Groundwater amphipods in Iceland: Population structure and phylogenetics
Etienne Kornobis
Dissertation submitted in partial fulfillment of a Philosophiae Doctor degree in Biology
Advisor: Dr. Snæbjörn Pálsson
PhD Committee: Pr. Bjarni K. Kristjánsson Pr. Jörundur Svavarsson
Opponents: Pr. Christophe Douady Dr. Guðmundur Guðmundsson
Faculty of Science School of Engineering and Natural Sciences University of Iceland Reykjavik, october 2011 Groundwater amphipods in Iceland: Population structure and phylogenetics
Dissertation submitted in partial fulfillment of a Philosophiae Doctor degree in Biology
Copyright c 2011 Etienne Kornobis All rights reserved
Faculty of Science School of Engineering and Natural Sciences University of Iceland Sæmundargötu 2 101 Reykjavik Iceland
Telephone: 525 4000
Bibliographic information: Etienne Kornobis, 2011, Groundwater amphipods in Iceland: Population struct- ure and phylogenetics, PhD dissertation, Faculty of Science, University of Ice- land, 166 pp.
ISBN 978-9935-9064-1-0
Printing: Háskólaprent Reykjavik, Iceland, October 2011 Abstract
Crymostygius thingvallensis and Crangonyx islandicus are two endemic species of groundwater amphipods which were recently discovered in Iceland. C. thing- vallensis is uncommon, but represents a new monotypic family Crymostygidae. C. islandicus is widespread in the geologically youngest parts of the island and belongs to a genus which is widely distributed in North America and Eurasia. Both species belong to the superfamily Crangonyctoidea, exclusively composed of freshwater species. Iceland is geographically isolated and was fully covered by an ice sheet during the last glacial maximum, 21 000 years ago, hence the Icelandic biota is characterized by extremely low endemism and low species diversity. Thus, the discovery of these two endemic freshwater species, on an island isolated in the midst of the Atlantic ocean raised questions about their colonization and their survival during the glacial periods of the Ice age. Sub- glacial refugia have been hypothesized to explain the occurrence of these two amphipods in Iceland which might have colonized the country via an ancient land bridge, now submerged, connecting Iceland and Greenland. In this thes- is, molecular and morphological markers were used to test this hypothesis. In addition, cryptic diversity, common in groundwater species, was assessed with mitochondrial and nuclear markers. Finally, the variation in secondary structure of the nuclear ITS regions has been characterized and its influence on the phylogeny reconstruction evaluated. The mitochondrial markers (COI and 16S) support that C. islandicus populations diverged within the island for up to five million years and consequently they survived repeated glaciati- ons in Iceland in subglacial refugia. The nuclear marker ITS1 is supporting, though weakly, the phylogeographic pattern observed with the mitochondrial markers. Though the populations of C. islandicus showed high genetic di- vergence, different species delimitation methods led to different conclusions about the potential species complex status of C. islandicus. The taxonomic status of both Icelandic species is supported by phylogenies based on nuclear markers (18S, 28S, 5.8S, ITS1 and ITS2). These phylogenies show high di- vergence between Crangonyx species from Eurasia and from North America. C. islandicus is more closely related genetically to Crangonyx species from
i North America, supporting the hypothesis of a colonization via Greenland. Morphological data led to a different pattern, C. islandicus clustering with species both from North America and Europe which might be due to conver- gent evolution in morphological traits. In addition, nuclear phylogenies show that Crangonyx, Synurella and Stygobromus genera are polyphyletic and might need taxonomic revision to restore their monophyly. The secondary structures of the ITS2 for Crangonyctoidea species are highly divergent from the common structure described for Metazoans and appeared to have little impact on the phylogenetic reconstruction.
ii Útdráttur
Crymostygius thingvallensis og Crangonyx islandicus eru tvær einlendar teg- undir grunnvatnsmarflóa sem fundust nýlega á Íslandi. C. thingvallensis er sjaldgæf og myndar hún nýja ætt marflóa. C. islandicus er útbreidd á gos- belti Íslands, og tilheyrir ættkvísl sem er útbreidd í N-Ameríku, Evrópu og Asíu. Báðar tegundirnar tilheyra yfirættinni Crangonyctoidea sem finnst ein- göngu í ferskvatni. Lífríki Íslands einkennist af lítilli tegundafjölbreytni og fáum einlendum tegundum. Hefur það verið rakið til landfræðilegrar einangr- unar Íslands og þess að landið var hulið jökli fyrir um 21000 árum. Fundur tveggja einlendra grunnvatnsmarflóa vakti því athygli og spurningar vöknuðu um hvernig þær höfðu borist til landsins. Tilgáta hefur verið sett fram um að þær hafi lifað af á Íslandi á ísöld og að þær hafi í raun fylgt Íslandi allt frá myndun þess eða mögulega numið land um landbrú milli Grænlands og Íslands. Í ritgerðinni er þessi tilgáta prófuð með aðferðum sem byggja á greiningu DNA sameinda. Annað meginviðfangsefni ritgerðarinnar er að greina flokkun teg- undanna út frá DNA sameindum og útlitseiginleikum marflónna. Auk þess var athugað hvort dulinn breytileiki greindist í hvatbera og kjarna DNA, en slíkt er algengt meðal grunnvatnstegunda. Breytileiki í annars stigs byggingu svokallaðra ITS DNA raða var einnig lýst og áhrif þess á flokkunartré tegund- anna athuguð. Breytileiki í hvatberagenunum COI og 16S RNA styðja að C. islandicus lifði af endurtekin kuldaskeið ísaldar undir jökli. Athugun á ITS1 innröðum í kjarna DNA gefur stuðning við þessa ályktun en sýnir mun minni aðgreiningu milli svæða en hvarberagenin. Erfðafræðileg aðgreining milli stofna C. islandicus gefur vísbendingu um ólíkar tegundir sé að ræða innan Íslands. Flokkunarleg staða beggja tegundanna byggð á RNA genum í kjarna DNA (18S, 28S, 5.8S, ITS1 og ITS2) styður fyrri flokkun byggða á útlitseinkennum. Flokkunartréið sýnir hinsvegar skýra aðgreiningu milli Crangnyx tegunda milli Evrasíu og N-Ameríku. C. islandicus er skyldari Crangonyx tegundum í Amer- íku, það styður landnám frá Grænlandi. Flokkunartré byggt á útlitseinkennum var þó öðruvísi, C. islandicus flokkaðist samkvæmt því bæði með tegundum frá N-Ameríku og Evrópu. Slíkt frávik gæti mögulega stafað af samhliða þróunar meðal útlitseiginleikanna. Flokkunartré byggð á kjarna DNA sýndi einnig að
iii ættkvíslirnar Crangonyx, Synurella og Stygobromus eru fjölstofna og að þörf sé að athuga frekar flokkun þeirra byggða á útlitseiginleikum. Annars stigs bygg- ing ITS2 meðal Crangonyctoidea tegunda víkur frá hinni almennu byggingu sem hefur verið lýst fyrir vefdýr (metazoa) og hafði hún lítil áhrif á byggingu ættartrésins.
iv List of Papers
The present thesis is based on three accepted papers and two currently under review. The papers will be referred in the text by their respective numbers as following:
• Paper 1: Kornobis E, Pálsson S, Kristjánsson BK, Svavarsson J (2010). Molecular evidence of the survival of subterranean amphipods (Arthro- poda) during Ice Age underneath glaciers in Iceland. Molecular Ecology 19: 2516-2530.
• Paper 2: Kornobis E, Pálsson S (2011). Discordance in variation of the ITS region and the mitochondrial COI gene in the subterranean amphi- pod Crangonyx islandicus. Journal of Molecular Evolution (in press), DOI: 10.1007/s00239-011-9455-2.
• Paper 3: Kornobis E, Pálsson S, Sidorov DA, Holsinger JR, Kristjáns- son BK (2011). Molecular taxonomy and phylogenetic affinities of two groundwater amphipods, Crangonyx islandicus and Crymostygius thing- vallensis, endemic to Iceland. Molecular Phylogenetics and Evolution 58: 527-539.
• Paper 4: Kornobis E, Pálsson S. (under review). Phylogenies of Cran- gonyctoidea species based on the ITS regions (ITS1 and ITS2).
• Paper 5: Kornobis E, Pálsson S, Svavarsson J. (under review). Classi- fication of Crangonyx islandicus based on morphological characters and comparison with molecular phylogenies.
v
Acknowledgments
It is my great pleasure to acknowledge all the persons which played a role, directly or indirectly, in the realization of this thesis. I had the opportunity during these four years to work in an outstanding environment and especially to be advised by a remarkable doctoral committee. I would like to greatly thank my advisor, Snæbjörn Pálsson, for his constant care, his bonhomie and his contagious interest in Science. I have been truly lucky to be supervised by such mentor and I wish to be able to keep on working with him. Thanks a lot also to Bjarni K. Kristjánsson for the awesome sampling trips, always pertinent suggestions and for creating the highest latitude beer club that I know. I have been delighted as well to work with Jörundur Svavarsson and I thank him for guiding me through the mysteries of Crustacean morphology. Thanks to Christophe Douady and Guðmundur Guðmundson for accepting to be my opponents for this thesis. I am grateful to the University of Iceland Research Fund and the Icelandic Research Council (Rannís) for their financial support. I am indebted to all the contributors of free softwares (GNU/Linux, Emacs, R, LATEX, Python, Inkscape, Gimp, Phylip ...) which were of great use during the realization of this thesis. Thanks also to all my lab mates which made this stay so cheerful. I would particularly like to thank Ubaldo, Eduardo, Dileepa, Sindri, Hlynur, Ehsan, Laurène, Baldur and many more. Thanks to Oli Patrick to have revived the cake club. I still feel sorry to have missed one of his french speeches. Finally, Kalina, my wingman who always was there to cheer me up, even when I was in “zombie mode”. It has been a pleasure to work with my first students in the lab, Virginia, Sonia, Daniel, Jaume and Cedric. I wish them the best for their ongoing career. Thanks to Niklas Zennström and Janus Friis for founding Skype, I think I owe them part of my mental sanity. I am deeply grateful to my parents, Lydie and Michel, for their constant wish to help and their support at so many levels. Their visits, calls and “Mon- day’s pdfs” were of great value during the most difficult times of this thesis. Merci pour tout ! I’m not even able to recall the exact number of times that my
vii brother came to Iceland. His visits and traveling tales always cheered me up. Pierrot, on choisit pas sa famille, çà tombe bien j’aurais pas fait mieux. Ben- jamin (aka Dr. Mathon), thank you my friend for the endless nights working and chilling out together through skype. You particularly animated the last moments of my thesis by your riddles and other puns. J’ai ri à tes blagues! Many thanks to Aurane, Déborah, David, Séverine, Fanny, Nanou, Nanette and to all my friends and family which were always present, one way or the other, during my short trips to France. Finally, mi Cari Mari, who faced once more the Icelandic winter to support me and patiently tried to convince me that everything I was doing was awesome. Celia, you are my most precious finding in Iceland. Takk Fyrir! I wish a joyful life to all of you people !
viii To my family,
ix
Contents
1 Introduction 1 1.1 The discovery of two endemic amphipod species in Iceland . . .1 1.2 A short Natural History of Iceland ...... 2 1.2.1 Subterranean fauna ...... 6 1.2.2 Glacial survival hypothesis ...... 6 1.3 The superfamily Crangonyctoidea ...... 8 1.4 Groundwater biodiversity ...... 8 1.5 Aims of the thesis ...... 11
2 Materials and Methods 13 2.1 Sampling ...... 13 2.2 Test of the subglacial refugia hypothesis and cryptic diversity . 16 2.3 Taxonomy and colonization of the two Icelandic amphipods . . 17 2.3.1 Background on phylogenetic reconstruction methods . . 18 2.3.2 Phylogeny based on morphological data ...... 20 2.4 Secondary structure of the ITS regions ...... 21 2.4.1 Ribosomal RNA secondary structure ...... 23 2.4.2 Characterization of the ITS regions, structures and phy- logenetic reconstruction ...... 23 2.4.3 ITS2 secondary structure and cryptic diversity . . . . . 24
3 Results and Discussion 25 3.1 The subglacial refugia hypothesis ...... 25 3.2 Cryptic diversity within C. islandicus ...... 27
xi 3.3 Phylogenetic status and colonization of the Icelandic amphipods 28 3.4 Taxonomy of the Crangonyctoidea and convergence in morphology 29 3.5 Evolution of the ITS regions ...... 31
4 Conclusion and Perspectives 33
5 Paper 1: Molecular evidence of the survival of subterranean amphipods (Arthropoda) during Ice Age underneath glaciers in Iceland 49
6 Paper 2: Discordance in variation of the ITS region and the mitochondrial COI gene in the subterranean amphipod Cran- gonyx islandicus 67
7 Paper 3: Molecular taxonomy and phylogenetic affinities of two groundwater amphipods, Crangonyx islandicus and Cry- mostygius thingvallensis, endemic to Iceland 81
8 Paper 4: Phylogenies of Crangonyctoidea species based on the ITS regions (ITS1 and ITS2) 99
9 Paper 5: Classification of Crangonyx islandicus based on mor- phological characters and comparison with molecular phyloge- nies 121
xii List of Figures
1.1 Microscopic photography of the two Icelandic amphipods . . . .2 1.2 The opening of the North Atlantic and the formation of Iceland3 1.3 Geology and volcanism of Iceland ...... 4 1.4 Distribution of the Crangonyctoidea ...... 9
2.1 Sampling localities of C. thingvallensis and C. islandicus in Ice- land ...... 14 2.2 Sampling localities for Crangonyctoidea and Niphargoidea species included in this study ...... 15 2.3 Crangonyx islandicus morphology ...... 21 2.4 The telson of two Crangonyx species...... 22 2.5 Location and structure of the ITS2 ...... 22 2.6 Description of the two main types of secondary structure . . . . 23
3.1 Genetic and geographic structure of C. islandicus populations in Iceland ...... 27 3.2 Phylogenetic relationships among Crangonyctoidea species. . . 29
7.1 Bayesian phylogenetic tree of crangonyctid species based on COI sequences ...... 97
xiii
List of Tables
1.1 Comparison in number of species and endemism between two oceanic islands: Iceland and Hawaï ...... 5
xv
Introduction
1.1 The discovery of two endemic amphipod species in Iceland
Recently, in years 2004 and 2006, two endemic species of groundwater am- phipods (Crustacea), Crymostygius thingvallensis and Crangonyx islandicus, were described from springs in lava fields in Iceland (Fig 1.1). The former rep- resents a new monotypic family Crymostygidae (Kristjánsson and Svavars- son, 2004) and the latter a new species of the genus Crangonyx (Svavarsson and Kristjánsson, 2006). Both species belong to the superfamily Cran- gonyctoidea which occurs worldwide (Väinölä et al., 2008). These findings are unique to Iceland as the Icelandic biota is characterized by low species diversity and low endemism (Sadler, 1999; Brochmann et al., 2003; Coope et al., 1986). As Iceland has been repeatedly covered by glaciers during the last Ice Age, most of the species currently present are considered to have colonized the island after the last glacial maximum around 21 000 years ago (e.g. Buck- land et al., 1986; Coope et al., 1986). As groundwater amphipods are known to have poor dispersal capacities (e.g. Lefébure et al., 2007; Trontelj et al., 2009), questions have been raised about their origin and the colonization of Iceland by these two endemic amphipod species. Iceland is an island geo- graphically isolated in the midst of the Atlantic Ocean and considering that the superfamily Crangonyctoidea is exclusively composed of freshwater species, largely restricted to subterranean habitats (Väinölä et al., 2008), it is unlikely that these two species recently evolved from marine species. Kristjánsson
1 2 Introduction
Figure 1.1: Microscopic photography of the two Icelandic amphipods: Cry- mostygius thingvallensis on the left and Crangonyx islandicus on the right. and Svavarsson (2007) raised the hypothesis that Crymostygius thingvallen- sis and Crangonyx islandicus colonized Iceland via a past land connection between Greenland and Iceland and survived repeated glaciations on the is- land in subglacial refugia. The present thesis evaluates this hypothesis and the taxonomic affinities of the two species using population genetics, phylogeogra- phy and phylogenetic methods based on various molecular and morphological characters.
1.2 A short Natural History of Iceland
In the midst of the North Atlantic Ocean, Iceland is a geologically young island which has been formed since the Miocene (i.e. between 23 and 5 million years ago) until now (Einarsson, 1994, p. 231). Its formation has been commonly explained by the presence of a volcanic hotspot i.e. a strong upwelling of magma in a mantle plume, which is now approximately situated beneath the Vatna- jökull glacier in Iceland (see Mihalffy et al., 2008). As the North American and Eurasian tectonic plates started to move apart around 55 million years ago, the mantel plume alimented important volcanism forming a land connection between Greenland and Scotland, known as the Greenland-Scotland transverse ridge (Fig 1.2, see Denk et al., 2011). Due to the continuous spreading of the plates, this ridge is now submerged and Iceland is the only part which re- mains above sea level. The subsistence of Iceland above sea level is due to the A short Natural History of Iceland 3
Scandinavia Spreading axis
Extinct spreading Greenland axis
Plateau basalts of continental origin Oceanic crust formed above Faeroes sea level Oceanic crust formed below sea level Miocene plateau basalts of Iceland
55 Ma 24 Ma 4 Ma
Figure 1.2: The opening of the North Atlantic and the formation of Iceland. Ma: Million years ago. Modified from Denk et al. (2011). combined volcanic production of the mantel plume and the mid-ocean ridge magmatism (see Geirsdóttir et al., 2007). A recent study suggested that the Greenland-Iceland part of the ridge was emergent at least until 10 million years ago while the Iceland-Scotland part submerged earlier (Poore, 2008). Fossil records indicate that plants migrated along this land bridge between Scotland and Iceland until 24 million years ago (Mya) and until 15 Mya or even 6 Mya between Greenland and Iceland (Grimsson et al., 2007; Denk et al., 2010). The geologically youngest part of the island is located at the boundaries of the tectonic plates in the middle of the volcanic active zone and the age of the geological formations are increasing with the distance from these bound- aries (Fig 1.3), the oldest being around 16 million years old (see Grimsson et al., 2007). The basaltic bedrocks originating before the Pleistocene i.e. before the onset of the last Ice Age, are now fairly impermeable and thus groundwater is scarce in the geologically oldest parts of Iceland (see Fig 1.3, Einarsson, 1994). Conversely, lava deposits occurring after the glaciations are highly porous and groundwater is plentiful in the youngest parts of the island (Einarsson, 1994). These areas are as well rich in fissures zones, which are mainly located at the tectonic plates boundaries (Fig 1.3) and are often characterized by high geothermal activity (Einarsson, 1994). Most of the early flora in Iceland was widespread in the northern hemi- sphere and no predominant colonization routes – either from North America 4 Introduction
Figure 1.3: Geology and volcanism of Iceland. 1: eruption fissures, 2: volca- noes. Age of the geological formations, 3: Tertiary, 4: late Pliocene and early Pleistocene, 5: late Pleistocene, 6: Holocene before settlement, 7: Holocene after settlement. The four red circles correspond to: a, Snæfellsness; and three main fissure zones: b, Þingvallavatn; c, Lakagígar; d, Mývatn. Modified from Einarsson (1994). A short Natural History of Iceland 5
Table 1.1: Comparison in number of species and endemism between two oceanic islands: Iceland and Hawaï. N species: number of species. After Sadler (1999).
Iceland Hawaï N species Endemics N species Endemics Insects 1245 0 7862 5237 Birds 80 0 274 60 Molluscs 35 0 1656 956 or Europe – can be inferred from fossil records (Grimsson et al., 2007). The warmth-demanding species present at that time became progressively extinct due to climate cooling (Rundgren and Ingólfsson, 1999; Einarsson, 1994) marking the beginning of the glaciation period, between 5 and 10 million years ago in the northern hemisphere (see for review Geirsdóttir et al., 2007). Iceland, located just south of the Arctic circle around latitude 65◦ North, has been repeatedly covered by an ice sheet during the glacial cycles of the last Ice Age which onset occurred around 2.75 million years ago (see for review Geirsdóttir et al., 2007). During the last glacial maximum, around 21 000 years ago, Iceland was almost completely covered by glaciers and only small ice-free areas may have been present along the coastal mountains (Andrews et al., 2000; Geirsdóttir et al., 2007). The last deglaciation started around 15 000 years ago while readvances of glaciers might have occurred in Iceland until 11 000 years ago (see Geirsdóttir et al., 2007). The combination of the short time since the retreat of the glaciers and the geographic isolation of Iceland explains the relatively low terrestrial species diversity and its low endemism (Table 1.1; Sadler, 1999; Brochmann et al., 2003; Coope et al., 1986). During the end of the last glacial period, approximately 12 500 years ago, the shoreline in Iceland was up to 100 meters above the current sea level. Though the melting of the glaciers resulted in an increase in global sea level, it caused also the rise of the island, free from the pressure due to the weight of the ice sheet (see Norðdahl et al., 2008). 6 Introduction
1.2.1 Subterranean fauna
The subterranean fauna in Iceland is poorly known so far and only three sty- gobiont species – species restricted to aquatic subterranean habitats – have been described. One, is the copepod Parastenocaris glacialis which is also known from groundwater in Europe (Juberthie et al., ress). The two oth- ers are the endemic amphipods Crymostygius thingvallensis and Crangonyx islandicus found in groundwater underneath lava fields no older than 10 000 years (Kristjánsson and Svavarsson, 2007; Kornobis et al., 2010). Cran- gonyx islandicus is relatively common and widespread throughout the geolog- ically youngest part of the island, the west and east volcanic zones around the boundaries of the tectonic plates (Svavarsson and Kristjánsson, 2006; Kornobis et al., 2010). Crymostygius thingvallensis is uncommon and has been found to date only at spring inlets in lake Þingvallavatn, Southwest Ice- land (Kristjánsson and Svavarsson, 2004), and in the stomach of an Arctic char (Salvelinus alpinus) in Jokulsargljúfur, East Iceland (Kristjánsson and Svavarsson, 2007; Egilsdóttir and Kristjánsson, 2008). Outside the east and west volcanic zones, in lava fields in the Snæfellsness peninsula (see Fig 1.3), neither of these species was found (see Kornobis et al., 2010). The Snæfellsness peninsula is outside the central volcanic zone and is mainly formed by older (see Fig 1.3) and less porous bedrocks which have been intruded by volcanic magma. Moreover this peninsula lacks fissure zones and its basaltic formations chemistry is different from the ones found in the central volcanic zone (Sigurdsson, 1970).
1.2.2 Glacial survival hypothesis
Most of the European biota, particularly temperate species, have survived dur- ing glacial maxima in refugia at lower latitudes, mostly located in the Iberic peninsula, Italy and the Balkans (Taberlet et al., 1998; Hewitt, 2000, 2004). Arctic species tend to exhibit low genetic diversity and shallow genetic clades, as a consequence of population contraction and expansion over the last glacial cycles (Hewitt, 2004). Nonetheless, a growing body of phylogeographic stud- ies supports northern cryptic refugia which might even have occurred within A short Natural History of Iceland 7 the supposed limits of the ice sheet (Stewart and Lister, 2001; Provan and Bennett, 2008). Two theories have been proposed to explain the origin of the Icelandic biota: the tabula rasa, i.e. total extinction of the species present in glaciated areas, and the glacial survival, i.e. the survival in located ice free refugia (see Ægisdóttir and Þórhallsdóttir, 2004; Denk et al., 2011). A vast ma- jority of plants occurring now in Iceland has an European origin (Einarsson, 1963; Rundgren and Ingólfsson, 1999). Plant species have been suggested to have survived the glaciations in ice free areas such as nunataks (e.g. Rund- gren and Ingólfsson, 1999) though the majority is known to have migrated from Europe and northern Eurasia after the last glacial maximum (Denk et al., 2011). According to Wares and Cunningham (2001), the genetic divergence observed between populations of the marine isopod Idotea balthica in Iceland and Europe supports a split older than 200 000 years and therefore the occur- rence of a marine refugium in southwestern Iceland. Nonetheless, geological data and the extent of the ice sheet during glaciation in this region are not supporting this hypothesis (Ingólfsson, 2009). These glacial survival depend on the occurrence of unglaciated regions which acted as refugia. Considering the extent of the ice sheet on Iceland during the last glacial maximum, the survival of the two Icelandic freshwater amphipods implies the occurrence of subglacial refugia (Kristjánsson and Svavarsson, 2007). Subglacial refugia have been as well hypothesized to explain geographic distribution of groundwa- ter amphipod species in North America (e.g. Holsinger, 1980; Koenemann and Holsinger, 2001) and Ireland (Hanfling et al., 2008). Nonetheless, the occurrence of unglaciated refugia and postglacial colonization from these adjacent ice free regions to explain current species distributions are commonly difficult to reject (e.g. Lefébure et al., 2007). The occurrence of the endemic species C. islandicus, widespread in a geographically isolated island which was repeatedly covered by ice during glaciations allows for testing the subglacial refugia hypothesis. 8 Introduction
1.3 The superfamily Crangonyctoidea
Crymostygius thingvallensis and Crangonyx islandicus belong to the freshwa- ter superfamily Crangonyctoidea. This superfamily encompasses 11 families and presents an unusual geographic distribution compared to other continental amphipods. Crangonyctoids are distributed worldwide (Fig 1.4) and present more species diversity in the Nearctic region than in the Palearctic (184 and 25 species respectively; Väinölä et al., 2008). Conversely, the Palearctic har- bors 70% of the total continental amphipod species diversity (Väinölä et al., 2008). The wide distribution of the exclusively freshwater superfamily Cran- gonyctoidea may be explained by the occurrence of a widespread freshwater ancestor before the breakup of Pangaea i.e. in the late Triassic and early Juras- sic (Holsinger, 1992). This is further supported by the fact that groundwater amphipod species appear to have poor dispersal capabilities (Lefébure et al., 2007; Trontelj et al., 2009). Among Crangonyctoids, the family Crangonyc- tidae is common and widespread in North America but occurs also in Eurasia, Iceland and North Africa (Fig 1.4). The large distribution of the Crangonycti- dae in the northern hemisphere suggests that their ancestor occupied Laurasia before its breakup around 55 Mya (Holsinger, 1992). The only known species from North Africa is Crangonyx africanus, recently described as endemic to Morocco (Messouli, 2006). Unfortunately, the well where the only sample of Crangonyx africanus was found, does not exist anymore. The genus Crangonyx is represented by 47 species, a vast majority (42 species) distributed in North America, three species in Eurasia, one in North Africa and one in Iceland.
1.4 Groundwater biodiversity
The groundwater have been described as “an unseen ocean beneath our feet” and the exhaustive inventory of their biodiversity is still at its beginnings (Danielopol et al., 2000). Groundwater represent 30% of the freshwater on Earth and 99% of the unfrozen continental waters (see Shiklomanov, 1993; Danielopol et al., 2003). Despite the abundance of groundwater habitats and due to their cryptic nature, groundwater species richness surveys have Groundwater biodiversity 9
Figure 1.4: Distribution of the Crangonyctoidea. The different families are: All: Allocrangonyctidae, Aus: Austrocrangonyctidae, Cry: Crymosty- gidae, fal: “falklandiellids”, Neo: Neoniphargidae, Par: Paracrangonyctidae, Per: Perthiidae, Phr: Phreatogammaridae, Pse: Pseudocrangonyctidae, Ste: Sternophysingidae. Modified from Väinölä et al. (2008). 10 Introduction been conducted only recently (e.g. the PASCALIS project in Europe, Gibert et al. 2005). The estimation of the species richness in such particular environ- ments faces major obstacles. Firstly, the nature of the groundwater habitats makes their access and sampling arduous and hydro-biological connections be- tween them are relatively difficult to assess. Secondly, subterranean species are subject to morphological stasis or convergent evolution: due to the similar selective pressures in groundwater habitats (absence of light, low nutrients and low oxygen, see Malard and Hervant 1999), different species often exhibit similar morphologies such as reduction or absence of eyes, elongation of other sensory appendages, loss of pigmentation. The abundance of morphological convergence in subterranean species hinders the clear delimitation of species and leads to an underestimation of the biodiversity when descriptions based solely on morphology are considered. This latter issue can be solved by the use of molecular taxonomy and DNA barcodes for species identification which have been extensively used in the past decades (e.g. Roe and Sperling, 2007; Lefébure et al., 2007; Seidel et al., 2009; Humphreys et al., 2010). Depletion and pollution by human activities such as agriculture, urban- ization, industry and tourism are widely threatening groundwater ecosystems (Danielopol et al., 2003). In addition, the abundance of endemism witnessed in groundwater species is making these species even more subject to extinc- tion. In the six regions of Europe covered by the PASCALIS survey, 41% of the aquatic subterranean species reported could be considered as endemic (Gibert et al., 2009). This highlights the need for a more precise estimation of the biodiversity in subterranean species, particularly using molecular markers, in order to define appropriate conservation areas and units. Aims of the thesis 11
1.5 Aims of the thesis
The main objective of this thesis is to describe the evolution of Icelandic groundwater amphipods. This was achieved through five different studies which addressed more specifically the following aims:
• To test the subglacial refugia hypothesis (Paper 1 and 2).
• To locate putative refugia in Iceland (Paper 1).
• To assess if the widespread populations of Crangonyx islandicus represent a single or a complex of species i.e. to detect cryptic diversity (Paper 1 and 2).
• To evaluate the species status of C. islandicus and the family status of C. thingvallensis by assessing the phylogenetic relationships among various Crangonyctoidea species (Paper 3 and 4).
• To evaluate the phylogeny of the Crangonyx genus based on morphologi- cal characters. Due to convergence which occurs in subterranean species the phylogenetic tree is compared with trees based on molecular data (Paper 5).
• To infer how C. thingvallensis and C. islandicus colonized Iceland (Paper 3 and 5).
• To characterize the secondary structure of the internal transcribed re- gions and evaluate its influence on the phylogeny of Crangonyctoidea species (Paper 2 and 4).
• To evaluate the behaviors of two commonly used barcode markers to assess variation in groundwater amphipods (Paper 1, 2, 4).
Materials and Methods
2.1 Sampling
To test the subglacial refugia hypothesis, locate their putative locations and detect cryptic diversity, genetic variation found in the widespread populations of C. islandicus was characterized. In order to do so, C. islandicus specimens were sampled from groundwater in 23 localities in Iceland (Fig 2.1). Samples were obtained by electrofishing in springs coming out from lava fields no older than 10 000 years. No specimens were found in springs of lava fields at the Snæfellsness peninsula in West Iceland (see Fig 1.3). The taxonomic status of the two Icelandic species and their colonization was investigated by assessing the phylogenetic relationships of C. islandicus and C. thingvallensis with other species from the superfamily Crangonyctoidea. C. thingvallensis is uncommon and was sampled from the only location known, in Lake Þingvallavatn (Fig 2.1). In addition, samples from eight Crangonyctoidea species were obtained from North America, Europe and Russia (Fig 2.2). DNA sequences from 17 additional species of Crangonyctoidea and Niphargoidea were obtained from Genbank (see Fig 2.2). The phylogenetic status of C. islandicus based on morphological charac- ters was studied by comparing with the classification of the North American Crangonyx species made by Zhang and Holsinger (2003) and by adding specimens of C. subterraneus and C. chlebnikovi from Eurasia (Fig 2.2).
13 14 Materials and Methods 66.5 Latitude (°N) .5 65.5 63.5 64
24 22 20 18 16 14
Longitude (°W)
Figure 2.1: Sampling localities of C. thingvallensis and C. islandicus in Iceland. Triangle: C. thingvallensis, circles: C. islandicus. The volcanic active zone is displayed in red. Glaciers are in blue. Longitudes and latitudes are shown in degrees on the x- and y-axes respectively. Sampling 15
10 12 13 6 11 1 7 8 14 2 3 15 4 9 5
Figure 2.2: Sampling localities for Crangonyctoidea and Niphargoidea species included in this study. Red circles represent samples obtained during this study and blue ones the species obtained from Genbank. Open circles cor- respond to several sampling locations. 1: Bactrurus brachycaudus, Bactrurus mucronatus, Bactrurus pseudomucronatus, Crangonyx forbesi, Crangonyx sp.; 2: Crangonyx pseudogracilis, Synurella sp.; 3: Crangonyx serratus, Stygobro- mus gracilipes, Stygobromus stegerorum; 4: Synurella dentata, 5: Crangonyx floridanus; 6: Crangonyx islandicus, Crymostygius thingvallensis; 7: Cran- gonyx subterraneus; 8: Niphargus plateaui; 9: Niphargus frasassianus; 10: Crangonyx subterraneus, Niphargus kochianus, Niphargus fontanus, Niphar- gus schellenbergi; 11: Synurella ambullans; 12: Crangonyx chlebnikovi; 13: Amurocrangonyx arsenjevi; 14: Procrangonyx primoryensis; 15: Pseudocran- gonyx korkishkoorum. 16 Materials and Methods
2.2 Test of the subglacial refugia hypothesis and cryp- tic diversity
The genetic variation among C. islandicus was characterized using mitochon- drial (COI and 16S in Paper 1) and nuclear markers (ITS1 and ITS2 in Paper 2). COI and 16S mitochondrial genes and the nuclear internal transcribed spacers 1 (ITS1) have been commonly used to estimate intraspecific variation and to describe cryptic diversity within species (e.g. in crustaceans Chu et al., 2001; Wattier et al., 2006; Audzijonyte and Väinölä, 2006; Lefébure et al., 2007; Hanfling et al., 2008). Genetic patterns may reflect the past history and distribution of species. Refugia are characterized by high genetic diversity though newly colonized areas usually present low diversity. Correlation between genetic and geo- graphic distances within species can reflect isolation by distance and past vi- cariance/dispersal events while the lack of correlation might indicate several colonization events from different origins. The genetic variation within Arctic and sub Arctic species are expected to be low due the ice sheet extent during the past glaciation and recent colonization events. Phylogeographic patterns and correlation between genetic and geographic distances are explored in or- der to infer if the occurrence of C. islandicus in Iceland is due to a single or several colonization events. The COI and 16S mitochondrial (mt) genes substi- tution rates have been calibrated for crustacean species using dated geological events (Knowlton et al., 1993; Knowlton and Weigt, 1998; Schubart et al., 1998; Cunningham et al., 1992; Stillman and Reeb, 2001). Accord- ing to the molecular clock hypothesis, the rate of amino acid or nucleotide substitution for a gene is approximately constant over evolutionary time (see Zuckerkandl and Pauling, 1965; Nei and Kumar, 2000). The COI and 16S mt genes variation and their calibrated substitution rates were used to provide time estimations for the split of C. islandicus populations which, in the case of a single colonization event, is giving an approximation of the time since the colonization of Iceland (Paper 1). These time estimates were then compared with the date of the last glacial maximum, when Iceland was cov- ered by glaciers. In Paper 1, a Bayesian approach implemented in the software Taxonomy and colonization of the two Icelandic amphipods 17
BEAST (Drummond and Rambaut, 2007) was used to estimate the time of divergence of C. islandicus populations. This software presents the advan- tage to compute simultaneously the phylogeny and the divergence dates for multi-locus data and a specific substitution model can be specified for each locus. The location of the subglacial refugia during the last glaciations was inves- tigated in Paper 1 by testing if COI and 16S genetic diversity was dependent on altitude, distance from the tectonic plate boundaries and age of the lava fields. Considering the weak dispersal capacities of subterranean amphipod species, the current populations with the most genetic diversity are expected to be located nearby the past subglacial refugia. Potential cryptic diversity within C. islandicus was detected by analyzing the variation of the COI and 16S genes and the ITS regions (Papers 1 and 2). The barcoding of life initiative proposed by Hebert et al. (2003) aims at the species identification of unknown specimens and enhancing the discovery of new species (see Moritz and Cicero, 2004). The COI and internal transcribed spacer 2 (ITS2) have been proposed as DNA barcodes and tools for species delimitation, of particular interest for species with high cryptic diversity. Their variation was used in this thesis to infer if C. islandicus represents a species complex. Genetic divergence thresholds for the COI gene (Hebert et al., 2004; Lefébure et al., 2006) and criteria based on the sequence variation and the secondary structure of the ITS2 (Müller et al., 2007) have been developed in order to delimit species and were used to detect putative cryptic species within C. islandicus.
2.3 Taxonomy and colonization of the two Icelandic amphipods
The taxonomic status of the two groundwater amphipods and their coloniza- tion were assessed by evaluating their phylogenetic relationships with other Crangonyctoidea species. Phylogenies were built based on the mitochondrial genes COI and 16S and the nuclear genes 18S, 28S, 5.8S, ITS1 and ITS2 (Pa- pers 3 and 4). By assessing the genetically closest relatives of the Icelandic 18 Materials and Methods species, it can be possible to infer if they colonized Iceland from the Nearctic or the Palearctic regions. For this purpose, morphological markers were used as well to assess which Crangonyx species was the closest morphologically to C. islandicus (Paper 5).
2.3.1 Background on phylogenetic reconstruction methods
Phylogenetics is the study of evolutionary relationships among species. The most common representation of these relationships is a phylogenetic tree, where the nodes correspond to putative ancestors and the tips of the branches are the organisms studied. These relationships can be inferred from various types of data and the phylogenies presented in this study are based on molecular (Papers 1, 2, 3, 4) and morphological ones (Paper 5). The phylogenetic re- constructions based on molecular data are known to be affected both by the alignment of DNA sequences and the phylogenetic tree building methods (e.g Saitou and Imanishi, 1989; Morrison and Ellis, 1997). Though it has been repeatedly shown that alignment techniques have at least as much effect as the tree building method, publications incorporating different alignment methods to build phylogenies are still scarce (see Morrison, 2008). We consequently used different alignment and tree building methods in this thesis and evaluated their influence on the topology of the trees in Paper 3.
Alignment methods
In order to reconstruct phylogenetic trees based on nucleotide sequences, it is necessary to assess which sites among the available sequences are homologous i.e. derived from a common ancestry. Numerous multiple sequences alignment techniques deal with this issue and several of them were used in Papers 2, 3 and 4. We attempted as well to evaluate the influence of these alignment techniques on the main conclusions of the Paper 3. The most common ap- proach in multiple alignment is to use a progressive alignment algorithm (see Nei and Kumar, 2000). The principle of this algorithm is to construct an approximate phylogenetic tree and build up the alignment by progressively adding sequences in the order specified by the tree (see Higgs and Attwood, Taxonomy and colonization of the two Icelandic amphipods 19
2005). This is the algorithm used by the program ClustalW (Larkin et al., 2007). Several improvements have been made in the field of multiple sequences alignment recently (see Morrison, 2009). The program Mafft (Katoh et al., 2002) is incorporating some of them, such as improving the alignment by using an iterative refinement method, and has been shown to perform better than ClustalW (see Morrison, 2009). Another software, RNAsalsa, uses both the secondary structure information (see section below on RNA secondary struc- ture) for adjusting and refining the sequences alignment and the sequence information contained in the alignment to refine the structure predictions in turn (Stocsits et al., 2009). RNAsalsa is particularly suited for alignment of RNA genes such as in Paper 3 and 4, since the divergence between the sequences may depend on the secondary structure of the molecule. The pro- gram FSA (Bradley et al., 2009) in Paper 3 is based on a recently developed statistical alignment method. Its principle is to build up the alignment sites per sites according to statistical pairwise estimations of homology (Bradley et al., 2009). Finally, the alignment of the ITS2 sequences was obtained with the program 4Sale (Seibel et al., 2008) which provides a scoring matrix spe- cific to this nuclear region and allows an alignment based on both secondary structures and nucleotide sequences simultaneously.
Phylogenetic methods
The commonly used methods to build phylogenies can be divided in two major groups: the methods based on distances and the ones based on characters (see Yang, 2006). In distance based methods, the phylogenetic tree is constructed with distances computed for all pairs of taxa. The neighbor joining algorithm (Saitou and Nei, 1987) is a distance based method which was used in Papers 4 and 5. The maximum parsimony, maximum likelihood and Bayesian methods are all character based (Yang, 2006), which means that every single site is treated independently. Trees selected according to the maximum parsimony methods are the one which require the smallest number of evolutionary changes among these characters. In the maximum likelihood method, the likelihood value of the tree is used to measure the fit of the tree to the characters states. The tree chosen will be the one with the highest likelihood. Finally, following 20 Materials and Methods the Bayesian method, the tree selected as the “true” tree is the one with the maximum posterior probability i.e. the probability that the tree is true given the data. The substitution model fitting best the data, is used to compute the pairwise distances for distance methods and to calculate the likelihood function for the Bayesian and the likelihood methods (Yang, 2006). In order to assess the influence of the reconstruction method on the tree, a Bayesian (with MrBayes, Ronquist and Huelsenbeck, 2003) and a max- imum likelihood method (with Phyml, Guindon and Gascuel, 2003) were applied and are presented in Paper 3. MrBayes software presents the main ad- vantages to allow partition of the data and to implement two different evolution models for paired and unpaired regions according to the secondary structure of rRNA genes. The software ProfdistS (Wolf et al., 2008) used in Paper 4 is based on a profile neighbor joining method which has been shown to provide better estimations of the deepest evolutionary relationships i.e. the basal nodes of the tree (Müller et al., 2004). The particular interest of ProfdistS for this study is the computation of phylogenetic trees based on both the sequence and the secondary structure of the ITS2 region. Trees based on ITS1 and complete data set (ITS1, 5.8S, ITS2) were reconstructed using a simple neighbor joining method for comparison with the tree obtained with the ITS2.
2.3.2 Phylogeny based on morphological data
Phylogenies based on morphological characters were reconstructed to check the taxonomic status of C. islandicus by assessing its morphological affinities with other Crangonyx species and to infer putative colonization routes (Paper 5). In their description of the species, Svavarsson and Kristjánsson (2006) suggested that C. islandicus might represent a new genus according to morpho- logical features unique to the species. The taxonomic status of C. islandicus as a species or a specific genus was evaluated with morphological characters previously defined by Zhang and Holsinger (2003) to assess the phyloge- netic relationships among North American Crangonyx species. The phylogeny was reconstructed by adding C. islandicus, C. subterraneus and C. chlebnikovi from Eurasia (Paper 5) and compared with the phylogenies based on molecular markers (Paper 3). The body parts used by Zhang and Holsinger (2003) Secondary structure of the ITS regions 21
Figure 2.3: Crangonyx islandicus morphology. Numbers correspond to anatomical parts of interest for the morphological study in Paper 5. 1: lo- cation of the mouthparts, Mandible and Maxilla (not visible on the drawing), 2: Maxilliped, 3: Gnathopod 1, 4: Gnathopod 2, 5: Pleopod 1, 6: Telson. to define the morphological characters are shown in Figure 2.3 and the Figure 2.4 shows one example of these characters. The phylogeny was reconstructed using both maximum parsimony and neighbor joining methods. Maximum par- simony was designed originally for morphological characters (Hennig, 1966) and was chosen in order to compare with the phylogenies already obtained by Zhang and Holsinger (2003).
2.4 Secondary structure of the ITS regions
The internal transcribed spacers, ITS1 and ITS2, are located respectively be- tween the 18S and 5.8S and between the 5.8S and the 28S ribosomal genes (Fig 2.5). These ITS regions and particularly their secondary structures have a crucial role in the processing of the pre rRNAs (Musters et al., 1990; van Nues et al., 1995). 22 Materials and Methods
Figure 2.4: The telson of two Crangonyx species with varying number of spines on each lobe. a: C. islandicus, less than 4 spines per lobe, b: C. dearolfi, more than 4 spines per lobes.
Figure 2.5: Location and structure of the ITS2. The excision site of the ITS2 is highlighted in red. The four main stem structures of the ITS2 are written I, II, III, IV. Modified from Keller et al. (2009) Secondary structure of the ITS regions 23
Figure 2.6: Description of the two main types of secondary structure. Hydrogen bonds between complementary bases are represented by the blue lines. A compensatory base change between sequence a and b is highlighted in green: the bases are both different but the hydrogen bond is conserved.
2.4.1 Ribosomal RNA secondary structure
Ribosomal RNA (rRNA) genes are synthesized in cells as single stranded and fold into a secondary structure which is essential to their function in the ribo- some. Their secondary structure and the ones from their internal transcribed spacers (ITS1 and ITS2) emerge from the presence of complementary bases along the sequence which form hydrogen bonds (Fig 2.6). The two main structures observed are stems (the paired regions) and loops (the unpaired ones). One of the most commonly used method to infer secondary structures of RNAs is based on the minimum free energy (see Mathews and Turner, 2006). Stems stabilize the structure and contribute to negative free energy, which value depends on the pairing of the bases found along the stem. Loops destabilize the structure and contribute to positive free energy. The secondary structure selected according to this method is the one with the minimal free energy. This method was used for secondary structure predictions in Papers 2, 3 and 4.
2.4.2 Characterization of the ITS regions, structures and phy- logenetic reconstruction
The variation of the ITS1 and ITS2 regions in terms of sequence and sec- ondary structure was estimated within C. islandicus (Paper 2) and between 24 Materials and Methods
Crangonyctoidea species (Paper 4). The number of stems in the ITS2 sec- ondary structure is variable among Eukaryotes but the stem regions II and III (Fig 2.5) are known to be common to essentially all of them (Coleman, 2007). In addition, a conserved hybridization region between the 5.8S and the 28S and its characteristic secondary structure (see Fig 2.5) have been used to delimit the ITS2 (Keller et al., 2009). This method was employed in Paper 2 and 4. Secondary structures obtained for the ITS2 in Paper 4 have been compared to the previously described model. Due to relatively high rate of evolution of the nucleotide sequences of the ITS1 and ITS2, these markers have been more commonly used to reconstruct phylogenies at the species and genus level. Nonetheless, recent methods have been developed for the ITS2 in order to reconstruct phylogenies at higher taxo- nomic levels by using both the nucleotide and structure sequence phylogenetic signal (e.g. Tippery and Les, 2008; Buchheim et al., 2011). One such method is implemented in the software ProfdistS (Wolf et al., 2008), which was used to reconstruct the phylogeny of Crangonyctoidea species (Paper 4).
2.4.3 ITS2 secondary structure and cryptic diversity
Compensatory base changes (CBC) are mutations occurring in both nucleotides of a stem structure, retaining the initial paired nucleotide bond (Fig 2.6). The occurrence of these CBC violates the assumption of independent evolution of the bases along the sequence and might affect phylogenetic reconstruction. Thus it is necessary to consider two different models of evolution for bases which are in stem and in loop regions. This was done for the phylogenies based on the 18S and 28S rRNA genes in Paper 3 with the software Mrbayes. In addition, CBC present in ITS2 sequences have been proposed as a species delimitation criteria (Müller et al., 2007) and their occurrence have been checked in Paper 2 to assess for potential cryptic species within C. islandicus. Results and Discussion
3.1 The subglacial refugia hypothesis
The occurrence of subglacial refugia in Iceland was tested in Paper 1 and 2. Based on the variation of the mt genes COI and 16S, the populations of C. is- landicus are highly structured genetically and geographically (Fig 3.1). Unlike most species at high latitudes and formerly glaciated areas (Hewitt, 2004), C. islandicus presents high genetic diversity and divergent genetic clades. The species is divided in six well supported monophyletic groups, which have di- verged from each other from about 430 000 years up to 5 million years ago. In addition, both COI and 16S mt genes show a clear isolation by distance of C. islandicus populations, genetic distances being highly correlated to the geographic distances. This supports a single colonization event of Iceland from ancestral populations of C. islandicus which diverged within the island after the colonization. Such patterns are unlikely to be observed from multiple col- onization events or in the case of ancestral polymorphism. As this divergence within C. islandicus is dated to up to 5 million years ago, their ancestors col- onized Iceland before the last glacial maximum and even before the onset of the last Ice Age, 2.75 million years ago (see Geirsdóttir et al., 2007). Only small ice free refugia at nunataks were possibly located in northwestern, north- ern and eastern Iceland during the last glacial maximum (see Geirsdóttir et al., 2007). Survival in such located ice free mountain peaks is unlikely for groundwater species and can hardly explain the phylogeographic structure ob- served among C. islandicus populations. Thus, the variation observed in mt genes COI and 16S provides a strong evidence for the occurrence of subglacial
25 26 Results and Discussion refugia in Iceland during the last Ice Age. This is the first documented example of a groundwater amphipod surviving underneath glaciers. The nuclear internal transcribed spacers showed conversely much less vari- ation within C. islandicus than the mtDNA sequences (Paper 2) and the ITS1 variation is only weakly supporting the groups defined by mtDNA variation (Fig 3.1). Nonetheless, ITS1 pairwise distances were correlated to COI dis- tances and to geographic distances supporting the isolation by distance in C. islandicus populations observed with the mt genes. As no calibration of the ITS1 and ITS2 substitution rates is available for crustacean species, it was not possible to estimate divergence times from the variation of the ITS region. The putative location of the subglacial refugia was investigated by evaluat- ing the correlation between genetic diversity and altitude, age of the lava fields and distance from the tectonic plates boundaries. The genetic diversity of the COI gene increases significantly with altitude and the less diverse populations were found at an altitude lower than 100 meters. This is reflecting recent colo- nization of previously submerged areas due to the past increase in sea level in Iceland, which was 100 meters above its current position about 12 500 years ago (Norðdahl et al., 2008). Interestingly, the genetic diversity was corre- lated with the distance from the tectonic plates boundaries. These boundaries are rich in fissure zones (see Fig 1.3) which are particularly common near the Þingvallavatn, Mývatn and Lakagígar areas (see Fig 1.3). The two most diverse phylogenetic groups (AA’ and D in Paper 1) within C. islandicus were found at Þingvallavatn and Lakagígar. Consequently, the genetic diversity patterns in C. islandicus populations suggest that fissure zones at the tectonic plates boundaries were potential refugia during the glacial periods of the Ice Age. The genetic divergence between C. islandicus populations is supporting the existence of several subglacial refugia. The six main monophyletic groups observed in the mtDNA phylogeny appear to have diverged before the last glacial maximum and hence at least six refuge areas have occurred in the corresponding geographic regions of Iceland during the last glaciations. Cryptic diversity within C. islandicus 27 a b c F A E
A'
B A' C C B D D E A F
Figure 3.1: Genetic and geographic structure of C. islandicus populations in Iceland. a: Sampling locations and their respective mtDNA groups (A, A’, B, C, D, E, F). b: Phylogenetic tree based on COI and 16S mt genes (modified from Paper 1). c: Network of ITS1 haplotypes (modified from Paper 2). Colors correspond in all three figures to the groups AA’, B, C, D, E, F.
3.2 Cryptic diversity within C. islandicus
According to the COI and 16S genes, C. islandicus presents high genetic di- versity and is structured in highly divergent clades. One clade from North Iceland (group F in Paper 1) diverged early from the rest of the species and is separated by 0.12 substitutions per site from the other clades for the COI gene. This is lower than the threshold of 0.16 subst/site defined by Lefébure et al. (2006) for the delimitation of Crustacean species, thus it is not supporting C. islandicus as a species complex. The ITS2 variation does not either support C. islandicus as a species complex, since no compensatory base change were observed among C. islandicus populations (Paper 2). Conversely, according to the species screening threshold defined by Hebert et al. (2004), the genetic distances between the F group and the other groups are ten times larger than the average intra-populations distances and therefore, the F group represents a provisional species, different from the other C. islandicus populations (Pa- per 1). No variation, in the morphological characters used in Paper 5, was observed between the A and F group defined by mtDNA variation in Paper 1. Further morphological and morphometric analysis are needed to conclude about the potential species complex status of C. islandicus. 28 Results and Discussion
3.3 Phylogenetic status of the Icelandic amphipods and the colonization of Iceland
Kristjánsson and Svavarsson (2007) suggested that C. thingvallensis and C. islandicus colonized Iceland from Greenland. Molecular phylogenies based on nuclear markers in this thesis are supporting this hypothesis (Paper 3 and 4). According to the phylogenies based on nuclear markers 18S and 28S in Paper 3, C. islandicus is more closely related to North American Crangonyx species. C. subterraneus and C. chlebnikovi (both from Eurasia) are highly divergent genetically from the other Crangonyx species indicating an ancient separation between North American and Eurasian Crangonyx species. Phylogenies in Pa- per 4, based on the ITS2 and the complete dataset (ITS1, 5.8S, ITS2), are as well supporting a ancient separation of C. islandicus from the two Eurasian Crangonyx species. Thus, it is more likely that C. islandicus colonized Iceland from the Nearctic which harbors most of the Crangonyx species, probably via past land contacts with Greenland. C. islandicus have most probably colo- nized via the Greenland-Iceland transverse ridge which connected Greenland and Iceland until 15 or even 10 Mya, before its submersion (see Grimsson et al., 2007; Denk et al., 2010). This supports further that C. islandicus colo- nized Iceland before the onset of the last Ice Age and had to survive repeated glaciations in Iceland. The morphological analysis in Paper 5 shows that C. islandicus is closely re- lated to C. subterraneus which is incongruent with the phylogenies based on the nuclear markers. The morphology tends to support a colonization of Iceland from Europe via the Iceland-Scotland part of the North Atlantic transverse ridge. Crangonyx species from Eurasia and from North America are highly di- vergent genetically and the absence of such split based on morphological data points to potential convergent evolution in morphological traits, particularly common in subterranean species, or to homoplasy in the characters used. Taxonomy of the Crangonyctoidea and convergence in morphology 29 a Crymostygius thingvallensis b Crangonyx islandicus Crymostygius thingvallensis Crangonyx forbesi
Crangonyx sp Crangonyx pseudogracilis
Crangonyx pseudogracilis Crangonyx islandicus Crangonyx floridanus
Synurella dentata Crangonyx forbesi Synurella sp
Amurocrangonyx arsenjevi Stygobromus gracilipes Synurella ambulans a-b Crangonyx chlebnikovi Bactrurus brachycaudus p<0.8 Bactrurus mucrunatus 0.8>p>0.95 Synurella ambulans p>0.95 Bactrurus pseudomucrunatus
Stygobromus mackini Stygobromus stegerorum
Stygobromus stegerorum 0.05 Crangonyx subterraneus Stygobromus gracilipes
Crangonyx subterraneus 0.01 Crangonyx chlebnikovi c
Crymostygius thingvallensis
Crangonyx islandicus
Stygobromus stegerorum Crangonyx subterraneus c
boot<80 Crangonyx chlebnikovi 80 Stygobromus gracilipes 0.01 Figure 3.2: Phylogenetic relationships among Crangonyctoidea species. a: Bayesian tree based on 18S gene, b: Bayesian tree based on 28S gene, c: Neighbor joining tree based on ITS1, 5.8S and ITS2. Posterior probabilities for trees a and b are displayed in box a-b. Bootstrap values for the tree c are displayed in box c. Crangonyx lineages are highlighted in red. 3.4 Taxonomy of the Crangonyctoidea and conver- gence in morphology The taxonomic status of the two Icelandic groundwater amphipods based on morphology (Kristjánsson and Svavarsson, 2004; Svavarsson and Kristjánsson, 2006) are relatively supported by the phylogenies based on 30 Results and Discussion the nuclear markers in Paper 3 and 4. C. thingvallensis is clearly divergent from the other Crangonyctoidea species included in this study for phylogenies based on the 18S, 28S, and the ITS regions. Though the family Crymosty- gidae appears to be a sister family of the Crangonyctidae, larger phylogenies encompassing the 11 families of the superfamily Crangonyctoidea are needed in order to reveal the phylogenetic relationships among these families. Phylo- genies based on 18S, 28S and 16S (Paper 3) show C. islandicus to be closely related to the Crangonyx species, supporting its taxonomic status. However, the distances observed for the 18S and 28S between C. islandicus and the other Crangonyx species are more similar to distances between genera than within other amphipod genera. This tends to indicate that C. islandicus fits poorly within the genus Crangonyx and should possibly be removed from that genus. Nonetheless, the tree based on morphological data in Paper 5 is not giving support to this taxonomic status. Based on morphology, C. islandicus clusters within Crangonyx in a separate branch with C. subterraneus, C. longicarpus, C. baculispina and C. packardi. Sequencing of these species could allow to test their relationships and provide an evaluation of the unique taxonomic status of C. islandicus. In all phylogenies based on nuclear markers (Paper 3 and 4), the species from the Crangonyctidae family clustered in a monophyletic group. This sup- ports the occurrence of a common ancestor of the family which was widespread in Laurasia before its breakup around 55 Mya, as hypothesized by Holsinger (1992). Based on 18S, 28S and ITS2 phylogenies, the Crangonyx genus is clearly polyphyletic, species from Eurasia being highly divergent from species in North America and in Iceland. Conversely, the morphological data is not supporting a clear split between species from these two geographic regions (Pa- per 5). The most likely hypothesis to explain this discrepancy is convergent evolution of the morphological traits, common among subterranean species. This might affect the characters used to reconstruct the phylogeny based on morphology and leads to an underestimation of the evolutionary distances be- tween North American and Eurasian species. Thus, the genus Crangonyx needs to be redescribed using a more integrative taxonomy (see Dayrat, 2005) i.e. by a combination of morphological, molecular and phylogeographic data. The Evolution of the ITS regions 31 genus Synurella is as well polyphyletic according to the phylogenies based on 18S and COI (Paper 3). Interestingly, Synurella species from North America are clustering with the North American Crangonyx and reciprocally, Synurella ambulans from Eurasia is more related to Eurasian Crangonyx than to other Synurella species. Finally, the monophyly of the genus Stygobromus is poorly supported by phylogenies based on 18S and 28S (depending on alignment meth- ods) and is not supported by the variation in ITS regions. Consequently, three genera among the family Crangonyctidae i.e. Crangonyx, Synurella and Stygo- bromus might need a taxonomic revision, combining morphological and genetic data in order to restore their monophyly. 3.5 Evolution of the ITS regions The variation of the ITS regions was investigated in this thesis within the species C. islandicus (Paper 2) and between species of the superfamily Cran- gonyctoidea (Paper 4). The ITS1 and ITS2 are two members of the rRNA gene family, which other members are 18S, 5.8S, 28S, ETS1, ETS2 and IGS (see Fig 2.5). The evolution of the RNA gene family is commonly explained by a model of concerted evolution. According to this model, each member of a gene family are evolving in a concerted manner and a mutation occurring in one member will either be purged or occasionally spread through all the mem- bers of the family by repeated unequal crossing-overs or gene conversions (Nei and Rooney, 2005). Concerted evolution leads to homogeneous copies within species but large variations between species (e.g. Brown et al., 1972). Paper 4 reports high variations in sequence and secondary structure among Crangonyc- tidae species whereas in Paper 2, low nucleotide variation was observed for the ITS1 and ITS2 regions within C. islandicus. Nonetheless, interesting patterns of evolution were found among C. islandicus populations. A duplication event occurred in the ITS1 region solely in the southern populations. The most likely hypothesis is a recent duplication, after the split of the northern and southern populations around 1 million year ago according to time estimates based on COI and 16S variations in Paper 1. The high divergence observed between the two duplicated regions is therefore pointing at a rapid divergent evolution 32 Results and Discussion where the duplicated gene may acquire a new function (Nei and Rooney, 2005), opposite to the homogenization effect of concerted evolution. The oc- currence of the duplication in all individuals from the southern populations might be due to the effect of such divergent evolution. In paper 4, the ITS2 sequences, lengths and secondary structures are found to be highly variable between amphipod species. This appears incongruent with previous descriptions of the ITS2 secondary structure as highly conserved and presenting structures nearly universal among Eukaryotes (Schultz et al., 2005; Coleman, 2007). The number of helices in the secondary structure, seen as relatively constant within different groups of organisms (see Cole- man, 2007), was as well variable among the amphipod species studied. These discrepancies do not appear to result from pseudogenes as the high conser- vation of the 5.8S and of the ITS2 excision site among the Crangonyctoidea species tend to support the functionality of the ITS2 region amplified. Further investigations of the ITS2 secondary structure variation among amphipods are needed to identify potential homologies and conserved motifs specific to the group. Conclusion and Perspectives The occurrence of the endemic groundwater amphipod C. islandicus in Iceland can be explained by an ancient colonization, which predates the onset of the last Ice Age, around 2.75 million years ago. Molecular data supports that the ancestor of C. islandicus colonized via past freshwater, or even groundwater, contacts between Greenland and Iceland occurring in the Greenland-Iceland land bridge which was still subaerial around 10 million years ago. C. islandi- cus thus survived repeated glacial cycles in Iceland. The genetic variation observed within C. islandicus populations is highly supporting the occurrence of several subglacial refugia located in fissure zones near the tectonic plate boundaries, probably maintained by the important geothermal activity of these areas. Though the hypothesis could not be tested for C. thingvallensis due to its restricted distribution, its only location in one of the main fissure area of Ice- land, lake Þingvallavatn, indicates that this species might have survived in the same refugia. This thesis demonstrates that the occurrence of subglacial refu- gia can explain phylogeographic patterns in groundwater amphipod species. Subglacial refugia have been hypothesized to explain biogeographic patterns in amphipod species in North America and Ireland. Considering the locations along the volcanic active zone in Iceland where the groundwater amphipods are found and the observed genetic patterns, fissure zones with groundwater possi- bly maintained by geothermal activity underneath glaciers may be required to sustain such refugia. Thus their occurrence could be limited to volcanic active regions. The survival of amphipod species in subglacial refugia indicates that a spe- cific community has thrived in these refugia during glaciations. The ground- 33 34 Conclusion and Perspectives water fauna in Iceland is still poorly described, only three species being known so far (Juberthie et al., ress). Further phylogeographic studies on groundwa- ter species in Iceland might give a better estimate of the biota which survived on the island. For example, several epibionts living on the cuticule of C. is- landicus have been observed during this study. Preliminary molecular analysis showed that some of these species might belong to the Ciliophora phylum. These organisms, tightly linked to C. islandicus, could have found the amphi- pod after the Ice Age or might as well have survived glaciations in Iceland. C. islandicus were found in some places very near to the surface and hence can probably receive epibionts from the limnic environment. The identifica- tion of these epibionts by morphological and molecular analysis could shed light on the time of their colonization. In addition, bacterial communities are known to live in subglacial lake underneath the Vatnajökull glacier in Iceland (Gaidos et al., 2004, 2009). This community appears to be dominated by chemotrophic bacteria (Gaidos et al., 2009). Such bacterias could have been primary producers in the ecosystems which survived glaciations in the fissure zones. The species of Crangonyctidae included in this thesis clustered in a mono- phyletic group in all nuclear phylogenies produced for this thesis and as ob- served earlier by Englisch et al. (2003). The occurrence of an ancestor of this family, widespread on the Laurasia continent is therefore supported. Nonethe- less, the family Crangonyctidae is composed of 7 extant genera i.e. Crangonyx, Synurella, Bactrurus, Stygobromus, Stygonyx, Amurocrangonyx and Lyurella (see Messouli, 2006; Sidorov and Holsinger, 2007), and adding more taxa to the phylogeny is needed in order to conclude about the monophyly of the family. Due to genetic linkage of the 18S, 28S and ITS2, other nuclear markers should be integrated in future studies. Nuclear phylogenies revealed three genera of the Crangonyctidae as poly- phyletic. Two of them, the genera Crangonyx and Synurella were separated into two highly divergent groups corresponding to Palearctic and Nearctic species. Consequently, these genera might need a taxonomic revision based on further molecular and morphological analysis. The morphological markers used did not lead to a distinction between Eurasian and North American Cran- Conclusion and Perspectives 35 gonyx species, while molecular phylogenies show an ancient divergence between the species of these two regions. This could be explained by convergent evo- lution, morphological stasis or homoplasy in the characters used, which lead to an underestimation of the divergence between species from Nearctic and Palearctic regions. The abundance of convergence or morphological stasis in morphological traits in subterranean species and the discordance between mor- phological and molecular affinities of Crangonyctidae species observed in this thesis highlight the need for a more integrative taxonomy (see Dayrat, 2005) for groundwater species. The split observed between Nearctic and Palearctic species of the Crangonyx and Synurella genera indicates that phylogeography is of particular interest in this integrative taxonomy approach. The phylogeo- graphic pattern observed in these two genera might correspond to the opening of the North Atlantic ocean which separated Nearctic and Palearctic regions around 55 million years ago. This could be tested by producing phylogenies of the Crangonyx and Synurella genera based on molecular markers with cal- ibrated substitution rates. Other genes than COI and 16S might have to be considered, since the variation of these genes observed in this thesis among Crangonyctidae species showed signs of saturation and gave poor phylogenetic resolution. In this study, various molecular markers were used at different taxonomic levels. Depending on the taxonomic level, molecular markers with a too low evolution rate will lead to low phylogenetic signal whereas a too high evolution rate will cause a saturation of the signal which both influence the robustness and accuracy of the gene trees reconstructed. The mt genes COI and 16S and the nuclear ITS1 are shown to be informative at the species level, while they gave poor resolution when used to construct phylogenies encompassing different families. Conversely, almost no variation was found with the 18S, 28S and ITS2 at the species level, but these markers were suited to build phylogenies at the family level. Interestingly, the variation pattern among C. islandicus populations is highly different for the two DNA barcodes COI and ITS2. COI sequences were highly variable and the three species delimitation thresholds used led to different conclusions on the occurrence of cryptic species within C. islandicus. Conversely, the ITS2 showed almost no variation. 36 Conclusion and Perspectives In order to test further the potential species complex of C. islandicus, morphometric analysis could be conducted. Such analysis have been shown to be potentially informative for species delimitation as well as for sex recognition (Wilhelm and Venarsky, 2009). This represents a particular interest for C. islandicus since only female individuals have been described so far, probably due to subtle morphological differences, biased sex ratios or different behaviors between sexes. The estimation of the groundwater biodiversity represents a challenge, both due to sampling difficulties and the abundance of cryptic diversity. As seen in this thesis, species considered congeneric might be highly divergent genetically and the use of integrative taxonomy is necessary to assess subterranean species diversity. A recent study showed that the use of morphospecies – species delim- ited solely by morphology – to assess the impact of the global climate change leads to an underestimation of the biodiversity loss and cryptic diversity should be considered in these estimations (Bálint et al., 2011). The increasing human water consumption, especially for agriculture, and the global climate change will have an certain affect on groundwater ecosystems (Danielopol et al., 2003) and better estimations of the groundwater biodiversity using molecular taxonomy are though needed. Furthermore, the identification of past refugia and the description of distribution changes of the genetic diversity over time are of primary concern to define future conservation units. Bibliography Andrews, J. T., A. E. Jennings, A. E. Sveinbjörnsdóttir, J. Hardardóttir, G. B. Kristjánsdóttir, G. Helgadóttir, J. P. M. Syvitski, A. Geirsdóttir, S. Schoolfield, L. Smith and K. Thors, 2000. The N and W Iceland Shelf: insights into Last Glacial Maximum ice extent and deglaciation based on acoustic stratigraphy and basal radiocar- bon AMS dates. Quaternary Science Reviews 19: 619–631. Audzijonyte, A. and R. Väinölä, 2006. Phylogeographic analyses of a circumarctic coastal and a boreal lacustrine mysid crustacean, and evidence of fast postglacial mtDNA rates. Molecular Ecology 15: 3287–301. Bálint, M., S. Domisch, C. H. M. Engelhardt, P. Haase, S. Lehrian, J. Sauer, K. Theissinger, S. U. Pauls and C. Nowak, 2011. Cryptic biodiversity loss linked to global climate change. Nature Climate Change 1: 313–318. Bradley, R. K., A. Roberts, M. Smoot, S. Juvekar, J. Do, C. Dewey, I. Holmes and L. Pachter, 2009. Fast statistical alignment. PLoS Com- putational Biology 5: 1–15. Brochmann, C., T. M. Gabrielsen, I. Nordal, J. Y. Landvik and R. Elven, 2003. Glacial Survival or tabula rasa? The History of North Atlantic Biota Revisited. Taxon 52: 417–450. Brown, D., P. Wensink and E. Jordan, 1972. A comparison of the ribo- somal DNA’s of and : the evolution of tandem genes. Journal of Molecular Biology 63: 57–73. 37 38 BIBLIOGRAPHY Buchheim, M. A., A. Keller, C. Koetschan, F. Förster, B. Mer- get and M. Wolf, 2011. Internal Transcribed Spacer 2 (nu ITS2 rRNA) Sequence-Structure Phylogenetics: Towards an Automated Reconstruction of the Green Algal Tree of Life. PLoS ONE 6: 1–10. Buckland, P. C., D. W. Perry, G. M. Gíslason and A. J. Dugmore, 1986. The pre-Landnám fauna of Iceland: a palaeontological contribution. Boreas 15: 173–184. Chu, K. H., C. P. Li and H. Y. Ho, 2001. The first internal transcribed spacer (ITS-1) of ribosomal DNA as a molecular marker for phylogenetic and population analyses in crustacea. Marine Biotechnology 3: 355–61. Coleman, A. W., 2007. Pan-eukaryote ITS2 homologies revealed by RNA secondary structure. Nucleic Acids Research 35: 3322–9. Coope, G., P. Moore and A. Gibbs, 1986. The Invasion and Colonization of the North Atlantic Islands: A Palaeoecological Solution to a Biogeographic Problem [and Discussion]. Philosophical Transactions of the Royal Society of London. B, Biological Sciences 314: 619–635. Cunningham, C. W., N. W. Blackstone and L. W. Buss, 1992. Evolution of king crabs from hermit crab ancestors. Nature 355: 539–542. Danielopol, D., C. Griebler, A. Gunatilaka and A. Notenboom, 2003. Present state and future prospects for groundwater ecosystems. En- vironmental Conservation 30: 104–130. Danielopol, D. L., P. Pospisil and R. Rouch, 2000. Biodiversity in groundwater: a large-scale view. Trends in Ecology & Evolution 15: 223– 224. Dayrat, B., 2005. Towards integrative taxonomy. Biological Journal of the Linnean Society 85: 407–415. Denk, T., F. Grimsson and R. Zetter, 2010. Episodic migration of oaks to Iceland: Evidence for a North Atlantic "land bridge" in the latest Miocene. American Journal of Botany 97: 276–287. BIBLIOGRAPHY 39 Denk, T., F. Grimsson, R. Zetter and L. A. Símonarson, 2011. Intro- duction to the nature and geology of Iceland. In Late Cainozoic Floras of Iceland, volume 35 of Topics in Geobiology, pp. 1–29. Springer Netherlands, Dordrecht. Drummond, A. J. and A. Rambaut, 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7: 214–222. Egilsdóttir, H. and B. Kristjánsson, 2008. Dvergbleikja í grennd við Jökulsá á Fjöllum. Náttúrufræðingurinn pp. 109–114. Einarsson, E., 1963. The elements and affinities of the Icelandic flora. In North Atlantic biota and their history, edited by Löve, A. and D. Löve, pp. 297–302. Pergamon, Oxford. Einarsson, T., 1994. Geology of Iceland. Rocks and Landscape. Mál & Men- ning, Reykjavik. Englisch, U., C. O. Coleman and J. W. Wägele, 2003. First observations on the phylogeny of the families Gammaridae, Crangonyctidae, Melitidae, Niphargidae, Megaluropidae and Oedicerotidae (Amphipoda, Crustacea), using small subunit rDNA gene sequences. Journal of Natural History 37: 2461–2486. Gaidos, E., B. Lanoil and T. Thorsteinsson, 2004. A viable microbial community in a subglacial volcanic crater lake, Iceland. Astrobiology 4: 327–345. Gaidos, E., V. Marteinsson, T. Thorsteinsson, T. Jóhannesson, A. R. Rúnarsson, A. Stefansson, B. Glazer, B. Lanoil, M. Skid- more, S. Han, M. Miller, A. Rusch and W. Foo, 2009. An oligarchic microbial assemblage in the anoxic bottom waters of a volcanic subglacial lake. The ISME journal 3: 486–97. Geirsdóttir, A., G. Miller and J. Andrews, 2007. Glaciation, erosion, and landscape evolution of Iceland. Journal of Geodynamics 43: 170–186. 40 BIBLIOGRAPHY Gibert, J., A. Brancelj, A. Camacho, F. Castellarini, C. De Broyer, L. Deharveng, M. Dole-Olivier, C. Douady, D. Galassi, F. Malard and Others, 2005. Groundwater biodiversity. Protocols for the assessment and conservation of aquatic life in the subsurface (PASCALIS): overview and main results. In Proceedings of an International Symposium on World Subterranean Biodiversity, December, pp. 39–51. Gibert, J., D. C. Culver, M.-J. Dole-Olivier, F. Malard, M. C. Christman and L. Deharveng, 2009. Assessing and conserving ground- water biodiversity: synthesis and perspectives. Freshwater Biology 54: 930– 941. Grimsson, F., T. Denk and L. Simonarson, 2007. Middle Miocene floras of Iceland - the early colonization of an island? Review of Palaeobotany and Palynology 144: 181–219. Guindon, S. and O. Gascuel, 2003. A Simple, Fast, and Accurate Algorithm to Estimate Large Phylogenies by Maximum Likelihood. Systematic Biology 52: 696–704. Hanfling, B., I. Douerello-Soler, L. Knight and G. Proudlove, 2008. Molecular studies on the Niphargus kochianus group (Crustacea: Am- phipoda: Niphargidae) in Great Britain and Ireland. Cave and Karst Science 35: 35–40. Hebert, P. D. N., A. Cywinska, S. L. Ball and J. R. DeWaard, 2003. Biological identifications through DNA barcodes. Proceedings of the Royal Society of London B: Biological Sciences 270: 313–21. Hebert, P. D. N., M. Y. Stoeckle, T. S. Zemlak and C. M. Francis, 2004. Identification of Birds through DNA Barcodes. PLoS Biology 2: 1657– 1663. Hennig, W., 1966. Phylogenetic Systematics. University of Illinois Press, Urbana. Hewitt, G., 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907–913. BIBLIOGRAPHY 41 Hewitt, G., 2004. Genetic consequences of climatic oscillations in the Qua- ternary. Philosophical Transactions of the Royal Society of London B 359: 183–195. Higgs, P. G. and T. K. Attwood, 2005. Bioinformatics and molecular evolution. Wiley-Blackwell Publishing. Holsinger, J., 1992. Sternophysingidae, a new family of Subterranean am- phipods (Gammaridea: Crangonyctoidea) from South Africa, with descrip- tion of sternophysinx calceola, new species, and comments on phylogenetic and biogeographic relationships. Journal of Crustacean Biology 12: 111–124. Holsinger, J. R., 1980. Stygobromus canadensis, a new subterranean amphi- pod crustacean (Crangonyctidae) from Canada, with remarks on Wisconsin refugia. Canadian Journal of Zoology 58: 290–297. Humphreys, W. F., T. Bradford, A. D. Austin, M. Adams and S. J. B. Cooper, 2010. DNA barcoding of stygofauna uncovers cryptic amphipod diversity in a calcrete aquifer in Western Australia’s arid zone. Molecular Ecology Resources 10: 41–50. Ingólfsson, A., 2009. A marine refugium in Iceland during the last glacial maximum: fact or fiction? Zoologica Scripta 38: 663–665. Juberthie, C., E. Kornobis and S. Pálsson, in press. Iceland. In ENCY- CLOPÆDIA BIOSPEOLOGICA, edited by Juberthie, C. and V. Decu, chapter Vol 1A, Eu, pp. 333–336. International Society for Subterranean Biology. Katoh, K., K. Misawa, K.-i. Kuma and T. Miyata, 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Research 30: 3059–66. Keller, A., T. Schleicher, J. Schultz, T. Müller, T. Dandekar and M. Wolf, 2009. 5.8S-28S rRNA interaction and HMM-based ITS2 annota- tion. Gene 430: 50–7. 42 BIBLIOGRAPHY Knowlton, N. and L. A. Weigt, 1998. New dates and new rates for diver- gence across the Isthmus of Panama. Proceedings of the Royal Society B: Biological Sciences 265: 2257–2263. Knowlton, N., L. A. Weigt, L. A. Solorzano, D. K. Mills and E. Bermingham, 1993. Divergence in proteins, mitochondrial DNA, and reproductive compatibility across the Isthmus of Panama. Science 260: 1629–32. Koenemann, S. and J. R. Holsinger, 2001. Systematics of the north amer- ican subterranean amphipo genus Bactrurus (Crangonyctidae). Beaufortia 51: 1–56. Kornobis, E., S. Pálsson, B. K. Kristjánsson and J. Svavarsson, 2010. Molecular evidence of the survival of subterranean amphipods (Arthropoda) during Ice Age underneath glaciers in Iceland. Molecular Ecology 19: 2516– 2530. Kristjánsson, B. K. and J. Svavarsson, 2004. Crymostygidae, a new family of subterranean freshwater gammaridean amphipods (Crustacea) recorded from subarctic Europe. Journal of Natural History 38: 1881–1894. Kristjánsson, B. K. and J. Svavarsson, 2007. Subglacial refugia in Ice- land enabled groundwater amphipods to survive glaciations. The American Naturalist 170: 292–6. Larkin, M. A., G. Blackshields, N. P. Brown, R. Chenna, P. A. McGettigan, H. McWilliam, F. Valentin, I. M. Wallace, A. Wilm, R. Lopez, J. D. Thompson, T. J. Gibson and D. G. Higgins, 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–8. Lefébure, T., C. J. Douady, M. Gouy and J. Gibert, 2006. Rela- tionship between morphological taxonomy and molecular divergence within Crustacea: Proposal of a molecular threshold to help species delimitation. Molecular Phylogenetics and Evolution 40: 435–447. Lefébure, T., C. J. Douady, F. Malard and J. Gibert, 2007. Testing dispersal and cryptic diversity in a widely distributed groundwater amphipod BIBLIOGRAPHY 43 (Niphargus rhenorhodanensis). Molecular Phylogenetics and Evolution 42: 676–86. Malard, F. and F. Hervant, 1999. Oxygen supply and the adaptations of animals in groundwater. Freshwater Biology 41: 1–30. Mathews, D. H. and D. H. Turner, 2006. Prediction of RNA secondary structure by free energy minimization. Current opinion in structural biology 16: 270–8. Messouli, M., 2006. The first African member of the Family Crangonyctidae, Crangonyx africanus, sp. nov.(Crustacea, Amphipoda) in the groundwaters of Western Morocco: systematics and biogeographical implications. Subter- ranean Biology 4: 67–78. Mihalffy, P., B. Steinberger and H. Schmeling, 2008. The effect of the large-scale mantle flow field on the Iceland hotspot track. Tectonophysics 447: 5–18. Moritz, C. and C. Cicero, 2004. DNA barcoding: promise and pitfalls. PLoS biology 2: 1529–1531. Morrison, D. and J. Ellis, 1997. Effects of nucleotide sequence align- ment on phylogeny estimation: A case study of 18S rDNAs of Apicomplexa. Molecular Biology and Evolution 14: 428–441. Morrison, D. a., 2008. A framework for phylogenetic sequence alignment. Plant Systematics and Evolution 282: 127–149. Morrison, D. A., 2009. Why Would Phylogeneticists Ignore Computerized Sequence Alignment? Systematic Biology 58: 150–158. Müller, T., N. Philippi, T. Dandekar, J. Schultz and M. Wolf, 2007. Distinguishing species. RNA 13: 1469–1472. Müller, T., S. Rahmann, T. Dandekar and M. Wolf, 2004. Accurate and robust phylogeny estimation based on profile distances: a study of the Chlorophyceae (Chlorophyta). BMC evolutionary biology 4: 20–26. 44 BIBLIOGRAPHY Musters, W., K. Boon, C. van der Sande, H. V. Heerikhuizen and R. J. Planta, 1990. Functional analysis of transcribed ribosomal DNA spacers of yeast. EMBO Journal 9: 3989–3996. Nei, M. and S. Kumar, 2000. Molecular evolution and phylogenetics. Oxford University Press, New York. Nei, M. and A. Rooney, 2005. Concerted and Birth-and-Death Evolution of Multigene Families. Annual Review of Genetics 39: 121–152. Norðdahl, H., O. Ingólfsson, H. G. Pétursson and M. Hallsdót- tir, 2008. Late Weichselian and Holocene environmental history of Iceland. Jökull 58: 343–364. Poore, R. H., 2008. Neogene epeirogeny and the Iceland plume. Ph.D. thesis, University of Cambridge, Cambridge, UK. Provan, J. and K. D. Bennett, 2008. Phylogeographic insights into cryptic glacial refugia. Trends in Ecology & Evolution 23: 564–571. Roe, A. D. and F. A. H. Sperling, 2007. Population structure and species boundary delimitation of cryptic Dioryctria moths: an integrative approach. Molecular ecology 16: 3617–33. Ronquist, F. and J. P. Huelsenbeck, 2003. MrBayes 3: Bayesian phylo- genetic inference under mixed models. Bioinformatics 19: 1572–1574. Rundgren, M. and O. Ingólfsson, 1999. Plant survival in Iceland during periods of glaciation ? Journal of Biogeography 26: 387–396. Sadler, J., 1999. Biodiversity on oceanic islands: a palaeoecological assess- ment. Journal of Biogeography pp. 75–87. Saitou, N. and T. Imanishi, 1989. Relative efficiencies of the Fitch- Margoliash, maximum-likelihood, minimum-evolution, and neighbor-joining methods of the phylogenetic tree construction. Molecular Biology and Evo- lution 6: 514–525. BIBLIOGRAPHY 45 Saitou, N. and M. Nei, 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–25. Schubart, C. D., R. Diesel and S. B. Hedges, 1998. Rapid evolution to terrestrial life in Jamaican crabs. Nature 393: 363–365. Schultz, J., S. Maisel, D. Gerlach, T. Müller and M. Wolf, 2005. A common core of secondary structure of the internal transcribed spacer 2 (ITS2) throughout the Eukaryota. RNA 11: 361–4. Seibel, P. N., T. Müller, T. Dandekar and M. Wolf, 2008. Synchronous visual analysis and editing of RNA sequence and secondary structure align- ments using 4SALE. BMC research notes 1: 91. Seidel, R. A., B. K. Lang and D. J. Berg, 2009. Phylogeographic analysis reveals multiple cryptic species of amphipods (Crustacea: Amphipoda) in Chihuahuan Desert springs. Biological Conservation 142: 2303–2313. Shiklomanov, I., 1993. World fresh water resources. In Water in Crisis. A Guide to the World’s Fresh Water Resources, edited by Gleick, P., pp. 13–24. Oxford University Press, New York, USA. Sidorov, D. A. and J. R. Holsinger, 2007. Amurocrangonyx, a New Genus of Subterranean Amphipod (Crangonyctidae) From the Russian Far East, With a Redescription of the Poorly Known Crangonyx Arsenjevi and Com- ments on Biogeographic Relationships. Journal of Crustacean Biology 27: 660–669. Sigurdsson, H., 1970. Structural origin and plate tectonics of the snaefellsnes volcanic zone, Western Iceland. Earth and Planetary Science Letters 10: 129–135. Stewart, J. and A. Lister, 2001. Cryptic northern refugia and the origins of the modern biota. Trends in Ecology \& Evolution 16: 608–613. Stillman, J. H. and C. A. Reeb, 2001. Molecular phylogeny of eastern Pacific porcelain crabs, genera Petrolisthes and Pachycheles, based on the 46 BIBLIOGRAPHY mtDNA 16S rDNA sequence: phylogeographic and systematic implications. Molecular Phylogenetics and Evolution 19: 236–245. Stocsits, R. R., H. Letsch, J. Hertel, B. Misof and P. F. Stadler, 2009. Accurate and efficient reconstruction of deep phylogenies from struc- tured RNAs. Nucleic Acids Research 37: 6184–93. Svavarsson, J. and B. Kristjánsson, 2006. Crangonyx islandicus sp. nov., a subterranean freshwater amphipod (Crustacea, Amphipoda, Crangonycti- dae) from springs in lava fields in Iceland. Zootaxa 1365: 1–17. Taberlet, P., L. Fumagalli and A. Wust-Saucy, 1998. Comparative phylogeography and postglacial colonization routes in Europe. Molecular Ecology 7: 453–464. Tippery, N. P. and D. H. Les, 2008. Phylogenetic analysis of the internal transcribed spacer (ITS) region in Menyanthaceae using predicted secondary structure. Molecular phylogenetics and evolution 49: 526–37. Trontelj, P., C. J. Douady, C. Fišer, J. Gibert, v. Gorički, T. Lefébure, B. Sket and V. Zakšek, 2009. A molecular test for cryptic diversity in ground water: how large are the ranges of macro-stygobionts? Freshwater Biology 54: 727–744. Väinölä, R., J. D. S. Witt, J. H. Bradbury, K. Jazdzewski, M. Grabowski and B. Sket, 2008. Global diversity of amphipods (Am- phipoda; Crustacea) in freshwater. Hydrobiologia 595: 241–255. van Nues, R. W., J. M. Rientjes, S. A. Morré, E. Mollee, R. J. Planta, J. Venema and H. A. Raué, 1995. Evolutionarily conserved structural elements are critical for processing of Internal Transcribed Spacer 2 from Saccharomyces cerevisiae precursor ribosomal RNA. Journal of Molecular Biology 250: 24–36. Wares, J. P. and C. W. Cunningham, 2001. Phylogeography and historical ecology of the North Atlantic intertidal. Evolution 55: 2455–69. Wattier, R. A., J. Beguet, M. Gaillard, J. C. Muller, L. Bollache and M. J. Perrot-Minnot, 2006. Molecular markers for systematic iden- tification and population genetics of the invasive Ponto-Caspian freshwater gammarid Dikerogammarus villosus (Crustacea, Amphipoda). Molecular Ecology Notes 6: 487–489. Wilhelm, F. M. and M. P. Venarsky, 2009. Variation in Gnathopod Mor- phology of Cave Amphipods and Its Use in Determination of Sex. Journal of Crustacean Biology 29: 26–33. Wolf, M., B. Ruderisch, T. Dandekar, J. Schultz and T. Müller, 2008. ProfDistS: (profile-) distance based phylogeny on sequence–structure alignments. Bioinformatics 24: 2401–2. Yang, Z., 2006. Computational Molecular Evolution. Oxford University Press, New York. Zhang, J. and J. R. Holsinger, 2003. Systematics of the Freshwater Amphi- pod Genus Crangonyx (Crangonyctidae) in North America. Virginia Museum of Natural History Memoirs. Zuckerkandl, E. and L. Pauling, 1965. Evolutionary divergence and con- vergence in proteins. In Evolving genes and proteins, edited by Bryson, V. and H. J. Vogel, pp. 97–166. Academic Press, New York. Ægisdóttir, H. and Þ. Þórhallsdóttir, 2004. Theories on migration and history of the North-Atlantic flora: a review. Jökull 54: 1–16. Paper 1 Molecular evidence of the survival of subterranean amphipods (Arthropoda) during Ice Age underneath glaciers in Iceland 49 Molecular Ecology (2010) 19, 2516–2530 doi: 10.1111/j.1365-294X.2010.04663.x Molecular evidence of the survival of subterranean amphipods (Arthropoda) during Ice Age underneath glaciers in Iceland ETIENNE KORNOBIS,* SNÆBJO¨ RN PA´ LSSON,* BJARNI K. KRISTJA´ NSSON† and JO¨ RUNDUR SVAVARSSON* *Department of Biology, University of Iceland, Askja, Sturlugata 7, 101 Reykjavik, Iceland, †Department of Aquaculture and Fish Biology, Ho´lar University College, Ha´eyri 1, 550 Sauða´rkro´kur, Iceland Abstract Two endemic groundwater arthropod crustacean species, Crangonyx islandicus and Crymostygius thingvallensis, were recently discovered on the mid-Atlantic volcanic island of Iceland. The extent of morphological differences from closest relatives, endemism, along with the geographic isolation of Iceland and its complete coverage by glaciers 21 000 years ago, suggests that these two species have survived glaciation periods in sub-glacial refugia. Here we provide strong support for this hypothesis by an analysis of mitochondrial genetic variation within Crangonyx islandicus. Our results show that the species is divided into several distinct monophyletic groups that are found along the volcanic zone in Iceland, which have been separated by 0.5 to around 5 million years. The genetic divergence between groups reflects geographic distances between sampling sites, indicating that divergence occurred after the colonization of Iceland. The genetic patterns, as well as the dependency of genetic variation on distances from the tectonic plate boundary and altitude, points to recent expansion from several refugia within Iceland. This presents the first genetic evidence of multicellular organisms as complex as crustacean amphipods which have survived glaciations beneath an ice sheet. This survival may be explained by geothermal heat linked to volcanic activities, which may have maintained favourable habitats in fissures along the tectonic plate boundary in Iceland during glaciations. Keywords: Crustacea, glaciation, phylogeography, sub-glacial refugia, subterranean, volcanism Received 25 January 2010; revision received 29 March 2010; accepted 3 April 2010 preceding expansion of populations after the Weichs- Introduction elian glaciation period. Conversely, the refugia are char- Pleistocene glaciations are known to have shaped both acterized by high taxonomic and genetic diversity the distribution and the genetic diversity of species at among and within regions (Hewitt 2004). high latitudes (Hewitt 2004). During glacial expansions, During the last glacial maximum (LGM) around populations either became extinct or managed to sur- 21 000 BP, Iceland, as well as most of northern Europe vive in ice-free refugia. High-latitude areas and for- and North America (Ehlers & Gibbard 2007), was merly glaciated northern parts of Europe and North almost completely covered by glaciers which extended America are, thus, characterized by low endemism and far offshore so that only small ice-free areas may have species diversity, shallow genetic clades and little diver- been present along coastal mountains, mainly in eastern sity within species (Sadler 1999; Hewitt 2004). This is and northern regions (Geirsdo´ttir et al. 2007). The low expected because of recent colonization and bottlenecks species diversity and lack of endemism in the current Icelandic terrestrial biota (e.g. Buckland et al. 1986) can Correspondence: Etienne Kornobis; E-mail: [email protected] be explained by the short time since the retreat of the