Researchers Find A Pattern In Evolution Of Lizard Groups

Using novel statistical and analytical techniques, a group headed by Jonathan Losos, Ph.D. and Allan Larson, Ph.D., both Washington University professors of biology in Arts & Sciences, examined two important dimensions in the evolutionary diversification of four groups of lizards: the ages of branching points on the evolutionary trees of the lizard groups and variation among branches in morphological (body) traits, such as limb length and head size.

The researchers found that the four lizard groups differed in both respects. For example, in Australian agamid lizards, a disproportionate number of branching events occur deep (early) in the evolutionary tree, whereas, at the other extreme, among the South American Liolaemus lizards, the branching points are evenly distributed throughout the tree.

Similarly, the distribution of morphological variation differed in the four groups. In the agamids, closely related species tend to be morphologically similar and distantly related species morphologically different, whereas this relationship is much weaker for Liolaemus . In both cases, the other two lizard groups were intermediate.

What was most surprising to the researchers was that, despite the many differences among the four lizard groups, a strong overall relationship was found between the two aspects of evolutionary diversification. "This correlation was a big surprise," said Losos. "The general worldview is that the history of each lineage is unique and due to varying circumstances so that no general pattern exists. Our findings tend to dispute that."

The results were published in the August 15 issue of Science magazine. The research was supported by the National Science Foundation.

To reach their conclusions, the researchers produced a detailed genetic phylogeny (think of a branching family tree) for all species and physical data for the different species body types. Co-author James A Schulte, Ph.D., a former member of the Losos and Larson laboratories, now a post-doctoral researcher at the Smithsonian Museum of Natural History, gathered much of the phylogenetic and morphological data. Washington university graduate student Luke Harmon, the lead author on the paper, created a statistical program that uses phylogenies derived from genetic information (DNA sequences) from each species to estimate the patterns of branching and morphological change in each group.

"We can use the shape of these phylogenies, or evolutionary trees, to make conclusions about evolution," said Harmon. "For example, we can compare the lengths of branches on these family trees to determine the rate that new species were formed. Deeper branches on trees connecting species indicate older branching, while shorter ones indicate more recent speciation. People have been studying these patterns with fossil evidence for some time, but there is not much of a fossil record for lizards and many other terrestrial groups. I think the results will help biologists understand adaptive radiation better."

"We tentatively explain our main finding using ecological and biogeographic theory," said Larson. "Ecological theory suggests that ecologically similar species are unable to coexist through long periods of time in the same geographic area. If lineages with different ecological adaptations arose early in the history of a group, their descendants could coexist geographically through long periods of evolutionary time by maintaining those differences. If a group does not establish ecologically disparate lineages early in its history, lineages produced at the tips of the tree are more likely to explore a wide diversity of ecological roles."

Lizards have been on the earth for more than 200 million years. There are as many lizard species as there are mammal species, and they make excellent models to study for evolution, said Losos, who has been studying Anolis and other lizards for more than a decade. He and his collaborators plan to analyze more lizard groups to see if the general patterns revealed in this report hold.