Related Concept Videos
Phylogeny
Variance
The standard deviation measures the spread in the same units as the data....
Reaction Rate
The mathematical representation of the change in the concentration of reactants and products, over time, is the rate...
Friedman Two-way Analysis of Variance by Ranks
Related Rates
Speciation Rates
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Clade-specific drivers of rapid vertebral evolution in squamate reptiles.
The Geometry of Macroevolution: Phenotypic Evolution on Non-Euclidean Manifolds.
Related Experiment Video
Updated: Feb 2, 2026

Decomposing the Variance in Reading Comprehension to Reveal the Unique and Common Effects of Language and Decoding
Published on: October 11, 2018
Phylogenies and Diversification Rates: Variance Cannot Be Ignored.
1Museum of Zoology and Department of Ecology and Evolutionary Biology, Biological Sciences Building, 1105 North University Avenue, University of Michigan, Ann Arbor, MI 48109, USA.
A new test for macroevolutionary models is flawed because it ignores sampling variation. Differences in evolutionary rate estimates are expected due to clade size, not model error. Biologists should account for variance in phylogenetic analyses.
Area of Science:
- Evolutionary Biology
- Phylogenetics
- Macroevolutionary Modeling
Background:
- Hierarchical macroevolutionary models are used to study evolutionary rates.
- A recent test in Evolution journal claims to assess model validity by comparing rate estimates at different phylogenetic levels.
- This test was applied to BAMM software, which estimates evolutionary rates from phylogenetic trees.
Purpose of the Study:
- To demonstrate that a proposed test for hierarchical macroevolutionary models is invalid.
- To show that observed differences in rate estimates are due to sampling variation, not model flaws.
- To highlight statistical and mathematical errors in the proposed testing framework.
Main Methods:
- Analysis of rate estimates from BAMM software applied to phylogenies of varying sizes.
- Statistical assessment of the impact of sampling variation on rate estimates.
- Mathematical critique of the proposed testing framework's first principles.
Main Results:
- Numerical differences in rate estimates between large clades and subclades are fully explained by sampling variation.
- Variance in evolutionary rate estimates is inversely proportional to clade size, especially for small clades.
- The proposed test relies on negative results stemming from low statistical power due to high variance.
Conclusions:
- The proposed test for hierarchical macroevolutionary models is invalid due to a failure to account for sampling variation.
- Differences in rate estimates do not necessarily indicate a flawed model, but can arise from statistical properties of small sample sizes.
- Biologists should incorporate standard statistical practices, particularly accounting for variance, when estimating evolutionary parameters.

