandic, O. Pavelic, D., Harzhauser, M., Zupanic, J., Reischenbacher, D., Sachsenhofer, R.F., Tadej, N., Vranjkovic, A. 2009. Depositional history of the Miocene Lake Sinj (Dinaride Lake System, Croatia): a long-lived hard-water lake in a pull-apart tectonic setting. Journal of Paleolimnology, 41, 431–452. Magyar, I., Radivojević, D., Sztanó, O., Synak, R., Ujszászi, K., Pócsik, M. 2013. Progradation of the paleo-Danube shelf margin across the Pannonian Basin during the Late Miocene and Early Pliocene. Global and Planetary Change, 103, 168–173. Mania, D. 1973. Paläoökologie, Faunenentwicklung und Stratigraphie des Eiszeitalters im mittleren Elbe-Saalegebiet auf Grund von Molluskengesellschaften. Geologie Beiheft 21 (78/79), 1–175. Martens, K. 1997. Speciation in ancient lakes (review). Trends in Ecology & Evolution, 12, 177–182. Moos, A. 1944. Neue Funde von Lymnaeiden, insbesondere von Valenciennesiiden im Pannon Kroatiens. Vjestnik Hrvatskog državnog geološkog zavoda i Hrvatskog državnog geološkog muzeja, 2 (3), 341–390. Neubauer, T.A., Mandic, O., Harzhauser, M. 2011. Middle Miocene freshwater mollusks from Lake Sinj (Dinaride Lake System, SE Croatia; Langhian). Archiv für Molluskenkunde, 140 (2), 201–237. Neubauer, T.A., Mandic, O., Harzhauser, M. 2013a. The Middle Miocene freshwater mollusk fauna of Lake Gacko (SE Bosnia and Herzegovina): taxonomic revision and paleoenvironmental analysis. Fossil Record, 16 (1), 77–96. Neubauer, T.A., Harzhauser, M., Kroh, A. 2013b. Phenotypic evolution in a fossil gastropod species lineage: evidence for adaptive radiation? Palaeogeography, Palaeoclimatology, Palaeoecology, 370, 117–126. Neubauer, T.A., Kroh, A., Harzhauser, M., Georgopoulou, E., Mandic, O. 2014a. Synopsis of valid species-group taxa for freshwater Gastropoda recorded from the European Neogene. ZooKeys, 435, 1–6.

17

Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2014b. The FreshGEN Database: Freshwater Gastropods of the European Neogene. Accessed at http://www.marinespecies.org/freshgen on 2015-10-15. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015. A gastropod based biogeographic scheme for the European Neogene freshwater systems. Earth- Science Reviews, 143, 98–116. Neubauer, T.A., Harzhauser, M., Mandic, O., Kroh, A., Georgopoulou, E. 2016a. Evolution, turnovers and spatial variation of the gastropod fauna of the late Miocene biodiversity hotspot Lake Pannon. Palaeogeography, Palaeoclimatology, Palaeoecology, 442, 84– 95. Neubauer, T.A., Mandic, O., Harzhauser, M. 2016b (online). The early Middle Miocene lacustrine gastropod fauna of Džepi, Bosnia and Herzegovina (Dinaride Lake System): high endemism in a small space. Bulletin of Geosciences, 91 (2). Økland, J., 1990. Lakes and Snails: Environment and Gastropoda in 1500 Norwegian lakes, ponds and rivers. UBS/Dr W. Backhuys, Oegstgeest, the Netherlands. Radoman, P. 1985. Hydrobioidea, a Superfamily of Prosobranchia (Gastropoda), II. Origin, Zoogeography, Evolution in the Balkans and Asia Minor. Monographs Institute of Zoology 1, Beograd. Rasser, M.W., 2014. Evolution in isolation: the Gyraulus species flock from Miocene Lake Steinheim revisited. Hydrobiologia, 739, 7–24. Reyjol, Y., Hugueny, B., Pont, D., Bianco, P.G., Beier, U., Caiola, N., Casals, F., Cowx, I., Economou, A., Ferreira, T., Haidvogl, G., Noble, R., de Sostoa, A., Vigneron, T., Virbickas, T. 2007. Patterns in species richness and endemism of European freshwater fish. Global Ecology and Biogeography, 16, 65–75. Ribera, I., Foster, G.N. & Vogler, A.P. 2003. Does habitat use explain large scale species richness patterns of aquatic beetles in Europe? Ecography, 26, 145–152. Sanko, A.F. 2007. Quaternary freshwater molluscs Belarus and neighbouring regions of Russia, Lithuania, Poland (field guide). Institute of Geochemistry and Geophysics, National Academy of Sciences, Belarus. (In Russian) Schlickum, W.R., Puisségur, J.-J. 1977. Die Molluskenfauna des Altpleistozäns von St. Bernard (Département Côte-d'Or). Archiv für Molluskenkunde 107 (4/6), 273–283. Schlickum, W.R., Puisségur, J.-J. 1978. Die Molluskenfauna der Schichten mit burgundinus und Pyrgula nodotiana von Montagny-les-Beaune (Dép. Côte-d'Or). Archiv für Molluskenkunde 109 (1/3), 1–26.

18

Settepassi, F., Verdel, U. 1965. Continental Quaternary mollusca of lower Liri Valley (Southern Latium). Geologica Romana, 4, 369–452. Strong, E.E., Gargominy, O., Ponder, W.F., Bouchet, P. 2008. Global diversity of gastropods (Gastropoda; Mollusca) in freshwater. Hydrobiologia, 595, 149–166. Szymanek, M. 2011. Climate oscillations of the Holsteinian (Mazovian) Interglacial recorded in shell morphometry of Viviparus diluvianus (Kunth, 1865) from eastern Poland. Quaternary International, 241, 143–159. Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M.J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J.H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29. Welter-Schultes, F.W. 2012. European non-marine molluscs, a guide for species identification. Planet Poster Editions, Göttingen. Wilke, T., Albrecht, C., Anistratenko, V.V., Sahin, S.K., Yildirim, M.Z. 2007. Testing biogeographical hypotheses in space and time: faunal relationships of the putative ancient Lake Eğirdir in Asia Minor. Journal of Biogeography, 34, 1807–1821. Willmann, R., 1981. Evolution, Systematik und stratigraphische Bedeutung der neogenen Süßwassergastropoden von Rhodos und Kos/Ägäis. Palaeontographica Abt. A 174, 10–235.

19

CHAPTER 2: An outline of the European Quaternary localities with freshwater gastropods: Data on geography and updated stratigraphy

Elisavet Georgopoulou1, 2, Thomas A. Neubauer1, Andreas Kroh1, Mathias Harzhauser1, Oleg Mandic1

1Geological-Paleontological Department, Natural History Museum Vienna, Burgring 7, 1010 Vienna, Austria 2Institute of Earth Sciences, University of Graz, Heinrichstrasse 26, 8010 Graz, Austria

2.1. Abstract

Fossil and subfossil freshwater gastropods of the European Quaternary have been studied extensively by numerous authors during the nineteenth and twentieth centuries. Despite the wealth of information, these papers usually focus on regional aspects. Geographic names, however, may have changed due to politics and the stratigraphy is outdated in many cases. Since then, numerous papers improved the taxonomic understanding of certain groups of freshwater gastropods and revised the often confused nomenclatural history. Nevertheless, no efforts were made to collect and summarize the data on a pan-European scale in order to compare them. This study focuses specifically on recording Pleistocene and Holocene localities of Europe and some adjacent Asian countries bearing freshwater gastropods, combined with updated stratigraphic information when available. This resulted in a dataset of 1473 localities, which may serve as a base for future analyses of Eurasian post-Pliocene freshwater systems and their biota.

Keywords: freshwater gastropods, database, geo-referenced points, Pleistocene, Holocene

2.2. Introduction

Due to their high preservation potential, shell-bearing gastropods are among the most common fossil and subfossil (<0.01 Ma) macro-organisms in Pleistocene and Holocene deposits (Sparks, 1961). Their shells provide valuable information on palaeoenvironmental evolution as geochemical archives (e.g., Leng & Marshall, 2004; Neubauer et al., 2014), as potential palaeoclimatic proxies (e.g., Szymanek et al., 2005; Sümegi et al., 2008; Szymanek,

20

2011, 2013, 2014) or as ecological indicators, since the ecology of many species is well understood (e.g., Welter-Schultes, 2012 and references therein). In Europe, Quaternary freshwater gastropod assemblages have been intensively studied during the last 150 years (e.g., Geyer, 1912; Ložek, 1964; Esu & Girotti, 1975; Alexandrowicz, 1999; Limondin-Lozouet et al., 2013). On a regional scale several studies tried to integrate areal and distribution data from different localities bearing Quaternary freshwater and/or terrestrial gastropods. Perhaps the most widely cited work is the comprehensive synopsis of Ložek (1964), who provided not only a complete list of the Quaternary localities of former Czechoslovakia, but also gave a detailed list of the species and their ecology. A significant addition to this effort is the study published by Hóračová et al. (2015) who assembled a complete and updated list of Holocene sites bearing molluscs in the Czech and Slovak Republics. Although shorter, but in the same spirit, Skompski (1989) published an outline of 107 sites in south-eastern Poland bearing fossil and subfossil gastropods. The list was composed of the site name, reference, the sediment type and the age of the deposit. Similarly, Alexandrowicz (2013) summarized the 154 localities with Late Glacial and Holocene lacustrine faunas in northern Poland, while the molluscs of the Eemian Interglacial in 35 sections across Poland were studied by Alexandrowicz & Alexandrowicz (2010). In north-western Europe, several non-marine molluscan faunas were reviewed in an attempt to shed light on the stratigraphic position of the Cromerian (Meijer & Preece, 1996). Moreover, the analysis of the distribution of the land snail Discus ruderatus by Dehm (1967) provided additional information on post-glacial localities including freshwater species in southern Germany. A substantial and up-to-date work on the Quaternary freshwater gastropods of Belarus with data on selected Lithuanian, Polish and Russian localities was recently published by Sanko (2007). Other notable papers with regional focus include the post-Pliocene non-marine Mollusca of Ireland and the East of England (Kennard & Woodward, 1917, 1922) and the Plio-Pleistocene and Holocene molluscs of Austria (Frank, 2006). These major works are a rich source of information on the freshwater assemblages of the European Quaternary. Nonetheless, at a pan-European scale, studies compiling and analysing geographic occurrences are missing. Despite the continuous effort to record fossil gastropod assemblages, inconsistencies and uncertainties concerning the stratigraphic ages hamper comparisons and analyses. For example, the use of the stratigraphic ages of the Holocene chronozones, differ depending on the study area and publication date (e.g., Ložek, 1964; Wojciechowski, 1999; Bitinas et al., 2002; Mouthon & Magny, 2014). Another point to

21 address is the difficulty to correlate regional stratigraphy with the international geological timescale (e.g., many Greek localities of the Damatria, Kritika, Kos, Phoka-Elia, Sefto, Stefanena, Tafi and Vokasia formations in Willmann, 1981; Böger, 1983). The problems of correlating literature-based data are to some extent purely conceptual due to changes in definitions of chronostratigraphic units. Hence, the inclusion of the formerly Pliocene Gelasian Stage into the Pleistocene in 2009 shifted the beginning of the Pleistocene from 1.806 myr to 2.588 myr (Gibbard et al., 2010; Gradstein et al., 2012) and resulted in potential pitfalls if informal terms such as “early Pleistocene” are interpreted from the literature. The purpose of this paper is to address the aforementioned need for an integrated study with consistent stratigrvaphy of the European Quaternary. To our knowledge this is a first attempt to gather all published and accessible information on Quaternary freshwater gastropods bearing localities in Europe and adjacent areas. Attached to this work we provide a fully geo-referenced dataset for 1473 literature-based fossil sites across Europe (see Supplementary material 2.1). This work is a substantial expansion of the dataset of Neubauer et al. (2015), which aimed for a comprehensive compendium of the Neogene (sensu Gradstein et al., 2012) localities of Europe with fossil freshwater gastropods. Since the age attributions are still ambiguous in some cases and geographic information is patchy for certain areas, this synopsis has to be considered as first step, which is open to improvements and additions.

2.3. Methods

Data were compiled after an elaborate review of the available literature on the Pleistocene and Holocene fossil and subfossil freshwater gastropod assemblages. The information on the sites and their stratigraphic age derives from original papers but is often updated according to current knowledge. The main sources were online sources (e.g., Biodiversity Heritage Library) as well as numerous journals and books stored at the Geological-Palaeontological and Zoological Departments of the Natural History Museum of Vienna, which were systematically screened for contributions on post-Pliocene freshwater gastropods. Furthermore, specialists on freshwater gastropods were contacted for copies of their publications when those were not available in any of the sources mentioned above. It is important to note that the list of localities provided in this work is not complete. Despite our efforts, not all literature sources were accessible, whereas others may have been unintentionally overlooked.

22

The main online source used to georeference the localities was Google™ Earth 7 software (version 7.1.2.2041). The study area is the European continent including Asia Minor and westernmost Asia (Turkmenistan). These regions were included due to their common palaeogeographic history (Rögl, 1998; Popov et al., 2004). In total, stratigraphic information was extracted from 421 publications (Supplementary material 2.1). The stratigraphic tables (Fig. 2.1, Fig. 2.2) are based on the global chronostratigraphic correlation table for the last 2.7 million years (Cohen & Gibbard, 2011). Accordingly, they are products of composition of several publications. In particular, the proposed age boundaries for the chronozones of the Holocene are largely based on Magny (1995), Litt et al. (2001), Bitinas et al. (2002) and Mouthon & Magny (2004, 2014). The Ages of Man are partly supplemented by Holzhauser et al. (2005) and Grosjean et al. (2007). The Baltic stages are after Hyvärinen (1984), Jensen et al. (1999), Damušytė (2009) and Wojciechowski (2011). The Postglacial is defined according to Dehm (1967), the Late Glacial is based on Bos et al. (2001) and Litt et al. (2001) and the Pleniglacial on Bos et al. (2001). For the Pleistocene, the standard Italian stages, the North West European stages and the British stages are derived from Cohen & Gibbard (2011). The Bavarian stages are taken from Doppler et al. (2011). The outline of the Polish stages was taken from Ber (2005) and Lindner et al. (2013). The ambiguous terms Diluvial/Ältere Kalktuffe and Alluvial/Jüngere Kalktuffe/Subrecent were assigned to Pleistocene (1.806- 0.0117 myr) and Holocene (0.0117-0 myr), respectively. The chronostratigraphic tables are as complete as possible, including the main units that were used to describe the age of the respective sites. Nevertheless, some terms (e.g., local names of Glacial-Interglacial stages) are omitted due to space constrains and readability.

23

Figure 2.1. Stratigraphic table of the Holocene with suggested boundaries for the climatic phases, the Ages of Man and the Baltic Stages. The Chronozones, the Ages of Man and the Baltic stages of the Latest Pleistocene are also included. The colour code corresponds to the colours in the online supplied kml-file. The dates are expressed in myr BP. For information about the sources see ´Methods´. Abbreviations used in the chart: S.s. = Sea stage.

24

Figure 2.2. Stratigraphic table of the Pleistocene based on the global chronostratigraphic correlation table for the last 2.7 million years (Cohen & Gibbard, 2011). The Marine Isotope Stages (MIS) follow Lisiecki & Raymo (2005). The colour code corresponds to the colours in the online supplied kml-file. The dates are expressed in myr BP. The first 0.1 Ma are not to scale. For information about the sources see ´Methods´. Abbreviations used in the chart: Bram. = Bramertonian, CI = Cromerian Interglacial, LG = Last Glacial, Mid. Pol. Com. = Middle Polish Complex, M/R I. = Mindel/Riss Interglacial, Nor. Pol. Com. = North Polish Complex, Pre-Past./Bav. = Pre-Pastonian/Baventian, R/W I. = Riss/Würm Interglacial.

2.4. Data design

In total, the list comprises 1487 entries on Quaternary localities, derived from publications of the nineteenth century up to the present. The stratigraphic time interval covered is 0-2.588 Ma. All localities are fully georeferenced, expect for 14 of which the geographic precision is restricted to country level and thus, they are not drawn on the map. In these cases the name of the site as provided in the publication could not be matched to a known geographic name.

25

The current dataset’s latitude ranges from c. 34° N to 68° N and the longitude runs from c. 10° W to 58° E (Fig. 2.3). The obvious scarcity of points northern than 59° N can be explained by the glacial erosion, while the lack of data for Eastern Europe is mainly due to the unavailability (to us) of regional publications. An overview of the geographic and stratigraphic dataset of the Quaternary localities is provided in the Supplementary material 2.1. It is accompanied by an online Keyhole Markup Language file (.kml) (see Supplementary material 2.2), which can be easily visualized in different Geographic Information Systems, such as Google™ Earth and ESRI® ArcGIS™. In addition, the placemarks of the kml-file are color-coded to correspond with each epoch (i.e., Holocene - dark blue, Pleistocene - light blue). Below a full explanation of each dataset field is given (following Neubauer et al., 2015):

Figure 2.3. European Pleistocene and Holocene localities with fossil and subfossil freshwater gastropods as recorded from the published literature. For complete data see Supplementary material 2.1.

ID. Database identification number of record. Locality. Name of the locality with indication of sample (when applicable). Locality is equal to the name of the parent administrative unit when precise data is unavailable. Administrative unit. Parent geographic category, e.g., the village or city the locality belongs to.

26

Country. The current country the locality is situated in. Alternate spellings. Historical, different or misspelled names attributed to the locality. Latitude. in decimal degrees, using the geographic coordinate system WGS 1984. Longitude. in decimal degrees, using the geographic coordinate system WGS 1984. Geographic precision. Estimate of accuracy of the coordinates (1 = 100 m confidence radius, 2 = 5 km confidence radius, 3 = 20 km confidence radius, 4 = only the political district is known, 5 = only the country is known). Stratigraphic age. Age of the deposits. Additional information is in brackets, when the absolute ages vary between different publications or palaeogeographic settings (e.g., former base of the Quaternary before 2009). When the age was uncertain or unavailable, the most probable stratigraphic age was assigned by the authors, often based on information from regional geology. The assigned age of a locality only corresponds to layers with the recorded freshwater gastropod assemblage of lacustrine, fluvial or brackish origin. Thus, the total stratigraphic range represented by a section might be larger. Age, lower boundary. Absolute age of lower boundary of given stratigraphic interval as derived from the literature. Age, upper boundary. Absolute age of upper boundary of given stratigraphic interval as derived from the literature. Localities with stratigraphic ages crossing the Pleistocene- Holocene or Pliocene-Pleistocene boundary were assigned to the epoch which constitutes the larger portion (e.g., Villafranchian is assigned to Pleistocene). This convention avoids duplicates and potential overlap with the list of Miocene-Pliocene localities by Neubauer et al. (2015). Epoch. The epoch the locality is assigned to. Reference. Publication reference for the stratigraphic information.

2.5. Acknowledgments

We are grateful to the following people for help with literature, stratigraphic classifications and/or other assistance: W.P. Alexandrowicz (AGH University of Science and Technology, Krakow), V.V. Anistratenko (Schmalhausen Institute of Zoology of the Ukrainian National Academy, Kiev), D. Esu (University of Rome), S. Herzog-Gutsch (Natural History Museum Vienna), J. Hupuczi ( University of Szeged), G. Iliopoulos (University of Patras), A. Kourgli (Natural History Museum Vienna), W. Kuijper (Universiteit Leiden), N. Limondin-Lozouet (French National Center for Scientific Research), R. Pouwer (Naturalis Biodiversity Center),

27

M.A. Salamon (University of Silesia), A. Sanko (Belarusian State University), M. Szymanek (University of Warsaw), P. Sümegi ( University of Szeged), F.P. Wesselingh (Naturalis Biodiversity Center), and K. Zágoršek (Praha). We also thank two anonymous reviewers for their comments and literature suggestions that significantly improved the manuscript. The project was financially supported by the Austrian Science Fund (FWF project no. P25365- B25: "Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients") and the European Commission's Research Infrastructure Action via the SYNTHESYS Programme.

2.6. References

Alexandrowicz, S.W., Alexandrowicz, W.P. 2010. Molluscs of the Eemian Interglacial in Poland. Annales Societatis Geologorum Poloniae, 80, 69–87. Alexandrowicz, W.P. 1999. Evolution of the malacological assemblages in north Poland during the Late Glacial and Early Holocene. Folia Quaternaria, 70, 39–69. Alexandrowicz, W.P. 2013. Late Glacial and Holocene molluscan assemblages in deposits filling palaeolakes in Northern Poland. Studia Quaternaria, 30, 5–17. Ber, A. 2005. Polish Pleistocene stratigraphy - A review of interglacial stratotypes. Netherlands Journal of Geosciences, 84, 61–76. Bitinas, A., Damušytė, A., Stančikaitė, M., Aleksa, P. 2002. Geological development of the Nemunas River Delta and adjacent areas, West Lithuania. Geological Quarterly, 46, 375–389. Böger, H. 1983. Stratigraphische und tektonische Verknüpfungen kontinentaler Sedimente des Neogens im Ägäis-Raum. Geologische Rundschau, 72, 771–814. Bos, J.A.A., Bohncke, S.J.P., Kasse, C., Vandenberghe, J. 2001. Vegetation and Climate during the Weichselian Early Glacial and Pleniglacial in the Niederlausitz, eastern Germany-macrofossil and pollen evidence. Journal of Quaternary Science, 16:269– 289. Cohen, K.M., Gibbard, P. 2011. Global chronostratigraphical correlation table for the last 2.7 million years. Subcommission on Quaternary Stratigraphy (International Commission on Stratigraphy), Cambridge, England. Damušytė, A. 2009. Late Glacial and Holocene subfossil mollusc shells on the Lithuanian Baltic Sea coast. Baltica, 22, 111–122.

28

Dehm, R. 1967. Die Landschnecke Discus ruderatus im Postglazial Süddeutschlands. Mitteilungen der Bayerischen Staatssammlung für Paläontologie und HistorischeGeologie, 7, 135–155. Doppler, G., Kroemer, E., Konrad, R., Wallner, J., Jerz, H., Grottenthaler, W. 2011. Quaternary Stratigraphy of Southern Bavaria. E & G Quaternary Science Journal, 60, 329–365. Esu, D., Girotti, O. 1975. La malacofauna continentale del Plio-Pleistocene dell’Italia centrale. I. Paleontologia. Geologica Romana, 13, 203–294. Frank, C. 2006. Plio-pleistozäne und holozäne Mollusken Österreichs. Teil 1. Mitteilungen der Prähistorischen Kommission, 62:1-395. Geyer, D. 1912. Die Molluskenfauna der diluvialen und postdiluvialen Kalktuffe des Diessener Tales, eine biologisch-geologische Studie. Mitteilungen der Geologischen Abteilung des K. Württembergischen Statistischen Landesamts, 9, 1–55. Gibbard, P.L., Head, M.J., Walker, M.J.C., and the Subcommission on Quaternary Stratigraphy. 2010. Formal ratification of the Quaternary System/Period and the Pleistocene Series/Epoch with a base at 2.58 Ma. Journal of Quaternary Science, 25, 96–102. Gradstein, F.M., Ogg, J.G., Schmitz, M.D., Ogg, G.M., 2012. The Geologic Time Scale 2012, 2 volumes. Elsevier, Oxford. Grosjean, M., Suter, P.J, Trachsel, M., Wanner, H. 2007. Ice-borne prehistoric finds in the Swiss Alps reflect Holocene glacier fluctuations. Journal of Quaternary Science, 22, 203–207. Holzhauser, H., Magny, M., Zumbühl, H.J. 2005. Glacier and lake-level variations in west- central Europe over the last 3500 years. The Holocene, 15, 789–801. Hóračová, J., Ložek, V., Juřičková, L. 2015. List of malacologically treated Holocene sites with brief review of palaeomalacological research in the Czech and Slovak Republics. Quaternary International, 357, 207–211. Hyvärinen, H. 1984. The Mastogloia stage in the Baltic Sea history: diatom evidence from southern Finland. Bulletin of the Geological Society of Finland, 56, 99–115. Jensen, J.B., Bennike, O., Witkowski, A., Lemke, W., Kuijpers, A. 1999. Early Holocene history of the southwestern Baltic Sea: the Ancylus Lake stage. Boreas, 28, 437–453. Kennard, A.S., Woodward, B.B. 1917. The Post-Pliocene non-marine Mollusca of Ireland. Proceedings of the Geologists' Association, 28, 109–190.

29

Kennard, A.S., Woodward, B.B. 1922. The Post-Pliocene non-marine Mollusca of the East of England. Proceedings of the Geologists' Association, 33, 104–142. Leng, M., Marshall, J.D. 2004. Palaeoclimate interpretation of stable isotope data from lake sediment archive. Quaternary Science Reviews, 23, 811–831. Limondin-Lozouet, N., Preece, R.C., Antoine, P. 2013. The Holocene tufa at Daours (Somme Valley, northern France): Malacological succession and palaeohydrological implications. Boreas, 42, 650–663. Lindner, L., Marks, L., Nita, M. 2013. Climatostratigraphy of interglacials in Poland: Middle and Upper Pleistocene lower boundaries from a Polish perspective. Quaternary International, 292, 113–123. Lisiecki, L.E., Raymo, M.E. 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography, 20:PA1003. Litt, T., Brauer, A., Goslar, T., Merkt, J., Bałaga, K., Müller, H., Ralska-Jasiewiczowa, M., Stebich, M., Negendank, J.F.W. 2001. Correlation and synchronisation of Late glacial continental sequences in northern central Europe based on annually laminated lacustrine sediments. Quaternary Science Reviews, 20, 1233–1249. Ložek, V. 1964. Quartärmollusken der Tschechoslovakei . Rozpravy Ústředního Ústavu Geologického 31, Praha. Magny, M. 1995. Paleohydrological Changes in Jura (France), and Climatic Oscillations around the North Atlantic from Allerød to Preboreal. Géographie physique et Quaternaire, 49, 401–408. Meijer, T., Preece, R.C. 1996. Malacological evidence relating to the stratigraphical position of the Cromerian, p. 53–82. In Turner, C. (ed.), The Early Middle Pleistocene in Europe. A.A. Balkema, Rotterdam, Brookfield. Mouthon, J., Magny, M. 2004. Malacological history of Lake Annecy (France): a comparison of Late Holocene (since 4700 BC) and present mollusc assemblages. Archiv für Hydrobiologie, 160, 555–573. Mouthon, J., Magny, M. 2014. Unprecedented changes in mollusc assemblages of the littoral zone in the two largest peri-Alpine French lakes since the last 4500 years. Journal of Limnology, 73, 20–30. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Wrozyna, C. 2014. Population bottleneck triggering millennial-scale morphospace shifts in endemic thermal-spring melanopsids. Palaeogeography, Palaeoclimatology, Palaeoecology, 414, 116–128.

30

Neubauer, T.A., Georgopoulou, E., Harzhauser, M., Mandic, O., Kroh, A., Esu, D. 2015. Synopsis of European Neogene localities with freshwater gastropod faunas with data on geography and updated stratigraphy. Palaeontologia Electronica, 18.1.3T:1-7; palaeo-electronica.org/content/2015/1153-neogene-freshwater-gastropods Popov, S.V., Rögl, F., Rozanov, A.Y., Steininger, F.F., Shcherba, I.G., Kovac M. 2004. Lithological-Paleogeographic maps of Paratethys. 10 Maps Late Eocene to Pliocene. Courier Forschungsinstitut Senckenberg, 250, 1–46. Rögl, F. 1998. Palaeogeographic considerations for Mediterranean and Paratethys Seaways (Oligocene to Miocene). Annalen des Naturhistorischen Museums in Wien, Series A, 99,279–310. Sanko, A.F. 2007. Quaternary freshwater molluscs of Belarus and neighbouring regions of Russia, Lithuania, Poland (field guide). Institute of Geochemistry and Geophysics, National Academy of Sciences, Belarus. (In Russian) Skompski, S. 1989. Role of malacologic investigations for stratigraphy of the Quaternary of southeastern Poland. Kwartalnik Geologiezny, 33, 525–540. Sparks, B.W. 1961. The ecological interpretation of Quaternary non-marine Molluscs. Proceedings of the Linnean Society of London, 172, 71–80. Sümegi, P., Gulyás, S., Jakab, G. 2008. Holocene paleoclimatic and paleohydrological changes in Lake Balaton as inferred from a complex quantitative environmental historical study of a lacustrine sequence of the Szigliget embayment. Documenta Praehistorica, 35, 33–43. Szymanek, M. 2011. Climate oscillations of the Holsteinian (Mazovian) Interglacial recorded in shell morphometry of Viviparus diluvianus (Kunth, 1865) from eastern Poland. Quaternary International, 241, 143–159. Szymanek, M. 2013. Palaeoecology of the Holsteinian lake in vicinity of Wilczyn (eastern Poland) based on molluscan studies. Geological Quarterly, 57, 637–648. Szymanek, M. 2014. Environmental changes of the Mazovian (Holsteinian /~MIS 11) palaeolake near Szymanowo (eastern Poland) in the light of malacological analysis. Acta Geologica Polonica, 64, 249–260. Szymanek, M., Nitychoruk, J., Trammer, J., Bińka, K. 2005. Influence of climate on the variability of snails of the genus Viviparus in deposits of the Holsteinian (Mazovian) Interglacial from Ortel Królewski, eastern Poland. Boreas, 43, 335–344. Welter-Schultes, F.W. 2012. European non-marine molluscs, a guide for species identification. Planet Poster Editions, Göttingen.

31

Willmann, R. 1981. Evolution, Systematik und stratigraphische Bedeutung der neogenen Süßwassergastropoden von Rhodos und Kos/Ägäis. Palaeon-tographica Abt. A, 174, 10–235. Wojciechowski, A. 1999. Late Glacial and Holocene lake-level fluctuations in the Kórnik- Zaniemyśl lakes area, Great Poland Lowland. Quaternary Studies in Poland, 16, 81– 101. Wojciechowski, A. 2011. Stages of the evolution of the South Baltic coast as recorded in the molluscan fauna. Journal of Coastal Research, ICS2011 Proceedings, 64, 711–715.

2.7. Supplementary material

2.7.1. Supplementary material 2.1

Complete list of the Pleistocene and Holocene localities. For explanation of the fields see ´Data design´. Available as PDF at palaeo-electronica.org/content/2015/1328-quaternary- gastropods.

2.7.2. Supplementary material 2.2

Interactive visualisation of the data points. Available as zipped file at palaeo- electronica.org/content/2015/1328-quaternary-gastropods.

32

CHAPTER 3: Distribution patterns of European lacustrine gastropods: a result of environmental factors and deglaciation history

Elisavet Georgopoulou1, 2, Thomas A. Neubauer1, Mathias Harzhauser1, Andreas Kroh1, Oleg Mandic1

1Geological-Paleontological Department, Natural History Museum Vienna, Burgring 7, 1010 Vienna, Austria; 2Institute of Earth Sciences, University of Graz, Heinrichstrasse 26, 8010 Graz, Austria

3.1. Abstract

Contemporary climate and deglaciation history have received strong support as drivers of species richness and composition for several European taxa. We explored the influence of these factors on patterns of species richness and faunal composition of 244 freshwater gastropod species from 898 European lakes. We evaluated the influence of late Pleistocene deglaciation and seven physiographical and climatic factors on gastropod distributions using multiple linear regression models. We investigated species beta diversity patterns and the influence of species dispersal abilities and/or environment on species composition between lake subsets with different deglaciation history. Contemporary factors and deglaciation history explain parts of variation in species richness across European lakes. Beta diversity analysis revealed moderate to high differences in species composition between the predefined groups. Patterns of species replacement and species loss indicate that lacustrine gastropod faunas of formerly glaciated areas are subsets of non-glaciated ones. Dispersal limitations and environmental gradients control patterns of beta diversity within different lake subsets. We find strong support that the distribution of European limnic gastropods, at least partially, carries the imprint of the last Ice Age. The differences in species richness and composition point towards a gradual, ongoing process of species recolonization after deglaciation.

Keywords: freshwater snails, species richness, beta diversity, Ice Age effects, contemporary predictors

3.2. Introduction

33

What shapes species richness over large geographical scales? This question has been addressed by a plethora of scientific works over the years testing a number of potential hypotheses (see Rahbek & Graves, 2001). Current environmental conditions are usually addressed as the primary determinant of broad-scale species richness patterns (e.g., Hawkins et al., 2003). Lately, the importance of historical events as a driving force in shaping spatial patterns of species distribution has received considerable attention (e.g., Hewitt, 1999, 2000; Graham et al., 2006; Araújo et al., 2008). Variations in glacial history of Europe have affected patterns of post-glacial recolonization in a range of species, with examples from beetles (Fattorini & Ulrich, 2012; Ulrich & Fattorini, 2013), fish (Reyjol et al., 2007), plants (Normand et al., 2011), mammals (Fløjgaard et al., 2011), reptiles and amphibians (Araújo et al., 2008). Species ecological traits along with climate and historical events have shaped the present-day latitudinal richness gradients of several European taxa (e.g., Hof et al., 2008; Fattorini & Ulrich, 2012). In particular, the relationship of European freshwater species richness in lentic habitats (i.e., standing waters including lakes) with latitude is apparently related to species dispersal abilities (Hof et al., 2008). Freshwater gastropods have received considerably less attention than other taxa, although they represent a rich clade with several hundred species described for Europe (de Jong, 2012). Only few studies have been carried out for few terrestrial (e.g., Harl et al., 2014) and freshwater (e.g., Cordellier & Pfenninger, 2008, 2009; Benke et al., 2011) gastropod taxa. The phylogeography and ecological niche modelling of Radix balthica (Linnaeus, 1758) revealed at least two cryptic refugia in north-western Europe during the late Pleistocene glaciations, while the species later expanded its range by tracking suitable habitats (Cordellier & Pfenninger, 2009). A similar methodological approach showed that Ancylus fluviatilis Müller, 1774 occupied northern European refugia and expanded its postglacial range by adapting to the warming conditions (Cordellier & Pfenninger, 2008). Other works addressed regional biogeographical (e.g., Radoman, 1985; Økland, 1990) and phylogeographical relationships in lake and spring gastropods (e.g., Wilke et al., 2007; Benke et al., 2011). Økland (1990) recorded only 27 gastropod species from over 1,000 Norwegian lakes, identifying that variation in species richness depends largely on lake characteristics, climate, historical immigration routes and topography. Apart from that, research often focused on the rich and highly endemic faunas of the ancient lakes of the Balkan Peninsula and Asia Minor (e.g., Albrecht et al. 2006; Wilke et al., 2007). Although the abovementioned studies on European freshwater gastropods allow for discerning differences in species distribution

34 and/or ecological adaptations, the biogeographical affinities of pan-European lacustrine gastropods are still poorly understood and a comprehensive approach is lacking. European lakes cover a wide climatic, altitudinal and historical range. Lakes in Scandinavia and the Alpine region are geologically young, mostly of post-Pleistocene origin, formed after the retreat of the glaciers following the Last Glacial Maximum (LGM, ca 20,000 years BP). In large parts of Scandinavia, the ice sheet started to retreat not before the onset of the Holocene (corresponding to the end of the Younger Dryas [YD], ca 11,000 years BP). On the contrary, south-eastern Europe is hotspot for long-lived lakes, some of which date back to the late Pleistocene and before (e.g., Lake Ohrid; Wagner et al., 2014b). The Iberian and Balkan Peninsulas, Italy and Asia Minor were hardly coved by ice at all, with the exception of the Pyrenees and a few restricted areas in the southeast (see Ehlers et al., 2011 for a detailed map). Naturally, in such a broad geographical range we would expect the distribution of lacustrine gastropods to reflect adaptations to different climatic regimes and hydrological characteristics. The present study, for the first time, explores broad-scale drivers of gastropod distribution in European lakes by addressing three main hypotheses. (1) Gastropod species richness across European lakes can be explained by physiographical and climatic factors. Considering previous studies, we expect a relation of species richness with latitude which may be driven by climatic gradients (see, e.g., Fattorini & Ulrich, 2012 and references therein) and/or species dispersal abilities (see Hof et al., 2008). Also, lakes are conceived as ecological islands and thus a relationship of area with species richness may be expected (e.g., Eadie et al., 1986; Lewis & Magnuson, 2000; Wagner et al., 2014a; Matthews et al., 2015; Neubauer et al., 2015). (2) Gastropod species richness and composition across European lakes are strongly affected by the late Pleistocene glaciations and particularly the timing of deglaciation. We expect an effect of late Pleistocene ice cover on the modern gastropod distributions, considering the evidence for the effect of the LGM on species contemporary distribution (e.g., see Araújo et al., 2008; Svenning et al., 2009; Normand et al., 2011). We hypothesize that lakes of common deglaciation history share a common biogeographical evolution and thus conform in their species compositions. Differences among such lake subsets would allow inferences on independent evolutionary histories and colonization patterns. (3) We expect species composition within the lake groups with common deglaciation history not to be uniform, but rather to reflect geographical and/or environmental distances. Pronounced differences in the factors affecting species composition would further suggest that the lake subsets show differences in their evolutionary histories.

35

3.3. Methods

3.3.1. Dataset

Data on species presences from 898 lakes throughout Europe were collected after an exhaustive literature search, covering 260 scientific papers, books and reports. Search strategy focussed on systematic and faunal studies, supplemented by library catalogue queries. Published faunal lists were then updated by additional records from systematic revisions and specialist studies (e.g., new species or new occurrences). For a complete list of bibliographical sources, see Table S3.1. To avoid taxonomic inconsistencies, we validated species names as well as genus and family classification using the following sources: Fauna Europaea (de Jong, 2012), Animalbase (Welter-Schultes, 2012), MolluscaBase (MolluscaBase, 2015), Glöer (2002) and Kantor et al. (2010). Only native species and subspecies, and two naturalized invaders, i.e., Potamopyrgus antipodarum (Gray, 1843) and Ferrissia fragilis (Tryon, 1863), were considered. Species were considered endemic only when they occurred in one lake in the current dataset (single lake endemics; SLE hereafter); otherwise species restricted to one lake group (for definition of a lake group see below) were considered as regionally confined. Dubious records were omitted from the analysis. The dataset contains 244 valid species and subspecies belonging to 77 genera and 13 families; the study area ranges from 36°N to 71°N and −10°E to 38°E (Fig. 3.1; Table S3.1).

3.3.2. Reconstruction of the ice sheets

To test the potential influence of the timing of deglaciations on lacustrine gastropod species richness and composition, we defined four lake groups (LGs) based on common deglaciation history and geography. Delimitation followed the glacial limits outlines of LGM and YD maxima as reconstructed by Ehlers et al. (2011). LG1 includes 205 lakes that were formed after the YD, i.e., are younger than 11,000 years. LG2 and LG3 contain lakes formed after the LGM, i.e., are younger than c. 18,000 years. Because the two regions represented by LG2 and LG3 are widely disjunct and had different geodynamic evolution we chose to analyse them separately (see Andrén et al., 2011 for LG2 and Kuhlemann, 2007; Sternai et al., 2012 for LG3). LG2 includes 334 lakes north of 50° latitude and LG3 includes 58 lakes of the Alpine region. Finally, LG4 is comprised of 301 lakes not covered by ice during the last Ice Age. Naturally, the distinction into these four lake subsets is only a coarse proxy for deglaciation history, but precise data on individual geological ages of the lakes are scarcely available.

36

Still, the division constrains the maximum potential age of the lakes in the first three groups. No well-defined age limitation is available for LG4, which combines lakes of different origins and evolutionary histories. Details on the species richness and composition of the four subsets are summarized in Tables 3.1 and S3.1.

Figure 3.1. Map showing the available data on freshwater gastropods in lakes across Europe. The lake groups (LGs) of similar deglaciation history are marked by the limits of the respective ice sheet following Ehlers et al. (2011). LG1 is delimited by the Younger Dryas ice sheet extension; LG2 and LG3 by the extensions during the Last Glacial Maximum in northern Europe and the Alps, respectively. Lakes outside the glacial limits belong to LG4. The geographical coordinate system used is WGS 1984.

37

Table 3.1. Summary statistics of gastropod species richness in the complete dataset and the four lake groups (LGs) 1–4. Lakes Species Mean species richness Richness range Genera Families

Complete dataset 898 244 (102) 7.33 ± 6.87 (7.17± 6.41) 1–66 (1–41) 77 (44) 13

LG1 205 39 (39) 4.17 ± 3.58 (4.17± 3.58) 1–25 (1–25) 22 (22) 9

LG2 334 74 (66) 8.87 ± 7.71 (8.85± 7.65) 1–42 (1–41) 32 (32) 11

LG3 58 53 (50) 10.36 ± 7.01 (10.31± 6.96) 1–27 (1–27) 28 (27) 10

LG4 301 216 (85) 7.20 ± 6.78 (6.76± 5.33) 1–66 (1–34) 75 (42) 13

The numbers in parentheses correspond to the summary statistics of species richness after the exclusion of single lake endemics. The number of families remained unchanged.

3.3.3. Predictor variables

We selected seven abiotic variables for each lake: (1) surface area, (2) longitude of lake centroid, (3) latitude of lake centroid, (4) lake altitude, (5) lake isolation, (6) annual precipitation and (7) annual mean temperature. Air temperature is considered a reliable proxy for lake temperature and has been used recently to study patterns of European lacustrine fish (Emmrich et al., 2014 and references therein). Polygons of the lakes’ outlines were reconstructed in Google™ Earth 7; surface area, longitude and latitude of the lakes’ centroid were calculated in ArcGIS 10 (Esri Inc., 1999–2010) using the feature “Calculate Geometry”. As a measure of isolation, we calculated the geographical distance to the nearest lake, using the “Near” tool in ArcGIS. In order to locate the nearest lake, we supplemented our dataset with the Global Lakes and Wetlands Database (GLWD) which contains over 23,000 polygons of European lakes and constitutes the best available source for lakes on a broad- scale (Lehner & Döll, 2004). The climatic variables were obtained from the WorldClim Database (Hijmans et al., 2005) in the spatial resolution of approximately 1 km2. To attribute the two climatic variables to the lakes, we used the “Zonal Statistics” and the “Sample” tools in ArcGIS. In addition, depth was only available for a small subset of the data (359 lakes). We performed a linear regression of depth and lake species richness to test for a potential relationship, but the correlation was very weak (R2 = 0.02, P = 0.006). The relationship did not change when depth was included in a multiple regression along with the other variables where the individual contribution of depth was negligible. Because of data unavailability for

38 the majority of investigated lakes and its minor effect on species richness, depth was excluded from further considerations below.

3.3.4. Data analysis

All parameters were log10-transformed to improve normality. We assessed normality assumptions for species richness and predictor variables using Q–Q plots. Prior to the analysis, we tested for collinearity among the variables, as its existence may bias the estimation of model parameters (Legendre & Legendre, 1998). To do so, we calculated the variance inflation factor (VIF) of predictor variables using the R package HH v. 3.1-14 (Heiberger, 2015) for the complete dataset and for each group separately. As a rule of thumb, VIF values greater than ten indicate the presence of multicollinearity (Quinn & Keough, 2002). To test for a possible effect of deglaciation history on species richness of the studied lakes, we constructed three linear models (LMs) for the complete dataset. The first model (LM1) included only the lake groups, i.e., a categorical variable with four levels of deglaciation history (LG1–LG4), the second model (LM2) included only a combination of the abiotic predictors, and the third model (LM3) included a combination of the abiotic predictors and the lake groups. Best variable combinations for LM2 and LM3 based on Akaike’s information criterion (AIC) were acquired via stepwise regressions using a backward elimination procedure. In LM3, we included the abiotic variables resulting from the stepwise regression of LM2. The three LMs were subsequently compared using analysis of variance (ANOVA) in order to test if they significantly differed from each other. Furthermore, using the seven abiotic variables, identical to LM2, we performed stepwise multiple regressions using a backward elimination procedure for each lake subset (LG1 to 2 LG4). The adjusted coefficient of determination (R adj) and the AIC were used to evaluate the resulting variable combinations. We measured the individual contributions of the remaining variables of the multiple regression models with hierarchical partitioning using the R package hier.part v. 1.0-4 (Walsh & Mac Nally, 2013). In addition, residuals of each final model were checked for spatial autocorrelation using Moran’s I index. Spatial autocorrelation is the lack of independence among observations that can lead to potential statistical biases (Legendre & Legendre, 1998; for in-depth discussion see also Diniz-Filho et al., 2003). Differences in mean species richness in the LGs were tested using one-way ANOVA with Dunnett–Tukey–Kramer pairwise multiple comparison post hoc tests adjusted for

39 unequal variances and sample sizes using the R package DTK v. 3.5 (Lau, 2013). Potential variation of species composition among the four LGs was analysed with the Jaccard family of dissimilarity measures for pairwise comparisons, namely βjac, βjtu and βjne, which partition species dissimilarity into species replacement without the influence of species richness (spatial turnover component) and species loss (nestedness-resultant component) (Baselga, 2010, 2012). The individual components, i.e., dissimilarity due to species replacement between regions (βjtu) and due to species loss (βjne), equal together βjac (Baselga, 2012). The dissimilarity measures were calculated using the R package betapart v. 1.3 (Baselga et al., 2013). Using these analyses, we aim to explore hypotheses (1) and (2). We used Mantel and partial Mantel tests to analyse compositional dissimilarity within each LG in relation to geographical distances between lakes and lakes’ environmental characteristics, corresponding to hypothesis (3). Lake perimeter, altitude, annual precipitation and annual mean temperature were selected as common environmental variables for all LGs. Temperature and precipitation were considered as proxies for climate, altitude for topography and perimeter for habitat availability (see Dehling et al., 2010). The use of Mantel and partial Mantel tests would allow us to control whether the compositional dissimilarity patterns (i.e., beta diversity) within each LG are structured by the lakes’ spatial position and/or environmental characteristics (for discussion see Tuomisto & Ruokolainen, 2006). Dissimilarity, geographical and environmental distance matrices were designed for each LG separately. The βjac dissimilarity matrices were based on the abovementioned measure for pairwise comparisons. We produced environmental distance matrices using Euclidean distances. Geographical distance matrices were produced in ArcGIS using the tool “Generate Near Table”, which calculates the geographical distance between each pair of lakes. Mantel and partial Mantel tests were implemented using the R package vegan v. 2.2-1 (Oksanen et al., 2015). Lake perimeter has been recently facilitated as an approximation for habitat availability (see Dehling et al., 2010), largely corresponding to the extent of shallow-water habitats to which most freshwater gastropods are confined to. However, because of its high collinearity with surface area, perimeter was not used in the linear models presented here. In addition, the varied proportion of single lake endemics per LG may have an effect on the results. Thus, the influences of habitat availability (approximated by lake perimeter) as well as the proportion of endemic species were tested in two separate sets of analyses and compared to the results presented here (see Appendix S3.1).

40

Statistical analyses were performed in R version 3.1.1 (R Core Team, 2014) and Moran’s I index statistics were calculated in SAM v.4.0 (Rangel et al., 2010).

3.4. Results

3.4.1. Gastropod community composition

LG1 was represented by 16% of the total number of species, with an average species richness of 4.17 per lake (Table 3.1). LG2, LG3 and LG4 were represented by 30.3, 21.7 and 88.5% of the species and showed an average species richness of 8.87, 10.36 and 7.2, respectively. Dominating family in all subsets except LG4 was the Planorbidae, with 15 (38.5%), 27 (36.5%) and 20 (37.7%) species in LG1, LG2 and LG3, respectively. LG4 was the only subset with representatives of all 13 families. In LG4, the dominating and most diverse family was the Hydrobiidae with 84 (38.9%) species. The average richness differed significantly across the LGs (ANOVA F3, 894 = 25.89; P < 0.0001), but not for all pairwise comparisons (Table S3.2). All pairs were significant (P < 0.01) except LG3–LG2 (P = 0.39). 102 species remained when SLE were excluded, but average species richness and quantitative differences between the LGs remained similar (ANOVA F3, 894 = 30.48; P < 0.0001; Tables 3.1, S3.2).

3.4.2. Patterns of species richness and composition

Given the lack of multicollinearity (VIF values <7.5), all variables were included in the models (Table S3.3). The spatial correlograms created for the model residuals of the LMs, the complete dataset and the four LGs, revealed low spatial autocorrelation in the first distance classes (Fig. S3.1). However, regression coefficients are not severely influenced by the presence of residual spatial autocorrelation within these ranges (see Hawkins et al., 2007). 2 The explanatory power of LM1 was significant but poor (F3, 894 = 27.56, R adj = 0.0816, P < 0.0001, Table S3.4). In LM2, latitude and altitude were excluded by the stepwise variable selection procedure, while the rest of the predictors explained only a moderate part of the 2 variation in species richness (F5, 892 = 72.75, R adj. = 0.2857, P < 0.0001, Table 3.2). LM3 included a combination of abiotic variables, LGs and interactions between lake characteristics 2 and deglaciation history (F20, 877 = 38.39, R adj = 0.4547, P < 0.0001, Table S3.4) and performed significantly better than LM1 and LM2 (both P < 0.0001). The three LMs indicate that variation in lacustrine gastropod richness of Europe is partially explained by

41 environmental, spatial characteristics, and deglaciation history. Multiple regressions showed that the determinants of gastropod species richness varied among the subsets (Fig. 3.2; Table 3.2; Fig. S3.2). In LG1, species richness increases significantly with decreasing latitude, altitude and precipitation and increasing area, accounting together for 32.36% of the variation; the main contributor here was precipitation (Fig. 3.2). Variation in species richness in LG2 was related negatively to latitude, altitude, precipitation and distance and positively to longitude and area. Here, the individual contribution of latitude was the highest, followed by precipitation. In LG3, latitude, temperature and precipitation were maintained during the stepwise removal. Almost 58% of the variation in species richness was explained by these variables, but only temperature showed a considerably high individual contribution. For LG4, latitude, area and temperature were maintained, accounting for 38.4% of the variation in species richness. Area was the predictor with the highest contribution. Pairwise beta diversity analysis revealed differences in species Figure 3.2. Partition of variance based on the multiple composition between the LGs (Table regression results of the complete dataset (CD) and each 3.3). Beta diversities between LG4 and lake group (LG). U, the unexplained variation; Main, the variation of the predictor with the highest individual each of the other LGs (0.78–0.85) were contribution, i.e., area (CD), precipitation (LG1), latitude distinctly higher than those for the (LG2), temperature (LG3) and area (LG4); Rest, the remaining combinations (0.46–0.47) variation attributed to the remaining predictors (for (Table 3.3). Species replacement was detailed results see Table 3.2). higher than species loss for the pairs LG1–LG3, LG2–LG3 and LG2–LG4 (Fig. 3.3; Table 3.3). For the remaining pairs (LG1–LG2, LG1–LG4 and LG3–LG4), beta diversity patterns were primarily caused by species loss (Fig. 3.3; Table 3.3).

Mantel tests showed that beta diversity (βjac) was positively correlated with geographical and environmental distance for all LGs, although the strength of the correlation varied (Table 3.4). Partial Mantel tests revealed that beta diversity was significantly correlated with geographical distance in LG1, LG2 and LG4 after the effect of environmental distance was partialled out. On the other hand, beta diversity was significantly correlated with environmental distance after the exclusion of the effect of geographical distance only for LGs 2–4.

42

Table 3.2 Variable sets affecting gastropod species richness of European lakes as proposed by stepwise multiple regressions (MR). 2 R adj P Lowest AIC ΔAIC Variables Est. MR P MR Ind.

Complete dataset 0.2857 <0.0001 -1943.15 2.27 Long 0.0851 0.0761 0.015

(LM2) Area 0.1025 <0.0001 0.109

Temp 0.6407 <0.0001 0.087

Prec -0.7780 <0.0001 0.074

Isol -0.0985 0.0013 0.005

LG1 0.3236 <0.0001 -530.46 5.35 Lat -4.9786 <0.0001 0.095

Alt -0.1790 <0.0001 0.049

Area 0.0245 0.141 0.006

Prec -0.9223 <0.0001 0.188

LG2 0.5511 <0.0001 -816.95 0.21 Lat -8.3673 <0.0001 0.230

Long 0.3053 <0.0001 0.052

Alt -0.1299 0.0005 0.017

Area 0.1153 <0.0001 0.094

Prec -0.4302 0.0115 0.134

Isol -0.1655 0.0031 0.032

LG3 0.5753 <0.0001 -161.13 4.78 Lat 10.5123 0.0131 0.047

Temp 1.4826 <0.0001 0.412

Prec -1.0471 0.0614 0.138

LG4 0.384 <0.0001 -761.6 6.21 Lat 3.2864 <0.0001 0.068

Area 0.1579 <0.0001 0.312

Temp 0.3299 0.112 0.010

2 2 Shown are the adjusted R (R adj) and P-values (P) of the MRs, Akaike’s information criterion (AIC), the difference between initial and final Akaike’s information criterion (ΔAIC), the estimate of the multiple regression (Est. MR), significance of the variables in the multiple regression (P MR) and the independent contribution for each variable as provided by hierarchical partitioning results (Ind.). Lat latitude (decimal degrees), Long longitude (decimal degrees), Alt altitude (metres), Area lake surface area (km2), Temp mean annual temperature (°C), Prec annual precipitation (mm), Isol Isolation (km).

43

Table 3 Pairwise beta diversity of the lake groups (LGs) 1–4 computed with the Jaccard dissimilarity measure (Baselga, 2012).

Groups βjac βjtu βjne LG1–LG2 0.47 0 0.47 LG1–LG3 0.46 0.30 0.16 LG1–LG4 0.85 0.23 0.62 LG2–LG3 0.47 0.29 0.18 LG2–LG4 0.78 0.44 0.34 LG3–LG4 0.80 0.26 0.54

βjac is the overall beta diversity, βjtu and βjne are the dissimilarity components referring to species replacement and species loss, respectively.

3.5. Discussion

3.5.1. Species richness

Historical constraints are a well-known determinant of large-scale richness patterns, although their influence can vary considerably among groups of organisms (e.g., Reyjol et al., 2007; Araújo et al., 2008; Svenning et al., 2009). The low to moderate explanatory power of the LMs and multiple regression models indicates that the selected physiographical and climatic factors could not efficiently predict lacustrine gastropod species richness, neither on a pan- European scale nor within the predefined subsets. Likewise, deglaciation history of lakes was a significant predictor but its explanatory power was also moderate. Nevertheless, we detected significant differences in species richness and composition among the lake subsets of different glacial history. Consistent with other studies (e.g., Svenning et al., 2009), we propose that additional forces, beyond those modelled here, are shaping lacustrine gastropod distribution and richness patterns across Europe. From a historical point of view, such factors may be distance from refuge locations, post-glacial migration patterns, human influence (cf. Svenning et al., 2009 and references therein) as well as the persistence and retreat pattern of the ice shields (Hawkins & Porter, 2003). On a local scale, gastropod species richness in lakes has been associated with chemical characteristics, lake surface area, altitude and geo-position, i.e., isolation of the lake (Aho, 1978; Økland, 1990; Lewis & Magnuson, 2000). In our study, area was a significant predictor on the pan-European scale, although the contribution to the explained variation was small as also shown for a subset of European pond gastropods (Swiss lakes) by Oertli et al.

44

(2002). Similarly, correlations with lake isolation and climate variables are significant, but with low explanatory power. On a larger scale, variation of taxonomic richness is usually linked to climate (Currie et al., 2004), whereas the impact of past conditions can still affect today’s distributions (Araújo et al., 2008). Although the classification of lakes into LGs according to their glacial history could not explain the variation in species richness (LM1), the explanatory power significantly improved when the effect of deglaciation history was added to the model (LM3). As a corroboration of this hypothesis, the explanatory power of the multiple regressions improved in comparison with the climatic model (LM2) and differences between species richness and composition emerged, when the faunas were divided into subsets of common glacial, hence biogeographical, history. Lakes of LG1 are presently distributed north of 56°N latitude. During the LGM and YD, however, this area was completely glaciated. None of the presently existing lakes were formed before the retreat of the glaciers after the YD and they are younger than 11,000 years. Many of those lakes are even younger, as Scandinavia was fully ice-free only after 9,800 years BP (Andrén et al., 2011). Occurrences of at least five subfossil freshwater gastropod species from deposits dated back to 9,000 years BP in southern central Norway (ca 59°N– 60°N) (Økland, 1990) and 9,500–6,700 years BP in northern Finland (ca 66°N–68°N) (Salmi, 1963; Vasari et al., 1963) indicate a rapid recolonization. Despite the high active- and passive-dispersal capabilities of freshwater gastropods (see Kappes & Haase, 2012 for a review), only few species have reached the northernmost lakes. A permanent establishment of more diverse gastropod faunas that far north is probably impeded by the extreme climatic conditions, resulting in prolonged freezing periods. The richest lakes in LG1 are near the southern margins of the former YD ice shield and are supposedly the oldest lakes of that group. Although we would expect fewer species to be tolerant to the boreal climate, temperature and precipitation did not prove to be dominant limiting factors for the establishment of rich lake faunas in LG1. Consequently, a relation of the low species richness compared to the other LGs with the young geological age of the lakes is likely. In other words, it is possible that many species presently confined to lower latitudes are still expanding their northern ranges. This conclusion is consistent with other studies, which proposed that certain European plant species are still expanding from their ice-age refugia (see Svenning et al., 2008 and references therein). Although characterized by comparable age constraints, the predictor variables yielded quite different results for LG2 and LG3 (Fig. 3.2; Table 3.2). The differences are likely related to the disjunct geographical location and the varied topographical and climatic

45 features of the two areas (Table S3.5). For LG2, the high species richness can be partly attributed to the fact that it includes a series of lakes and freshwater lagoons around the Baltic Sea as well as two of the largest lakes of the dataset (Ladoga and Onega). Our results are in agreement with Dehling et al. (2010), who found that lentic species richness is the highest in the peri-Baltic region. The peri-Baltic lakes are theoretically easily accessible to freshwater gastropods as the area is relatively flat with well-connected and extended lake-wetland systems and many outflows of major rivers. Only temperature yielded a high individual contribution in the geographically limited LG3 (Fig. 3.2). Lakes attributed to that group show an extensive altitudinal range (62–2,362 m a.s.l.). Surprisingly, however, altitude was not found as important predictor here as it was for other geographical regions (e.g., Zhao et al., 2006). Although LG3 was glaciated during the LGM, still pockets of land free of ice existed, wherefore it is possible that gastropods had survived the glacial period in marginal alpine areas (Cordellier & Pfenninger, 2009; Harl et al., 2014). Moreover, most of the lakes in LG3 are situated in peri-Alpine valleys and were almost immediately formed after the retreat of the ice shields (e.g., Lake Geneva; Favre, 1927). Therefore, the higher species richness of LGs 2 and 3 compared to LG1 might be explained by their prolonged geological history and, accordingly, the longer timeframe available for recolonization. The lower explanatory power of the multiple regression model for LG4 can be linked to the rather heterogeneous composition of lakes in terms of origin and faunal evolution. In comparison with LGs1–3, the lakes in LG4 do not share a common (glacial) history and are scattered across Europe (Fig. 3.1), making this group an artificial unit. Average species richness is lower than for LG2 and LG3, but local richness includes the highest values in the whole dataset, owed to the diverse faunas of the Balkan lakes (e.g., Lake Ohrid with 66 species). Unlike in the first three LGs, surface area was the strongest predictor and accounted for 31.2% of variation in species richness. Lakes of LG4 were lesser influenced by the Ice Ages and their faunal history often dates further back in time, providing more time for species accumulation. Accordingly, those lakes are probably closer to faunal equilibrium in regards to surface area, while those of LGs1–3 are still being recolonized and may not have yet reached the stage where area is a limiting factor for species richness (Eadie et al., 1986; Wagner et al., 2014a). Clarification of species richness patterns within this lake group is beyond the scope of the present study.

3.5.2. Species composition

46

It has been previously shown that beta diversity patterns can be linked to historical processes (Graham et al., 2006). Thus, we would expect a more detailed insight into the differentiation between and within the LGs as provided by the beta diversity analysis. The selection of a richness-independent index allows conclusions on the differences in gastropod composition among the four LGs. Our results show that among the glacially affected LGs1–3 the variation in species richness is moderate and differences between LG2 and LG1 to LG3 are mainly due to species replacement. This is closely related to the influence of the geographical position of the respective LGs. Both Alpine and northern European lakes are inhabited by generalist species with wide environmental tolerances (e.g., Økland, 1990; Sturm, 2007 and references therein). In addition, LGs1–2 contain a number of typical boreal taxa as well as species associated with brackish or coastal waters (see Økland, 1990; Welter-Schultes, 2012). The moderate dissimilarity between LG1 and LG2 is mainly a result of nestedness, indicating species loss towards the north (see also Baselga, 2008); hence, not all species in LG2 could yet colonize LG1. Not surprisingly, the gastropod faunas of the three glacially constrained LGs, where species distributions are a result of post-glacial recolonization, is more similar to each other than to LG4 (Table 3.3). The faunal similarity among LGs1–3 is more likely explained by the dominance of pulmonates, i.e., Planorbidae and Lymnaeidae. Consistent with the Holocene fossil record for Scandinavian freshwater gastropods (e.g., Salmi, 1963; Vasari et al., 1963; Bennike et al., 1998), our results indicate that pulmonates are the major colonizers of newly formed lakes, as they are generally characterized by good dispersal abilities, broad environmental tolerance and are successful colonizers of new environments (see Strong et al., 2008; Kappes & Haase, 2012 and references therein; see species examples in Cordellier & Pfenninger, 2008, 2009). Thus, widespread gastropods mainly live in LGs1–3, i.e., lakes that formed after the termination of the late Pleistocene glaciations. These species could have reached the LGs1–3 not only from southern but also central and northern European refugia (Cordellier & Pfenninger, 2008, 2009) located in the area of LG4. This hypothesis is supported by the beta diversity analysis, indicating species loss as the main component explaining dissimilarity between LGs1 and 3 and LG4; for LG2 beta diversity is about equally partitioned into species loss and replacement. This implies that gastropod faunas of formerly glaciated (typically less diverse) areas are partly nested within non-glaciated ones (compare cf. Baselga, 2008), conforming to hypothesis (2). On the other hand, our results for LG4, e.g., high number of SLE (see Table 3.1), are in agreement with studies suggesting that

47 species with narrow ranges are mainly restricted to the southern parts of Europe, partly because of their poor ability to adjust to contemporary climate (see Araújo et al., 2008 and references therein) or their weak dispersal abilities (see Fattorini & Ulrich, 2012). In support of hypothesis (3), the Mantel and partial Mantel tests (Table 4) indicate that within the LGs compositional dissimilarity of lakes is related to geographical and environmental features, demonstrating that both the species’ dispersal history and response to environmental characteristics of the lakes shape the observed beta diversity patterns. In particular, the number of shared species in lakes of LG1 is independent of the lakes’ environmental characteristics but apparently reflect geographical distances between the lakes. Thus, species composition in LG1 is random but spatially autocorrelated due to dispersal limitation of species (see Tuomisto & Ruokolainen, 2006 and references therein). Although lakes in LG1 are characterized by widespread species, differences in the mode of dispersal, e.g., active versus passive (see Kappes & Haase, 2012), in combination with different post- glacial recolonization routes (e.g., Refseth et al., 1998; Tollefsrud et al., 2008), time from deglaciation (see Andrén et al., 2011) and the presence of geographical barriers may explain our findings. The partial Mantel tests showed that environmental distances are more important than geographical distances in LG2. This result is not entirely unexpected; the limnic phase of the Baltic Sea in the early Holocene (see Andrén et al., 2011) and the high connectivity of present lakes facilitated dispersal of species largely unhindered by geographical distance (or boundaries) in LG2. Similarly, species composition in lakes of LG3 is strongly correlated to environmental gradients. Our results agree with a previous study suggesting the colonization patterns of freshwater gastropods in Alpine lakes to depend on the species’ ecological preferences (Sturm, 2007). The lack of correlation between species composition and geographical distance may root in the complex Alpine topography. Lakes located in small geographical distance from each other are often isolated by high mountain ranges, posing major constraints on gastropod dispersal. Species composition in LG4, in contrast, was mainly correlated to geographical distances. The high number of SLE as well as species with low dispersal capability, e.g., habitat specialists such as some hydrobiids (Strong et al., 2008), could explain this pattern.

3.5.3. Synthesis

Our results indicate that the modern distribution of lacustrine gastropods in Europe is still shaped by terminal Pleistocene deglaciation history. Differences in selected predictor

48 variables between LGs suggest that there is no fixed factor controlling richness patterns across Europe. A latitudinal gradient is present in all four LGs but not in the complete dataset, with precipitation and latitude being the strongest predictors for the northernmost lake groups (LG1 and LG2). The results of the general linear models partly support hypotheses (1) and (2). More precisely, the moderate explanatory power of all models, i.e., LMs and multiple regression models for the LGs separately, and the absence of clear geographical patterns suggest that post-glacial recolonization is still ongoing and lake faunas in LGs1–3 are not in a stable equilibrium yet, supporting previous conclusions on species richness patterns of European freshwater animals in general (Dehling et al., 2010). Deglaciation history and related processes are proposed as principal causes; lentic species, in contrast to lotic ones, are closer to equilibrium (Dehling et al., 2010). Overall, based on the multiple regression results, we suggest that additional forces beyond those modelled here are responsible for species richness—not only at European scale but also within the LGs. Other aspects such as habitat diversity, trophic status, total hardness, pH value, human impact and interaction with other species (e.g., predator–prey relationships), which potentially may explain additional parts of the variation in species richness, could not be considered in the present study, largely due to data deficiency on a pan-European scale. The present patterns are considerably affected by gastropod dispersal, which is mostly limited to passive dispersal, e.g., introduction via waterfowl, large mammals or anthropogenic vectors (Kappes & Haase, 2012; van Leeuwen et al., 2012, 2013). Therefore, any historical signal in their distribution pattern can potentially be rapidly obscured. Investigation of the impact of transport by aquatic birds needs detailed analyses of their migratory pathways (Reynolds et al., 2015) and interactions with specific gastropod groups. Differences in richness between the LGs is probably a combination of the timing of deglaciation during the late Pleistocene, differences in post-glacial recolonization and dispersal capabilities of lacustrine gastropods (see also Hof et al., 2008; Dehling et al., 2010). Furthermore, the beta diversity between the LGs is in fact still very high in the light of the many thousand years of possible migration and introduction. The overall high dissimilarity between LGs1–3 and LG4 is related to the presence of many regionally confined species in the latter group, even with the exclusion of the SLE (see Appendix S3.1). As those lakes were not covered by ice during the last Ice Age, faunal development was less constrained. Thus, some of them could have acted as refugia during the glacial periods or as speciation centres (e.g., Albrecht et al., 2006).

49

The correlation of geographical and/or environmental distances with beta diversity showed species distribution patterns are not uniform within the LGs, which is in line with hypothesis (3). The varied degree of correlation with either geographical or environmental distances suggests differences in the lakes’ faunal histories. All the correlations, however, fully confirmed our expectancies, corresponding to differences in the time elapsed since deglaciation, the LG’s geographical location, and the current environmental and topographical conditions. The current study is an important step towards understanding large-scale patterns of the limnic gastropod fauna of Europe. As broad-scale studies on faunal patterns in aquatic ecosystems are generally quite restricted (Heino, 2011) and the impact of deglaciations on the evolution and dispersal of invertebrate taxa are inadequately known, we expect this contribution to stimulate investigations on the historical components explaining distribution patterns of other European freshwater biotas as well.

3.6. Acknowledgments

We thank: R.M. Albuquerque de Matos, V.V. Anistratenko, N. Bonada, P. Djursvoll, Z. Féher, M. Jeffries, G.S. Karaman, A. Kołodziejczyk, P. Nõges, M. Özbek, P. Raposeiro, M. Rieradevall, K. Schniebs, T. Timm, G. Urbanič, M.V. Vinarski and M. Zeki Yildirim for their assistance with literature review. We also thank A. Oikonomou, K. Rijsdijk, E. Tjørve and anonymous reviewers for valuable suggestions. The project was financially supported by the Austrian Science Fund (FWF Project No. P25365-B25: “Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients”).

3.7. References

Aho, J., 1978. populations and the equilibrium theory of island biogeography. II. Relative importance of chemical and spatial variables. Annales Zoologici Fennici, 15, 155–164. Albrecht, C., Trajanovski, S., Kuhn, K., Streit, B., Wilke, T. 2006. Rapid evolution of an ancient lake species flock: freshwater limpets (Gastropoda: Ancylidae) in the Balkan Lake Ohrid. Organisms, Diversity & Evolution, 6, 294–307.

50

Andrén, T., Björck, S., Andrén, E., Conley, D., Lambeck, K., Zillén, L., Anjar, J. 2011. The Development of the Baltic Sea Basin During the Last 130 ka. In Harff, J., Björck, S., Hoth, P. (eds), The Baltic Sea Basin as a Natural Laboratory. Springer, Berlin, 75–97. Araújo, M.B., Nogues-Bravo, D., Diniz-Filho, J.A.F., Haywood, A.M., Valdes, P.J., Rahbek, C. 2008. Quaternary climate changes explain diversity among reptiles and amphibians. Ecography, 31, 8–15. Baselga, A., 2008. Determinants of species richness, endemism and turnover in European longhorn beetles. Ecography, 31, 263–271. Baselga, A., 2010. Partitioning the turnover and nestedness components of beta diversity. Global Ecology and Biogeography, 19, 134–143. Baselga, A., 2012. The relationship between species replacement, dissimilarity derived from nestedness, and nestedness. Global Ecology and Biogeography, 21, 1223–1232. Baselga, A., Orme, D. Villeger, S., De Bortoli, J., Leprieur, F. 2013. betapart: Partitioning Beta Diversity into Turnover and Nestedness Components. R package version 1.3. http://CRAN.R-project.org/package=betapart. Benke, M., Brändle, M., Albrecht, C., Wilke, T. 2011. Patterns of freshwater biodiversity in Europe: lessons from the spring snail genus Bythinella. Journal of Biogeography, 38, 2021–2032. Bennike, O., Lemke, W., Jensen, J.B. 1998. Fauna and flora in submarine early-Holocene lake-marl deposits from the southwestern Baltic Sea. The Holocene, 8, 353–358. Cordellier, M., Pfenninger, M. 2008. Climate-driven range dynamics of the freshwater limpet, Ancylus fluviatilis (Pulmonata, Basommatophora). Journal of Biogeography, 35, 1580–1592. Cordellier, M., Pfenninger, M. 2009. Inferring the past to predict the future: climate modelling predictions and phylogeography for the freshwater gastropod Radix balthica (Pulmonata, Basommatophora). Molecular Ecology, 18, 534–544. Currie, D. J., G. G. Mittelbach, H. V. Cornell, R. Field, J.-F. Guégan, B. A. Hawkins, D. M. Kaufman, J. T. Kerr, T. Oberdorff, E. O’Brien & J. R. G. Turner, 2004. Predictions and tests of climate based hypotheses of broad-scale variation in taxonomic richness. Ecology Letters, 7, 1121–1134. de Jong, Y.S.D.M. 2012. Fauna Europaea Version 2.5. http://www.faunaeur.org. Dehling, D.M., Hof, C., Brändle, M., Brandl, R. 2010. Habitat availability does not explain the species richness patterns of European lentic and lotic freshwater animals. Journal of Biogeography, 37, 1919–1926.

51

Diniz-Filho, J.A.F., Bini, L.M., Hawkins, B.A. 2003. Spatial autocorrelation and red herrings in geographical ecology. Global Ecology and Biogeography, 12, 53–64. Eadie, J.McA, Hurly, T.A., Montgomerie, R.D., Teather, K.L. 1986. Lakes and rivers as islands: species-area relationships in the fish faunas of Ontario. Environmental Biology of Fishes, 15, 81–89. Ehlers, J., Gibbard, P.L., Hughes, P.D. 2011. Quaternary Glaciations – Extent and Chronology, 1st ed. Elsevier, Amsterdam. Emmrich, M., Pédron, S., Brucet, S., Winfield, I.J., Jeppesen, E., Volta, P., Argillier, C., Lauridsen, T.L., Holmgren, K., Hesthagen, T., Mehner, T. 2014. Geographical patterns in the body-size structure of European lake fish assemblages along abiotic and biotic gradients. Journal of Biogeography, 41, 2211–2233. Esri Inc., 1999–2010. ArcGIS for Desktop: Release 10. Environmental Systems Research Institute, Redlands, CA. Fattorini, S., Ulrich, W. 2012. Spatial distributions of European Tenebrionidae point to multiple postglacial colonization trajectories. Biological Journal of the Linnean Society, 105, 318–329. Favre, J., 1927. Les mollusques post-glaciaires et actuels du bassin de Genève. Mémoires de la Société de Physique et d’Histoire Naturelle de Genève, 40, 171–434. Fløjgaard, C., Normand, S., Skov, F., Svenning, J.-C. 2011. Deconstructing the mammal species richness pattern in Europe – towards an understanding of the relative importance of climate, biogeographic history, habitat heterogeneity and humans. Global Ecology and Biogeography, 20, 218–230. Glöer, P., 2002. Die Tierwelt Deutschlands, 73. Teil: Die Süßwassergastropoden Nord- und Mitteleuropas. Bestimmungsschlüssel, Lebensweise, Verbreitung. ConchBooks, Hackenheim. Graham, C.H., Moritz, C., Williams, S.E. 2006. Habitat history improves prediction of biodiversity in rainforest fauna. Proceedings of the National Academy of Sciences of the United States of America, 103, 632–636. Harl, J., Duda, M., Kruckenhauser, L., Sattmann, H., Haring, E. 2014. In search of glacial refuges of the land snail Orcula dolium (Pulmonata, Orculidae) – an integrative approach Using DNA sequence and fossil data. PLoS One 9: e96012. Hawkins, B.A., Porter, E.E. 2003. Relative influences of current and historical factors on mammal and bird diversity patterns in deglaciated North America. Global Ecology and Biogeography, 12, 475–481.

52

Hawkins, B.A., Field, R., Cornell, H.V., Currie, D.J., Guegan, J.F., Kaufman, D.M., Kerr, J.T., Mittelbach, G.G., Oberdorff, T., O’Brien, E.M., Porter, E.E., Turner, J.R.G. 2003. Energy, water, and broad-scale geographic patteerns of species richness. Ecology, 84, 3105–3117. Hawkins, B.A., Diniz-Filho, J.A.F., Bini, L.M., De Marco, P., Blackburn, T.M. 2007. Red herrings revisited: spatial autocorrelation and parameter estimation in geographical ecology. Ecography, 30, 375–384. Heiberger, R.M. 2015. HH: Statistical Analysis and Data Display: Heiberger and Holland. R package version 3.1-14. http://CRAN.R-project.org/package=HH. Heino, J. 2011. A macroecological perspective of diversity patterns in the freshwater realm. Freshwater Biology, 56, 1703–1722. Hewitt, G. 1999. Post-glacial re-colonization of European biota. Biological Journal of the Linnean Society, 68, 87–112. Hewitt, G. M. 2000. The genetic legacy of the Quaternary ice ages. Nature, 405, 907–913. Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A. 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology, 25, 1965–1978. Hof, C., Brändle, M., Brandl, R. 2008. Latitudinal variation of diversity in European freshwater animals is not concordant across habitat types. Global Ecology and Biogeography, 17, 539–546. Kantor, Y.I., Vinarski, M.V., Schileyko, A.A., Sysoev, A.V. 2010. Catalogue of the continental mollusks of Russia and adjacent territories. Version 2.3.1. http://www.ruthenica.com/documents/Continental_Russian_molluscs_ver2-3-1.pdf. Kappes, H., Haase, P. 2012. Slow, but steady: dispersal of freshwater molluscs. Aquatic Sciences, 74, 1–14. Kuhlemann, J. 2007. Paleogeographic and paleotopographic evolution of the Swiss and Eastern Alps since the Oligocene. Global and Planetary Change, 58, 224–236. Lau, M.K. 2013. DTK: Dunnett–Tukey–Kramer Pairwise Multiple Comparison Test Adjusted for Unequal Variances and Unequal Sample Sizes. R package version 3.5. http://CRAN.R-project.org/package=DTK. Legendre, P., Legendre, L. 1998. Numerical Ecology, 2nd ed. Elsevier, Amsterdam. Lehner, B., Döll, P. 2004. Development and validation of a global database of lakes, reservoirs and wetlands. Journal of Hydrology, 296, 1–22.

53

Lewis, D.B., Magnuson, J.J. 2000. Landscape spatial patterns in freshwater snail assemblages across Northern Highland catchments. Freshwater Biology, 43, 409–420. Matthews, T.J., Guilhaumon, F., Triantis, K.A., Borregaard, M.K., Whittaker, R.J. 2015. On the form of species–area relationships in habitat islands and true islands. Global Ecology and Biogeography. doi:10.1111/geb.12269. MolluscaBase. 2015. http://www.molluscabase.org. Accessed 21 Nov 2015. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Kroh, A., Mandic, O., 2015b. Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots. Proceedings of the National Academy of Sciences of the United States of America 112, 11478–11483. Normand, S., Ricklefs, R.E., Skov, F., Bladt, J., Tackenberg, O., Svenning, J.-C. 2011. Postglacial migration supplements climate in determining plant species ranges in Europe. Proceedings of the Royal Society B, 278, 3644–3653. Oertli, B., Auderset Joye, D., Castella, E., Juge, R., Cambin, D., Lachavanne, J.-B. 2002. Does size matter? The relationship between pond area and biodiversity. Biological Conservation, 104, 59–70. Økland, J. 1990. Lakes and Snails: Environment and Gastropoda in 1500 Norwegian Lakes, Ponds and Rivers. UBS/Dr W. Backhuys, Oegstgeest. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H. 2015. vegan: Community Ecology Package. R package version 2.2-1. http://CRAN.R-project.org/package=vegan. Quinn, G.P., Keough, M.J. 2002. Experimental design and data analysis for biologists. Cambridge University Press, Cambridge. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org. Radoman, P. 1985. Hydrobioidea, a Superfamily of Prosobranchia (Gastropoda), II. Origin, Zoogeography, Evolution in the Balkans and Asia Minor. Monographs Institute of Zoology 1, Beograd. Rahbek, C., Graves, G.R. 2001. Multiscale assessment of patterns of avian species richness. Proceedings of the National Academy of Sciences of the United States of America, 98, 4534–4539. Rangel, T.F., Diniz-Filho, J.A.F., Bini, L.M. 2010. SAM: a comprehensive application for Spatial Analysis in Macroecology. Ecography, 33, 46–50.

54

Refseth, U.H., Nesbø, C.L., Stacy, J.E., Vøllestad, L.A., Fjeld, E., Jakobsen, K.S. 1998. Genetic evidence for different migration routes of freshwater fish into Norway revealed by analysis of current perch (Perca fluviatilis) populations in Scandinavia. Molecular Ecology, 7, 1015–1027. Reyjol, Y., Hugueny, B., Pont, D., Bianco, P.G., Beier, U., Caiola, N., Casals, F., Cowx, I., Economou, A., Ferreira, T., Haidvogl, G., Noble, R., de Sostoa, A., Vigneron, T., Virbickas, T. 2007. Patterns in species richness and endemism of European freshwater fish. Global Ecology and Biogeography, 16, 65–75. Reynolds, C., Miranda, N.A.F., Cumming, G.S. 2015. The role of waterbirds in the dispersal of aquatic alien and invasive species. Diversity and Distributions, 21, 744–754. Salmi, M. 1963. On the subfossil Pediastrum algae and molluscs in the Late-Quaternary sediments of Finnish Lapland. Archivum Societatis Zoologicae Botanicae Fennicae Vanamo, 18, 105–120. Sternai, P., Herman, F., Champagnac, J.-D., Fox, M., Salcher, B., Willett, S.D. 2012. Pre- glacial topography of the European Alps. Geology, 40, 1067–1070. Strong, E.E., Gargominy, O., Ponder, W.F., Bouchet, P. 2008. Global diversity of gastropods (Gastropoda; Mollusca) in freshwater. Hydrobiologia, 595, 149–166. Sturm, R. 2007. Freshwater molluscs in mountain lakes of the Eastern Alps (Austria): relationship between environmental variables and lake colonization. Journal of Limnology, 66, 160–169. Svenning, J.-C., Normand, S., Skov, F. 2008. Postglacial dispersal limitation of widespread forest plant species in nemoral Europe. Ecography, 31, 316–326. Svenning, J.-C., Normand, S., Skov, F. 2009. Plio-Pleistocene climate change and geographic heterogeneity in plant diversity–environment relationships. Ecography, 32, 13–21. Tollefsrud, M.M., Kissling, R., Gugerli, F., Johnsen, Ø., Skrøppa, T., Cheddadi, R., van der Knaap, W.O., Latałowa, M., Terhürne-Berson, R., Litt, T., Geburek, T., Brochmann, C., Sperisen, C. 2008. Genetic consequences of glacial survival and postglacial colonization in Norway spruce: combined analysis of mitochondrial DNA and fossil pollen. Molecular Ecology, 17, 4134–4150. Tuomisto, H., Ruokolainen, K. 2006. Analyzing or explaining beta diversity? Understanding the targets of different methods of analysis. Ecology, 87, 2697–2708. Ulrich, W., Fattorini, S. 2013. Longitudinal gradients in the phylogenetic community structure of European Tenebrionidae (Coleoptera) do not coincide with the major routes of postglacial colonization. Ecography, 36, 1106–1116.

55 van Leeuwen, C.H.A., van der Velde, G., van Lith, B., Klaassen, M. 2012. Experimental quantification of long distance dispersal potential of aquatic snails in the gut of migratory birds. PLoS One 7: e32292. van Leeuwen, C.H.A., Huig, N., van der Velde, G., van Alen, T.A., Wagemaker, C.A.M., Sherman, C.D.H., Klaassen, M., Figuerola, J. 2013. How did this snail get here? Several dispersal vectors inferred for an aquatic invasive species. Freshwater Biology, 58, 88–99. Vasari, Y., Vasari, A., Koli, L. 1963. Purkuputaanlampi, a calcareous mud series from Kuusamo, North East Finland. Archivum Societatis Zoologicae Botanicae Fennicae Vanamo, 18, 96–104. Wagner, C.E., Harmon, L.J., Seehausen, O. 2014a. Cichlid species-area relationships are shaped by adaptive radiations that scale with area. Ecology Letters, 17, 583–592. Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M.J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J.H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29. Walsh, C., Mac Nally, R. 2013. hier.part: Hierarchical Partitioning. R package version 1.0-4. http://CRAN.R-project.org/package=hier.part. Welter-Schultes, F.W. 2012. European non-marine molluscs, a guide for species identification. Planet Poster Editions, Göttingen. Wilke, T., Albrecht, C., Anistratenko, V.V., Sahin, S.K., Yildirim, M.Z. 2007. Testing biogeographical hypotheses in space and time: faunal relationships of the putative ancient Lake Eğirdir in Asia Minor. Journal of Biogeography, 34, 1807–1821. Zhao, S., Fang, J., Peng, C., Tang, Z., Piao, S. 2006. Patterns of fish species richness in China’s lakes. Global Ecology and Biogeography, 15, 386–394.

3.8. Supplementary material

3.8.1. Appendix S3.1

To test for the influence of habitat availability and endemic species we included in the current Appendix further analyses carried out in two different datasets. The results of these were subsequently compared to the original dataset of the manuscript (OD). The first set (Set 1) is

56 almost identical to the OD, but surface area was replaced by lake perimeter. Surface area and lake perimeter were never included in the same model because of high collinearity (Table S3.1.1). Lake perimeter was reconstructed like surface area in Google™ Earth 7 and was used as a surrogate for habitat availability (see Dehling et al., 2010 and references therein). In the second set (Set 2) the predictor variables are the same as the OD, but here single lake endemics (SLE) were removed in order to evaluate the effect of endemism on the differences between the lake groups. For both sets we implemented stepwise multiple regressions (Tables S3.1.2, S3.1.3) and evaluated the spatial autocorrelation of the regressions’ residuals using Moran`s I index as for the OD. For each lake group, the resulting regression models of the two sets were compared to the model of the OD using ANOVA. Additionally, the beta diversity analysis was repeated for Set 2 in order to test how endemic species affect species composition (Table S3.1.4) Regarding both Sets low or no spatial autocorrelation was observed in the models’ residuals (Fig. S3.1.1, S3.1.2). The comparison of Set1 with the OD revealed that none of models differs significantly from the original one except for the model for LG4 (F = 10.841, P = 0.001). In all models, the contribution of perimeter was similar to that of surface area and was prominent only in LG4 (Table S3.1.2). This provides indication that habitat availably is an insufficient predictor of lacustrine gastropod richness suggesting that lake faunas in a pan- European scale and within the formerly glaciated areas are not in equilibrium (see also Dehling et al., 2010). However, faunal evolution seems to be stronger linked to habitat availability in lakes not directly affected by glaciations (LG4). This is likely affected by the higher longevity of some of these lakes (e.g. Balkan lakes). To disentangle, however, the faunal patterns within LG4 additional information are required. The comparison of the OD with Set 2 did not yield significant differences and thus we conclude that the presence of the SLE does not alter the patterns of species richness. Accordingly, differences in species composition were not strongly influenced by the exclusion of the SLE but the pairwise beta diversity was lower (compare Table 3 and Table S3.1.4). The balance between species loss and species turnover between the LGs remained the same indicating that the patterns of beta diversity of European lacustrine gastropods are independent of SLE.

References

57

Baselga, A. 2012. The relationship between species replacement, dissimilarity derived from nestedness, and nestedness. Global Ecology and Biogeography, 21, 1223–1232. Dehling, D.M., Hof, C., Brändle, M., Brandl, R. 2010. Habitat availability does not explain the species richness patterns of European lentic and lotic freshwater animals. Journal of Biogeography, 37, 1919–1926.

Table S3.1.1 Multicollinearity test results for the complete dataset and the four lake groups (LGs) 1–4 for Set 1. Variation inflation factor greater than 10 indicates increased multicollinearity. Lat Long Alt Area Perim Temp Prec Isol Complete dataset 2.95 1.45 1.59 36.31 36.84 2.86 1.41 1.14 LG1 3.06 3.23 1.41 40.93 41.26 3.02 2.57 1.24 LG2 7.05 2.39 1.41 35.81 36.15 7.28 4.27 1.14 LG3 2.17 2.82 5.35 103.14 98.20 4.68 1.67 1.47 LG4 2.72 1.51 2.50 36.87 36.64 4.16 1.12 1.32 Lat, latitude (decimal degrees), Long, longitude (decimal degrees), Alt, altitude (metres), Perim, lake perimeter (km), Temp, mean annual temperature (°C), Prec, annual precipitation (mm) and Isol, Isolation (km).

58

Table S3.1.2 Predictor variables affecting gastropod species richness of European lakes as proposed by stepwise 2 2 multiple regressions (MR) for Set 1. Shown are the adjusted R and P-values (P) of the MRs (R adj), Akaike’s information criterion (AIC), the difference between initial and final Akaike’s information criterion (ΔAIC), the estimate of the multiple regression (Est. MR), significance of the variables in the multiple regression (P MR) and the independent contribution for each variable as provided by hierarchical partitioning results (Ind.). 2 R adj P Lowest AIC ΔAIC Variables Est. MR P MR Ind. Complete dataset 0.2679 <0.0001 -1920.02 1.99 Lat -0.4053 0.1363 0.031 Long 0.0896 0.0677 0.014 Perim 0.1628 <0.0001 0.095 Temp 0.5819 <0.0001 0.060 Prec -0.7413 <0.0001 0.068 Isol -0.0960 0.0027 0.006 LG1 0.3248 <0.0001 -530.8 5.37 Lat -5.0253 <0.0001 0.095 Alt -0.1800 <0.0001 0.049 Perim 0.0412 0.144 0.007 Prec -0.9181 <0.0001 0.187 LG2 0.5448 <0.0001 -811.3 0 Lat -6.5865 <0.0001 0.155 Long 0.2889 <0.0001 0.058 Alt -0.1060 0.0116 0.014 Perim 0.1868 <0.0001 0.090 Temp 0.4459 0.1263 0.065 Prec -0.6920 0.0017 0.143 Isol -0.1487 0.0081 0.029 LG3 0.5852 <0.0001 -161.58 4.7 Lat 9.8998 0.0183 0.043 Perim 0.0751 0.1362 0.132 Temp 1.2999 <0.0001 0.328 Prec -0.9282 0.0958 0.111 LG4 0.3637 <0.0001 -752.81 5.92 Lat 2.7703 <0.0001 0.065 Perim 0.2837 <0.0001 0.303 Lat, latitude (decimal degrees), Long, longitude (decimal degrees), Alt, altitude (metres), Perim, lake perimeter (km), Temp, mean annual temperature (°C), Prec, annual precipitation (mm) and Isol, Isolation (km).

59

Table S3.1.3 Predictor variables affecting gastropod species richness of European lakes as proposed by stepwise 2 2 multiple regressions (MR) for Set 2. Shown are the adjusted R and P-values (P) of the MRs (R adj), Akaike’s information criterion (AIC), the difference between initial and final Akaike’s information criterion (ΔAIC), the estimate of the multiple regression (Est. MR), significance of the variables in the multiple regression (P MR) and the independent contribution for each variable as provided by hierarchical partitioning results (Ind.). 2 R adj P Lowest AIC ΔAIC Variables Est. MR P MR Ind. Complete dataset 0.2789 <0.0001 -1954.63 3.15 Long 0.0725 0.129 0.014 Area 0.0994 <0.0001 0.105 Temp 0.6171 <0.0001 0.083 Prec -0.7854 <0.0001 0.076 Isol -0.1072 0.0004 0.006 LG1 0.3236 <0.0001 -530.46 5.35 Lat -4.9786 <0.0001 0.095 Alt -0.1790 <0.0001 0.049 Area 0.0245 0.141 0.006 Prec -0.9223 <0.0001 0.187 LG2 0.5505 <0.0001 -817.33 0.22 Lat -8.3529 <0.0001 0.231 Long 0.3035 <0.0001 0.052 Alt -0.1302 0.0004 0.017 Area 0.1146 <0.0001 0.093 Prec -0.4331 0.0109 0.134 Isol -1656 0.0030 0.032 LG3 0.5765 <0.0001 -161.65 4.67 Lat 10.7108 0.0112 0.05 Temp 1.4717 <0.0001 0.410 Prec -1.0605 0.0572 0.139 LG4 0.3856 -785.18 7.16 Lat 3.4185 <0.0001 0.085 Area 0.1503 <0.0001 0.297 Temp 0.2855 0.153 0.010 Lat, latitude (decimal degrees), Long, longitude (decimal degrees), Alt, altitude (metres), Area, lake surface area (km2), Temp, mean annual temperature (°C), Prec, annual precipitation (mm) and Isol, Isolation (km).

Table S3.1.4 Pairwise beta diversity of the lake groups (LGs) 1–4 computed with the Jaccard dissimilarity measure (Baselga, 2012) when single lake endemics are excluded. βjac is the overall beta diversity, βjtu and βjne are the dissimilarity components referring to species replacement and species loss, respectively.

Groups βjac βjtu βjne LG1–LG2 0.41 0 0.41 LG1–LG3 0.44 0.31 0.13 LG1–LG4 0.62 0.23 0.39 LG2–LG3 0.39 0.22 0.17 LG2–LG4 0.46 0.33 0.13 LG3–LG4 0.50 0.18 0.32

60

Figure S3.1.1 Moran’s I correlograms for the residual of multiple regression models for the complete dataset and lake groups (LGs) 1–4 for Set 1. Equal number of pairs in default distance classes is selected. Geographic distances are measured in kilometres.

61

Figure S3.1.2 Moran’s I correlograms for the residual of multiple regression models for the complete dataset and lake groups (LGs) 1–4 for Set 2. Equal number of pairs in default distance classes is selected. Geographic distances are measured in kilometres.

62

3.8.2. Supplementary figures and tables

63

Figure S3.1. Moran’s I correlograms for the residuals of multiple regression models for the complete dataset, lake groups (LGs) 1–4 and linear models (LMs) 1 and 3. Equal number of pairs in default distance classes is selected. Geographic distances are measured in kilometres. Note that the correlogram for the residuals of the multiple regression model for the complete dataset corresponds also to the correlogram for the residuals of LM2.

Complete dataset

64

LG1

65

LG2

66

LG3

67

LG4

Figure S3.2. Univariate relationships between species richness and seven predictor variables, i.e., area, perimeter, longitude, latitude, altitude, isolation, annual precipitation and annual mean temperature, for the complete dataset (LM2) and lake groups (LGs) 1–4. All variables were log10-transformed to improve normality.

68

Table S3.1. Distribution of species in the studied European lakes, the predefined lake groups (LGs) and the physiographic and climatic data for each lake. Species presences are marked with 1, latitude in decimal degrees, longitude in decimal degrees, altitude in metres, lake surface area in km2, lake perimeter in km, mean annual temperature in °C, annual precipitation in mm and lake isolation in km. For all analyses data where log10- transformed. References used to extract the species information are listed in alphabetical order. (Due to size matters this table is available only at http://link.springer.com/article/10.1007/s10750-016-2713- y#SupplementaryMaterial)

Table S3.2 Pairwise comparisons between average species richness of the four lake groups (LGs 1–4). Results of the Dunnett–Tukey–Kramer pairwise multiple comparison post hoc tests are adjusted for unequal variances and sample sizes. Difference represents the difference of the means for each pair. All pairwise comparisons of LGs show significant difference (P–adjusted < 0.01) in species richness except for the pair LG3–LG2. Values between parentheses indicate differences in species richness between the LGs when single lake endemics were removed. Significance remained unchanged. Difference P–adjusted LG2 – LG1 4.71 (4.68) <0.0001 LG3 – LG1 6.20 (6.14) <0.0001 LG4 – LG1 3.03 (2.59) <0.0001 LG3 – LG2 1.49 (1.46) n.s. LG4 – LG2 -1.68 (-2.09) <0.01 LG4 – LG3 -3.17 (-3.55) <0.01 n.s. not significant at a P < 0.05 level

Table S3.3 Multicollinearity test results for the complete dataset and the four lake groups (LGs) 1–4. Variation inflation factor greater than 10 indicates increased multicollinearity. Lat Long Alt Area Temp Prec Isol Complete dataset 2.90 1.44 1.58 1.13 2.84 1.41 1.12 LG1 3.06 3.08 1.37 1.10 3.01 2.57 1.24 LG2 7.04 2.38 1.40 1.28 7.05 4.15 1.09 LG3 2.00 2.47 5.20 1.86 4.46 1.57 1.29 LG4 2.71 1.49 2.48 1.56 4.15 1.11 1.31 Lat, latitude (decimal degrees), Long, longitude (decimal degrees), Alt, altitude (metres), Area, lake surface area (km2), Temp, mean annual temperature (°C), Prec, annual precipitation (mm) and Isol, Isolation (km).

69

Table S3.4 Results of the linear models (LMs) 1 and 3. The first model (LM1) included only the lake groups, i.e., a categorical variable with four levels of deglaciation history (LG1 to LG4). The third model (LM3) included a combination of the abiotic predictors and the lake groups. Shown are the adjusted R2 and P-values 2 (P) of the LMs (R adj), Akaike’s information criterion (AIC), the difference between initial and final Akaike’s information criterion (ΔAIC), the estimate of the LM (Est.) and significance of the variables in the LM (P LM). 2 R adj P Lowest AIC ΔAIC Variables Est. P LM LM1 0.0816 <0.0001 LG2 0.2687 <0.0001 LG3 0.3965 <0.0001 LG4 0.2196 <0.0001 LM3 0.4547 <0.0001 -2170.77 1.88 LG2 -2.5080 0.0057 LG3 -3.5879 0.1634 LG4 -1.3412 0.2208 Long -0.4067 0.0930 Area 0.0243 0.1778 Temp 0.3714 <0.0001 Prec -1.2145 <0.0001 Isol -0.0947 0.0014 LG2*Long 0.5752 0.0213 LG3*Long 1.1664 0.0684 LG4*Long 0.2568 0.3177 LG2*Area 0.0911 0.0001 LG3*Area 0.0235 0.5546 LG4*Area 0.1317 <0.0001 LG2*Temp 1.2524 <0.0001 LG3*Temp 0.9282 0.0057 LG2*Temp -0.9369 <0.0001 LG2*Prec -0.1893 0.3858 LG3*Prec 0.2322 0.7412 LG4*Prec 1.1032 <0.0001 Long, longitude (decimal degrees), Area, lake surface area (km2), Temp, mean annual temperature (°C), Prec, annual precipitation (mm), Isol, Isolation (km) and LG, lake group.

70

Table S3.5 Summary statistics of eightphysiographic and climatic factors for lakes in the complete dataset and the four lake groups (LGs) 1–4 in Europe. Completedataset LG1 LG2 LG3 LG4 Mean (Range) Mean (Range) Mean (Range) Mean (Range) Mean (Range) Lat 52.13 (36.50–70.47) 62.70 (56.14–70.47) 56.74 (51.83–70.41) 46.97 (44.42–48.07) 46.79 (36.50–54.08) Long 14.10(-9.59–37.58) 14.56(-4.63–31.73) 12.89(-9.59–36.13) 11.06 (5.86–15.08) 15.72 (-8.92–37.59) Alt 252.73 (-4–2450) 217.68 (4–1117) 81.96 (-4–423) 788.67 (62–2362) 362.82 (-4–2450) Area 66.01 (0.000085–17640.38) 50.78 (0.003–5580.32) 126.66 (0.0006–17640.38) 37.80 (0.002–578.78) 14.51 (0.000085–660.01) Perim 32.80 (0.04–2032.89) 45.43 (0.30–1364.48) 46.30 (0.09–2032.89) 26.02 (0.178–286.28) 10.53 (0.04–244.24) Temp 7.05 (-3–18.6) 3.51 (-3–8.6) 6.64 (-2–9.6) 7.22 (-0.8–13) 9.89 (0.1–18.6) Prec 815.94 (352–2420) 874.79 (400–2420) 840.14 (447–2080) 1188.14 (839–1559) 677.27 (352–1678) Isol 3.42 (0–88.75) 2.37 (0–17.76) 1.97 (0–31.24) 6 (0.14–31.23) 5.25 (0.01–88.75) Lat, latitude (decimal degrees), Long, longitude (decimal degrees), Alt, altitude (metres), Area, lake surface area (km2), Perim, lake perimeter (km), Temp, mean annual temperature (°C),Prec, annual precipitation (mm)and Isol, Isolation (km).

71

CHAPTER 4: Beginning of a new age: How did freshwater gastropods respond to the Quaternary climate change in Europe?

Elisavet Georgopoulou 1, 2, Thomas A. Neubauer 1, Mathias Harzhauser 1, Andreas Kroh 1, Oleg Mandic 1

1Geological-Paleontological Department, Natural History Museum Vienna, Burgring 7, 1010 Vienna, Austria 2Institute of Earth Sciences, University of Graz, Heinrichstrasse 26, 8010 Graz, Austria

4.1. Abstract

While the fossil freshwater gastropods of the European Quaternary are relatively well documented, biogeographical studies on a continental scale are lacking. Here we explored the spatial and temporal patterns of freshwater gastropods from six distinct time intervals of the Quaternary, i.e., Gelasian, Calabrian, Ionian, Last Interglacial, Last Glacial and Holocene. We investigated for differences in species richness and calculated turnover rates among the predefined time intervals. In order to test for the magnitude of the Plio-Pleistocene turnover, we included Late Pliocene records for that purpose as well. In addition, we quantified nestedness of species geographical ranges using the nestedness metric based on overlap and decreasing fill (NODF). We counted 398 freshwater gastropods from 1129 localities across Europe throughout the Quaternary. Among the studied time intervals, species richness differed significantly, mainly because of the intervals’ duration. For the Early Pleistocene, increased species richness is related to highly endemic aquatic systems like the long-lived lakes Bresse and Tiberino. The rich Middle to Late Pleistocene and Holocene assemblages are mostly linked to fluvial and/or lacustrine systems with short temporal durations. The decreasing turnover rate and increasing nestedness over time demonstrate the gradual loss of provincialism during the Quaternary. The Plio-Pleistocene boundary coincided with a major turnover event, related to the demise of long-lived lakes and their rich, endemic faunas at the end of the Pliocene. For subsequent intervals, little or no turnover was observed. At the same time, nestedness increased paralleling the retreat of long-lived lakes and a declining speciation rate. For the Late Pleistocene and Holocene, the increase of nestedness indicates the prevalence of generalist species with wide geographical distributions. The Holocene

72 patterns of species’ geographical ranges can be partly attributed to repeated cycles of glaciation and deglaciation in the late Quaternary.

Keywords: Pleistocene, Holocene, species richness, species turnover, nestedness, range- through approach, geographical ranges

4.2. Introduction

The Quaternary Period (2.588 to 0 myr; Gibbard et al., 2010), short in a geological context, is characterized by successive glacial episodes and distinctly lower temperatures compared to the preceding Neogene Period (see Lisiecki & Raymo, 2005). During this time, Europe attained its present shape although extensive changes in shorelines and inland waters occurred mainly because of glacio-isostatic movements and tectonic activity (e.g., Zagwijn, 1989; Gábris & Nádor, 2007; Stocchi et al., 2005; Antonioli et al., 2009). Changes in species diversity patterns during the Quaternary are usually affiliated with historical features like the recurrent ice ages (e.g., Hewitt, 1999; Dynesius and Jansson, 2000; Jansson & Dynesius, 2002, Colwell & Rangel, 2010) and in particular the effect of the Last Ice Ages (e.g., Hewitt, 1999, 2000). Especially the effect of palaeoenvironmental conditions on diversity and distribution of large mammal communities during the Early and Middle Pleistocene (Kahlke et al., 2011) and the link between humans and the extinction of the megafauna in the Late Quaternary has received considerable attention (Sandom et al., 2014 and references therein). Quaternary ice house conditions affected different groups in different ways as, for example, shown by European insect faunas, which do not show high extinction rates during the Quaternary but rather display shifts in their geographical ranges (Coope, 1994). Similarly, changes in species geographical ranges, i.e. changes in local species richness and composition, have been recorded for the Late Pleistocene and Holocene land snail faunas of north-western Europe (Limondin-Lozouet & Preece, 2014). As shown by Neubauer et al., (2015b), the freshwater gastropods of Europe experienced a major turnover and diversity severely declined at the Plio-Pleistocene boundary. Deteriorating climate and the loss of long-lived lakes are recognized as main explanations of this event (Neubauer et al., 2015b). While overall trends as well as the distribution of freshwater gastropods during the Quaternary are reasonably documented on a regional scale (e.g., Kennard & Woodward, 1917, 1922; Ložek, 1964; Settepassi & Verdel, 1965; Esu & Girotti, 1975; Mania, 1973; Alexandrowicz, 1999; Sanko, 2007), a detailed breakdown of trends in freshwater gastropod

73 biodiversity and biogeography is entirely missing on a larger scale. Beyond that, a limited number of articles deal with regional biogeographical patterns, time intervals or taxa, focusing mostly on the terrestrial gastropods (e.g., Meijer & Preece, 1996; Limondin-Lozouet & Preece, 2014). In contrast, the biogeographical relationships of Neogene lacustrine gastropods on a pan-European scale have recently been investigated (Harzhauser & Mandic, 2008; Neubauer et al., 2015a). These works report a trend towards increasing provincialism during the Neogene, resulting from the formation of large, long-lived lakes with highly endemic faunas. By the onset of the Pleistocene, however, nearly all of these major hotspots had vanished. A recent investigation of modern gastropod distribution in European lakes indicates that lacustrine snails still carry the imprint of the Last Ice Ages (Georgopoulou et al., 2016). Therefore, we seek to demonstrate diversity trends, temporal turnover events and biogeographical patterns of the European freshwater gastropod fauna throughout the Pleistocene and Holocene, i.e., the last 2.6 myr, and so provide the missing link between the well-resolved patterns of Neogene and modern freshwater biogeography. For the current work, a comprehensive dataset of the occurrences of freshwater gastropods of the European Quaternary was compiled. We document changes in species richness and composition among predefined time intervals. In addition, the geographical ranges of species are calculated and examined for potential nested patterns over time. This is the first detailed study of Quaternary biogeography of freshwater faunas on a continental scale.

4.3. Methods

4.3.1. Data evaluation and temporal subdivision

Occurrence data of freshwater gastropods of the European Quaternary were obtained after a comprehensive literature search (see Table S4.1). Prior to any analyses, species names were validated using updated scientific sources (Glöer, 2002; de Jong, 2012; Welter-Schultes, 2012; Neubauer et al., 2014) in order to avoid taxonomical inconsistences (i.e., different opinions on synonymy, taxonomic level and spellings). The geographical frame is the European continent extending as far as the eastern shores of the Caspian Sea and the Volga Upland. The preservation issues (especially because of glacial erosion) however, constrained data availability in certain regions like Scandinavia. The dataset was divided into six time intervals (TIs hereafter): Gelasian (2.588–1.806 myr),

74

Calabrian (1.806–0.781 myr), Ionian (Middle Pleistocene; 0.781–0.126 myr), Last Interglacial (0.126–0.115 myr), Last Glacial (0.115–0.0117 myr) and Holocene (0.0117–0 myr). These divisions were chosen to fit standard stratigraphic schemes, facilitate comparability and to pinpoint the main glacial event of the Late Pleistocene (i.e., the Last Glacial). All localities extracted from the literature were assigned to one of the six TIs based on their stratigraphic age as attributed by Georgopoulou et al. (2015) (see Table S4.2). For localities with age ranges crossing a TI’s age boundary, attribution to a TI is based on the longer partial interval. Localities with a too coarse stratigraphic attribution (e.g., Early Pleistocene) were excluded from the analyses.

4.3.2. Species richness

To correct for oversampling for each TI, we created 100×100 km equal-area grid cells in

ArcGIS 10 (Esri Inc., 1999–2010) in Behrmann projection. For each TI, the number of species was calculated for each grid cell including localities. As suggested by Kreft & Jetz (2010), the selected grid size reflects the explicit character of species distribution records and maintains their spatial accuracy, while sampling errors related to variation in area sizes are avoided. To assess the effect of a potential collection bias in the dataset, we performed a correlation test between total species richness and number of grid cells with occurrence data per TI. The null hypothesis is that differences in species richness between the TIs are not a result of sampling intensity. To evaluate the effect of time a second correlation test between the TIs’ total species richness and their temporal duration was performed. All correlations were assessed using non-parametric Spearman’s rank test. Differences in average species richness between the defined TIs were investigated using ANOVA with a Dunnett–Tukey– Kramer pairwise multiple comparison post hoc test adjusted for unequal variances and sample sizes.

4.3.3. Species temporal turnover and range-through assumption

Species turnover among TIs was modelled with the βjtu dissimilarity measure following Baselga (2012). This index was selected because it measures species replacement (i.e., turnover) independently from differences in species richness (Baselga, 2012). To evaluate the turnover rate at the Pliocene–Pleistocene boundary, we included the Late Pliocene (Piacenzian) European freshwater gastropod species from the dataset of Neubauer et al. (2015b). In order to correct for missing records, we adopted the range-through assumption,

75 i.e., all species are assumed to have continuously existed from their first to last recorded occurrence (Hammer & Harper, 2006). To avoid biasing the real first and last appearance dates of the species by using only Quaternary records, we completed the current data by integrating Neogene records based on the dataset of Neubauer et al. (2015b) and by including distribution records of modern species (Georgopoulou et al., 2016). The analysis was repeated a second time excluding singletons, i.e., species present in only one TI, in order to evaluate their effect on the temporal turnover. All analyses were carried out in R 3.1.1 (R Development Core Team 2014) using packages DTK (Lau, 2013) and betapart (Baselga et al., 2013).

4.3.4. Range sizes and nestedness

We investigated the geographical range patterns of the species throughout the Quaternary using nestedness analysis. Nestedness, i.e., that smaller assemblages are proper subsets of larger assemblages, is a commonly recognised pattern in natural systems (Almeida-Neto et al., 2008). The assessment of species geographical ranges using nestedness analysis may significantly help to disentangle the distribution patterns of the freshwater gastropods of the European Quaternary.

Each species’ geographical range was measured in ArcGIS 10 (Esri Inc., 1999–2010) using the Minimum Bounding Geometry Tool. A species geographical range was defined as the enclosed area of the polygon delineated by the species presences. To measure nestedness a presence-absence matrix is required, where sites represent rows and species represent columns (see Almeida-Neto et al., 2008). Nestedness was measured with the “nestedness metric based on overlap and decreasing fill” (NODF), which is related to matrix fill but insensitive to matrix shape and size (Almeida-Neto et al., 2008). The significance of the observed nestedness for each TI was assessed using the standardized measure Z. Z-values were calculated from 100 null matrices where the probability of an assigned presence is proportional to the corresponding row and column totals of the original matrix (Strona et al., 2014). For Z-values higher than 1.64 there is significant nestedness at a 95% confidence interval (see Strona et al., 2014). Nestedness analysis was implemented with ND (Strona & Fattorini, 2014). Species range sizes were plotted against the cumulative number of species in order to observe if the proportion of restricted to widespread species ranges changes throughout the

76

Quaternary. Non-parametric Spearman’s rank test was used to explore whether the number of first and last occurrences and TI duration may account for nestedness across the TIs.

4.4. Results

In total, we included the records of 398 species and subspecies, classified into 78 genera and 16 families, from 1129 localities of the European Quaternary (Fig. 4.1, Table 4.1). The number of grid cells with occurrence data varied between 28 in the Gelasian and 160 in the Holocene (Fig. 4.2). No significant collection bias was detected between the TIs, as the numbers of grid cells per TI was not significantly correlated with species richness (rho = - 0.42, P = 0.4). Actual species richness and richness based on the range-through assumption were higher in TIs with longer duration (rho = 0.94, P = 0.005 and rho = 0.86, P = 0.02 respectively). ANOVA revealed significant differences in average species richness across the defined Quaternary time intervals, but not for all pairwise comparisons (ANOVA

F5,421=5.063; P < 0.0001, Table S4.3). In particular, average species richness which was highest during the Last Interglacial, dropped significantly to the Last Glacial and increased markedly to the Holocene (Table S4.3).

TABLE 4.1. Summary statistics for the 1129 Quaternary localities. Numbers of singletons, i.e. species that occur only in one time interval, for the range-through approach are given in parentheses. For the complete dataset see Table S4.1. Holocene Last Glacial Last Interglacial Ionian Calabrian Gelasian

Species richness 146 63 59 150 162 152

Genus richness 54 34 33 51 55 41

Range-through richness 175 (47) 139 (3) 132 (0) 189 (29) 235 (12) 238 (40)

No. of localities 487 107 86 189 168 93

No. of occupied grid cells 158 58 50 84 49 28

Mean grid cell richness 13.10 7.88 13.80 12.52 13.06 8.00

Interval duration (myr) 0.0117 0.1033 0.011 0.655 1.025 0.782

No. of first occurrences 47 7 9 42 55 80

No. of last occurrences 175 0 11 66 88 40

77

Figure 4.1. Localities with Quaternary freshwater gastropods for the six time intervals Gelasian, Calabrian, Ionian, Last Interglacial, Last Glacial and Holocene.

Dissimilarity due to species replacement (βjtu) varied from 0 between the Last Interglacial and the Last Glacial to 0.91 between the Piacenzian and the Holocene (Table 4.2). The lack of species replacement between the Last Interglacial and Last Glacial reflects the high similarity in species composition of the two TIs. Considering consecutive TIs, species turnover was exceptionally high between the Piacenzian and the Gelasian (0.82), followed by the Gelasian–Calabrian (0.38) and Calabrian–Ionian (0.36), while it remained low for the remaining pairs, i.e., Ionian–Last Interglacial (0.13) and Last Glacial–Holocene (0.15). The exclusion of singletons lowered the compositional dissimilarity between the pairwise comparisons and average βjtu dropped from 0.52 ± 0.30 to 0.35 ± 0.26.

78

TABLE 4.2. Species turnover rates between the six Quaternary TIs and the Piacenzian, computed with the βjtu index of species turnover. Numbers in parentheses refer to turnover rates with singletons removed. Turnover rates between stratigraphically consecutive TIs are shown in bold. Turnover rates between Quaternary TIs are marked with grey. Holocene Last Glacial Last Interglacial Ionian Calabrian Gelasian

Last Glacial 0.15 (0)

Last Interglacial 0.11 (0.06) 0 (0)

Ionian 0.51 (0.18) 0.21 (0.17) 0.13 (0.13)

Calabrian 0.58 (0.33) 0.35 (0.32) 0.29 (0.29) 0.36 (0.15)

Gelasian 0.71 (0.55) 0.59 (0.58) 0.56 (0.56) 0.62 (0.52) 0.38 (0.17)

Piacenzian 0.91 (0.75) 0.89 (0.75) 0.88 (0.75) 0.90 (0.67) 0.87 (0.40) 0.82 (0)

79

Figure 4.2. Maps showing the geographical distribution of species richness for the six TIs based on 100×100 km grids (Behrmann projection). Species richness per grid cell is colour-coded, with equal scale across all figures (based on the grid cell richness of the Calabrian). The number indicated by the arrow in the richness bar corresponds to the maximum species richness per grid cell of the respective TI.

Species ranges in all TIs were significantly nested (Table 4.3, Fig. S4.1). According to the NODF metric, the Last Interglacial was the most nested TI (NODF = 90.395), followed by the Last Glacial (NODF = 82.926), the Holocene (NODF = 72.357), the Ionian (NODF = 53.028), the Calabrian (NODF = 51.755) and the Gelasian (NODF = 22.42). Small-ranged species prevail in the Early Pleistocene (Gelasian and Calabrian) in contrast to the Late Pleistocene (Last Interglacial and Last Glacial), where most of the species are wide-ranged (Fig. 4.3). NODF was negatively correlated with the number of first occurrences (rho = -0.89, P = 0.02) and TI duration (rho = -0.89, P = 0.02), indicating that species geographical ranges are less nested in TIs with increased speciation events or longer duration. There was no correlation between NODF and the number of last occurrences. Richness, turnover and nestedness trends can be seen in Fig. 4.4.

80

TABLE 4.3. Summary results of nestedness and summary statistics of species ranges for the six TIs. NODF, measure of total matrix nestedness; Z, Z-score; Fill (%), percentage of matrix fill; mean SR, mean species ranges in km2. All Z-values were highly significant (P<0.001). TI NODF Z Fill (%) Mean SR

Holocene 72.357 60.363 34.5 949278

Last Glacial 82.926 19.567 54.6 506674

Last Interglacial 90.395 17.286 73.2 498896

Ionian 53.028 28.096 29.8 550727

Calabrian 51.755 44.893 23.4 215226

Gelasian 22.42 12.406 11.3 36752

4.5. Discussion

Following the latest phase of the Pliocene with ca. 450 species (after Neubauer et al., 2015b), European freshwater gastropod diversity continuously declined through the Pleistocene. With the onset of the Holocene, diversity again recovered and approached its present magnitude of over 500 lacustrine species (Neubauer et al., 2015b). The patterns of species richness and turnover and related processes in the intermediate Pleistocene and Holocene intervals have not been examined in detail as yet.

4.5.1. Patterns of species richness

The diversity of species during the Quaternary differed significantly, mainly as a result of TI duration. The majority of the Gelasian species was connected to a handful of diverse and often highly endemic aquatic systems, e.g., the fluvio-lacustrine Bresse Basin in south- eastern France with 45 species, the East Anglian Crag in eastern England with 21 species, Lake Kos in Greece with 41 species and Lake Tiberino in central Italy with 39 species (e.g., see Schlickum & Puisségur, 1978; Willmann, 1981; Neubauer et al., 2015b). The species-rich lakes Bresse and Tiberino persisted into the Calabrian (Schlickum & Puisségur, 1977, 1978; Esu & Girotti, 1975) and qualify as long-lived lakes (following Gorthner, 1992). Additionally, during the Calabrian new centres of diversity evolved in the Hungarian fluvial

81 basins and fluvio-lacustrine systems in south-western Germany and the Netherlands. On the Iberian Peninsula and in eastern Europe, only small aquatic bodies with less diverse faunas were present during the Early Pleistocene. Throughout the Middle Pleistocene (Ionian), Late Pleistocene and Holocene, local richness was mostly linked to smaller water bodies with short temporal durations. Examples of such fluvial and/or lacustrine systems with highly diverse fauna for the Middle Pleistocene can be found in central-eastern England (Kennard & Woodward, 1922; Horton et al., 1992) and south-western Germany (Bibus & Rähle, 2003). Beyond that, several long-lived lakes existed at that time, like Lake Ohrid in the Balkans (Wagner et al., 2014), the Black Sea (Wesselingh, 2007) and the Caspian Sea (Dumont, 1998), for which information on freshwater gastropod faunas is still scarce prior to the Holocene. Lake Lirino in central Italy may also qualify as a long-lived lake given its quite diverse fauna consisting of 33 species (Settepassi & Verdel, 1965) and the well-constrained temporal duration of ca. 200 kyr (Manzi et al., 2010). Despite the continuous climatic deterioration, an expansion of the climatic periodicity from 41 kyr to 100 kyr between 0.9–0.4 Ma resulted in longer warm (interglacial) or cold (glacial) intervals of climatic stability up to the Late Pleistocene (Lisiecki & Raymo, 2005). Thus, the climatic stability may partly explain the increased species diversity (Jansson & Dynesius, 2002 and references therein), which in our case was observed during the Middle Pleistocene. Exceptionally rich assemblages during the Last Interglacial were found in central Germany (Mania, 1973), southern and eastern England (Sparks, 1959, 1964; West & Sparks, 1960), and Belarus (Sanko, 2007). It should, however, be noted that data for the Last Interglacial was only available for a restricted latitudinal zone, which included parts of central and northern Europe. Despite the increased average grid richness (see Table 4.1), the Last Interglacial was characterised by the lowest species diversity of all TIs. During the Last Glacial, both species diversity and local richness were restricted. The fossil deposits of this TI indicate the existence of small, short-lived water bodies inhabited by few but widespread species (e.g., Hoek et al., 1999; Apolinarska & Ciszewska, 2006; Kossler & Strahl, 2011; Sümegi et al., 2011). In addition, temperature had considerably declined (see Lisiecki & Raymo, 2005) and the advancing ice sheets from the middle Weichselian to the Last Glacial Maximum had covered northern Europe (Mangerud et al., 2004). Local richness maxima were detected above or near the permafrost line (for the permafrost limits see Frenzel et al., 1992), while the few available southern localities were much poorer. Although fossil data of freshwater gastropods from the southern regions of Europe is incomplete, the

82 existence of poor southern localities suggests that local richness is not a function of temperature alone (see Fig. 4.2 and Table S4.2). As shown by a series of molecular analyses (e.g., Cordellier & Pfenninger, 2008, 2009; Benke et al., 2009), refugia for freshwater gastropods existed during the Late Pleistocene glaciations in northern and central Europe. During the Holocene, examples of rich gastropod assemblages were associated with freshwater deposits in Belarus (Sanko, 2007), Belgium (Kuijper, 2006) and the Netherlands (Meijer, 1988). The three main reasons for the observed high species richness during the Holocene are A) following the ice sheet retreat, the increased availability of suitable habitats in postglacially developing lakes triggered successive recolonization of formerly glaciated areas (Georgopoulou et al., 2016); B) the resulting increase in speciation rates, mostly in southern European species (see Hewitt, 1999; 2000); and C) the higher preservation potential for the geologically young deposits.

4.5.2. Temporal turnover

A substantial species turnover (0.82) marked the Pliocene–Pleistocene boundary, constituting one of the major tipping points in European freshwater gastropod diversity. Species replacement was entirely controlled by singletons (Table 2), which were mainly associated with long-lived lakes. The strong species turnover at the end of the Pliocene resulted from the demise of large, long-lived and diverse lakes such as Dacia, Slavonia and Transylvania and the disappearance of their highly endemic faunas. A comparable pattern, albeit to a lesser extent, characterised the Gelasian–Calabrian boundary. It corresponds mostly to the diminution of aquatic basins, such as Lake Kos, the East Anglian Crag and Patras Graben (see Doutsos et al., 1988), and the appearance of Calabrian species such as Microcolpia wuesti (Meijer, 1990); a species characteristic of the Bavelian stage (1.2–0.86 Ma) but with no association to such long-lived lakes (Meijer & Preece, 1996). Species turnover between the Calabrian and the Ionian was to a large extent but not solely, caused by the presence of singletons. Typical Calabrian faunas, such as those of the Lakes Bresse and Tiberino, were replaced in the Ionian by faunas of aquatic basins in southern and eastern Europe like Lirino (Liri valley, Italy), Larymna-Malesina (Atalanti basin, Greece), Don (Don river basin, Russia), and the Black Sea, some of which may qualify as long-lived lakes sensu Gorthner (1992). Climate conditions during the Early-Middle Pleistocene transition have been regarded as a potential cause of faunal turnover in western Palaearctic mammals (Kahlke et al., 2011), and may have also contributed to the freshwater gastropod turnover.

83

At the end of the Middle Pleistocene, a number of genera and species typical for the Neogene to early Quaternary disappeared from Europe’s inland aquatic systems, e.g., the genera Neumayria (see Girotti, 1972) and Pseudodianella (see Neubauer et al., 2013) and species like Borysthenia goldfussiana (Wüst, 1901), Lithoglyphus jahni (Urbański, 1975), Tanousia runtoniana (Sandberger, 1880) and T. stenostoma (Nordmann in Madsen and Nordmann, 1901) (see Meijer and Preece, 1996). In addition, few modern taxa first appeared during the ensuing Last Interglacial [e.g., Bythinella austriaca (Frauenfeld, 1857); Benke et al., 2009, 2011]. The rather low turnover rate was unaffected by singletons, reflecting the lack of endemic systems during these periods of time. As some of the abovementioned species have been linked with temperate climatic conditions (see Meijer & Preece, 1996), their disappearance may be linked to the further deteriorating climate. However, despite the fundamental climate changes following the transition into the Last Glacial, there was no species turnover between the Last Interglacial and Last Glacial. Turnover between the Last Glacial and the Holocene in turn was moderate and again strictly related to the presence of singletons. Consequently, climate cannot be regarded as the main driving force for species turnover in the Quaternary. Although climatic fluctuations apparently affected the spatial distribution of modern lacustrine gastropods (Georgopoulou et al., 2016), their effect on temporal turnover seems to be limited. As a general rule, turnover rates are controlled by rare species, mostly related to the existence and demise of long-lived lakes. This implies that there was little temporal turnover in “common aquatic systems”, like rivers or short-lived lakes, which comprise the vast majority of Quaternary freshwater deposits.

4.5.3. Nested patterns of species ranges

All of our matrices were significantly nested, albeit not perfectly ordered, and nestedness generally increased from the Gelasian to the Holocene. Nestedness can be influenced by several mechanisms such as area, local speciation, immigration and palaeogeography (Wright, 1998; Sfenthourakis et al., 1999; Kougioumoutzis et al., 2014; Simaiakis & Strona, 2015). Moreover, because of unfavourable or unstable climatic conditions (see Dynesius & Jansson, 2000 and references therein), distribution ranges of species may change, resulting from speciation, local extirpation or migration (see Gaston & Blackburn, 2000 for discussion).

84

On account of the endemic radiations within the Early Pleistocene long-lived lakes, a prevailing number of geographically limited species occur during the Gelasian. The narrow geographical ranges of species (based on one or few clamped presences) rarely coincide, indicating that on a European scale the biogeographical affinities of freshwater gastropods were low and a common palaeogeographical history is unlikely. This is consistent with other studies suggesting that the absence of a common palaeogeographical history explains a low degree of nestedness (Kougioumoutzis et al., 2014; Simaiakis & Strona, 2015). In addition, the high number of first occurrences, i.e., speciation events, corroborates the low nestedness (see Wright et al., 1998; Sfenthourakis et al., 1999). Nestedness increased during the subsequent TIs. The intermediate degree of nestedness observed for the Calabrian and the Ionian on the one hand results from the relatively high number of speciation events, partly related to long-lived lakes (see above), which decrease nestedness (see Wright et al., 1998; Sfenthourakis et al., 1999; Kougioumoutzis et al., 2014). On the other hand, Calabrian to Ionian species Figure 4.3. Proportion of species’ geographical ranges of have wider ranges than before (see Fig. gastropod species per TI. The vertical axis represents the percentage of the study area covered by a species range. 4.3, Table 4.3), which naturally have Geographical ranges of all TIs were scaled to the greater chances of being nested. Thus, maximum range [Radix labiata (Rossmässler, 1835), Last the negative effect of speciation on Interglacial] to visualize the relative differences in between nestedness is mitigated by the gradual the TIs. expansion of species’ geographical ranges. An exceptionally high degree of nestedness was found for the Last Interglacial, the Last Glacial and the Holocene when species ranges continued to expand (see Fig. 4.3). According to Briers (2003), resource availability and species’ ability to exploit them affect range sizes of freshwater gastropods. The increased proportion of wide-ranged species during the Late Pleistocene suggests that the majority of surviving species were generalists, able to explore a considerable amount of resources. The high degree of nestedness is additionally explained by the low number of first occurrences during the Late Pleistocene (Wright et al.,

85

1998; Sfenthourakis et al., 1999). Particularly regarding the Last Interglacial, range expansion as a response to favourable climatic conditions during interglacial phases (Rousseau, 1989; Limondin-Lozouet & Preece, 2014) is another plausible explanation. During the Holocene, nestedness was lower than in the Late Pleistocene, but still above average. The highly nested patterns are not entirely surprising considering the homogenised distribution of modern lacustrine gastropods resulting from the recolonization of glaciated areas after deglaciation starting around 19 kyr BP (Clark et al., 2009; Neubauer et al., 2015b; Georgopoulou et al., 2016). In that way, the Holocene nested patterns of species ranges mirror the patterns observed for the Last Interglacial, considering the fact that the Holocene is also an interglacial. The lower degree of nestedness for the Holocene may have resulted from the higher number of first occurrences as well as the gradually increasing aridity during the Holocene, which might have limited the geographical ranges of the species, as proposed by Limondin-Lozouet & Preece (2014) for terrestrial gastropods.

4.5.4. Conclusions

The incomplete nature of the fossil record naturally affected our results to some extent. Particularly the massive erosion by advancing and retreating ice sheets constrained the availability of surface outcrops and preserved faunas in northern Europe (Molnar, 2004; Sternai et al., 2012). Nevertheless, the lack of a significant correlation between total species richness and the number of the grid cells per TI (rho = -0.42, P = 0.4) suggests that the patterns observed were not driven by data availability. Moreover, actual species richness and richness based on the range-through assumption were highly correlated (rho = 0.94, P = 0.005), indicating that the dataset is reasonably complete and probably little affected by sampling bias.

86

Our results point towards a gradual loss of the Neogene provincialism during the Quaternary. After the demise of the big lacustrine diversity hotspots at the end of the Pliocene only few, smaller long-lived lakes persist into the Early Pleistocene. The continuous decline of such systems – today only Lake Ohrid and the Caspian Sea exist in Europe– paralleled the decline of endemic species and provincialism per se as well as lowering temporal turnover (see Fig. 4.4). Thus climate should be considered a secondary force in shaping species turnover in the Quaternary. Independent from that, climate change coupled with repeated cycles of glaciation and deglaciation affected the gastropods’ geographical ranges. The nestedness analysis herein indicated, that the Holocene (and present-day) pattern is a result of recolonization of formerly glaciated areas following deglaciation after the Last Glaciation Maximum ca. 19 kyr BP (Clark et al., 2009). Very likely, the climatically Figure 4.4. Richness, turnover and nestedness trends induced local extirpation and following throughout the Quaternary. Richness, turnover and nestedness refer to the species richness based on the range- recolonization was an iterative process through assumption, the values of the βjtu index of species shaping freshwater gastropod turnover (solid black line), the values of the βjtu index of distribution throughout the glacial species turnover after the exclusion of singletons (dotted cycles. The high proportion of species grey line) and the values of the NODF metric, respectively. with wide geographical ranges in the Notice that the species turnover after the exclusion of singletons is very low. Hol, Holocene; LG, Last Glacial; Late Pleistocene and Holocene indicate Lint, Last Interglacial; Ion, Ionian; Cal, Calabrian; Gel, that the majority of surviving species Gelasian. were generalists. The increase of nestedness during the Quaternary and especially its high value for the Last Interglacial corroborate this hypothesis, although the limited temporal resolution and time averaging hampers a more precise breakdown for earlier intervals.

4.6. Acknowlegedements

87

This contribution was financed by the Austrian Science Fund and is part of the FreshGen- project (FWF project no. P25365-B25: “Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients”).

4.7. References

Apolinarska, K., Ciszewska, M. 2006. Late Glacial and Holocene lacustrine molluscs from Wielkopolska (central Poland) and their environmental significance. Acta Geologica Polonica, 56 (1), 51–66. Alexandrowicz, W.P. 1999. Evolution of the malacological assemblages in north Poland during the Late Glacial and Early Holocene. Folia Quaternaria, 70, 39–69. Almeida-Neto, M, Guimarães, P, Guimarães, P.R.Jr, Loyola R.D., Ulrich, W. 2008. A consistent metric for nestedness analysis in ecological systems: reconciling concept and measurement. Oikos, 117, 1227–1239. Antonioli, F., Ferranti, L., Fontana, A., Amorosi, A., Bondesan, A., Braitenberg C., Dutton A., Fontolan, G., Furlani, S., Lambeck, K., Mastronuzzi, G., Monaco, C., Spada, G., Stocchi, P. 2009. Holocene relative sea-level changes and vertical movements along the Italian and Istrian coastlines. Quaternary International, 206, 102-133. Baselga, A. 2012. The relationship between species replacement, dissimilarity derived from nestedness, and nestedness. Global Ecology and Biogeography, 21, 1223–1232. Baselga, A., Orme, D., Villeger, S., De Bortoli, J., Leprieur, F. 2013. betapart: Partitioning beta diversity into turnover and nestedness components. R package version 1.3. http://CRAN.R-project.org/package=betapart. Benke, M., Brändle, M., Albrecht, C., Wilke, T. 2009. Pleistocene phylogeography and phylogenetic concordance in cold-adapted spring snails (Bythinella spp.). Molecular Ecology, 18, 890–903. Benke, M., Brändle, M., Albrecht, C., Wilke, T. 2011. Patterns of freshwater biodiversity in Europe: lessons from the spring snail genus Bythinella. Journal of Biogeography, 38, 2021–2032. Bibus, E., Rähle, W. 2003. Stratigraphische Untersuchungen an molluskenführenden Terrassensedimenten und ihren Deckschichten im mittleren Neckarbecken (Württemberg). Eiszeitalter und Gegenwart, 53, 94–113. Briers, R.A. 2003. Range size and environmental calcium requirements of British freshwater gastropods. Global Ecology & Biogeography, 12, 47–51.

88

Clark, P.U., Dyke, A.S., Shakun, J.D., Carlson, A.E., Clark, J., Wohlfarth, B., Mitrovica, J.X., Hostetler, S.W., McCabe, A.M. 2009. The Last Glacial Maximum. Science, 325 (5941), 710–714. Colwell, R.K., Rangel, T.F. 2010. A stochastic, evolutionary model for range shifts and richness on tropical elevational gradients under Quaternary glacial cycles. Philosophical Transactions of the Royal Society B, 365, 3695–3707. Coope, G.R. 1994. The response of insect faunas to glacial-interglacial climatic fluctuations. Philosophical Transactions of the Royal Society B, 344, 19–26. Cordellier, M., Pfenninger, M. 2008. Climate-driven range dynamics in the freshwater limpet Ancylus fluviatilis (Pulmonata, Basommatophora). Journal of Biogeography 35, 1580– 1592. Cordellier, M., Pfenninger, M. 2009. Inferring the past to predict the future: climate modelling predictions and phylogeography for the freshwater gastropod Radix balthica (Pulmonata, Basommatophora). Molecoluar Ecology, 18, 534–544. Doutsos, T. Kontopoulos, N., Poulimenos, G. 1988. The Corinth-Patras rift as the initial stage of continental fragmentation behind an active island arc (Greece). Basin Research, 1, 177–190. Dumont, H.J. 1998. The Caspian Lake: History, biota, structure, and function. Limnology and Oceanography, 43 (1), 44-52. Dynesius, M., Jansson, R. 2000. Evolutionary consequences of changes in species’ geographical distributions driven by Milankovitch climate oscillations. Proceedings of the National Academy of Sciences of the United States of America, 97, 9115–9120. Esri Inc. 1999–2010. ArcGIS for Desktop: release 10. Environmental Systems Research Institute, Redlands, CA. Esu, D., Girotti, O. 1975. La malacofauna continentale del Plio-Pleistocene dell’Italia centrale. I. Paleontologia. Geologica Romana 13, 203–294. Frenzel, B., Pésci, M., Velichko, A.A. 1992. Atlas of Paleoclimates and Paleoenvironments of the Northern Hemisphere. Late Pleistocene–Holocene. Hungarian Academy of Sciences/G. Fischer, Budapest/Jena. Gábris, G., Nádor, A. 2007. Long-term fluvial archives in Hungary: response of the Danube and Tisza rivers to tectonic movements and climatic changes during the Quaternary: a review and new synthesis. Quaternary Science Reviews, 26, 2758–2782. Gaston, K.J., Blackburn, T.M. 2000. Pattern and process in macroecology. Blackwell Science, Oxford.

89

Georgopoulou, E., Neubauer, T.A., Kroh, A., Harzhauser, M., Mandic, O. 2015. An outline of the European Quaternary localities with freshwater gastropods: Data on geography and updated stratigraphy. Palaeontologia Electronica, 18.3.48A, 1–9. Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016. Distribution patterns of European lacustrine gastropods – a result of environmental factors and deglaciation history. Hydrobiologia, 775, 69–82. Gibbard, P.L., Head, M.J., Walker, M.J.C., and the Subcommission on Quaternary Stratigraphy. 2010. Formal ratification of the Quaternary System/Period and the Pleistocene Series/Epoch with a base at 2.58 Ma. Journal of Quaternary Science, 25, 96–102. Girotti, O. 1972. Il genere Neumayria Stefani 1877 (Gastropoda, Prosobranchia). Geologica Romana, 11, 115–136. Glöer, P. 2002. Die Tierwelt Deutschlands, 73. Teil: Die Süßwassergastropoden Nord- und Mitteleuropas. Bestimmungsschlüssel, Lebensweise, Verbreitung. ConchBooks, Hackenheim. Gorthner, A. 1992. Bau, Funktion und Evolution komplexer Gastropodenschalen in Langzeit- Seen Mit einem Beitrag zur Paläobiologie von Gyraulus "multiformis" im Steinheimer Becken. Stuttgarter Beiträge zur Naturkunde Serie B, 190, 1–173. Hammer, Ø., Harper, D.A.T. 2006. Paleontological Data Analysis. Blackwell Publishing, Oxford. Harzhauser, M., Mandic, O. 2008. Neogene lake systems of Central and South-Eastern Europe: faunal diversity, gradients and interrelations. Palaeogeography, Palaeoclimatology Palaeoecology, 260, 417–434. Hewitt, G.M. 1999. Post-glacial re-colonization of European biota. Biological Journal of the Linnean Society, 68, 87–112. Hewitt, G. 2000. The genetic legacy of the Quaternary ice ages. Nature, 405, 907–913. Hoek, W.Z., Bohncke, S.J.P., Ganssen, G.M., Meijer, T. 1999. Lateglacial environmental changes recorded in calcareous gyttja deposits at Gulickshof, southern Netherlands. Boreas, 28, 416–432. Horton, A., Keen, D.H., Field, M.H., Robinson, J.E., Coope, G.R., Currant, A.P., Graham, D.K., Green, C.P., Phillips, L.M. 1992. The Hoxnian Interglacial Deposits at Woodston, Peterborough. Philosophical Transactions of the Royal Society of London B, 338, 131–164.

90

Jansson, R., Dynesius, M. 2002. The fate of clades in a world of recurrent climatic change: Milankovitch oscillations and evolution. Annual Review of Ecology, Evolution and Systematics, 33, 741–777. de Jong, Y.S.D.M. 2012. Fauna Europaea version 2.5. http://www.faunaeur.org. Kahlke, R.-D., García, N., Kostopoulos, D.S., Lacombat, F., Lister, A.M., Mazza, P.P.A., Spassov, N., Titov, V.V. 2011. Western Palaearctic palaeoenvironmental conditions during the Early and early Middle Pleistocene inferred from large mammal communities, and implications for hominin dispersal in Europe. Quaternary Science Reviews, 30, 1368–1395. Kennard, A.S., Woodward, B.B. 1917. The Post-Pliocene non-marine Mollusca of Ireland. Proceedings of the Geologists’ Association, 28, 109–190. Kennard, A.S., Woodward, B.B. 1922. The Post-Pliocene non-marine Mollusca of the East of England. Proceedings of the Geologists’ Association, 33, 104–142. Kossler, A., Strahl, J. 2011. The Late Weichselian to Holocene succession of the Niedersee (Rügen, Baltic Sea) – new results based on multi-proxy studies. E&G Quaternary Science Journal, 60 (4), 434–454. Kougioumoutzis, K., Simaiakis, S.M., Tiniakou, A. 2014. Network biogeographical analysis of the central Aegean archipelago. Journal of Biogeography, 41, 1848–1858. Kreft, H., Jetz, W. 2010. A framework for delineating biogeographic regions based on global species distributions. Journal of Biogeography, 37, 2029–2053. Kuijper, W.J. 2006. Flora en fauna in en rond een Scheldegeul bij Kallo op het einde van het atlanticum (Beveren, prov. Oost-Vlaanderen). Relicta, 1, 29–48. Lau, M.K. 2013. DTK: Dunnett-Tukey-Kramer Pairwise Multiple Comparison Test Adjusted for Unequal Variances and Unequal Sample Sizes. R package version 3.5. http://CRAN.R-project.org/package=DTK. Limondin-Lozouet, N., Preece, R.C. 2014. Quaternary perspectives on the diversity of land snail assemblages from northwestern Europe. Journal of Molluscan Studies, 80, 224– 237. Lisiecki, L.E., Raymo, M.E. 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20, PA1003, doi:10.1029/2004PA001071. Ložek, V. 1964. Quartärmollusken der Tschechoslovakei. Rozpravy Ústředního Ústavu Geologického 31, Praha. Mangerud, J., Jakobsson, J., Alexanderson, H., Astakhov, V., Clarke, G.K.C., Henriksen, M., Hjort, C., Krinner, G., Lunkka, J.-P., Möller, P., Murray, A., Nikolskaya, O.,

91

Saarnisto, M., Svendsen, J.I. 2004. Ice-dammed lakes and rerouting of the drainage of northern Eurasia during the last glaciation. Quaternary Science Reviews, 23, 1313– 1332. Mania, D. 1973. Paläoökologie, Faunenentwicklung und Stratigraphie des Eiszeitalters im mittleren Elbe-Saalegebiet auf Grund von Molluskengesellschaften. Geologie Beiheft, 21 (78/79), 1–175. Mania, D. 1978. Die Molluskenfauna aus den Travertinen von Burgtonna in Thüringen. Quartärpaläontologie, 3, 69–85. Manzi, G., Magri, D., Milli, S., Palombo, M.R., Margari, V., Celiberti, V., Barbieri, M., Barbieri, M., Melis, R.T., Rubini, M., Ruffo, M., Saracino, B., Tzedakis, P.C., Zarattini, A., Biddittu, I. 2010. The new chronology of the Ceprano calvarium (Italy). Journal of Human Evolution, 59, 580–585. Meijer, T. 1988. Mollusca from the Borehole Zuurland-2 at Brielle, the Netherlands (an interim report). Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie, 25 (1), 49–60. Meijer, T, Preece, R.C. 1996. Malacological evidence relating to the stratigraphical position of the Cromerian. In: Turner, C. (eds), The Early Middle Pleistocene in Europe, A.A. Balkema, Brookfield, Rotterdam, pp. 53–82. Molnar, P. 2004. Late Cenozoic increase in accumulation rates of terrestrial sediment: how might climate change have affected erosion rates? Annual Review of Earth and Planetary Sciences, 32, 67–89. Neubauer, T.A., Mandic, O., Harzhauser, M., Hrvatović, H. 2013. A new Miocene lacustrine mollusc fauna of the Dinaride Lake System and its palaeobiogeographic, palaeoecologic, and taxonomic implications. Palaeontology, 56 (1), 129–156. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2014. The FreshGEN Database: Freshwater Gastropods of the European Neogene. Accessed at http://www.marinespecies.org/freshgen on 2015-10-15. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015a. A gastropod-based biogeographic scheme for the European Neogene freshwater systems. Earth Science Reviews, 143, 98–116. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Kroh, A., Mandic, O. 2015b. Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots. Proceedings of the National Academy of Sciences of the United States of America, 112, 11478–11483.

92

R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org. Rousseau, D.D. 1989. Reponses des malacofaunes terrestres quaternaires aux contraintes climatiques en Europe septentrionale. Palaeogeography, Palaeoclimatology, Palaeoecology, 69, 113–124. Sandom, C., Faurby, S., Sandel, B., Svenning, J.-C. 2014. Global late Quaternary megafauna extinctions linked to humans, not climate change. Proceedings of the Royal Society B, 281, 20133254. Sanko, A.F. 2007. Quaternary freshwater molluscs Belarus and neighbouring regions of Russia, Lithuania, Poland (field guide). Institute of Geochemistry and Geophysics, National Academy of Sciences, Belarus. (In Russian) Schlickum, W.R., Puisségur, J.-J. 1977. Die Molluskenfauna des Altpleistozäns von St. Bernard (Département Côte-d'Or). Archiv für Molluskenkunde 107 (4/6), 273–283. Schlickum, W.R., Puisségur, J.-J. 1978. Die Molluskenfauna der Schichten mit Viviparus burgundinus und Pyrgula nodotiana von Montagny-les-Beaune (Dép. Côte-d'Or). Archiv für Molluskenkunde, 109 (1/3), 1–26. Settepassi, F., Verdel, U., 1965. Continental Quaternary mollusca of lower Liri Valley (Southern Latium). Geologica Romana 4, 369–452. Sfenthourakis, S., Giokas, S., Mylonas, M. 1999. Testing for nestedness in the terrestrial isopods and snails of Kyklades islands (Aegean archipelago, Greece). Ecography, 22, 384–395. Simaiakis, S.M., Strona, G. 2015. Patterns and processes in the distribution of European centipedes (Chilopoda). Journal of Biogeography, 42, 1018–1028. Sparks, B.W. 1964. The distribution of Non-Marine Mollusca in the Last Interglacial in south-east England. Proceedings of the Malacological Society of London, 36 (1), 7– 25. Sparks, B.W., West, R.G. 1959. The Palaeoecology of the Interglacial Deposits at Histon Road, Cambridge. Eiszeitalter und Gegenwart Quaternary Science Journal, 10 (1), 123–143. Sternai, P., Herman, F., Champagnac, J.-D., Fox, M., Salcher, B., Willett, S.D. 2012. Preglacial topography of the European Alps. Geology, 40, 1067–1070. Stocchi, P., Spada, G., Cianetti. S. 2005. Isostatic rebound following the Alpine deglaciation: impact on the sea level variations and vertical movements in the Mediterranean region. Geophysical Journal International, 162, 137–147.

93

Strona, G., Galli, P., Seveso, D., Montano, S., Fattorini, S. 2014. Nestedness for Dummies (NeD): a user-friendly web interface for exploratory nestedness analysis. Journal of Statistical Software, 59, 1–9. Strona, G, Fattorini, S. 2014. NeD – Nestedness for Dummies. URL http://purl.oclc.org/ned. Sümegi, P., Lócskai, T., Hupuczi, J. 2011. Late Quaternary palaeoenvironment and palaeoclimate of the Lake Fehér (Fehér-tó) sequence at Kardoskút (South Hungary), based on preliminary mollusc records. Central European Journal of Geosciences, 3 (1), 43–52. Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M.J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J.H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29. Welter-Schultes, F.W. 2012. European non-marine molluscs, a guide for species identification. Planet Poster Editions, Göttingen. Wesselingh, F.P. 2007. Long-Lived Lake Molluscs as Island Faunas: a bivalve perspective. In: Renema, W. (Ed.), Biogeography, Time, and Place: Distributions, Barriers, and Islands. Springer, Dordrecht, pp. 275–314. West, R.G., Sparks, B.W. 1960. Coastal Interglacial deposits of the English Channel. Philosophical transactions of the Royal Society of London, B 243, 95–133. Willmann, R. 1981. Evolution, Systematik und stratigraphische Bedeutung der neogenen Süßwassergastropoden von Rhodos und Kos/Ägäis. Palaeontographica Abt. A, 174, 10–235. Wright, D.H., Patterson, B.D., Mikkelson, G.M., Cutler, A., Atmar, W. 1998. A comparative analysis of nested subset patterns of species composition. Oecologia, 113, 1–20. Zagwijn, W.H. 1989. The Netherlands during the Tertiary and the Quaternary: A case history of Coastal Lowland evolution. Geologie en Mijnbouw, 68, 107–120.

4.8. Supplementary material

Table S4.1. List of species presence data, together with genus and family attribution, locality name, GPS coordinates, stratigraphic unit name, proposed age boundaries, time interval (TI) and bibliographical references. Asterisks in TI mark localities with assigned ages crossing an intervals’ age boundary. (Due to size matters this table is not provided here)

94

Table S4.2. List of localities with indication of species richness, GPS coordinates, proposed age boundaries, stratigraphic unit name, Epoch (Pleistocene–Holocene) and time interval (TI). Asterisks in TI mark localities with assigned ages crossing an intervals’ age boundary. (Due to size matters this table is not provided here)

Table S4.3. ANOVA results of pairwise comparisons between average species richness per grid cell of the six time intervals using the Dunnett–Tukey–Kramer pairwise multiple comparison post hoc test which is adjusted for unequal variances and sample sizes. P-values are indicated on the upper right triangle of the table. Differences of the means for each pair are on the lower left triangle of the table. Significant pairwise comparisons are marked bold. Holocene Last Glacial Last Interglacial Ionian Calabrian Gelasian Holocene 0 1 1 1 0.04 Last Glacial -5.22 0.01 0.02 0.02 1 Last Interglacial 0.7 5.92 0.96 1 0.05 Ionian -0.58 4.64 -1.28 1 0.15 Calabrian -0.04 5.18 -0.74 0.54 0.13 Gelasian -5.14 0.08 -5.84 -4.56 -5.1

Figure S4.1. Nestedness original and packed matrices for the six time intervals.

95

CHAPTER 5: Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots

Thomas A. Neubauer1, Mathias Harzhauser1, Elisavet Georgopoulou1, 2, Andreas Kroh1, Oleg Mandic1

1Geological-Paleontological Department, Natural History Museum Vienna, Burgring 7, 1010 Vienna, Austria 2Institute of Earth Sciences, University of Graz, Heinrichstrasse 26, 8010 Graz, Austria

5.1. Abstract

Continental aquatic species richness hotspots are unevenly distributed across the planet. In present-day Europe, only two centers of biodiversity exist (Lake Ohrid on the Balkans and the Caspian Sea). During the Neogene, a wide variety of hotspots developed in a series of long-lived lakes. The mechanisms underlying the presence of richness hotspots in different geological periods have not been properly examined thus far. Based on Miocene to Recent gastropod distributions, we show that the existence and evolution of such hotspots in inland- water systems are tightly linked to the geodynamic history of the European continent. Both past and present hotspots are related to the formation and persistence of long-lived lake systems in geological basins or to isolation of existing inland basins and embayments from the marine realm. The faunal evolution within hotspots highly depends on warm climates and surface area. During the Quaternary icehouse climate and extensive glaciations, limnic biodiversity sustained a severe decline across the continent and most former hotspots disappeared. The Recent gastropod distribution is mainly a geologically young pattern formed after the Last Glacial Maximum (19 kyr) and subsequent formation of postglacial lakes. The major hotspots today are related to long-lived lakes in preglacially formed, permanently subsiding geological basins.

Keywords: biogeography, hotspot evolution, freshwater gastropods, Cenozoic, species-area relationship

5.2. Significance

96

To our knowledge, this study is the first investigation of the evolution of species richness hotspots in continental aquatic systems. We demonstrate the development of European richness hotspots over the last 23 myr based on a comprehensive dataset combining recent and fossil occurrences of gastropod species. We show that changes in species richness patterns can be related to geodynamic and climatic processes. The addition of tectonics, geological time, and spatial scales to ecology and climate is essential for understanding hotspot development in general. These insights also provide a foundation to explain the modern, uneven distribution of species richness as a whole. The pattern for Recent European faunas is a geologically young phenomenon, triggered by the ice sheet retreat after the Last Glacial Maximum.

5.3. Introduction

The term “hotspot” is variably defined in the literature. Hotspots may characterize regions with particularly high species richness, high levels of endemism, high numbers of rare or threatened species, or refer to the intensity of threat (Reid, 1998; Orme, 2005). In this paper, we specifically deal with species richness hotspots and their evolution over geological time. Species richness patterns and changes therein have been tightly linked to climatic processes (Orme, 2005; Mittelbach et al., 2007; Hof et al., 2008; Ezard et al., 2011). In some cases, however, climate may only be a secondary or indirect cause for varying species richness (Graham et al., 2006). Regional climate conditions can be strongly influenced by tectonic processes causing closing or opening of seaways or orogenesis (Champagnac et al., 2012; Godard et al., 2014). Such configurations determine oceanic and atmospheric circulation dynamics and, hence, climatic dynamics (Micheels et al., 2010, Quan et al., 2014). A good example is the triggering of the Indian monsoon at the beginning of the Neogene by the collision of the Indian subcontinent with Eurasia and the uplift of the Himalaya (Cook et al., 2010). Today, half of the world's human population is affected by this particular climate regime that was initiated more than 20 myr BP (Cook et al., 2010). Paleogeographic settings may affect faunal distributions indirectly (through climate), but also directly. The initiation and cessation of continental basins, along with the related development and persistence of freshwater and brackish environments, controls the possibilities for dispersal and evolutionary radiations for nonmarine biota. Lacustrine basins that are tectonically and ecologically stable on geological time scales provide opportunities for settlement and diversification (Wesselingh 2007; Neubauer et al., 2015). In various cases

97 biodiversity maxima might be related to a basin's geographic position, whereas climatic and physiographic parameters may influence the species richness and composition. Diversification patterns have been linked to tectonic evolution for various settings (Kohn & Fremd, 2008; Renema et al., 2008; Hoom et al., 2010; Keith et al., 2013), yet such a relation has not been established for inland aquatic biota. Freshwater and brackish biota are ideal research objects because they have limited possibilities for dispersal (van Leeuwen et al., 2012) and must therefore overcome environmental pressure in their habitats. In particular, species originating from intralacustrine radiations in long-lived lakes rarely colonize other lakes (Wesselingh 2007; Schön & Martens, 2004). We analyzed the distributions of Miocene to Recent European nonmarine gastropods to identify the dynamics and drivers of species richness hotspots through time. We tested for correlations of species richness trends, large-scale faunal turnovers, and biodiversity maxima to climatic and to physiographic parameters, as well as geodynamic setting. The specific aim of this paper is to unravel the evolution and history of the modern distributions.

5.4. Materials and methods

The data are based on an extensive literature research and derive from more than 400 publications on Miocene to Recent continental aquatic gastropod faunas (Dataset S5.1). Due to preservation issues, the fossil record is slightly biased toward lacustrine systems (Neubauer et al., 2015), which is why we chose to include only lacustrine faunas for Recent times. Only freshwater to brackish environments were considered in this study. This approach yielded a total of 2,785 species-group taxa (species and subspecies) from 5,414 localities. Although the general geographic frame is Europe, we integrated Turkey and marginal territories of Azerbaijan, Georgia, and Kazakhstan into the analysis because biogeographic entities do not adhere to political borders. For the fossil faunas, no data were available for latitudes north of 52° due to erosion of former sediments by the advancing ice sheet during the Pleistocene. The heat maps were produced in ESRI ArcGIS 10.0 (Fig. 5.1). An equal-area grid of 100-km cell size was created using the Behrmann projection to correct for oversampling of certain areas. The available faunas were separated into six discrete temporal units: Early Miocene (23.03–15.97 myr), Middle Miocene (15.97–11.62 myr), Late Miocene (11.62–5.333 myr), Pliocene (5.333–2.588 myr), Pleistocene (2.588–0.00117 myr), and Recent. For each cell containing localities, the number of species was calculated. For each time slice, an ordinary Kriging algorithm (100-km cell size, 200-km interpolation radius) was performed on the grid-

98 specific species richness. The coloration gradient is equal in all maps, defined based on the maximum range of species numbers (Late Miocene). The maps are intended as graphical visualization supporting intuitive recognition of diversity hotspots and should be interpreted carefully, because the interpolation method tends to exaggerate actual richness toward the margins. For details on taxonomic and stratigraphic treatment, constraints on area and duration data for the lakes and regression analyses, see Supplementary materials and methods and Figs. S5.3 and S5.4.

5.5. Results

5.5.1. Shifting hotspots through time

As measures of diversity we chose a number of proxies: (i) the total number of species per million years as a proxy for the overall trend, (ii) the maximum number of species per environment as an indicator for individual systems, (iii) the mean number of species of all environments per million years as a measure of the average trend, and (iv) origination, extinction, turnover rates and Whittaker's β diversity (Hof et al., 2008) between million year bins to estimate the degree of faunal changes. The rise and demise of richness hotspots is shown in Figs. 5.1 and 5.2. In general, the total number of species increases constantly over time (linear correlation between million year bin and number of species: r = 0.866, P < 0.001). This trend is occasionally interrupted by marked breakdowns, e.g., in the Middle Miocene and Pleistocene (Fig. 5.2). A similar picture is presented by the maximum number of species per environment, which peaked in the Late Miocene to Pliocene and declined toward the Pleistocene. Marked episodes of faunal turnover occurred during the Middle Miocene (15 myr), the Late Miocene (9 myr), the Early Pliocene (5 myr), and the Late Pliocene (3 myr). Particularly, the Pliocene events are characterized by extremely high extinction rates. Preservation and sampling biases constrain the reconstructions of past biodiversity to some extent. As the fossil record of continental aquatic systems is commonly biased toward large, long-lived lakes (Neubauer et al., 2015), which typically hold the most species, the potential bias for detecting biodiversity hotspots is expected to be minor.

99

Figure 5.1. Heat maps for the six selected temporal units, using an equal-area grid of 100-km cell size in Behrmann projection. Colors are based on the ordinary Kriging algorithm with 100-km cell size and 200-km interpolation radius performed on grid-specific species richness. Coloration is equal in all maps, scaled to the maximum grid richness of 218 species (Late Miocene, Lake Pannon).

Early Miocene. This period is characterized by a relatively low average species richness (Fig. 5.1). Three small hotspots of about equal magnitude and geographical extent exist, corresponding to major geological basins present during this interval (Fig. 5.3). The westernmost center is within the Aquitaine Basin (Biteau et al., 2006), which is famous for its rich marine Cenozoic mollusk faunas. Additionally, a rich freshwater to brackish gastropod fauna is recorded from marginal deposits (Table 5.1 and Table S5.1). A second hotspot is located in the German Hanau-Wetterau Basin along the Upper Rhine Graben, whose origins date back to the Early Oligocene (Kümmerle & Radtke, 2012). Due to partial geographical restriction and cutoff from the North Sea Basin, brackish and freshwater environments

100 developed between the Late Oligocene and the early Middle Miocene (Berger et al., 2005; Kümmerle & Radtke, 2012). Maximum diversity was reached during the Early Miocene. A third, smaller hotspot existed in the south German North Alpine Foreland Basin and is also a relic of a former sea. With the retreat of the western branch of the Paratethys Sea in the late Early Miocene, small brackish basins formed within the North Alpine Foreland Basin (Berger et al., 2005; Reichenbacher et al., 2013). The Upper Brackish Water Molasse (UBWM) is characterized by a rich, endemic gastropod fauna, whose evolution around 18 myr BP boosted β diversity. From this time onward, β diversity remained at a high level of 0.6–0.7. Later diversity fluctuations are minor and largely parallel main turnover events. The overall diversity increase in the earliest Miocene coincides with a trend toward higher temperature and humidity after the comparatively dry and cool Oligocene (Zachos et al., 2001; Mosbruggeret al., 2005).

101

Table 5.1. Late Cenozoic freshwater and brackish gastropod species richness hotspots, reflecting lakes with more than 40 species. Number Maximum of species number of Interval of Maximum temporal Type of Lake/Basin across coevally maximum range of environment entire present richness [Ma] environment [Ma] duration species* Ohrid freshwater lake 68 65 0 1.5-0 Caspia brackish lake 105 92 0 0.88-0† Curonian Lagoon brackish lagoon 42 42 0 0.0135-0 Tiberino freshwater lake 42 32 1 3.1-1.55 Kos freshwater lake 41 26 1 3.0-1.4 fluvio-lacustrine Bresse system 64 44 2 4.5-1.5 Rioni fluvial plain 56 52 3 5.5-1.6 Slavonia freshwater lake 163 145 3 4.0-2.6 Transylvania freshwater lake 78 78 4 3.8-0.8 brackish to Dacia freshwater lake 303‡ 159 4 8.6-2.6 Metohia§ freshwater lake 70 36 5 6.04-2.588 brackish Galați embayment 133 131 7 8.6-4.4 Bresse-Valence fluvial plain 60 56 8 10.0-8.0 Pannon brackish lake 605 248 10 11.6-4.0 brackish Soceni embayment 51 51 12 12.7-12.3 Bakony wetlands (?) 87 48 13 17.5-12.5 Steinheim freshwater lake 42 42 13 15-13.9 Sinj freshwater lake 58 50 15 18.0-15.0 Drniš freshwater lake 43 43 15 15.7-15.0 Upper Brackish Water Molasse Brackish lake 40 37 16 16.6-17.6 brackish wetlands Aquitaine (?) 59 29 20 28.1-11.6{ brackish Hanau-Wetterau embayment 49 15 21 28.1-16.1{ *Per one million years for fossil lakes. †Time from last marine transgression. ‡Excludes peri-marine Sarmatian assemblages.

107

§Excludes Middle Miocene assemblage, which represents a separate lake development. {Poorly constrained, freshwater/brackish conditions not continuous throughout. For a more detailed version of the table with references for the age models, see Table S5.1.

Middle Miocene. Overall species richness in Europe increased toward a maximum at the Middle Miocene Climate Optimum (MCO) (Zachos et al., 2001; Böhme et al., 2011). This peak is followed by the first big Miocene turnover phase with simultaneously increased origination and extinction rates. Turnover is related to the rise and demise of individual lake basins. Many of the characteristic Early Miocene faunas decline, mostly as a result of deteriorating settings in the Aquitaine and Hanau-Wetterau basins. At the same time, a hotspot emerged on the Balkan Peninsula, comprising several freshwater lake basins in Croatia and Bosnia and Herzegovina (Fig. 5.3) that form the Dinaride Lake System (De Leeuw et al., 2012). These isolated intramontane lakes formed as a result of the folding of the Dinaride Mountains (De Leeuw et al., 2012) and accommodated rich and highly endemic mollusk faunas in the early Middle Miocene (Neubauer et al., 2015; Harzhauser & Mandic, 2008). Decreasing global (Zachos et al., 2001) and regional (Mosbrugger et al., 2005) temperatures, as well as decreasing regional precipitation (Böhme et al., 2011) around 14 myr BP, coincide with a strong decrease of total species richness and individual richness per lake, reflecting the ongoing continentalization of the Dinaride lakes (De Leeuw et al., 2012). Afterward, species richness throughout Europe rebounded in a prolonged (5 myr), gradual rise in the total number of species and in the maximum richness per lake, paralleling increased origination rates. Simultaneous faunal diversifications occurred in the late Middle Miocene Bakony wetlands near Budapest (Hungary), the Paratethyan Soceni wetland fringes (Romania), and the small Lake Steinheim (Germany).

Late Miocene. With the final retreat of the Central Paratethys by the onset of the Late Miocene, all of the previous hotspots disappeared. However, a huge new hotspot developed in Lake Pannon in the Pannonian Basin at that time, hosting large faunal radiations (Harzhauser & Mandic, 2008; Magyar et al., 2013; ter Borgh et al., 2013) (Fig. 5.3). Its individual species richness reached its maximum at 9 myr, coinciding with the lake's maximum extent and an overall peak in European diversity. That peak is followed by the second major Miocene turnover phase, characterized by elevated extinction rates. From ∼9 myr onward, the diversity of Lake Pannon declined together with its shrinking size. A small

108 peak at 7 myr reflects the evolution of faunas in the semienclosed Dacian Basin and adjacent Galati Seaway in southeastern Romania, Moldova, and Ukraine, both of which yielded brackish to marginally freshwater conditions at that time (Jipa & Olariu, 2009). An additional, smaller hotspot developed in the Bresse–Valence graben system in southeastern France, coinciding with its successive isolation from the Mediterranean during the Late Miocene and the development of brackish to fluvio-lacustrine conditions (Sissingh, 2001). In addition, several rich faunas are known from the Iberian Peninsula (e.g., Lake Calatayud- Teruel) (Neubauer et al., 2015). Most of these, however, are poorly studied on the species level and few of these could be included in the present dataset. Their potential status as hotspot requires further taxonomic investigations of the local faunas.

Pliocene. The total number of species reached a temporary maximum, mostly because of the presence of the rich fauna of Lake Dacia (Jipa & Olariu, 2009; Stoica et al., 2013). With the isolation of the basin in the Late Dacian/Middle Zanclean, ∼4 myr, freshwater conditions started to develop (Jipa & Olariu, 2009), offering suitable conditions for mollusk settlement and evolution. Two biodiversity peaks are present in the Pliocene, followed by major faunal turnovers. The peak at 5 myr represents the final phase of the Lake Pannon fauna, which is associated with a high extinction rate, and the small but rich Lake Metohia in Kosovo. A second peak at 3 myr corresponds to the highly diverse fauna of Lake Pannon's successor, Lake Slavonia (Harzhauser & Mandic, 2008; Magyar et al., 2013), and the fauna of the fluvial plains along Rioni Bay in the eastern Black Sea depression, which had brackish conditions at that time (Popov et al., 2004) (Fig. 5.3). This biodiversity peak, corresponding to the mid-Pliocene warm period (Brigham-Grette et al., 2013; Koenig et al., 2014), is followed by the extinction of more than 300 species, representing the biggest turnover event in the late Cenozoic.

Pleistocene. The successive richness decline early in the Pleistocene coincides with global and regional cooling (Zachos et al., 2001, Brigham-Grette et al., 2013). The total number of species drops to less than one half of the Pliocene maximum. With the demise of long-lived lakes by the end of the Pliocene, such as Dacia, Slavonia, and Transylvania lakes, the maximum number of species per lake also declines (Fig. 5.2). Former mega hotspots vanished and gave way to dispersed small hotspots such as Lake Bresse in France, Lake Tiberino in Italy, and Lake Kos in Greece. Beyond that, several low-diversity centers formed in central Europe, none of which can be attributed to a bigger paleo-lake or an evolving

109 geological basin. The pattern is to some extent biased by the varied availability of outcrops and time-averaging of glacial and interglacial deposits. Preliminary data suggest that the Caspian Sea, today a very large, brackish lake, likely formed a hotspot already during the Pleistocene, but our knowledge on early stage faunas is limited. The picture for the Pleistocene is not well resolved in detail. Nevertheless, the low overall diversity compared with the Pliocene is evident.

Recent. With the onset of the Ice Ages and the associated glaciations of large parts of northern Europe and the Alpine region (Mangerud et al., 2004; Clark et al., 2009; Andrén et al., 2011), many of the former lakes and their faunas vanished. Most extant lakes and their faunas emerged after the Last Glacial Maximum (20,000–19,000 yr BP) (Clark et al., 2009). Two areas with high species numbers develop in the Caspian Sea and in a series of lakes in the southern Balkans, i.e., Ohrid, Prespa, Mikri Prespa, Pamvotis, and Trichonis (4 Hauswald et al., 2008; Albrecht et al., 2012; Wagner et al., 2014b). Their corresponding lake basins were already present before the Pleistocene. Another area of relatively high biodiversity encompasses the Curonian Lagoon, a low-brackish lake close to the Baltic Sea. Contrary to the two other hotspots, this area only contains geologically young lakes (<19 kyr). The high diversity is likely related to the lacustrine-brackish development of the Baltic Sea in the Late Pleistocene to earliest Holocene (Andrén et al., 2011) and to the following migrations between surrounding lakes.

110

Figure 5.2. Species richness trends across time. From Left to Right: mean number of species per lake (dotted line); maximum number of species per lake (dashed line); total number of species (solid line); number of originations (dotted line); number of extinctions (dashed line); faunal turnover rate (dash-dotted line; summed number of originations and extinctions); total number of species with range-through assumption (solid line); Whittaker’s β diversity (solid line); deep-sea oxygen isotope record (5 pt moving average) as proxy for global temperature (Zachos et al., 2001); schematic indication of important climatic events (MCO, Middle Miocene Climate Optimum; MPWP, Mid-Pliocene warm period). Horizontal bars denote marked faunal turnover events. Note the logarithmized number of species was used for the first three curves, as otherwise trends of mean species richness per lake would be indiscernible.

5.6. Discussion

5.6.1. What drives species richness trends and patterns?

Using different datasets and resolutions (locality based, environment based, and overall richness per million years; Tables S5.2–S5.4), we carried out regression analyses to test for relations of species richness with geographic, physiographic, and climatic parameters. Although species richness is known to vary across geographic settings in general, particularly latitude (Allen & Gillooly, 2006; Mittelbach et al., 2007; Mannion et al., 2014), our data indicate very weak geographic relations throughout all time intervals (Table S5). In contrast, species richness trends over time are correlated with climate changes. Linear regressions of richness and δ18O as proxy for global temperature yielded highly significant correlations for 2 total number of species (log10-transformed, r = 0.514, P < 0.001; with range-through 2 assumption: r = 0.465, P < 0.001), maximum richness per environment (log10-transformed, r2 = 0.245, P = 0.016), number of originations (r2 = 0.470, P < 0.001), number of extinctions (r2 = 0.189, P = 0.038), and turnover rate (r2 = 0.433, P = 0.001) (Tables S5.2 and S5.3 and Fig. S5.1). Global cooling and pronounced glaciations, paired with the decline of the Paratethyan long-lived lakes by the end of the Pliocene, dramatically reduced the availability of biotopes for lacustrine gastropods. Throughout the Neogene, continental aquatic species richness maxima in Europe correspond to climate maxima (e.g., MCO and MPWP; Fig. 5.2).

111

Additionally, we performed regression analyses for species richnesses of 30 selected systems to assess potential correlations with latitude, longitude, lake surface area and temporal duration (Table S5.4). Linear ordinary least-square regressions indicated a statistically significant association only for area (log-log, r2 = 0.489, P < 0.001; Fig. S5.2 and Tables S5.6–S5.18). A multiple regression including all parameters did not improve the model (Supplementary materials and methods and Tables S5.6–S5.18). The species-area relationship serves as a common explanatory model for varied numbers of species in geographically restricted environments, such as lakes or islands, and has been confirmed by many theoretical and empirical studies (Whittaker et al., 2007; Hof et al., 2008; Franzén et al., 2012; Wagner et al., 2014a).

112

Figure 5.3. European species richness hotspots in relation to geodynamic development. The Miocene-Pliocene palinspastic maps follow latest reconstructions (Neubauer et al., 2015, Brigham-Grette et al., 2013). Pleistocene Recent maps are created in European equidistant- conic projection to ease comparison with the palinspastic reconstructions. The boundaries of the Scandinavian and Alpine ice sheets represent their maximum extent during the Late Glacial Maximum (Ehlers et al., 2011). Hotspots with larger stratigraphic ranges are indicated for the time interval of their maximum richness. Numbers refer to the total number of species across a system’s full duration. CP, Central Paratethys; UBWM, Upper Brackish Water Molasse.

5.6.2. Geodynamics as a primary driver of hotspot development

It is crucial to distinguish between the faunal development of a hotspot and its mere existence. The regression analyses indicate that faunal diversification correlates with climate and surface area, confirming earlier studies (Graham et al., 2006; Mittelbach et al., 2007; Hof et al., 2008; Ezard et al., 2011). However, the presence of hotspots and their spatial distribution are rarely discussed. Here, we demonstrate that the shifting presence of species richness hotspots through time is tightly linked to the development of geological basins accommodating long-lived, stable freshwater or brackish environments (Fig. 5.3). Particularly large, long-lived lakes such as the Late Miocene Lake Pannon or Late Miocene-Pliocene Lake Dacia presented environments that were stable across geological timescales and offered a great variety of habitats. In these and other such systems, intralacustrine speciation gave rise to many hundreds of species over time (Geary et al., 2008; Hauswald et al., 2008; Albrecht et al., 2012). This process created diversities far above the average of typical short-lived systems such as most modern lakes. Conversely, not every basin with freshwater habitats offered opportune conditions for settlement and evolution. Thus, the availability of a persisting, stable geological basin providing continual freshwater or brackish environments is a prerequisite for hotspot evolution for aquatic gastropods. Further faunal evolution is mainly controlled by climatic factors and surface area (Fig. S5.1). Hence, large-scale biodiversity patterns through time rise and fall with the presence of large lakes. As the regression analyses indicate, species numbers are not directly related to a system's temporal duration per se.

5.6.3. The cradle of modern faunas

113

The two greatest centers of continental aquatic biodiversity today, the Caspian Sea and Lake Ohrid, both reflect long-lived geological lakes dating back at least into the Pleistocene (Van Baak et al., 2013; Wagner et al., 2014b). The great majority of other lakes comprise low diversities and share similar species compositions. This situation differs greatly from the Miocene and Pliocene distributions, which were characterized by a centralization of biodiversity into few, highly diverse long-lived lakes. Despite the low number of aquatic environments available for past time slices compared with the huge number for the most recent interval, the earlier time slices still preserve comparable levels of biodiversity (Fig. 5.2 and Tables S5.2 and S5.3). The discrepancy between Miocene-Pliocene and Pleistocene-Recent distribution patterns is partly explained by deteriorating climates during the Quaternary. Together with the disappearance of long-lived Paratethyan lakes at the end of the Pliocene, global cooling and large-scale glaciations dramatically reduced the availability of suitable habitats. Ice sheet retreat after the Last Glacial Maximum triggered the formation of thousands of lakes in the successively emerging glacial depressions and valleys. The pattern for the Recent faunas is consequently a very young phenomenon. Although current hotspots are confined to long- lived, geologically induced lakes outside the reach of the Pleistocene glacial sheet and permafrost belt, most present distributions reflect immigration events to postglacial lakes after deglaciation starting about 19,000 yr BP (Clark et al., 2009). The here proposed scheme is certainly not restricted to European continental aquatic systems and not only to the late Cenozoic. The rise and demise of species richness hotspots through time is tightly related to regional tectonic phases.

5.7. Acknowledgments

We thank a great many colleagues for assistance with literature research and/or stratigraphic classifications, which markedly improved the data: J. Albesa, D. Alba, P. Anadón, V. V. Anistratenko, I. Casanovas-Vilar, A. Engelbrecht, D. Esu, M. Gross, G. Haszprunar, S. Herzog-Gutsch, R. Macaleţ, I. Magyar, N. Krstić, G. Mas Gornals, V. A. Prysjazhnjuk, M. W. Rasser, and D. Vasilyan. We thank M. Stachowitsch and A. Neubauer for linguistic amendments. The comments of three reviewers and the editor greatly improved the work. The study was financed by the Austrian Science Fund (FWF; Project P25365-B25: “Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients”).

114

5.8. References

Albrecht, C., Hauffe, T., Schreiber, K., Wilke, T. 2012. Mollusc biodiversity in a European ancient lake system: lakes Prespa and Mikri Prespa in the Balkans. Hydrobiologia, 682, 47–59. Allen, A.P., Gillooly, J.F. 2006. Assessing latitudinal gradients in speciation rates and biodiversity at the global scale. Ecology Letters, 9 (8), 947–954. Andrén, T., Björck, S., Andrén, E., Conley, D., Lambeck, K., Zillén, L., Anjar, J. 2011. The Development of the Baltic Sea Basin During the Last 130 ka. In Harff, J., Björck, S., Hoth, P. (eds), The Baltic Sea Basin as a Natural Laboratory. Springer, Berlin: 75–97. Berger, J-P, Reichenbacher, B., Becker, D., Grimm, M., Grimm, K., Picot, L., Storni, A., Pirkenseer, C., Derer, C., Schaefer, A. 2005. Paleogeography of the Upper Rhine Graben (URG) and the Swiss Molasse Basin (SMB) from Eocene to Pliocene. International Journal of Earth Sciences, 94 (4), 697–710. Biteau, J.-J., Le Marrec, A., Le Vot, M., Masset, J.-M. 2006. The Aquitaine Basin. Petroleum Geoscience, 12 (3), 247–273. Böhme, M., Winklhofer, M., Ilg, A. 2011. Miocene precipitation in Europe: Temporal trends and spatial gradients. Palaeogeography, Palaeoclimatology, Palaeoecology, 304 (3-4), 212–218. Brigham-Grette, J., Melles, M., Minyuk, P., Andreev, A., Tarasov, P., DeConto, R., Koenig, S., Nowaczyk, N., Wennrich, V., Rosén, P., Haltia, E., Cook, T., Gebhard, C., Meyer- Jacob, C., Snyder, J. and Herzschuh, U. 2013. Pliocene warmth, polar amplification, and stepped Pleistocene cooling recorded in NE Arctic Russia. Science, 340 (6139), 1421–1427. Champagnac, J.-D., Molnar, P., Sue, C., Herman, F. 2012. Tectonics, climate, and mountain topography. Journal of Geophysical Research, 117 (B2):B02403. Clark, P.U., Dyke, A.S., Shakun, J.D., Carlson, A.E., Clark, J., Wohlfarth, B., Mitrovica, J.X., Hostetler, S.W., McCabe, A.M. 2009. The Last Glacial Maximum. Science, 325 (5941), 710–714. Cook, E.R., Anchukaitis, K.J., Buckley, B.M., D’Arrigo, R.D., Jacoby, G.C., Wright, W.E. 2010. Asian monsoon failure and megadrought during the last millennium. Science 328 (5977), 486–489.

115

De Leeuw, A., Mandic, O., Krijgsman, W., Kuiper, K., Hrvatović, H. 2012. Paleomagnetic and geochronologic constraints on the geodynamic evolution of the Central Dinarides. Tectonophysics, 530-531, 286–298. Ehlers, J., Gibbard, P.L., Hughes, P.D. 2011. Quaternary Glaciations – Extent and Chronology, 1st ed. Elsevier, Amsterdam. Ezard, T.H.G., Aze, T., Pearson, P.N., Purvis, A. 2011. Interplay between changing climate and species’ ecology drives macroevolutionary dynamics. Science, 332 (6027), 349– 351. Franzén, M., Schweiger, O., Betzholtz, P.-E. 2012. Species-area relationships are controlled by species traits. PLoS One 7 (5):e37359. Geary, D.H., Staley, A.W., Müller, P., Magyar, I. 2002. Iterative changes in Lake Pannon Melanopsis reflect a recurrent theme in gastropod morphological evolution. Paleobiology, 28 (2), 208–221. Godard, V., Bourlès, D.L., Spinabella, F., Burbank, D.W., Bookhagen, B., Fisher, G.B., Moulin, A., Léanni, L. 2014. Dominance of tectonics over climate in Himalayan denudation. Geology, 42 (3), 243–246. Graham, C.H., Moritz, C., Williams, S.E. 2006. Habitat history improves prediction of biodiversity in rainforest fauna. Proceedings of the National Academy of Sciences of the United States of America, 103 (3), 632–636. Harzhauser, M, Mandic, O. 2008. Neogene lake systems of Central and South-Eastern Europe: Faunal diversity, gradients and interrelations. Palaeogeography, Palaeoclimatology, Palaeoecology, 260 (3-4): 417–434. Hauswald, A.-K., Albrecht, C., Wilke, T. 2008. Testing two contrasting evolutionary patterns in ancient lakes: Species flock versus species scatter in valvatid gastropods of Lake Ohrid. Hydrobiologia, 615 (1), 169–179. Hof, C., Brändle, M., Brandl, R. 2008. Latitudinal variation of diversity in European freshwater animals is not concordant across habitat types. Global Ecology and Biogeography, 17 (4), 539–546. Hoorn, C., Wesselingh, F.P., ter Steege, H., Bermudez, M.A., Mora, A., Sevink, J., Sanmartín, I., Sanchez-Meseguer, A., Anderson, C.L., Figueiredo, J.P., Jaramillo, C., Riff, D., Negri, F.R., Hooghiemstra, H., Lundberg, J., Stadler, T., Sarkinen, T., Antonelli, A. 2010. Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. Science, 330 (6006), 927–931.

116

Jipa, D.C., Olariu, C. 2009. Dacian Basin: Depositional Architecture and Sedimentary History of a Paratethys Sea (GeoEcoMar, Bucharest, Hungary). Keith, S.A., Baird, A.H., Hughes, T.P., Madin, J.S., Connolly, S.R. 2013. Faunal breaks and species composition of Indo-Pacific corals: The role of plate tectonics, environment and habitat distribution. Proceedings of the Royal Society B, 280 (1763): 20130818. Koenig, S.J., DeConto, R.M., Pollard, D. 2014. Impact of reduced Arctic sea ice on Greenland ice sheet variability in a warmer than present climate. Geophysical Research Letters, 41 (11), 3934–3943. Kohn, M.J., Fremd, T.J. 2008. Miocene tectonics and climate forcing of biodiversity, western United States. Geology, 36 (10), 783–786. Kümmerle, E., Radtke, G. 2012. Die Fossilien des Tertiärmeeres im Hanauer Becken. Jahresberichte der Wetterauischen Gesellschaft für die gesamte Naturkunde zu Hanau, 162, 59–77. Magyar, I., Radivojević, D., Sztanó, O., Synak, R., Ujszászi, K., Pócsik, M. 2013. Progradation of the paleo-Danube shelf margin across the Pannonian Basin during the Late Miocene and Early Pliocene. Global and Planetary, 103, 168–173. Mangerud, J., Jakobsson, J., Alexanderson, H., Astakhov, V., Clarke, G.K.C., Henriksen, M., Hjort, C., Krinner, G., Lunkka, J.-P., Möller, P., Murray, A., Nikolskaya, O., Saarnisto, M., Svendsen, J.I. 2004. Ice-dammed lakes and rerouting of the drainage of northern Eurasia during the Last Glaciation. Quaternary Science Reviews, 23 (11-13), 1313–1332. Mannion, P.D., Upchurch, P., Benson, R.B.J., Goswami, A. 2014. The latitudinal biodiversity gradient through deep time. Trends in Ecology & Evolution, 29 (1), 42–50. Micheels, A., Bruch, A.A., Eronen, J., Fortelius, M., Harzhauser, M., Utescher, T., Mosbrugger, V. 2010. Analysis of heat transport mechanisms from a Late Miocene model experiment with a fully-coupled atmosphere-ocean general circulation model. Palaeogeography, Palaeoclimatology, Palaeoecology, 304 (3-4), 337–350. Mittelbach, G.G., Schemske, D.W., Cornell, H.V., Allen, A.P., Brown, J.M., Bush, M.B., Harrison, S.P., Hurlbert, A.H., Knowlton, N., Lessios, H.A., McCain, C.M., McCune, A.R., McDade, L.A., McPeek, M.A., Near, T.J., Price, T.D., Ricklefs, R.E., Roy, K., Sax, D.F., Schluter, D., Sobel, J.M., Turelli, M. 2007. Evolution and the latitudinal diversity gradient: Speciation, extinction and biogeography. Ecology Letters, 10 (4), 315–331.

117

Mosbrugger, V., Utescher, T., Dilcher, D.L. 2005. Cenozoic continental climatic evolution of Central Europe. Proceedings of the National Academy of Sciences of the United States of America, 102 (42), 14964–14969. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015. A gastropod based biogeographic scheme for the European Neogene freshwater systems. Earth Science Reviews, 143, 98–116. Orme, C.D.L., Davies, R.G., Burgess, M., Eigenbrod, F., Pickup, N., Olson, V.A., Webster, A.J., Ding, T.-S., Rasmussen, P.C., Ridgely, R.S., Stattersfield, A.J., Bennett, P.M., Blackburn, T.M., Gaston, K.J., Owens, I.P.F. 2005. Global hotspots of species richness are not congruent with endemism or threat. Nature, 436 (7053), 1016–1019. Popov, S.V., Rögl, F., Rozanov, A.Y., Steininger, F.F., Shcherba, I.G., Kovac, M. 2004. Lithological-Paleogeographic maps of Paratethys. 10 Maps. Late Eocene to Pliocene. Courier Forschungsinstitut Senckenberg, 250, 1–46. Quan, C., Liu, Y.S., Tang, H., Utescher, T. 2014. Miocene shift of European atmospheric circulation from trade wind to westerlies. Scientific Reports, 4:5660. Reichenbacher, B., Krijgsman, W., Lataster, Y., Pippèrr, M., Van Baak, C.G.C., Chang, L., Kälin, D., Jost, J., Doppler, G., Jung, D., Prieto, J., Abdul Aziz, H., Böhme, M., Garnish, J., Kirscher, U., and Bachtadse, V. 2013. A new magnetostratigraphic framework for the Lower Miocene (Burdigalian/Ottnangian, Karpatian) in the North Alpine Foreland Basin. Swiss Journal of Geosciences, 106 (2), 309–334. Reid, W.V. 1998. Biodiversity hotspots. Trends in Ecology & Evolution, 13 (7), 275–280. Renema, W., Bellwood, D.R., Braga, J.C., Bromfield, K., Hall, R., Johnson, K.G., Lunt, P., Meyer, C.P., McMonagle, L.B., Morley, R.J., O’Dea, A., Todd, J.A., Wesselingh, F.P., Wilson, M.E.J., Pandolfi, J.M. 2008. Hopping hotspots: Global shifts in marine biodiversity. Science, 321 (5889), 654–657. Schön, I., Martens, K. 2004. Adaptive, pre-adaptive and non-adaptive components of radiations in ancient lakes: A review. Organisms Diversity & Evolution, 4 (3), 137– 156. Sissingh, W. 2001. Tectonostratigraphy of the West Alpine Foreland: Correlation of Tertiary sedimentary sequences, changes in eustatic sea-level and stress regimes. Tectonophysics, 333 (3-4), 361–400. Stoica, M, Lazăr, I., Krijgsman, W., Vasiliev, I., Jipa, D., Floroiu, A. 2013. Paleoenvironmental evolution of the East Carpathian foredeep during the late

118

Miocene-early Pliocene (Dacian Basin; Romania). Global and Planetary Change, 103, 135–148. ter Borgh, M., Vasiliev, I., Stoica, M., Knežević, S., Matenco, L., Krijgsman, W., Rundić, L., Cloetingh, S. 2013. The isolation of the Pannonian basin (Central Paratethys): New constraints from magnetostratigraphy and biostratigraphy. Global and Planetary Change, 103, 99–118. Van Baak, C.G.C., Vasiliev, I., Stoica, M., Kuiper, K.F., Forte, A.M., Aliyeva, E., Krijgsman, W. 2013. A magnetostratigraphic time frame for Plio-Pleistocene transgressions in the South Caspian Basin, Azerbaijan. Global and Planetary Change, 103, 119–134. van Leeuwen, C.H.A., van der Velde, G., van Lith, B., Klaassen, M. 2012. Experimental quantification of long distance dispersal potential of aquatic snails in the gut of migratory birds. PLoS One 7 (3):e32292. Wagner, C.E., Harmon, L.J., Seehausen, O. 2014a. Cichlid species-area relationships are shaped by adaptive radiations that scale with area. Ecology Letters, 17, 583–592. Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M.J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J.H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29. Wesselingh, F.P., 2007. Long-Lived Lake Molluscs as Island Faunas: a bivalve perspective. In: Renema, W. (Ed.), Biogeography, Time, and Place: Distributions, Barriers, and Islands. Springer, Dordrecht, pp. 275–314. Whittaker, R.J., Fernández-Palacios, J.M. 2007. Island Biogeography. Ecology, evolution, and conservation (Oxford Univ Press, Oxford, UK). Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science, 292 (5517), 686–693.

5.9. Supplementary material

5.9.1. Data setup

A thorough taxonomic and systematic check was conducted before the analysis. Only species undoubtedly assigned to freshwater or brackish environments were included. Localities with stratigraphic ranges exceeding 3 myr were excluded to avoid temporal biases. The data were

119 critically examined and adopted after being checked by taxonomic specialists (author team and several colleagues mentioned in the acknowledgments). Doubtful records and obvious misidentifications were excluded. Moreover, taxonomic history was considered, and changes to faunal lists by revisions of earlier works were implemented accordingly. The systematics of all taxa was updated according to latest published concepts (Neubauer et al., 2014a, b). The data derive from many publications of different research periods and schools. The taxonomic treatment is therefore not fully homogeneous. Nonetheless, we expect no systematic bias between faunas of the different time slices analyzed because most researchers worked on several intervals. Moreover, both extant and fossil species are almost solely based on shell characters. The extant species are therefore not more finely split than fossil forms. To achieve a well-based stratigraphical framework, all of the stratigraphic ages of the localities were checked and updated where necessary, following latest published concepts (Neubauer et al., 2015b).

5.9.2. Assessment of area and temporal duration

Information on the sizes of fossil freshwater systems is very limited. We collected data for 30 fossil and recent systems for which temporal durations are known and reliable reconstructions for area calculation are available (Neubauer et al., 2015a). These data are usually based on outcrop area of associated sediments or boundaries of geological basins (Neubauer et al., 2015a). In addition, we attempted to reconstruct areas of fossil systems based on the distribution of localities (if more than two localities per system). With the Minimum Bounding Geometry tool in ArcGIS 10.0, we created polygons from the localities and calculated their areas. We tested for the fit of the reconstructions by comparing newly calculated areas with available ones. The calculated and real areas were strongly different (Student t test: t = 3.359, P = 0.002). The deviations from real sizes were enormous, with a mean deviation of 55.77% and a range of 0.05–302.07%. Because of the overall poor potential of the reconstructions, we could not apply this approach to enhance our dataset and therefore use the restricted one with 30 systems for the ordinary least-squares regression analyses. These cover a wide range of different geographic regions, geological ages, areas, and temporal durations. We did not include more recent lakes, whose areas are well known, because of limited data on temporal duration. Moreover, most modern lake faunas are geologically very young, representing colonizations of postglacially formed lakes: they may not yet have reached an equilibrial balance of species gains (immigration, speciation) and

120 species losses (emigration, extinction) (eadie et al., 1986; Losos & Schluter, 2000; Reyjol et al., 2007). The species-area relationship for extant systems might therefore be severely downscaled. We tested for this effect, including faunas (more than five species) of all recent lakes for which data on surface area are available in the linear regression analysis. The species-area relationship was still significant but distinctly lower (log-log, r2 = 0.287, P < 0.001).

5.9.3. Climate data and regression analyses

As proxy for large-scale climate change, we used global oxygen data (Zachos et al., 2001). To facilitate comparison, the mean δ18O value was calculated for each million years bin. Data on European precipitation and temperature are available for certain time intervals and geographic regions (Mosbrugger et al., 2005; Böhme et al., 2011; Quan et al., 2014), but are still too patchy and inconsistent to be used in regression analyses. To avoid a bias of the estimation of the model parameters, we tested for collinearity among the predictor variables and calculated the variance inflation factor (VIF) of the variables. VIF values greater than 10 indicate the presence of multicollinearity (Quinn & Keough, 2002). Moreover, we tested for correlation among the predictor variables. The fit of the multiple regression model including all parameters was evaluated with a stepwise regression using a backward elimination procedure (Mac Nally, 2000). The adjusted 2 coefficient of determination (R adj) and the Akaike information criterion (AIC) were used to evaluate the resulting variable combinations. The results of the multiple regression and the tests for correlation and multicollinearity among the predictor variables are given in Tables S5.5–S5.18 and Fig. S5.4. Statistical analyses were performed in Past 2.17c (Hammer et al., 2001) and R 3.1.3 (R Core Team, 2014).

5.9.4. Potential effect of sampling bias

Sampling bias can seriously distort estimates of biodiversity (Smith et al., 2012). In our dataset, species richness is significantly correlated with the number of systems per million years bin (r = 0.487, P = 0.02). Even when excluding the much better sampled modern faunas, the correlation persists (r = 0.445, P = 0.04). This result is not surprising, considering that possibilities for diversification increase with the number of systems available. As discussed in the main text, the presence of aquatic systems as such is a main driver of biodiversity during time. To evaluate potential biases through spatially and temporally

121 uneven sampling, we tested for a relationship between sampling effort (estimated as the mean number of localities sampled per system and time slice) and diversification metrics (i.e., species richness, number of originations, number of extinctions and turnover rate; Fig. S3). The lack of any significant correlation suggests that the biodiversity signal in our dataset is not strongly distorted by sampling.

5.9.5. References

Böhme, M., Winklhofer, M., Ilg, A. 2011. Miocene precipitation in Europe: Temporal trends and spatial gradients. Palaeogeography, Palaeoclimatology, Palaeoecology, 304 (3-4), 212–218. Eadie, J.McA, Hurly, T.A., Montgomerie, R.D., Teather, K.L. 1986. Lakes and rivers as islands: species-area relationships in the fish faunas of Ontario. Environmental Biology of Fishes, 15, 81–89. Hammer, Ø, Harper, D.A.T., Ryan, P.D. 2001. Past: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontol Electronica, 4 (1), 1–9. Losos, J.B., Schluter, D. 2000. Analysis of an evolutionary species-area relationship. Nature, 408 (6814), 847–850. Mac Nally, R. 2000. Regression and model-building in conservation biology, biogeography and ecology: The distinction between—and reconciliation of—‘predictive’ and ‘explanatory’ models. Biodiversity and Conservstion, 9 (5), 655–671. Mosbrugger, V., Utescher, T., Dilcher, D.L. 2005. Cenozoic continental climatic evolution of Central Europe. Proceedings of the National Academy of Sciences of the United States of America, 102 (42), 14964–14969. Neubauer, T.A., Kroh, A., Harzhauser, M., Georgopoulou, E., Mandic, O. 2014a. Synopsis of valid species-group taxa for freshwater Gastropoda recorded from the European Neogene. ZooKeys, 435, 1–6. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2014b. The FreshGEN Database: Freshwater Gastropods of the European Neogene. Accessed at http://www.marinespecies.org/freshgen on 2015-10-15. Neubauer, T.A., Georgopoulou, E., Kroh, A., Harzhauser, M., Mandic, O., Esu, D. 2015a. Synopsis of European Neogene freshwater gastropod localities: Updated stratigraphy and geography. Palaeontologia Electronica 18.1.21A:1-7.

122

Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015b. A gastropod based biogeographic scheme for the European Neogene freshwater systems. Earth-Science Reviews, 143, 98–116. Quan, C., Liu, Y.S., Tang, H., Utescher, T. 2014. Miocene shift of European atmospheric circulation from trade wind to westerlies. Scientific Reports, 4:5660. Quinn, G.P., Keough, M.J. 2002. Experimental design and data analysis for biologists. Cambridge University Press, Cambridge. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org. Reyjol, Y., Hugueny, B., Pont, D., Bianco, P.G., Beier, U., Caiola, N., Casals, F., Cowx, I., Economou, A., Ferreira, T., Haidvogl, G., Noble, R., de Sostoa, A., Vigneron, T., Virbickas, T. 2007. Patterns in species richness and endemism of European freshwater fish. Global Ecology and Biogeography, 16, 65–75. Smith, A.B., Lloyd, G.T., McGowan, A.J. 2012. Phanerozoic marine diversity: Rock record modelling provides an independent test of large-scale trends. Proceedings of the Royal Society B, 279 (1746), 4489–4495. Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science, 292 (5517), 686–693.

123

5.9.6. Supplementary figures and tables

Figure S5.1. Plots of all significant linear regression analyses of species richness and turnover parameters with mean δ18O per million years. In the first plot, past biodiversity is approximated by raw species counts, whereas in the second plot, it is based on a range through assumption (i.e., that all taxa are supposed to have been continuously existing from their first to last appearance).

124

Figure S5.2. Plots of all linear regression analyses of species richness with surface area, duration, latitude (centroid), and longitude (centroid) for the 30 selected systems. The regression line is only indicated for the sole statistically significant correlation.

Figure S5.3. Correlation between sampling effort (estimated as the mean number of localities sampled per system and time slice) and diversification metrics.

Figure S5.3. Correlation between surface area and duration for the 30 selected systems (both axes log10- transformed).

125

Table S5.1. Extended table of species richness hotspots for European non-marine gastropods since Miocene times.

Number of Max. Max. Interval species number of temporal Countries of max. Age Lake/Basin Type of environment across coevally range of Reference richness (ISO code) entire present environment [myr] duration species [myr]

Recent Ohrid AL, MK freshwater lake 68 65 0 1.5-0 Wagner et al. (2014)

Recent Caspian Sea AZ, IR, KZ, RU, TM brackish lake 105 92 0 0.88-0 Van Baak et al. (2013)

Gelumbauskaitė & Recent Curonian Lagoon LT brackish lagoon 42 42 0 0.0135-0 Šečkus (2005)

estimated from locality Pleistocene Tiberino IT freshwater lake 42 32 1 3.1-1.55 ages

Pleistocene Kos GR freshwater lake 41 26 1 3.0-1.4 Böger (1983)

fluvio-lacustrine estimated from Sissingh Pleistocene Bresse FR 64 44 2 4.5-1.5 system (2001) and locality ages

Pliocene Rioni GE fluvial plain 56 52 3 5.5-1.6 Popov et al. (2004)

constrained by Krstić Pliocene Slavonia BA, HR, ?HU, RS freshwater lake 163 145 3 4.0-2.6 (2003) and Magyar et al. (2013)

constrained by Chalot- Pliocene Transylvania RO freshwater lake 78 78 4 3.8-0.8 Prat & Girbacea (2009)

brackish to freshwater Jipa & Olariu (2009) and Pliocene Dacia BG, MD, RO, UA 303 159 4 8.6-2.6 lake Andreescu et al. (2013)

126

Pliocene Metohia KV freshwater lake 70 36 5 6.04-2.588 Elezaj et al. (2010)

Late Miocene Galați MD, UA brackish embayment 133 131 7 8.6-4.4 Jipa & Olariu (2009)

Late Miocene Bresse-Valence FR fluvial plain 60 56 8 10.0-8.0 Sissingh (2001

AT, BA, CZ, HR, Late Miocene Pannon brackish lake 605 248 10 11.6-4.0 Magyar et al. (2013) HU, RO, RS, SI, SK

Middle Miocene Soceni RO brackish embayment 51 51 12 12.7-12.3 Lukeneder et al. (2011)

Middle Miocene Bakony HU wetlands (?) 87 48 13 17.5-12.5 Kókay (2006)

basin constrained by Ries impact, top by Middle Miocene Steinheim DE freshwater lake 42 42 13 15-13.9 mammal zone MN7 (Tütken et al., 2009)

Middle Miocene Sinj HR freshwater lake 58 50 15 18.0-15.0 De Leeuw et al. (2010)

Middle Miocene Drniš HR freshwater lake 43 43 15 15.7-15.0 Neubauer et al. (2015)

Upper Brackish Water Reichenbacher et al. Early Miocene DE brackish lake 40 37 16 16.6-17.6 Molasse (2013)

Cahuzac & Janssen Early Miocene Aquitaine FR brackish wetlands (?) 59 29 20 28.1-11.6 (2010)

Kümmerle & Radtke Early Miocene Hanau-Wetterau DE brackish embayment 49 15 21 28.1-16.1 (2012)

References

127

Andreescu, I, Codrea, V., Lubenescu, V., Munteanu, T., Petculescu, A., Ştiucǎ, E., Terzea, E. 2013. New developments in the Upper Pliocene-Pleistocene stratigraphic units of the Dacian Basin (Eastern Paratethys), Romania. Quaternary International, 284, 15–29. Böger, H. 1983. Stratigraphische und tektonische Verknüpfungen kontinentaler Sedimente des Neogens im Ägäis-Raum. Geologische Rundschau, 72, 771–814. Cahuzac, B., Janssen, A.W. 2010. Eocene to Miocene holoplanktonic Mollusca (Gastropoda) of the Aquitaine Basin, southwest France. Scripta Geologica, 141, 1–193. Chalot-Prat, F., Girbacea, R. 2000. Partial delamination of continental mantle lithosphere, uplift-related crust-mantle decoupling, volcanism and basin formation: a new model for the Pliocene-Quaternary evolution of the southern East-Carpathians, Romania. Tectonophysics 327, 83–107. De Leeuw, A., Mandic, O., Vranjković, A., Pavelić, D., Harzhauser, M., Krijgsman, W., Kuiper, K.F. 2010. Chronology and integrated stratigraphy of the Miocene Sinj Basin (Dinaride Lake System, Croatia). Palaeogeography, Palaeoclimatology, Palaeoecology, 292, 155–167. Elezaj, Z., Tmava, A., Pashko, P., Vaso, P. 2010. The Neogene of the Dukagjini Basin. Oltenia. Studii şi comunicări. Ştiinţele Naturii, 26 (1), 291–295. Gelumbauskaitė, L.Ž., Šečkus, J. 2005. Late-Glacial-Holocene history in Curonian Lagoon (Lithuanian sector). Baltica, 18, 77–82. Jipa D.C., Olariu, C. 2009. Dacian Basin: Depositional Architecture and Sedimentary History of a Paratethys Sea (GeoEcoMar, Bucharest, Hungary). Kókay, J. 2006. Nonmarine mollusc fauna from the Lower and Middle Miocene, Bakony Mts., W Hungary. Geologica Hungarica, Series Palaeontologica, 56, 1–196. Krstić, N. 2003. Paludinian beds of Voivodina. Chronostratigraphie und Neostratotypen. Neogen der Zentrale Paratethys, Bd. X, Pl2. Romanien, eds Papaianopol, I., Marinescu, F., Krstić, N., Macaleţ, R. (Editura Academiei Române, Bucuresti, Hungary), pp 77–81.

128

Kümmerle, E., Radtke, G. 2012. Die Fossilien des Tertiärmeeres im Hanauer Becken. Jahresberichte der Wetterauischen Gesellschaft für die gesamte Naturkunde zu Hanau, 162, 59–77. Lukeneder, S., Zuschin, M., Harzhauser, M., Mandic, O. 2011. Spatiotemporal signals and palaeoenvironments of endemic molluscan assemblages in the marine system of the Sarmatian Paratethys. Acta Palaeontologica Polonica, 56, 767–784. Magyar, I., Radivojević, D., Sztanó, O., Synak, R., Ujszászi, K., Pócsik, M. 2013. Progradation of the paleo-Danube shelf margin across the Pannonian Basin during the Late Miocene and Early Pliocene. Global and Planetary Change, 103, 168–173. Neubauer, T.A., Georgopoulou, E., Kroh, A., Harzhauser, M., Mandic, O., Esu, D. 2015. Synopsis of European Neogene freshwater gastropod localities: Updated stratigraphy and geography. Palaeontologia Electronica 18.1.21A:1-7. Popov, S.V., Shcherba, I.G., Ilyina, L.B., Nevesskaya, L.A., Paramonova, N.P., Khondkarian, S.O., Magyar, I. 2004. Late Miocene to Pliocene palaeogeography of the Paratethys and its relation to the Mediterranean. Palaeogeography, Palaeoclimatology, Palaeoecology, 238, 91–106. Reichenbacher, B., Krijgsman, W., Lataster, Y., Pippèrr, M., Van Baak, C.G.C., Chang, L., Kälin, D., Jost, J., Doppler, G., Jung, D., Prieto, J., Abdul Aziz, H., Böhme, M., Garnish, J., Kirscher, U., and Bachtadse, V. 2013. A new magnetostratigraphic framework for the Lower Miocene (Burdigalian/Ottnangian, Karpatian) in the North Alpine Foreland Basin. Swiss Journal of Geosciences, 106 (2), 309–334. Sissingh, W. 2001. Tectonostratigraphy of the West Alpine Foreland: Correlation of Tertiary sedimentary sequences, changes in eustatic sea- level and stress regimes. Tectonophysics, 333 (3-4), 361–400. Tütken, T., Vennemann, T.W., Janz, H., Heizmann, E.P.J. 2006. Palaeoenvironment and palaeoclimate of the Middle Miocene lake in the Steinheim basin, SW Germany: A reconstruction from C, O, and Sr isotopes of fossil remains. Palaeogeography, Palaeoclimatology, Palaeoecology, 241, 457–491. Van Baak, C.G.C., Vasiliev, I., Stoica, M., Kuiper, K.F., Forte, A.M., Aliyeva, E., Krijgsman, W. 2013. A magnetostratigraphic time frame for Plio-Pleistocene transgressions in the South Caspian Basin, Azerbaijan. Global and Planetary Change, 103, 119–134.

129

Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M.J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J.H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29.

Table S5.2. Biodiversity and climate parameters per million years, together with results of linear regressions between them. log(mean log(maximum log(total no. of Age no. of no. of log(total no. no. of no. of no. of turnover beta mean δ18O no. of species species, range- (myr) systems localities of species) species per originations extinctions rate diversity per lake) through corr.) lake)

0 1236 2402 4.079463602 2.723455672 0.675545894 1.963787827 2.723455672 405 0 405 0

1 30 225 3.928239075 2.326335861 0.740837878 1.505149978 2.385606274 88 119 207 0.74899

2 31 149 3.694177215 2.403120521 0.79740771 1.643452676 2.46686762 121 138 259 0.55269

3 30 231 3.324486486 2.64738297 0.786241885 2.161368002 2.703291378 224 333 557 0.71593

4 17 276 3.033952096 2.550228353 0.860262668 2.201397124 2.650307523 214 166 380 0.58198

5 22 280 2.975627158 2.63447727 0.972048674 2.240549248 2.717670503 205 289 494 0.75827

6 10 170 2.986449275 2.437750563 1.173690979 2.26245109 2.617000341 95 97 192 0.58298

7 11 111 2.788151659 2.465382851 1.072397519 2.117271296 2.648360011 153 126 279 0.61484

8 7 89 2.85012987 2.411619706 0.930110939 2.303196057 2.555094449 119 67 186 0.68727

9 12 136 2.764802632 2.531478917 0.990601274 2.380211242 2.619093331 160 176 336 0.62876

10 10 175 2.665426357 2.45331834 0.705112473 2.394451681 2.565847819 161 112 273 0.56731

130

11 11 98 2.57 2.365487985 1.142712728 2.1430148 2.445604203 151 72 223 0.57752

12 19 116 2.481593407 2.220108088 0.755856102 1.755874856 2.322219295 98 82 180 0.69849

13 7 26 2.324157783 2.025305865 1.017719209 1.681241237 2.195899652 54 45 99 0.63971

14 12 41 2.077706422 1.755874856 0.653653258 1.255272505 2.096910013 18 22 40 0.66871

15 28 123 1.719639175 2.315970345 0.792169971 1.698970004 2.376576957 144 131 275 0.77273

16 29 165 1.73729927 2.086359831 0.601727061 1.568201724 2.170261715 91 54 145 0.68389

17 14 92 1.761444444 1.86332286 0.716293597 1.397940009 1.934498451 55 29 84 0.71282

18 6 34 1.941568627 1.662757832 0.800302528 1.342422681 1.72427587 20 22 42 0.73109

19 2 22 1.997727273 1.477121255 1.024609011 1.447158031 1.62324929 6 9 15 0.47368

20 6 40 2.05 1.69019608 0.878995501 1.462397998 1.72427587 20 17 37 0.41772

21 12 97 2.057272727 1.707570176 0.787521991 1.230448921 1.72427587 36 20 56 0.44

22 8 10 1.994285714 1.278753601 0.357166562 0.954242509 1.278753601 0 2 2 0.57143

Table S5.3. Results of linear regressions (dependent variable: mean δ18O). Slope a Intercept b Error a Error b r r² p log(total no. of species) 0.42481 1.0708 0.090142 0.24226 0.71693 0.514 0.000 log(mean no. of species per lake) 0.039896 0.73248 0.058693 0.15774 0.14673 0.022 0.504 log(maximum no. of species per lake) 0.3047 0.99516 0.1166 0.31337 0.49536 0.245 0.016 log(total no. of species, range-through corr.) 0.41007 1.2063 0.095932 0.25782 0.68211 0.465 0.000

131 no. of originations 91.556 -123.37 21.236 57.074 0.68522 0.470 0.000 no. of extinctions 54.708 -49.727 24.715 66.424 0.43494 0.189 0.038 turnover rate 146.26 -173.09 36.543 98.212 0.65783 0.433 0.001 Whittaker's beta diversity -0.07225 0.78903 0.049031 0.13177 -0.30612 0.094 0.155

Table S5.4. Data on geography, surface area, duration, and species richness for the 30 selected lake systems. Reconstructed area Lake Latitude N (centroid) Longitude E (centroid) Area [km²] Duration [Ma] Species richness (ArcGIS) Balaton 46.86882707 17.73482561 579.6972923 0.015 34 Bresse 46.64092713 5.243419294 9651.035536 275.700633 1.8 35 Dacia (Dacian) 45.131743 26.29519 98730.88432 115426.918 4.5 167 Dacia (Maeotian) 45.370621 26.633002 96812.0658 53088.3301 2.56 69 Dacia (Pontian) 45.159783 26.160647 83299.8677 54091.3841 3.9 138 Denizli 37.93372043 29.07008047 1307.017076 470.848094 8.5 14 Drniš 43.836238 16.263161 24.885221 7.1715132 0.7 43 Gacko 43.13967 18.544173 39.574537 3.02637294 0.7 12 Granada 37.10994772 -3.832040476 931.429776 230.476311 1.7 18 Groisenbach 47.5431035 15.2700923 22.62960789 0.2 12 Kosovo 42.58965138 21.06555387 920.4434807 290.095 1.34 24 Kupres 43.985708 17.215151 65.401531 0.4 23 Le Locle 47.07394 6.779174 11.382504 0.00605282 0.5 18 Metohia (Pontian) 42.582151 20.664146 438.970413 368.8574 1.34 39 Metohia (Romanian) 42.501379 20.543718 1805.04745 362.813118 3.452 38 Nördlinger Ries 48.8851942 10.56374635 438.2377922 53.3537384 1.2 6 North Bohemian Lake 50.40579215 13.55768328 1303.91645 245.217576 4.6 10

132

Ohrid 41.03679747 20.71629463 356.3727148 1.5 66 Pannon (Portaferrian) 46.00639 19.276999 116125.6729 140675.434 7.6 402 Pannon (Serbian) 46.49107 20.339897 237458.4771 205989.723 2.6 319 Prespa 40.89733898 21.02059952 272.6810725 80.7731824 0.11 23 Randeck Maar 48.57554622 9.525521728 1.076680058 0.3 4 Sinj 43.694693 16.68236 131.796903 27.3018974 3 57 Slavonia 45.40435541 18.76501759 24243.24577 27437.0653 1.6 163 Sofia 42.73182081 23.38492028 908.399597 1.5 6 Šoštanj 46.38209687 15.07099883 13.86935183 1.77348121 0.5 9 Steinheim 48.68595531 10.06957468 7.321887955 1.1 42 Tiberino 42.79615554 12.43924877 544.43449 1644.57263 0.25 42 Transylvania 45.83983475 25.75057558 2221.11568 827.018242 3 78 Turiec 48.95905161 18.87375748 305.8967089 5.5 5 Note: Reconstructed areas are only available for systems based on at least three localities.

133

Supplementary Table S5.5. Results of linear regressions of latitude/longitude on species richness.

Early Miocene

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a 0.012402 Slope a 0.0031616 Intercept b -0.24022 Intercept b 0.32129 SE a 0.0067717 SE a 0.0028786 SE b 0.32177 SE b 0.029421 χ2 65.181 χ2 65.459 r 0.081469 r 0.048961 r2 0.0066372 r2 0.0023972 t statistic 1.8314 t statistic 1.0983 p(uncorrel) 0.067628 p(uncorrel) 0.27259 Permutat. p 0.0659 Permutat. p 0.273

Middle Miocene

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a -0.00052913 Slope a 0.0028392 Intercept b 0.42557 Intercept b 0.35532 SE a 0.0058754 SE a 0.0020233 SE b 0.27131 SE b 0.0372 χ2 69.968 χ2 69.677 r -0.0041585 r 0.064661 r2 1.73E-05 r2 0.004181 t statistic -0.090058 t statistic 1.4033 p(uncorrel) 0.92828 p(uncorrel) 0.1612 Permutat. p 0.9253 Permutat. p 0.1626

Late Miocene

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a 0.014639 Slope a -0.0025216 Intercept b -0.18801 Intercept b 0.51671 SE a 0.0044965 SE a 0.0014283 SE b 0.20189 SE b 0.030576 χ2 210.75 χ2 212.16 r 0.097037 r -0.052795 r2 0.0094162 r2 0.0027874 t statistic 3.2556 t statistic -1.7654 p(uncorrel) 0.0011656 p(uncorrel) 0.077772 Permutat. p 0.0011 Permutat. p 0.0745

134

Pliocene

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a 0.011427 Slope a -0.00034255 Intercept b 0.022405 Intercept b 0.53398 SE a 0.0067339 SE a 0.0024031 SE b 0.29741 SE b 0.057362 χ2 106.72 χ2 107.25 r 0.070107 r -0.0059036 r2 0.004915 r2 3.49E-05 t statistic 1.6969 t statistic -0.14255 P (uncorrel) 0.090241 P (uncorrel) 0.8867 Permutat. p 0.0914 Permutat. p 0.8873

Pleistocene

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a 0.022858 Slope a -0.0029997 Intercept b -0.41719 Intercept b 0.71073 SE a 0.0025898 SE a 0.0011016 SE b 0.12342 SE b 0.02119 χ2 154.42 χ2 165.71 r 0.27437 r -0.087681 r2 0.075277 r2 0.0076879 t statistic 8.8263 t statistic -2.7229 P (uncorrel) 5.06E-18 P (uncorrel) 0.0065885 Permutat. p 0.0001 Permutat. p 0.0063

Recent

Latitude vs. Species richness Longitude vs. Species richness (log) (log) Slope a -0.00019113 Slope a 0.0053645 Intercept b 0.65017 Intercept b 0.56512 SE a 0.0013841 SE a 0.0010302 SE b 0.072566 SE b 0.01829 χ2 212.44 χ2 208.05 r -0.0038553 r 0.14386 r2 1.49E-05 r2 0.020697 t statistic -0.13809 t statistic 5.2072 P (uncorrel) 0.89019 P (uncorrel) 2.23E-07 Permutat. p 0.8899 Permutat. p 0.0001

135

Table S5.6. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Multiple regression (residuals). Min 1Q Median 3Q Max -0.76953 -0.19613 -0.03409 0.26864 0.71155

Table S5.7. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Multiple regression (coefficients). Estimate Std.-Error t-value p (Intercept) 0.852794 1.11083 0.768 0.449851 Latitude -0.004296 0.02362 -0.182 0.857153 Longitude 0.003199 0.011555 0.277 0.78417 Log (area) 0.277281 0.06679 4.152 0.000335 Log (duration) -0.094437 0.144378 -0.654 0.519022 Residual SE: 0.3964 on 25 df. Multiple R2: 0.4999. Adjusted R2: 0.4199. F-statistic: 6.247 on 4 and 25 df, P value: 0.001255.

Table S5.8. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Backward stepwise regression.

Start: AIC=-50.98 log_richness vs. Latitude + Longitude + log_area + log_duration Df Sum-of-Sq RSS AIC - Latitude 1 0.0052 3.9344 -52.943 - Longitude 1 0.01205 3.9412 -52.891 -log_duration 1 0.06724 3.9964 -52.474 3.9292 -50.983 - log_area 1 2.70883 6.638 -37.252

Step: AIC=-52.94 log_richness vs. Longitude + log_area + log_duration Df Sum-of-Sq RSS AIC - Longitude 1 0.01413 3.9485 -54.836 -log_duration 1 0.06674 4.0011 -54.439 3.9344 -52.943 - log_area 1 2.7184 6.6528 -39.185

136

Step: AIC=-54.84 log_richness vs. log_area + log_duration Df Sum-of-Sq RSS AIC -log_duration 1 0.0626 4.0111 -56.364 3.9485 -54.836 - log_area 1 3.2721 7.2206 -38.728

Step: AIC=-56.36 log_richness vs. log_area Df Sum-of-Sq RSS AIC 4.0111 -56.364 - log_area 1 3.8456 7.8567 -38.195

Table S5.9. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of area (log) and species richness (residuals). Min 1Q Median 3Q Max -0.74911 -0.20629 0.01587 0.25435 0.64959

Table S5.10. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of area (log) and species richness (coefficients). Estimate Std.-Error t-value p (Intercept) 0.74503 0.16037 4.646 7.30E-05 log_area 0.26442 0.05104 5.181 1.69E-05 Residual SE: 0.3785 on 28 df. Multiple R2: 0.4895. Adjusted R2: 0.4712. F-statistic: 26.84 on 1 and 28 df, P value: 69e-05.

Table S5.11. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of duration (log) and species richness (residuals). Min 1Q Median 3Q Max -0.96077 -0.33832 0.04408 0.29739 0.92478

137

Table S5.12. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of duration (log) and species richness (coefficients). Estimate Std.-Error t-value p (Intercept) 1.47606 0.09348 15.789 1.8E-15 log_duration 0.24811 0.15797 1.571 0.128 Residual SE: 0.5078 on 28 df. Multiple R2: 0.08097. Adjusted R2: 0.04815. F-statistic: 2.467 on 1 and 28 df, P value: 0.1275.

Table S5.13. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of latitude and species richness (residuals). Min 1Q Median 3Q Max -0.83124 -0.39529 0.05352 0.25798 1.12894

Table S5.14. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of latitude and species richness (coefficients). Estimate Std.-Error t-value p (Intercept) 2.22714 1.39368 1.598 0.121 Latitude -0.01634 0.03103 -0.527 0.603 Residual SE: 0.5271 on 28 df. Multiple R2: 0.000981. Adjusted R2:- 0.02555. F-statistic: 0.2774 on 1 and 28 df, P value: 0.6026.

Table S5.15. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of longitude and species richness (residuals). Min 1Q Median 3Q Max -0.86593 -0.39292 0.02322 0.28008 1.06126

Table S5.16. Multiple and linear regressions of latitude, longitude, area (log), and duration (log) with species richness for 30 selected systems: Linear regression of longitude and species richness (coefficients). Estimate Std.-Error t-value p (Intercept) 1.06843 0.24197 4.415 0.000137 Longitude 0.02462 0.01294 1.902 0.067509 Residual SE: 0. 4985 on 28 df. Multiple R2: 0. 1144. Adjusted R2:- 0. 08279. F-statistic: 3.617 on 1 and 28 df, P value: 0.06751.

138

Table S5.17. Test for multicollinearity of predictor variables. Latitude Longitude Area Duration 1.024413 1.260245 1.561088 1.370624

Table S5.18. Correlation table of predictor variables. Latitude Longitude Area Duration

Latitude 0.434 0.595 0.709 Longitude -0.148 0.016 0.130 Area -0.101 0.438 0.004 Duration -0.071 0.282 0.516 r values are given in lower triangle, P values in upper triangle; significant correlations are marked bold. Only moderate correlations exist between area and duration as well as area and longitude (Fig. S4).

Dataset S5.1. Full dataset used for the present study, including geographic and stratigraphic information, lake affiliation (if applicable), and species occurrences per locality/sample (Due to size matters this table is provided at http://www.pnas.org/content/112/37/11478?tab=ds)

139

CHAPTER 6: Conclusions

The importance of climatic and geodynamic processes as driving forces in shaping spatial patterns of species distribution has been stressed several times in the past (e.g., Hewitt, 2000; Reyjol et al., 2007; Neubauer et al., 2015). Our findings both on recent and freshwater gastropods of European lacustrine systems of the last 23 myr strongly support this view. During the Neogene, several examples of autochthonous radiations and extinctions of indigenous lacustrine faunas (e.g., Harzhauser & Mandic, 2008; Rasser, 2014) paralleling the rise and demise of long-lived lakes, show exactly the paramount importance of such lakes for the evolution of freshwater gastropod faunas. The longevity and stability provided by the geological basins accommodating long-lived lakes assist to the establishment of hotspots for aquatic gastropods and increased provinciality (Fig. 6.1; see Neubauer et al., 2015). An excellent example of such a mega-hotspot of aquatic gastropods in Europe is Lake Pannon in

Figure 6.1. Total and endemic species richness of the biogeographic provinces proposed by Neubauer et al. (2015) for the Early Miocene, Middle Miocene, Late Miocene and Pliocene. The increased number of regional endemics plays a decisive role in the observed high degree of provincialism. Figure modified after Neubauer et al. (2015). the Pannonian Basin with an estimated duration of 5.8 myr (after Harzhauser & Mandic, 2008 and references therein) and 74% endemic species (after Neubauer et al., 2015). Further faunal

140 diversification in such lakes is related to climate and surface area of the lakes (see Fig. 6.2; e.g., Graham et al., 2006; Hof et al., 2008; Neubauer et al., 2015). However, global cooling and successive cycles of glacial events (Zachos et al., 2001) along with the decline of the stable geological basins facilitating such long-lived lakes, lead to their demise and subsequently the reduction of available biotopes for lacustrine gastropods (e.g., see Neubauer et al., 2016). By the end of the Pliocene, the continual presence of several long-lived lakes was terminated e.g., Lake Slavonia (Krstić, 2003; Magyar et al., 2013), Lake Dacia (Jipa & Olariu, 2009; Andreescu et al., 2013) and Lake Metohia (Elezaj et al., 2010). Subsequently, a major turnover event marked the transition to the Early Pleistocene.

Figure 6.2. Species-area regression curves for Late Miocene and Pliocene lakes. The regressions indicate a strong positive relationship between species richness and surface area of long-lived lakes. Figure modified after Neubauer et al. (2015).

At the beginning of the Quaternary, it is mainly the remaining long-lived and highly endemic aquatic systems like lakes Bresse (Schlickum & Puisségur, 1977, 1978) and Tiberino (Esu & Girotti, 1975) that account for increased species richness. However, the recorded freshwater gastropod richness is several times smaller in magnitude compared to the preceding Neogene period (Fig. 6.3). From Middle to Late Pleistocene and Holocene the most diverse assemblages of freshwater gastropods are mostly linked to fluvial and/or lacustrine systems with short temporal durations (e.g., Ložek, 1964; Szymanek, 2011; Alexandrowicz, 2013; Limondin- Lozouet et al., 2013). The few new species that appear e.g., Lithoglyphus jahni Urbański, 1975 and Parafossarulus crassitesta (Brömme, 1883) (see Meijer & Preece, 1996), are widespread and not associated with any known long-lived lakes. Thus, the high degree of

141

Neogene provincialism is gradually lost and species with wide geographical ranges become more common. The decline of long-lived lakes parallels the increase of generalist species. At the same time, the role of climate is secondary in species temporal turnover; between the Last Interglacial and Last Glacial species composition in freshwater systems of Europe remains unchanged. Modern European lakes, except for few exceptions e.g., Lake Ohrid (Wagner et al., 2014), are younger than 21 kyr (Cohen, 2003). The link of modern lacustrine distributions to the immediate past is strong. During the last interglacial phases of the Quaternary i.e., Last Interglacial and Holocene, recolonization and range expansion of generalist species was rather fast, although modern lake faunas are far from reaching equilibrium (see also Dehling et al., 2010).

Figure 6.3. Total species richness calculated for the six studied Neogene and Quaternary time intervals. Highest gastropod species richness was recorded in the Late Miocene. Total species richness calculated partly after Neubauer et al. (2015) and Georgopoulou et al. (2016).

Following the retreat of the last ice sheet, the newly formed lakes up to the northern most edge of Europe were resettled by gastropods in a remarkably short time (see Økland, 1990 and references therein). Several of them are in fact so rich that a small hotspot of gastropod diversity around the Baltic region can be discerned. Still, lakes at the south-eastern tips of Europe are very diverse, inhabited by a plethora of endemic species. Thus, beta diversity between lakes of different deglaciation history is quite prominent. Furthermore, species composition is variably controlled by dispersal limitations and environmental gradients, also reflecting differences in faunal history of the lakes.

142

This PhD dissertation was oriented towards exploring and understanding the processes that shaped freshwater gastropod patterns in European lacustrine systems during the last 23 myr. Future work should be focused on exploring other aspects such as passive and active dispersal of species (e.g., Kappes & Haase, 2012; van Leeuwen et al., 2012, 2013) and human impact (e.g., Albrecht et al., 2014) that have also been acknowledged as crucial for species distribution.

6.1. References

Albrecht, C. Föller, K., Clewing, C., Hauffe, T., Wilke, T. 2014. Invaders versus endemics: alien gastropod species in ancient Lake Ohrid. Hydrobiologia, 739, 163–174. Alexandrowicz, W.P. 2013. Late Glacial and Holocene molluscan assemblages in deposits filling palaeolakes in Northern Poland. Studia Quaternaria, 30, 5–17. Andreescu I, Codrea, V., Lubenescu, V., Munteanu, T., Petculescu, A., Ştiucǎ, E., Terzea, E. 2013. New developments in the Upper Pliocene-Pleistocene stratigraphic units of the Dacian Basin (Eastern Paratethys), Romania. Quaternary International, 284, 15–29. Cohen, A. S. 2003. Paleolimnology: the history and evolution of lake systems. Oxford University Press. Dehling, D.M., Hof, C., Brändle, M., Brandl, R. 2010. Habitat availability does not explain the species richness patterns of European lentic and lotic freshwater animals. Journal of Biogeography, 37, 1919–1926. Elezaj, Z., Tmava, A., Pashko, P., Vaso, P. 2010. The Neogene of the Dukagjini Basin. Oltenia. Studii şi comunicări. Ştiinţele Naturii, 26 (1), 291–295. Esu, D., Girotti, O. 1975. La malacofauna continentale del Plio-Pleistocene dell’Italia centrale. I. Paleontologia. Geologica Romana, 13, 203–294. Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016. Distribution patterns of European lacustrine gastropods – a result of environmental factors and deglaciation history. Hydrobiologia, 775, 69–82. Graham, C.H., Moritz, C., Williams, S.E. 2006. Habitat history improves prediction of biodiversity in rainforest fauna. Proceedings of the National Academy of Sciences of the United States of America, 103, 632–636. Hof, C., Brändle, M., Brandl, R. 2008. Latitudinal variation of diversity in European freshwater animals is not concordant across habitat types. Global Ecology and Biogeography, 17, 539–546.

143

Harzhauser, M., Mandic, O. 2008 Neogene lake systems of Central and South-Eastern Europe: Faunal diversity, gradients and interrelations. Palaeogeography, Palaeoclimatology, Palaeoecology, 260, 417–434. Hewitt, G. M., 2000. The genetic legacy of the Quaternary ice ages. Nature, 405, 907–913. Jipa D.C., Olariu, C. 2009. Dacian Basin: Depositional Architecture and Sedimentary History of a Paratethys Sea (GeoEcoMar, Bucharest, Hungary). Kappes, H., Haase, P. 2012. Slow, but steady: dispersal of freshwater molluscs. Aquatic Sciences, 74, 1–14. Krstić, N. 2003. Paludinian beds of Voivodina. Chronostratigraphie und Neostratotypen. Neogen der Zentrale Paratethys, Bd. X, Pl2. Romanien, eds Papaianopol, I., Marinescu, F., Krstić, N., Macaleţ, R. (Editura Academiei Române, Bucuresti, Hungary), pp 77–81. van Leeuwen, C.H.A., van der Velde, G., van Lith, B., Klaassen, M. 2012. Experimental Quantification of Long Distance Dispersal Potential of Aquatic Snails in the Gut of Migratory Birds. PLoS ONE 7: e32292 van Leeuwen, C.H.A., Huig, N., van der Velde, G., van Alen, T.A., Wagemaker, C.A.M., Sherman, C.D.H., Klaassen, M., Figuerola, J. 2013. How did this snail get here? Several dispersal vectors inferred for an aquatic invasive species. Freshwater Biology, 58, 88–99. Limondin-Lozouet, N., Preece, R.C., Antoine, P. 2013. The Holocene tufa at Daours (Somme Valley, northern France): Malacological succession and palaeohydrological implications. Boreas, 42 (3), 650–663. Ložek, V. 1964. Quartärmollusken der Tschechoslovakei. Rozpravy Ústředního Ústavu Geologického, 31, 1–374. Magyar, I., Radivojević, D., Sztanó, O., Synak, R., Ujszászi, K., Pócsik, M. 2013. Progradation of the paleo-Danube shelf margin across the Pannonian Basin during the Late Miocene and Early Pliocene. Global and Planetary Change, 103, 168–173. Meijer, T., Preece, R.C. 1996. Malacological evidence relating to the stratigraphical position of the Cromerian. 53–82. In: Turner, C. (eds) The Early Middle Pleistocene in Europe. A.A. Balkema, Rotterdam, Brookfield. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015. A gastropod- based biogeographic scheme for the European Neogene freshwater systems. Earth- Science Reviews, 143, 98–116.

144

Neubauer, T.A., Harzhauser, M., Mandic, O., Kroh, A., Georgopoulou, E., 2016. Evolution, turnovers and spatial variation of the gastropod fauna of the late Miocene biodiversity hotspot Lake Pannon. Palaeogeography, Palaeoclimatology, Palaeoecology, 442, 84– 95. Økland, J., 1990. Lakes and Snails: Environment and Gastropoda in 1500 Norwegian lakes,

ponds and rivers. UBS/Dr W. Backhuys, Oegstgeest, the Netherlands.

Rasser, M.W., 2014. Evolution in isolation: the Gyraulus species flock from Miocene Lake Steinheim revisited. Hydrobiologia, 739, 7–24. Reyjol, Y., Hugueny, B., Pont, D., Bianco, P.G., Beier, U., Caiola, N., Casals, F., Cowx, I., Economou, A., Ferreira, T., Haidvogl, G., Noble, R., de Sostoa, A., Vigneron, T., Virbickas, T. 2007. Patterns in species richness and endemism of European freshwater fish. Global Ecology and Biogeography, 16, 65–75. Schlickum, W.R., Puisségur, J.-J. 1977. Die Molluskenfauna des Altpleistozäns von St. Bernard (Département Côte-d'Or). Archiv für Molluskenkunde, 107 (4/6), 273–283. Schlickum, W.R., Puisségur, J.-J. 1978. Die Molluskenfauna der Schichten mit Viviparus burgundinus und Pyrgula nodotiana von Montagny-les-Beaune (Dép. Côte-d'Or). Archiv für Molluskenkunde, 109 (1/3), 1–26. Szymanek, M. 2011. Climate oscillations of the Holsteinian (Mazovian) Interglacial recorded in shell morphometry of Viviparus diluvianus (Kunth, 1865) from eastern Poland. Quaternary International, 241, 143–159. Wagner, B., Wilke, T., Krastel, S., Zanchetta, G., Sulpizio, R., Reicherter, K., Leng, M. J., Grazhdani, A., Trajanovski, S., Francke, A., Lindhorst, K., Levkov, Z., Cvetkoska, A., Reed, J.M., Zhang, X., Lacey, J. H., Wonik, T., Baumgarten, H., Vogel, H. 2014. The SCOPSCO drilling project recovers more than 1.2 million years of history from Lake Ohrid. Scientific Drilling, 17, 19–29. Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292 (5517), 686–693.

145

Curriculum vitae

Full name: Elisavet Georgopoulou Date of Birth: 28th of November 1987 Place of Birth: Thessaloniki, Greece Nationality: Greek Marital status: Single

Education

May 2013 – Jul. 2016: PhD candidate of the University of Graz, Institute of Earth Sciences & project researcher at the Natural History Museum of Vienna (FWF-Project P25365-B25: "Freshwater systems in the Neogene and Quaternary of Europe: Gastropod biodiversity, provinciality, and faunal gradients") Nov. 2012: Master of Science in "Management of Terrestrial & Marine Resources" University of Crete, Department of Biology. Mar. 2010: Bachelor of Science in "Environmental Biology And Management Of Biological Recourses" University of Crete, Department of Biology. Mar. 2009 – Jul. 2009: 5-month attendance in the University of Genova, Department of Biology through the Erasmus program of international mobility. Jun. 2013: High school final examination and graduation

Grants & awards

2015: SYNTHESYS grant for a visit to Naturalis Biodiversity Center (19th-31st January 2015, Leiden, Netherlands) 2014: Travel grant of the Malacological Society of London for attending the 7th Congress of the European Malacological Societies (7th-11th September 2014, Cambridge, UK). 2013: Travel grant of the Unitas Malacologica for attending the World Congress of Malacology (21st-28th July 2013, Azores, Portugal).

146

2012: Award from ANEK Lines for excellence in Master of Science "Management of Terrestrial & Marine Resources". 2012: Award from the Greek State Scholarships Foundation (IKY) within the framework of the Erasmus program of international mobility (Practical Internship). 2011: Travel grant of the Malacological Society of London for attending the 6th Congress of the European Malacological Societies, (18th-22nd July 2011, Vitoria-Gasteiz, Spain). 2010: Scholarship of Balassi Institute (Hungarian Scholarship Board Office) for attending summer courses of Hungarian language and culture (18th July-1st August 2010). 2009: Award from the Greek State Scholarships Foundation (IKY) within the framework of the Erasmus program of international mobility (Studies).

Other research experience

Feb. 2013 – Mar. 2013: External partner in the LIFE project Innovative Actions Against Illegal Poisoning in EU Mediterranean Pilot Areas (Natural History Museum of Crete). May 2012-July 2012: Practical Internship in ITC (Faculty of Geo-Information Science and earth Observation- University of Twente) within the framework of the Erasmus program of international mobility. Oct. 2007- Dec. 2012: Assistant in the invertebrate laboratory (Gastropods) in the Natural History Museum of Crete (Natural History Museum of Crete).

Abstracts in scientific conferences & meetings

Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016. European Quaternary: Insights from freshwater gastropods. RCMNS Interim Colloquium 2016, Croatian Geological Survey Limnogeology Workshop, Zagreb, Croatia. (Oral presentation)

147

Harzhauser, M., Neubauer, T.A., Kroh, A., Georgopoulou, E., Mandic, O. 2016. History of European lake systems – evolution, geodynamics, and climate change. RCMNS Interim Colloquium 2016, Croatian Geological Survey Limnogeology Workshop, Zagreb, Croatia. Neubauer, T.A., Georgopoulou, E., Harzhauser, M., Mandic, O., Kroh, A. 2016. Predictors of shell size in long-lived lake gastropods RCMNS Interim Colloquium 2016, Croatian Geological Survey Limnogeology Workshop, Zagreb, Croatia. Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016. Biogeographical patterns of European freshwater gastropods during the Quaternary. 1st International Meeting of Early-stage Researchers in Palaeontology, Alpuente, Spain. (Oral presentation) Georgopoulou E., Neubauer T.A., Harzhauser M., Kroh A. & Mandic O. 2015. From Past to Present: What do fossil snails tell us about the evolution of European aquatic hotspots? Abstracts of Nobis 9. Eggenburg, Austria. (Oral presentation) Georgopoulou E., Simaiakis S.M., Neubauer, T.A., Harzhauser M., Kroh A. & Mandic O. 2015. Lakes as islands: Biogeographic patterns of gastropods in Balkan and Anatolian lakes. Abstracts of 13th ICZEGAR. Heraklion, Greece. (Oral presentation) Akkari N., Georgopoulou E. & Simaiakis S.M. 2015. Indicator Species Analysis (ISA) unveils remarkable ubiquitous and specialist centipedes (Myriapoda, Chilopoda) along the bioclimatic gradient in Tunisia. Abstracts of 13th ICZEGAR. Heraklion, Greece. Georgopoulou E., Kougioumoutzis K., Vasilakopoulos P., Simaiakis S.M. & Trigas P. 2015. Biodiversity in the central Aegean archipelago: the effect of geography, geology and climate. Abstracts of 13th ICZEGAR. Heraklion, Greece. (Poster) Neubauer T.A., Georgopoulou E., Harzhauser M., Kroh A. & Mandic O. 2015. European freshwater biogeography for the last 23 million years. Abstracts of SIAL 7. Windsor, Canada. Harzhauser M., Neubauer T.A., Kroh A., Georgopoulou E. & Mandic O. 2015. A world of lakes – European Neogene freshwater systems. Abstracts of Neogene of the Paratethyan Region. 6th Workshop on the Neogene of Central and South-Eastern Europe. RCMNS Interim Colloquium. Orfü, Hungary. Neubauer T.A., Harzhauser M., Georgopoulou E., Kroh A. & Mandic, O. 2015. Development of freshwater biodiversity during the late Cenozoic: impact of geodynamics and climate on hotspot formation. Abstracts of Neogene of the

148

Paratethyan Region. 6th Workshop on the Neogene of Central and South-Eastern Europe. RCMNS Interim Colloquium. Orfü, Hungary. Neubauer T.A., Harzhauser M., Georgopoulou E., Kroh A. & Mandic O. 2015. Biodiversity and paleobiogeography of the European freshwater Neogene: trends, hotspots and faunal turnovers. Geophysical Research Abstracts, EGU2015. Vienna, Austria. Georgopoulou E., Neubauer T.A., Harzhauser M., Kroh A. & Mandic O. 2015. Biodiversity patterns of freshwater snails in European lakes. Abstracts of IBS 2015. Bayreuth, Germany. (Poster with lightning talk) Neubauer T.A., Harzhauser M., Kroh A., Georgopoulou E. & Mandic O. 2015. Establishing a paleobiogeographic framework for European Neogene freshwater systems. Abstracts of IBS 2015. Bayreuth, Germany. Georgopoulou E., Neubauer T.A., Harzhauser M., Kroh A. & Mandic O. 2014. European freshwater gastropods: major patterns of biodiversity and biogeographical processes. Abstracts of Nobis 8. Munich, Germany. (Oral presentation) Harzhauser M., Neubauer T.A., Georgopoulou E. & Wrozyna C. 2014. Morphospace shifts in endemic thermal-spring melanopsids or how an endangered species could lose its species status. Abstracts of Nobis 8. Munich, Germany. Neubauer T.A., Harzhauser M., Kroh A., Georgopoulou E. & Mandic O. 2014. Paleobiogeographic regions for European Neogene freshwater systems: a trend toward increasing provincialism. Abstracts of Nobis 8. Munich, Germany. Georgopoulou E., Neubauer T.A., Harzhauser M., Kroh A. & Mandic O. 2014. Exploring patterns of freshwater gastropod diversity in European lakes. Abstracts of the 7th Congress of the European Malacological Societies, Cambridge, United Kingdom. (Oral presentation) Georgopoulou E., Neubauer T.A., Harzhauser M., Kroh A. & Mandic O. 2014. The FreshGEN Project: Europe's Neogene and Quaternary lake gastropod diversity - aims, scope and first results. Abstracts of the 7th Congress of the European Malacological Societies, Cambridge, United Kingdom. (Poster) Neubauer T.A., Harzhauser M., Georgopoulou E., Kroh A. & Mandic O. 2014. The FreshGEN Project: Freshwater Gastropods of the European Neogene - aims, scope and first results. Abstracts of GeoFrankfurt 2014. Frankfurt, Germany. Neubauer T.A., Harzhauser M., Georgopoulou E., Kroh A. & Mandic O. 2014. A geo- referenced database for Pan-European Neogene freshwater mollusk-bearing localities. Abstracts of PANGEO 2014. Graz, Austria.

149

Simaiakis, S.M., Georgopoulou, E., Djursvoll, P. 2014. Biogeographical patterns in the distribution of European centipedes. Abstracts of the 16th International Congress of Myriapodology. Olomouc, Czech Republic. Neubauer T.A., Georgopoulou E., Harzhauser M., Mandic O. & Kroh A. 2014. Europe's Neogene and Quaternary lake gastropod diversity - a statistical approach. Geophysical Research Abstracts, EGU2014. Vienna, Austria. Georgopoulou E. Neubauer T.A., Harzhauser M., Mandic O. & Kroh A. 2013. European freshwater gastropods of Neogene and Quaternary lakes. Abstracts of Nobis 7. Vienna, Austria. (Oral presentation) Harzhauser M., Neubauer T., Mandic O., Georgopoulou E. & Kroh A. 2013. The FreshGen - a unique database on Freshwater Gastropods of the European Neogene. Abstracts of the GEEWEC2013. Smolenice, Slovakia. Neubauer T.A., Harzhauser M., Mandic O., Georgopoulou E. & Kroh A. 2013. Freshwater th gastropods of the European Neogene – setup of a new database. Abstracts of the 14 RCMNS Congress. Istanbul, Turkey. Probonas M., Georgopoulou E., Nikolopoulou S., Baxevani K. & Xirouchakis S. 2013. Innovative Actions Against Illegal Poisoning in Protected Areas of Crete. Abstracts of the S.R.P.A. Mittersill, Austria. Georgopoulou E., Vardinoyannis K. & Mylonas M. 2013. Spatial patterns of Cretan land snails. Abstracts of the WCM. Azores, Portugal. (Oral presentation) Neubauer T.A., Harzhauser M., Mandic O., Georgopoulou, E. & Kroh A. 2013. Freshwater gastropod diversity of European lake systems over the last 23 million years - state of the art and outlook. Abstracts of the WCM. Azores, Portugal. Georgopoulou E., Mylonas M. & Vardinoyannis K. 2011. Review of Helicigona and Campylaea (Gastropoda: Pulmonata) in Greece. Abstracts of the 6th CEMS. Vitoria- Gasteiz, Spain. (Oral presentation) Georgopoulou E. 2011. Review of the two related genera Campylaea and Helicigona (Gastropoda: Helicidae) in Greece. Abstracts of the 13th A.M.P.P. Chania, Greece. (Poster) Georgopoulou E., Vardinoyannis K. & Mylonas M. 2010. Systematics of Campylaea and Helicigona (Gastropoda: Helicidae) in mainland Greece. Abstracts of the 5th HCE. Patra, Greece. (Poster)

150

List of scientific publications

Neubauer, T.A., Georgopoulou, E., Harzhauser, M., Mandic, O., Kroh, A. 2016 (in press). Predictors of shell size in long-lived lake gastropods. Journal of Biogeography. doi: 10.1111/jbi.12777 Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. Beginning of a new age: How did freshwater gastropods respond to the Quaternary climate change in Europe? (Quaternary Science Reviews, in revision) Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2016 (published online). Distribution patterns of European lacustrine gastropods – a result of environmental factors and deglaciation history. Hydrobiologia, 775, 69–82. Georgopoulou, E., Djursvoll, P., Simaiakis, S.M. 2016. Predicting species richness and distribution ranges of centipedes at the northern edge of Europe. Acta Oecologica, 74, 1–10. doi: 10.1016/j.actao.2016.03.006. Neubauer, T.A., Harzhauser, M., Mandic, O., Georgopoulou, E., Kroh, A. 2016. Paleobiogeography and historical biogeography of the non-marine caenogastropod family Melanopsidae. Palaeogeography, Palaeoclimatology, Palaeoecology, 444, 124– 143. Neubauer, T.A., Harzhauser, M., Mandic, O., Kroh, A., Georgopoulou, E.2016. Evolution, turnovers and spatial variation of the gastropod fauna of the late Miocene biodiversity hotspot Lake Pannon. Palaeogeography, Palaeoclimatology, Palaeoecology, 442, 84– 95. Harzhauser, M., Mandic, O., Neubauer, T.A., Georgopoulou, E., Hassler, A. 2016. Disjunct distribution of the Miocene limpet-like freshwater gastropod genus Delminiella. Journal of Molluscan Studies, 82: 129–136. Georgopoulou, E., Neubauer, T.A., Harzhauser, M., Kroh, A., Mandic, O. 2015. An outline of the European Quaternary localities with freshwater gastropods: Data on geography and updated stratigraphy. Palaeontologia Electronica, 18.3.48A, 1–9. Harzhauser, M., Neubauer, T.A., Georgopoulou, E., Esu, D., D'Amico, C., Pavia, G., Giuntelli, P., Carnevale, G. 2015. Late Messinian continental and Lago-Mare gastropods from the Tertiary Piedmont Basin, NW Italy. Bollettino della Società Paleontologica Italiana. Bollettino della Società Paleontologica Italiana, 54, 1–53. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Kroh, A., Mandic, O. 2015. Tectonics, climate, and the rise and demise of continental aquatic species richness hotspots.

151

Proceedings of the National Academy of Sciences of the United States of America, 112 (37), 11478–11483. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2015. A gastropod-based biogeographic scheme for the European Neogene freshwater systems. Earth-Science Reviews, 143, 98–116. Neubauer, T.A., Georgopoulou, E., Kroh, A., Harzhauser, M., Mandic, O., Esu, D. 2015. Synopsis of European Neogene freshwater gastropod localities: updated stratigraphy and geography. Palaeontologia Electronica, 18.1.3T, 1-7. Neubauer, T.A., Kroh, A., Harzhauser, M., Georgopoulou, E., Mandic, O. 2015. Taxonomic note on problematic Neogene European freshwater Gastropoda. Annalen des Naturhistorischen Museums in Wien, Serie A, 117, 95–100. Harzhauser, M., Neubauer, T.A., Georgopoulou, E., Harl, J. 2014. The Early Miocene (Burdigalian) mollusc fauna of the North Bohemian Lake (Most Basin). Bulletin of Geosciences, 89, 819–809. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Wrozyna, C. 2014. Population bottleneck triggering millennial-scale morphospace shifts in endemic thermal-spring melanopsids. Palaeogeography, Palaeoclimatology, Palaeoecology, 414, 116–128. Neubauer, T.A., Harzhauser, M., Georgopoulou, E., Mandic, O., Kroh, A. 2014. Replacement names and nomenclatural comments for problematic species-group names in Europe's Neogene freshwater Gastropoda. Zootaxa, 3785 (3), 453–468. Neubauer, T.A., Harzhauser, M., Kroh, A., Georgopoulou, E., Mandic, O. 2014. Replacement names and nomenclatural comments for problematic species-group names in Europe's Neogene freshwater Gastropoda. Part 2. ZooKeys, 429, 13–46. Neubauer, T.A., Kroh, A., Harzhauser, M., Georgopoulou, E., Mandic, O. 2014. Synopsis of valid species-group taxa for freshwater Gastropoda recorded from the European Neogene. ZooKeys, 435, 1–6.

152