Effects of discordance between species and gene trees on phylogenetic diversity conservation

TL;DR

This paper investigates how discordance between gene trees and species trees affects biodiversity prioritization using the Fair Proportion index, revealing conditions where rankings agree or differ and demonstrating variability in real data.

Contribution

It provides a theoretical and empirical analysis of the impact of gene-species tree discordance on phylogenetic diversity prioritization.

Findings

For certain tree shapes, gene and species tree rankings coincide.

For trees with five or more leaves, discordance can lead to different rankings.

Empirical data shows variability in prioritization due to gene tree discordance.

Abstract

Phylogenetic diversity indices such as the Fair Proportion (FP) index are frequently discussed as prioritization criteria in biodiversity conservation. They rank species according to their contribution to overall diversity by taking into account the unique and shared evolutionary history of each species as indicated by its placement in an underlying phylogenetic tree. Traditionally, phylogenetic trees were inferred from single genes and the resulting gene trees were assumed to be a valid estimate for the species tree, i.e., the "true" evolutionary history of the species under consideration. However, nowadays it is common to sequence whole genomes of hundreds or thousands of genes, and it is often the case that conflicting genealogical histories exist in different genes throughout the genome, resulting in discordance between individual gene trees and the species tree. Here, we analyze the effects of gene and species tree discordance on prioritization decisions based on the FP index. In particular, we consider the ranking order of taxa induced by (i) the FP index on a species tree, and (ii) the expected FP index across all gene tree histories associated with the species tree. On one hand, we show that for particular tree shapes, the two rankings always coincide. On the other hand, we show that for all leaf numbers greater than or equal to five, there exist species trees for which the two rankings differ. Finally, we illustrate the variability in the rankings obtained from the FP index across different gene tree and species tree estimates for an empirical multilocus mammal data set.