Towards comprehensive phylogenies: examples within the (Mammalia)

A thesis submitted for the degree of D.Phil. at the University of Oxford

Olaf R. P. Bininda-Emonds

Merton College Trinity Term 1998 O. R. P. Bininda-Emonds D.Phil. thesis Merton College, University of Oxford Trinity Term 1998

Towards comprehensive phylogenies: examples within the Carnivora (Mammalia)

Abstract

Using case studies from the mammalian order Carnivora, I investigate some of the issues, implications, methodologies, and uses within systematic biology of more complete phylogenies. Chapter 1 reviews the benefits (increased phylogenetic accuracy; more powerful tests of a wider range of hypotheses) and costs (increased analysis times) of larger phylogenies and ways to construct them by combining existing phylogenetic information. Chapter 2 examines the implications of one shortcut used for phylogenetic analysis at higher taxonomic levels, that of representing supraspecific groups as single terminal taxa. Using hypothetical examples and a complete phylogeny of the true seals (family Phocidae), I show that representing a monophyletic group by the states of its common ancestor minimizes errors compared to two other techniques. For paraphyletic taxa, no technique was satisfactory. Chapter 3 investigates the mechanics and biases of matrix representation with parsimony analysis (MRP), a new method of combining phylogenetic information that appears to combine the benefits of more established techniques. This is done as a prelude to Chapter 4, where I use MRP to combine 274 partial estimates of carnivore phylogeny into the first fully dated composite tree of all extant carnivore . In Chapter 5, I divide the 274 carnivore source trees according to 1) data source, 2) study size, 3) study age, and 4) the tree selection criteria used to generate them to investigate the impact of these factors on inferences of carnivore phylogeny. Composite trees derived using MRP did not differ significantly within or among factors, or between families, indicating that all methodologies provide the same estimate of carnivore phylogeny. In Chapter 6, I use the method of independent contrasts together with my carnivore tree to test the widely held, but poorly justified, claim that adaptations to the aquatic habitat functionally separate aquatic from non-aquatic carnivores in general and from the remaining species (fissipeds) in particular. My analysis of 22 morphological, life history, and physiological variables revealed that the differences have been overstated: only limited differences exist between aquatic and non-aquatic carnivores and none between pinnipeds and fissipeds. Acknowledgements ii

Acknowledgements

As always, an absolute myriad of people helped to contribute to this thesis, both directly and indirectly. Each of the major chapters is preceded by a specific set of acknowledgements, so I’ll just take the time here to scribble out the more general ones. First off, I’d like to thank my supervisor, Paul Harvey, for allowing a “tall, harmless Canadian” to come to Oxford and for giving me the support and freedom to write “novels” about things that were decidedly non-molecular and even more decidedly non-viral. Additional support and encouragement were provided by the following, many of who appear as co-authors on the finished products: Harold Bryant (Chapters 2 and 3), John Gittleman (Chapters 4 and 6), Colleen Kelly (Chapter 6), Andy Purvis (Chapter 4), and Tony Russell (Chapter 2). Apart from continually rebuking me for my “chatty” writing style, they help me to shape the ideas and arguments in each chapter and often awakened me to the fact that other equally valid perspectives to some problems exist. Some late thanks go to my examiners, Robert Scotland and Rod Page, first for accommodating my compressed time frame for holding the viva (although it was rather moot in the end — thanks Business Post!) and then for helping to put the finishing touches on the thesis. Now by my reckoning, the number of people that have passed through the revolving door of the 01 group over the past three and a half years numbers at least 30. And those are only the ones directly concerned with our “little” corner of it. So, rather than add yet another appendix, I’ll limit myself to specifically thanking those that I begged, borrowed, or stole advice, friendship, or pints from over that time: Tim Barraclough, Rob McCall, Sophia Kossida (“Olaf, I ask you …”), Mike Charleston (a late, but very welcome addition), Lindell Bromham, Oliver Pybus, Nicholas Grassly, Helen Steers, Fernley Symons, Paul Bollyky, Mike Worobey, Andrew Rambaut, Vitali Proutski (Russian translations while-U-wait), the list goes on … Special thanks go to Chris Bond and Gill Partridge who again helped me in areas my Mac wouldn’t go (e.g., FAXes, application forms, …), helped me to remain blissfully ignorant of the bureaucratic workings of the department, and were just stars in general. Acknowledgements iii

Of course, life does extend beyond work (well, sometimes), and many, many people made Oxford a great place to be. Again, to forestall another appendix, I’ll thank you anonymously en masse. Having said that, I will single out Steve Barge and Tom MacFaul, two friends who made Oxford just that little bit better. I’d be remiss if I didn’t mention the two Mansfield Road cats, Bob and Jack, the only carnivores I had any extensive field experience with while I was in Oxford. Sadly, both no longer belong to the group of extant carnivores. Funding for this project was provided by an NSERC Postgraduate B Scholarship, a United Kingdom Overseas Research Scholarship, a pair of Alberta Heritage Sir James Lougheed Awards of Distinction, Merton College and the Committee for Graduate Studies for a number of travel grants, and my next paying job (for the student loans). For cheerfully writing kind things on my behalf for this seemingly endless stream of grant applications, I thank Tony Russell, Harold Bryant, Michael Cavey, and Paul Harvey. And, speaking of Merton, I’d like to thank the College for its enlightened attitude towards graduate students. Despite everyone’s continual gripes during mealtimes, Merton gave me no reason to complain during my time here and, if anything, made me feel rather spoiled. The staff were always helpful, cheerful, and incredibly friendly. Thanks to you all. I can’t say enough good things about Paul Ong. Despite his predilections for dodgy predictions and a rather bizarre definition of the word “auspicious,” he kept me going through the one bad stretch and made me enjoy the good times even more. I always looked forward to our long talks about this, that, and everything else. A true friend. There’s no better way to put it than by simply saying thanks. Finally, I’d like to thank my family for all their tremendous love, support, and understanding throughout the years and their running of trans-Atlantic favours for me over the past few years. This thesis is dedicated to them, particularly my parents who can finally say “Our son, the doctor” (although I don’t know if this is quite what they had in mind way back when). Table of contents iv

Table of contents

Abstract ii

Acknowledgements iii

Chapter 1: General introduction 1 Larger phylogenies — benefits and problems 1 Possible solutions 3 Combining phylogenetic information 4 Mammalian carnivores as a case study 6 Structure of this thesis 7

Chapter 2: Supraspecific taxa as terminals in cladistic analysis: implicit assumptions of monophyly and a comparison of methods 10 Synopsis 11 Introduction 12 The issues 14 Assumptions of monophyly 14 Representing supraspecific taxa 15 Analysis 18 Hypothetical examples 18 Representing monophyletic groups 19 Representing paraphyletic groups 22 Overview 25 Phocid examples 25 Background and methodology 25 Monophyly examples 29 Paraphyly examples 31 Levels of hom