Tuesday, February 2, 2010

Rotten Fish and the Chordate Tree of Life

The origin of vertebrates is part of a larger story involving the evolution of a more inclusive group of kin, the chordates that along with vertebrates include urochordates (tunicates) and cephalochordates (lancelets). Recent decades have been an exciting time for our understanding of chordate evolution, with the discovery and evolutionary analysis of important chordate fossils that date to the Cambrian (542 to 488 million years ago). On the basis of distinct morphological features preserved in the fossils, scientists have been able to integrate these extinct lineages into phylogenetic hypotheses of chordates, extinct early vertebrate lineages, and extant vertebrates.

In a very clever study published last week in Nature, Sanson et al. allowed specimens of lancelets and larval lampreys (ammocoetes) to decay in artificial seawater in the laboratory. The anatomical features that are the most important for distinguishing clades in the phylogeny were the most prone to decay, while the least informative (pleisiomorphic) characters were more resistant to decay. The results are striking and have important implications for the evolutionary interpretation of early chordate and vertebrate fossils. The bias in anatomical decay will result in placement of derived lineages further down the stem of the phylogeny, a phenomena the authors refer to 'stem-ward slippage.' As a result, placement of chordate and vertebrate fossil taxa on the stem of their respective phylogenies may be an artifact of character preservation. As such, the phylogenetic position of fossil taxa lacking derived features, but possessing character states of stem chordates or vertebrates, may be suspect. A video showing time lapse photography of a rotting lamprey is available on Nature's youtube channel. Picture of the cleared and stained ammocoete from www.biology.uco.edu/PersonalPages/CButler/.


Mike Keesey said...

Or, Decomposition Decapitulates Phylogeny.

Susan Perkins said...

I love it, MIke! Good one!

Mason Posner said...

Very cool study. My students and I read these papers in my Vertebrate Biology course, which comes back around in the Fall. I am looking forward to incorporating this new paper in Nature.

Any sense of how one would know if taphonomy is affecting the characters in a particular fossil? This study suggests that we should be careful when coding characters for phylogenetic analysis, but if you are being careful are there ways to know if character states are due to decomposition?