Phylogenetic Relationships of Major Lineages of Cypriniformes Actinopterygii: Teleostei Based on Rib

1
Phylogenetic Relationships of Major Lineages of
Cypriniformes (Actinopterygii Teleostei) Based
on Ribosomal RNA Sequence Data

• Nicholas J. Gidmark
• Department of Fisheries, Wildlife, and
  Conservation Biology
• Bell Museum of Natural History
• University of Minnesota

2
Cypriniformes

• Over 3,400 species
• Native to Africa, Asia, Europe, North America
• Almost entirely restricted to freshwater
• Occupy a wide variety of habitats

3
Phylogenetic position

• Fink Fink 1981
• Kinethmoid
• Enlarged fifth ceratobranchial
• Teeth on fifth ceratobranchial ankylosed to the
  bone

4
Kinethmoid
5
Enlarged fifth ceratobranchial
6
Teeth on fifth ceratobranchial ankylosed to bone
7
Teeth acting against basioccipital
8
Phylogenetic position
Gymnotiformes
Siluriformes
Characiformes
Cypriniformes
Gonorynchiformes
Fink Fink 1981
9
Cypriniform relationships
Gyrinocheilidae
10
Consensus
Gyrinocheilidae
Balitoridae
Cobitidae
Catostomidae
Cyprinidae
11
Question
How do taxon sampling and analytical method
change our hypotheses of cypriniform phylogeny?
12
Starting point

• Harris and Mayden (2001) dataset
• 12S, Valine, 16S Ribosomal RNA dataset
• 22 taxa (mostly catostomids)
• 2854 nucleotides
• All Cypriniform families represented
• Parsimony methods
• Tree rooted with
• Carassius and Cyprinus

13
Methods, taxon sampling

• Added 113 taxa to Harris Mayden (2001)
• Ingroups
• 99 Cyprinids (Simons et al. 2003)
• 1 Balitorid (GenBank)
• 1 Cobitid (GenBank)
• Outgroups
• 5 Siluriforms (GenBank)
• 3 Characiforms (GenBank)
• 2 Gonorhynchiforms (GenBank)
• 2 Gymnotiforms (GenBank)

14
Methods, analytical I

• ModelTest
• Best model of sequence evolution
• GTR I ?
• Bayesian inference
• 1 million generations

15
Methods, analytical II

• Data partitions
• Embedded helixes
• Exposed helixes
• Loops
• Bulges
• Unpaired bases

16
Analysis method alone
Gyrinocheilidae
Catostomidae
Cobitidae
Baltoridae
Cyprinidae
Bayesian inference
Parsimony
17
Increased sampling
Gyrinocheilidae
Cyprinidae
Catostomidae
Catostomidae
Cobitidae
Cobitidae
Baltoridae
Baltoridae
Cyprinidae
Gyrinocheilidae
Outgroups
Outgroups
Bayesian inference
Parsimony
18
Bottom line
Cyprinidae
Cyprinidae
Balitoridae
Balitoridae
Cobitidae
Cobitidae
Catostomidae
Catostomidae
Gyrinocheilidae
Gyrinocheilidae
Cyprinidae
Gyrinocheilidae
Balitoridae
Balitoridae
Cobitidae
Cobitidae
Catostomidae
Catostomidae
Cyprinidae
Gyrinocheilidae
19
Pharyngeal dentition
20
Key adaptation?
Gyrinocheilidae
Balitoridae
Cobitidae
Catostomidae
Cyprinidae
21
Conclusions

• Intense taxon sampling alone does not change
  hypotheses of phylogeny
• Analytical method alone does not change
  hypotheses of phylogeny
• Only when increased taxon sampling and analytical
  method are combined do differences emerge

22
Conclusions II

• Is this different topology any closer to the true
  tree?
• Greater sampling
• minimizes long branches
• Greater sampling polarizes changes
• Objectively roots tree
• More complex analysis method
• Accounts for rate heterogeneity
• Accounts for multiple nucleotide
  substitutions/site
• Statistical node support

23
Still need to complete

• Full set of partitions
• Run ModelTest on each partition
• Partitioned Bayesian analyses
• Partitioned likelihood analyses

24
Acknowledgments

• Andrew Simons
• Liz Johnson
• Phillip Harris and Rick Mayden
• Susan Weller
• George Weiblen
• GenBank contributors

25
ModelTest Partition models

• Embedded helixes
• Model selected
• Exposed helixes
• Model selected
• Loops
• Model selected
• Bulges
• Model selected
• Unpaired bases
• Model selected

26
Analysis method alone I
Moxostoma
Scartomyzon
Thorburnia
Hypentilium
Erimyzon
Minytrema
Catostomus
Xyrauchen
Cycleptus
Carpiodes
Ictiobus
Myxocyprinus
Parsimony
Bayesian inference