Estimating rates and patterns of diversification with incomplete sampling: a case study in the rosids

Area of Science:

• Evolutionary Biology
• Ecology
• Phylogenetics

Background:

• Large-scale phylogenies are increasingly used for species diversification estimation.
• Common methods often suffer from incomplete species coverage and sparse backbone representation.
• Presumed phylogenetic placements are frequently used for unsampled species.

Purpose of the Study:

• To examine the effects of incomplete species sampling on diversification estimation.
• To provide constructive suggestions for ecologists and evolutionary biologists.
• To evaluate the robustness of different phylogenetic methods to sampling biases.

Main Methods:

• Used empirical case studies with rosids and Cucurbitaceae phylogenies.
• Compared results from large phylogenies, smaller supermatrices, and synthetic trees.
• Simulated random and representative taxon sampling, analyzing parametric (RPANDA, BAMM) and semiparametric (DR) methods.

Main Results:

• Incomplete sampling had idiosyncratic and significant impacts on diversification estimates.
• Speciation rates were either depressed or inflated by random and representative sampling schemes, varying by method.
• The BAMM method demonstrated the least sensitivity to moderate levels of missing taxa.

Conclusions:

• Caution is advised against uncritical modeling of missing taxa using taxonomic data for poorly sampled trees.
• Summary backbone trees and data sets with high representative bias should be used cautiously.
• Explicit sampling methodologies are crucial for robust macroevolutionary studies.