1. Character-based DNA barcoding for identifying conservation units in Odonates J. Rach 1, R. DeSalle 2, I.N. Sarkar 2, B. Schierwater 1,2 & H. Hadrys 1, 3 1 ITZ- Ecology & Evolution, TiHo Hannover, Germany 2 Division of Invertebrate Zoology, American Museum of Natural History, New York, USA 3 Dept. Ecology & Evolutionary Biology, Yale University, New Haven, USA

2. Thank you to: DAWB (CBOL)/DIMACS Sandra Giere Antonia Wargel Janne Timm Linn Groeneveld Nadine Habekost Kai Kamm DFG & BMBF

3. Character-based DNA barcoding: A rapid and reliable method for the identification of conservation units in dragonflies

4. Contents 1.Introduction: - Why barcoding dragonflies? - Why character-based DNA barcoding? - Which genetic marker is appropriate ? 2.Methods - Character-based DNA barcoding 3.Case studies I. Species identification II. Discrimination of conservation units 4. Conclusions & Future prospects

5. 1. Introduction: Why barcoding dragonflies? Odonata (demonstrator system): -Small insect order -Model organisms for ecology and evolution -Wide range of habitat specificity (generalists / specialists) -Fast respond to environmental changes

6. 1. Introduction: Why barcoding dragonflies? - Prime indicators for all types of fresh water ecosystems Terrestrial Aquatic Increasing importance for conservation management

7. 1. Introduction: Why barcoding dragonflies? - Wing veneation: requires a lot of experience - Colours: Bright colours of males fade quickly after death; females of same genus inconspiciuous - Ecological and behavioural patterns: difficult and time-consuming - Larvae: discrimination often impossible Identification through phenotypic traits is difficult: ♀♀

8. 1. Introduction: Why barcoding dragonflies? If phenotypic traits do not serve  Need of genetic approaches! How to get DNA non-invasive: Rapid and reliable identification of dragonflies valuable for conservation management: Exuvia Middle leg (Hadrys et al. 1992)

9. 1. Introduction: Why character-based DNA barcoding? - High intraspecific genetic variability (e.g. geographical clusters) can hinder assignment of unknown samples to their species - Distances between species often lower than within species - Thresholds cannot be defined (might lead to overestimated biodiversity) Distance approaches can be misleading:

10. 1. Introduction: Why character-based DNA barcoding? Identification at any taxonomic level Diagnostic characters useful for DNA barcoding: Character-based DNA barcodes for species and single populations

11. 1. Introduction: Which genetic marker is appropriate? Has not been applied for Odonates before: CO1 (cytochrome c oxidase 1) supposed to be appropriate for DNA barcoding of most animal groups:  Search for conserved primer sequences  Optimization of PCR conditions  Test for suitability

12. 1. Introduction: Which genetic marker is appropriate? - Sequences easy to obtain and analyse - Detection of geographical patterns - Identification of conservation units ND1 (NADH dehydrogenase subunit 1) is a suitable marker: Cryptic speciation in Trithemis stictica

13. 2. Methods: Character-based DNA barcoding - PCR with gene specific primers - Sequencing (MegaBACE 500) - Alignment (MUSCLE) - NJ tree based on Kimura-2-parameter (K2P) distances (PAUP) 1. Standard Methods

14. 2. Methods: Character-based DNA barcoding - Search for diagnostic characters by application of CAOS algorithm - Development of perl scripts to assist further analyses - Selection of nucleotide positions for final DNA barcodes by eye 2. Establishment of character-based DNA barcodes:

15. 2. Methods: Character-based DNA barcoding 2. Establishment of character-based DNA barcodes: 0 1 10 01 I. Phylogenetic Tree III. Find unique combinations of character states II. Search for characteristic attributes with CAOS algorithm

16. 2. Methods: Character-based DNA barcoding - Pure (Pu): Exist in all elements of a group but not in alternate group Types of characteristic attributes (CAs): - Private (Pr): Only present in some members of a group but absent from alternate group - Simple (s): At a single nucleotide position - Compound (c): combination of states  sPu and sPr CAs shared by at least 80% of members of a group were used (Filtered by diagViewer)

17. 2. Methods: Character-based DNA barcoding 1. “BarcodeFilter”: sorts out non-relevant nodes Analyses were assisted by a set of perl scripts: Nodes within species cluster are not relevant for barcoding species

18. 2. Methods: Character-based DNA barcoding 2. “BarcodeMaker”: Convertion of “diagViewer- attributes file” into tab delimited file importable to Microsoft Excel:

19. 2. Methods: Character-based DNA barcoding 3. “BarcodeHistMaker”:Counting numbers of CAs at each nucleotide position (selection of sequence fragment with highest number of CAs:

20. Case studies (Study I) Case Study I: Species identification 842 ND1 sequences (65 species) - Suitability of ND1 for DNA barcoding - Applicability of the CAOS algorithm for character-based DNA barcoding

21. Case studies (Study I) NJ tree based on K2P distances: Overview tree: ND1 sequence of one individual of each species Overlap of species cluster

22. Case studies (Study I) Results: Character-based DNA barcodes Unique combinations of character states at 13 selected nucleotide position

23. Case studies (Study I) NJ tree based on K2P distances: Overview tree: ND1 sequence of one individual of each species Overlap of species cluster

24. Case studies (Study I) Results: Character-based DNA barcodes Family Aeshnidae: Combination of character states shared by two or more species  Additional analysis with CAOS algorithm

25. Case studies (Study I) Results: Character-based DNA barcodes for Aeshnids - Search for diagnostic characters within whole ND1 fragment  better resolution

26. Case studies (Study I) NJ tree based on K2P distances: Overview tree: ND1 sequence of one individual of each species Overlap of species cluster

27. Case studies (Study I) Results: Character-based DNA barcodes - Combination of character states shared by several individuals of Calopteryx splendens (cs) and of Calopteryx virgo (cv) - No diagnostic characters found through additional analysis with CAOS algorithm Hybridisation Wrong identification Recent radiation

28. Case studies (Study I) Summary: Case study I - 60 of 65 species distinguishable through unique combinations of character states within ND1 fragment - ND1 suitable - Diagnostic characters easily found by application of the CAOS algorithm

29. Case studies (Study II) Case Study II: Discrimination of conservation units Subset of Case study I; 122 ND1 sequences (9 species) + 101 CO1 sequences (same 9 species) - Suitability of CO1 for DNA barcoding - Ability of both markers to discrimininate conservation units

30. Case studies (Study II) NJ trees based on K2P distances CO1 ND1

31. Case studies (Study II) Results: Character-based DNA barcodes unique combinations of character states at 11 selected nucleotide positions of CO1 fragment  CO1 also suitable

32. Case studies (Study II) Results: Identification of populations  Combination of CO1 and ND1 to improve identification success CO1ND1 X

33. Case studies (Study II) Results: Identification of conservation units CO1 ND1 Pb77 Pb113 Pb78 Pb79

34. Case studies (Study II) Results: Identification of cryptic species Populations: Tst94: Tst118: Tst119: Pb128: CO1 ND1 * One individual of Tst119 shares a combination of character states with 6 individuals of Tst128 Popa Falls Zebra River Kwando Tst128 Tst119 Tst118 Tst94 = Kenya

35. Case studies (Study II) Summary: Case study II -All nine species distinguishable through unique combinations of character states within ND1 and CO1 fragments -Both markers suitable -Character-based DNA barcodes established for conservation units of several species

36. Conclusions Rapid Reliable - Application of CAOS algorithm - Assignment of samples through a few nucleotide positions - Discrete characters - Combination of ND1 and CO1 increases success - DNA barcodes for several conservation units Character-based approaches are :

37. Conclusions Character-based DNA barcoding: A rapid and reliable method for the identification of conservation units in dragonflies !

38. Future Prospects Next steps: - More species - More individuals of some species - Development of data base - Character-based DNA barcodes for genera - Application of character-based DNA barcodes  Identification of adults, exuvia and larvae  Long-time monitoring

39. Thank you !!