Systematics/Phytogeography / Taxonomie/ Section
Olmstead, Richard , Tank, David C. .
Over-precision in Molecular Dating.
Dating nodes on phylogenetic trees using calibrations based on the fossil record often results in unrealistically precise dates. This is due to a number of reasons often beyond the control of the investigator. The problem is often, but not always, acknowledged, though rarely addressed in published studies. Typically, when any error estimate is presented, it is based on the variance in the estimated substitution rate, which is a product of the estimation method. However, more substantive error often is associated with the calibration points used to produce the estimates. The use of multiple calibration points scattered around a tree does not necessarily reduce the variance in estimates for the minimum ages of nodes. Rather, having multiple calibration points for an individual node permits an estimate of the variance for that node in a way that multiple single points for several nodes cannot. A second source of over-precision is an error in logic. Providing an error estimate for a node based on the variance in substitution rate, or based on multiple calibration points on the tree, does not provide an upper and lower bound on the time of origin of the node, but rather upper and lower bounds on the minimum age estimate of that node. We illustrate some of the problems with providing precise molecular dating estimates and call for new methods that accommodate multiple sources of error in molecular dating.
Log in to add this item to your schedule
1 - University of Washington, Department of Biology and Burke Museum, Box 355325, Seattle, WA, 98195, USA
2 - Yale University, Peabody Museum of Natural History, Botany Division, P.O. Box 208118, New Haven, CT, 06520-8118, USA
Presentation Type: Oral Paper:Papers for Sections
Location: Room 1/Woodward
Date: Monday, July 28th, 2008
Time: 3:45 PM