Phylogenetic Networks: A Review of Methods to Display Evolutionary History

Abstract

Phylogenetic analysis attempts to reconstruct the genealogical history of evolutionary change in biological organisms. If the genealogy is complex, involving so-called horizontal evolutionary processes (such as recombination, hybridization, introgression and horizontal gene transfer) then an evolutionary network is required in order to graphically represent the history. Empirical examples of such networks have been used since the 1750s but only rarely. They fell out of favor from the late 1800s, when phylogenetic trees, which can represent only so-called vertical evolutionary processes (transfer of hereditary information directly from parent to offspring), were introduced to represent the Tree of Life. However, in the past 20 years there has been increased interest in using networks, as the evolutionary importance of horizontal processes has become increasingly more apparent. Unfortunately, there are currently few automated methods available, although this is an area of active algorithmic development. In this review, I discuss the development of both trees and networks as icons (or metaphors) for displaying phylogenetic relationships, to clarify some misunderstandings. I then provide an overview of the current approaches to using networks for the study of reticulate evolutionary relationships, explaining how the reticulation processes are detected based on the genetic patterns (or fingerprints) they produce. Finally, I review the current empirical use of evolutionary networks for displaying reticulate evolutionary histories. Due to the limitations of the current methods, many empirical networks have been produced manually or by modifying the output of a computer program.