Doctoral Thesis
Refine
Document Type
- Doctoral Thesis (4) (remove)
Has Fulltext
- yes (4)
Is part of the Bibliography
- no (4)
Keywords
- Phylogenie (4) (remove)
Institute
This thesis draws a comprehensive picture about the radiation and diversification of truncatelloidean gastropods across the south pacific. It covers three more specifc studies focussing on the Truncelloideans from Fiji, Vanuatu and New Caledonia, respectively. And a conclusive analysis that combines the results of the three more specific studies and enhances them using species from the Austral Islands, Lord Howe Island, the Indonesian island Sulawesi as well as several species from New Zealand and Australia. Molecular phylogenies were calculated using four nuclear gene fragments (ITS2; 18S rRNA; 28S rRNA and Histone 3) besides the mitochondrial COI and 16S rRNA. Further molecuular data was used to calculate dated phylogenies, perform ancestral range reconstructions and develop a modified molecular barcoding approach.
Chaetognaths are a fascinating taxon with unique features and a great impact on marine food webs as primary predators of zooplankton. Their phylogenetic position has been subject to many speculations ever since their discovery and even contemporary phylogenomic methods have not yet been able to suggest a stable hypothesis on their phylogenetic position within the Bilateria. Neuroanatomical studies may contribute new aspects to this discussion. This study aims to provide new insights into the chaetognath nervous system using a fresh set of methods to determine characters for a phylogenetic discussion. The method of choice in this case was immunohistochemistry combined with confocal microscopy. Experiments were conducted with a host of antibodies. The most effective target antigenes were RFamides (a family of neuropeptides), synapsins (synaptic proteins), tyrosinated tubulin (a cytoskeletal element, especially in neurites) and BrdU (bromodeoxyuridin, a proliferation marker). Each of those markers was of great use in highlighting certain aspects of the nervous system. A fresh look at the development of juvenile chaetognaths shortly after hatching revealed that the ventral nerve center (VNC) is developing earlier than the brain and that the production of neurotransmitters has already started at hatching. Specifically, some neurons exhibit RFmide-like immunoreactivity (ir). Neurogenesis continues for about five days after hatching and the mode of division in the neuronal stemcells is asymmetrical. In adult chaetognaths, the brain is divided into a stomatogastric anterior and a sensory posterior neuropil domain. It contains a set of individually identifiable neurons that exhibit RFamide-like ir. The study highlights the interspecific variation of brain architecture between representatives of spadellids and sagittids. The VNC consists of two lateral bands of somata that flank a central neuropil. Within the VNC exists a serial arrangement of neurons with RFamide-like ir. A variety of other neurotransmitters and related substances are also present in both, the brain and the VNC. More interspecific differences and similarities were explored in another part of the study, comparing even more different chaetognath species and focusing on the VNC and its internal structure. The two species of Krohnitta have an unusual distribution of nuclei that is not clearly separated into two lateral bands like in other species. Many of the sagittid species exhibit a striation pattern of the neuropil that is mostly absent in other groups and some of their nerve nets show varying degrees of order as opposed to the rather disorganized nerve net in other groups. In addition, immunohistochemical methods were applied to several specimens of Gnathostomula sp. in order to test one of the many hypotheses about the chaetognaths phylogenetic position, a sister-group relationship to gnathostomulids. A comparison between the two taxa, taking into account also other gnathifera and platyhelminthes, makes a sistergroup relationship between chaetognaths and gnathostomulids very unlikely. In conclusion, chaetognaths remain in an enigmatic phylogenetic position and likely branched off close to the deuterostome/protostome split.
The goal of this thesis was to study the systematic relationships within the superfamily Sylvioidea (Aves: Passeriformes) in general and within the closely related families Acrocephalidae and Locustellidae in particular, by means of DNA sequences. Sylvioidea itself and families therein were the focus of many studies based as well on morphological characters as on DNA. Due to their morphological similarity and their presumably rapid radiation most studies failed to solve relationships between sylvioidean families and also demarcations of single families and relations within are still in progress. In this study, an enlargement of previous datasets, both taxa and number of DNA sequences, and more sophisticated analysis methods were used to improve the resolution in Sylvioidea, Acrocephalidae and Locustellidae. In addition, the applicability of barcoding in Acrocephalidae was tested. The monophyly of Sylvioidea could be supported and the families Paridae and Remizidae, which were sometimes still included, clustered among the outgroup taxa. Four families, Nicatoridae, Panuridae, Alaudidae, and Macrosphenidae constitute basal splits within Sylvioidea. The division of the former sylviid/timaliid clade in five families, Sylviidae, Leiothrichidae, Pellorneidae, Timaliidae, and Zosteropidae was supported. Scotocerca, Erythrocercus, and Hylia, previously supposed to be members of Cettiidae, were shown not to belong to this family. As the three genera are also morphologically and ecologically different, they were here proposed to be elevated to family rank, Scotocercidae, Erythrocercidae and Hyliidae, respectively. The family Acrocephalidae consisted of the four genera, Nesillas, Acrocephalus, Hippolais, and Chloropeta. In the analysis for this thesis, the latter three appeared to be non-monophyletic. One Acrocephalus species, A. aedon was sister to a clade containing four species of Hippolais as well as two out of three Chloropeta species. They were all merged in the genus Iduna, based on the DNA evidence and shared morphological and ecological characters. Iduna had priority over Hippolais or Chloropeta according to the International Code of Zoological Nomenclature. The one remaining Chloropeta species (C. gracilirostris) had to be renamed to Calamonastides as Chloropeta was no longer available for this taxon. Seven genera were included in the re-analysis of the family Locustellidae: Locustella, Bradypterus, Megalurus, Dromaeocercus, Schoenicola, Cincloramphus, and Eremiornis. Apart from the monotypic genera Dromaeocercus and Eremiornis and Schoenicola, of which only one species was included, the remaining genera were found to be non-monophyletic. One clade contained all Locustella species, Megalurus pryeri and all Asian/Oriental Bradypterus species. All species in this clade were synonymized with Locustella, as the type species of Locustella was included, whereas the type species of Bradypterus fell in a different clade. Therefore, the remaining African Bradypterus species retained their genus name, and Dromaeocercus was renamed to Bradypterus as it clustered within Bradypterus. Cincloramphus, intermingling with the remaining Megalurus species, was synonymized with the latter. Barcoding, growing in popularity for delimiting species, was tested in its applicability for Acrocephalidae. Fourteen taxa currently recognized as full species would fall under the 2% threshold of sequence divergence proposed for delimiting species using the mitochondrial cytochrome b gene. It was also shown that it is important to clarify which part of a DNA sequence is used, as different parts can give different results regarding the 2% threshold. In addition, the choice of “complete deletion” or “pairwise deletion” in calculating genetic distances is important, if incomplete are sequences used.
Ziel meiner Arbeit war es, die evolutionären Beziehungen innerhalb und zwischen den verschiedenen Arten der Möwen (Laridae) genauer zu untersuchen. Der Großteil der Untersuchungen in dieser Arbeit basiert auf DNA-Sequenzen - mitochondriale Regionen sowie nukleare Intronequenzen. Bei einem molekulare n Ansatz wie in meiner Arbeit ist es von enormer Wichtigkeit, einen umfassenden und nicht zu kleinen Datensatz zu behandeln. Dabei wurde auch darauf geachtet, dass die ausgewählten Sequenzen homolog sind und das Alignment robust ist. Meine Arbeit gliedert sich in sechs Schwerpunkte, auf die ich nun näher eingehen möchte. 1. Phylogenie der Möwen Die vorliegende Arbeit erreichte das gesetzte Ziel einer verbesserten Phylogenierekonstruktion in den Laridae und zeigt deutlich die Mängel der bisherigen molekularer Studien (mit zu wenigen Taxa oder zu kleinen und uninformativen Datensätzen). Sicher bestätigt werden kann in dieser Studie die Unterteilung in eine basale Möwengruppe, bestehend aus sieben Gattungen, sowie der Gattung Larus mit sechs voneinander genetisch differenzierten Gruppen. Eine gute Stützung erfahren alle Gruppen der Larus-Gattung. Schwerer ist aber erwartungsgemäß die genauere Erstellung der Verwandtschaftsbeziehungen der jüngsten Taxa. Zu ihrer Abgrenzung werden weitere Marker benötigt. Entdeckt wurde in der Studie ein Signal (Deletion in den LDH - Sequenzen), das entscheidend zur Bestimmung der Gruppenmitglieder der basalen, nicht-Larus Möwengattungen beiträgt. 2. AFLP-Untersuchung in der Gruppe der Großmöwen Bei der von Vos et al. (1995) entwickelten Methode der AFLP (engl. für amplified fragment length polymorphism)-Analyse ist kein Vorwissen der untersuchten Gen(om)sequenz notwendig. Es gelang mit der AFLP-Untersuchung dieser Arbeit die sieben untersuchten Großmöwentaxa voneinander autosomal zu differenzieren und drei mitochondrial biphyletisch auftretenden Taxa (argentatus, hyperboreus und marinus) zu näher zu charakterisieren. Die Eismöwe (hyperboreus) erhielt ihre Clade 1 - Haplotypen von argentatus-Individuen aus Nordeuropa und die Mantelmöwe (marinus) ihre Clade 2 - Haplotypen von nordamerikanischen Arten, vermutlich smithsonianus. Die europäischen Silbermöwen (argentatus) zeigen beide mitochondrialen Clades in allen untersuchten Kolonien mit einem geographischen Gradienten in deren Verteilung. Hier scheinen Vorläufer der Heringsmöwen ihre Clade 2 Mitochondriengenome in die argentatus-Populationen eingebracht zu haben, die anschließend in einer sekundären Ausbreitungswelle über das vollständige Verbreitungsgebiet verteilt wurden. Autosomal erscheinen sogar vier Genlinien, die auf noch mehr Ausbreitungswellen verweisen. 3. Populationsstudien in Dominikanermöwen (L. dominicanus) Nach einer Publikation von Jiguet (2002) werden bei Dominikanermöwen vier Unterarten unterschieden. Die in dieser Arbeit ermittelten Sequenzen der Gene Cyt b, ND 2 und HVR I zeigen eine klare Differenzierung der untersuchten Kolonien. Die Ursprünge der Dominikanermöwen liegen demnach in Südafrika. Von dort erfolgte die Besiedlung von Argentinien, der Kerguelen-Inseln und der Antarktis in mehreren Ausbreitungswellen. In Chile wurde der südamerikanische Kontinent in einem sehr rezenteren Migrationsereignis zum zweiten Mal kolonisiert. Die dort gefundenen Haplotypen sind den südafrikanischen noch sehr ähnlich. Am jüngsten sind die Populationen Neuseelands und der Chatham-Inseln. 4. Populationsstudie in der Sturmmöwe (L. canus) Ganz anders zeigte sich die genetische Differenzierung für dieselben Gene bei der Sturmmöwe (L. canus) und ihren phänotypisch deutlich unterscheidbaren vier Unterarten. Im mitochondrialen Netzwerk bilden die paläarktischen Taxa canus, heinei und kamtschatschensis eine panmiktische Population. Anders das vierte Taxon brachyrhynchus. Dieses nordamerikanische Taxon unterscheidet sich mitochondrial signifikant von den paläarktischen Individuen. 5. und 6. SNP-Analyse in Großmöwen und Ausblick auf geplante weiterführende Untersuchungen Das Detektieren variabler Nukleotidpositionen (Punktmutationen), die SNPs genannt werden, ist von grundlegender Bedeutung für die weitere Untersuchung der molekularen Evolution. In Rahmen dieser Arbeit wurden 32000 Fragmente mittels der CROPS-Analyse untersucht, dabei wurden in 7400 variablen Fragmenten 11000 SNPs gefunden, 24000 Fragmenten ließen keinerlei genetische Variationen erkennen. Somit zeigt sich in eine Rate von einer variablen Position (SNP) in ~500 Nukleotiden, was mit denen in Säugetieren und Menschen vergleichbar ist. Zukünftig mit diesem umfangreichen Basiswissen eine groß angelegte SNP-Typisierung geplant mit dem Ziel autosomale und sexchromosomale SNPs vergleichend zu analysieren. Des Weiteren können die SNP-Daten auch mit mitochondrialen Daten verglichen werden.