| Presentation Detail
Macroevolution Soul, Laura [1], Lloyd, Graeme [1], Friedman, Matt [1]. How good is good enough? - Taxonomy and phylogeny in evolutionary analyses. The fossil record contains a wealth of information on macroevolutionary patterns and corresponding processes that have acted over the history of life on Earth. Early observations of these trends inspired development of many of the comparative methods now available to quantitatively assess evolutionary signals, which draw upon estimated phylogenetic relationships between species. Unfortunately for paleontologists specifically, and evolutionary biologists more broadly, many groups with the most extensive and richest fossil records have not been subjected to formal phylogenetic analysis, although mature taxonomies are often available for these clades. This raises two key questions about the applications of comparative approaches to paleontological questions. First, what is the sensitivity of tree-based analyses to uncertainty in evolutionary relationships? Second, can taxonomic information be a useful (i.e., not misleading) substitute for an explicit phylogenetic hypothesis? In order to address these issues we collected data for more than 20 animal clades (vertebrate and invertebrate) that include fossil representatives, for which a recent cladogram and pre-cladistic taxonomies were available. We quantified the difference between their implied phylogenies using pairwise distances, under the assumption that the modern cladogram was correct. We measured a series of commonly used, phylogenetically explicit parameters (phylogenetic conservatism [Blomberg's K], inferred divergence times, phylogenetic generalised least squares, mode of evolution and phylogenetic clustering of extinction [Fritz and Purvis' D]) and determined the degree to which they co-varied between taxonomic and cladistic trees. Results show that distances between taxa in a taxonomy are strongly correlated with their equivalent distance in a phylogeny, and any difference is significantly less (p<0.05) than in random topologies that have been time-scaled with the same stratigraphic data. Each analytical method investigated produces results that co-vary strongly when phylogenetic and taxonomic frameworks are used, particularly those methods which focus on changes in traits (phylogenetic conservatism: R2>0.8; PGLS: R2>0.89; and mode of evolution: 92% correct). Taxonomies for inclusive clades perform better than those at low taxonomic rank, likely as a consequence of being more highly nested. This suggests that under particular circumstances, taxonomies are indeed 'good enough' to be used in place of a formal cladistic solution, and are unlikely to be misleading relative to this standard. This potentially opens a large and previously inaccessible section of the fossil record to interrogation within an explicit comparative framework, which will help to test many classical macroevolutionary hypotheses that have been based on groups for which formal phylogenetic hypotheses remain lacking.
Log in to add this item to your schedule
1 - University of Oxford, Earth Sciences, South Parks Road, Oxford, OX1 3AN, UK
Keywords:
comparative methods
phylogeny
Fossil record
macroevolution
taxonomy.
Presentation Type: Regular Oral Presentation
Session: 75
Location: Cotton B/Snowbird Center
Date: Sunday, June 23rd, 2013
Time: 2:45 PM
Number: 75006
Abstract ID:825
Candidate for Awards:W.D. Hamilton Award for Outstanding Student Presentation |
