High-Throughput Biodiversity Assessment – Powers and Limitations of Meta-Barcoding
Total Page:16
File Type:pdf, Size:1020Kb
High-throughput biodiversity assessment – Powers and limitations of meta-barcoding Hochdurchsatzerfassung von Biodiversität – Stärken und Grenzen von Meta-barcoding Doctoral thesis for a doctoral degree at the Graduate School of Life Sciences, Julius-Maximilians-Universität Würzburg, Section Integrative Biology submitted by Wiebke Sickel from Oranienburg Würzburg, 2016 Submitted on: …………………………………………………………..…….. Office stamp Members of the Promotionskomitee: Chairperson: Prof Dr Thomas Müller Primary Supervisor: Dr Alexander Keller Supervisor (Second): Prof Dr Ingolf Steffan-Dewenter Supervisor (Third): Prof Dr Jörg Schultz Date of Public Defence: …………………………………………….………… Date of Receipt of Certificates: ………………………………………………. Affidavit I hereby confirm that my thesis entitled 'High-throughput biodiversity assessment - powers and limitations of meta-barcoding' is the result of my own work. I did not receive any help or support from commercial consultants. All sources and / or materials applied are listed and specified in the thesis. Furthermore, I confirm that this thesis has not yet been submitted as part of another examination process neither in identical nor in similar form. Würzburg, 07 October 2016 Place, Date Signature Eidesstattliche Erklärung Hiermit erkläre ich an Eides statt, die Dissertation 'Hochdurchsatzerfassung von Biodiversität - Stärken und Grenzen von Meta-barcoding' eigenständig, d.h. insbesondere selbständig und ohne Hilfe eines kommerziellen Promotionsberaters, angefertigt und keine anderen als die von mir angegebenen Quellen und Hilfsmittel verwendet zu haben. Ich erkläre außerdem, dass die Dissertation weder in gleicher noch in ähnlicher Form bereits in einem anderen Prüfungsverfahren vorgelegen hat. Würzburg, 07. Oktober 2016 Ort, Datum Unterschrift Acknowledgements I am highly grateful to my three supervisors, Dr Alexander Keller, Prof Dr Ingolf Steffan-Dewenter and Prof Dr Jorg¨ Schultz. Firstly, for giving me the opportunity to pursue a PhD; secondly for encouraging me to develop and conduct my own research projects; and thirdly for their constant support and advice during these last three years. I would like to thank the rest of the Molecular Biodiversity Group for valuable discussions, helping me out when needed and generally for creating a great working environment; thank you Gudrun Grimmer, Markus J. Ankenbrand, Anna Voulgari-Kokota, and the Bachelor and Master students Annette Bran- del, Mira Becker, Rebecca Balles and Jonas Stelz. Many thanks go to my collaborators Dr Giulia Zancolli, Dr Ivonne Meuche, Dr Andrea Holzschuh, Dr Frank Forster,¨ Prof Dr T. Ulmar Grafe, Dr Dieter Mahsberg, Dr Stephan Hartel¨ and Jonathan Lanzen. I would also like to thank the staff of the Graduate School of Life Sciences. Further thanks go to the following people who so far have not been named: Niklas Terhoeven of the Computational Evolutionary Biology Group within the Centre for Computational and Theoretical Biology, Dr Simone Rost and Jens Graf¨ of the Human Genetics Department, Janina Kay of the Depart- ment for Neurobiology and Genetics as well as my family and (other) friends. vii Summary Traditional species identification based on morphological characters is labo- rious and requires expert knowledge. It is further complicated in the case of species assemblages or degraded and processed material. DNA-barcoding, species identification based on genetic data, has become a suitable alterna- tive, yet species assemblages are still difficult to study. In the past decade meta-barcoding has widely been adopted for the study of species commu- nities, due to technological advances in modern sequencing platforms and because manual separation of individual specimen is not required. Here, meta-barcoding is put into context and applied to the study of bee-collected pollen as well as bacterial communities. These studies provide the basis for a critical evaluation of the powers and limitations of meta-barcoding. Ad- vantages identified include species identification without the need for expert knowledge as well as the high throughput of samples and sequences. In microbiology, meta-barcoding can facilitate directed cultivation of taxa of in- terest identified with meta-barcoding data. Disadvantages include insuffi- cient species resolution due to short read lengths and incomplete reference databases, as well as limitations in abundance estimation of taxa and func- tional profiling. Despite these, meta-barcoding is a powerful method for the analysis of species communities and holds high potential especially for au- tomated biomonitoring. viii Zusammenfassung Traditionelle Methoden der Identifizierung von Organismen anhand von mor- phologischen Merkmalen sind arbeits- und zeitaufwendig und benotigen¨ Ex- pertenkenntnisse der Morphologie. Weitere Probleme liegen in der Anal- yse von Artgemeinschaften und prozessiertem Material. DNA-barcoding, Artbestimmung anhand von genetischen Merkmalen, hat sich als Alterna- tive herausgebildet, jedoch sind Artgemeinschaften nach wie vor schwierig zu analysieren. Im vergangenen Jahrzehnt wurde meta-barcoding zur Anal- yse von Artgemeinschaften entwickelt; insbesondere durch die Weiteren- twicklung moderner Sequenziergerate¨ und da eine Auftrennung der Organ- ismen innerhalb einer Gemeinschaft nicht mehr notwendig ist. In der vor- liegenden Arbeit wurde zunachst¨ ein Uberblick¨ uber¨ meta-barcoding er- stellt. Die Methode wurde dann fur¨ die Analyse von Bienen-gesammeltem Pollen und Bakteriengemeinschaften angewandt. Diese Studien bilden eine gute Basis, um die Vor- und Nachteile von meta-barcoding kritisch zu be- werten. Vorteile beinhalten unter anderem, dass Organismen bestimmt werden konnen,¨ ohne dass Expertenkenntnisse notwendig sind, sowie der hohe Durchsatz von Proben und Sequenzen. In der Mikrobiologie kann meta-barcoding eine gerichtete Kultivierung von Bakterien erleichtern, die durch meta-barcoding als Zielorganismen indentifiziert wurden. Nachteile finden sich in der manchmal noch unzureichenden Unterscheidung nah ver- wandter Arten aufgrund von kurzen Sequenzlangen¨ und luckenhaften¨ Ref- erenzdatenbanken, sowie Einschrankungen¨ in der Abschatzung¨ von Abun- danzen und Funktionen der Organismen innerhalb der Artgemeinschaft. Trotz dieser Problematiken ist meta-barcoding eine leistungsstarke Methode fur¨ die Analyse von Artgemeinschaften und ist besonders vielversprechend fur¨ automatisiertes Bio-Monitoring. ix Table of contents Acknowledgements vii Summary viii Zusammenfassung ix I. Introduction1 II. Publications7 1. Metabarcoding put into context 10 P.1. DNA-Metabarcoding - ein neuer Blick auf organismische Di- versitat...............................¨ 11 2. Pollen analysis 15 P.2. Increased efficiency in identifying mixed pollen samples by meta-barcoding with a dual-indexing approach . 16 P.3. Standard method for identification of bee pollen mixtures through meta-barcoding . 26 3. Bacterial communities 38 P.4. Reptiles as Reservoirs of Bacterial Infections: Real Threat or Methodological Bias? . 39 P.5. Bacterial Diversity and Community Structure in Two Bornean Nepenthes Species with Differences in Nitrogen Acquisition Strategies . 46 III. Discussion 63 x Appendix 88 Bibliography 88 Abbreviations 89 List of Figures 91 List of Tables 92 Author Contributions 93 Curriculum Vitae 99 xi Part I. Introduction 1 Part I Introduction Morphological species identification: A central aspect of biology with limitations The classification of a specimen to a species remains a central aspect of biology (Wiens and Servedio 2000). Its application ranges from system- atic biology and ecology to conservation biology (Wiens and Servedio 2000; Balakrishnan 2005). It can also become important in food safety (Woolfe and Primrose 2004) and law enforcement (Ogden et al. 2009). Tradition- ally, species identifications are based on morphological characters (Wiens and Servedio 2000; Balakrishnan 2005). Sometimes, other aspects are in- cluded, such as behaviour (Balakrishnan 2005). However, there are many situations in which these aspects are not feasible or simply impossible, for example, when the specimen has been processed in some way, which is the case in gut contents (Soininen et al. 2009; Valentini et al. 2009; Pompanon et al. 2012) or in traditional Chinese medicine (Yip et al. 2007; Li et al. 2011; Coghlan et al. 2012). One other example, where classification based on morphological charac- ters is not sufficient, is pollen analysis, which traditionally utilises light mi- croscopy (Mullins and Emberlin 1997). The pollen grains of closely-related plant species very closely resemble one another, so often the lowest taxo- nomic level that can be identified is plant family (Williams and Kremen 2007; Galimberti et al. 2014). In addition, pollen grain classification is very labori- ous (Galimberti et al. 2014) and requires expert knowledge of the respective bioregion the pollen was collected in (Keller et al. 2015). In the case of bacteria, species identification is further complicated by the need to bring them into culture to study them in detail (Handelsman and Smalla 2003). However, cultivation in standard media only captures a low amount of bacterial diversity to study (Handelsman and Smalla 2003). Ad- ditionally, diversity of morphological characters is limited in bacteria (Han- delsman and Smalla 2003), which means that numerous tests of bacterial morphology and physiology are needed to describe a bacterial specimen further and to classify it (Gerner-Smidt et al. 1991; Mata et al. 2002; Ed- berg et al. 1986). This is very