Taxon sampling and seed plant phylogeny

Area of Science:

• Evolutionary Biology
• Phylogenetics
• Molecular Evolution

Background:

• Phylogenetic relationships among seed plants, particularly the Gnetales, remain contentious.
• Previous phylogenetic studies have often relied on limited taxon sampling, potentially leading to varied hypotheses.

Purpose of the Study:

• To investigate how varying taxon sampling affects phylogenetic inference in seed plants.
• To assess the impact of matrix size and analytical methods on tree topology consistency.

Main Methods:

• Applied parsimony and Bayesian analyses to 38-taxon and 80-taxon rbcL matrices for seed plants.
• Exchanged terminals to create ten matrices of each size for comparative analysis.
• Focused on the placement of the Gnetales within the seed plant phylogeny.

Main Results:

• Different taxon samples produced mutually contradictory tree topologies, even for strongly supported groups.
• Increasing taxon sampling improved consistency for unweighted parsimony but not for other methods (parsimony with third positions excluded, Bayesian analysis).
• Bayesian analysis, especially with the general time-reversible model, showed reduced consistency with increased taxon sampling.

Conclusions:

• Resolving deep evolutionary relationships with few taxa can result in unreliable phylogenetic conclusions.
• Insufficient taxon sampling may explain the diversity of proposed seed plant phylogenies.
• Reanalyzing restricted data sets with alternative taxon selections is recommended to ensure topological consistency.