Will there ever be a tree of life that systematists can agree on?

Abstract

Since the concept of a "Tree of Life" was raised by Charles Darwin, researches in this field have not only contributed to our understanding of phylogenetic relationships among taxa, but also significantly accelerated the development of related subjects in biological science. Evolutionary biologist Dobzhansky once remarked that "nothing makes sense in biology except in the light of evolution", which has been largely echoed by later biologists. Indeed, reconstruction of an accurate phylogeny of the living world is very important for biological classification and nomenclature, and also crucial to elucidate the origin and diversification of life. We have experienced three major phases for Tree of Life reconstruction in the past century. Prior to the 1990s, taxonomists published classification systems that were largely dependent on morphological characters. DNA sequencing technology facilitated by the development of polymerase chain reaction (PCR) techniques has allowed systematists to reconstruct phylogenetic relationships using molecular data. More recently, the rapid development of next-generation sequencing tools has brought the Tree of Life to a phylogenomic era by enabling the construction of phylogenies using hundreds or thousands of loci from organellar and nuclear genomes. However, significant conflicts have been detected in phylogenies of various organisms with the large increase in the number of loci used for phylogenetic analyses. Given the level of conflict in some data sets, some researchers have begun to doubt the accuracy and congruence of the Tree of Life and its applications in related biological fields. So, will there ever be a Tree of Life that systematists can agree on? In this paper, we highlight three reasons why researchers cannot retrieve a totally congruent tree that reflects the real evolutionary history of life. This is despite significant improvements in morphological, molecular, and statistical methods and is analogous to our inability to restore a collapsed building, even when all bricks and other building materials remain. (i) Sampling limitations: we cannot sample all the species in the world because a large percentage of species have become extinct throughout Earth's history and many species are currently facing extinction or have not yet been recognized by scientists (especially life in the oceans); (ii) Biological processes: hybridization/introgression, incomplete lineage sorting, gene duplication and/or loss, horizontal gene transfer and other biological events that have occurred during evolutionary history have frequently resulted in gene tree heterogeneity; and (iii) Systematic biases and models for tree reconstruction: phylogenetic noise in the data such as evolutionary saturation and compositional bias can lead to incorrect phylogenies and any algorithms for reconstructing phylogenetic trees cannot absolutely simulate the real processes of organic evolution. Furthermore, biological factors attributed to discordance can become even more complicated when we reconstruct phylogenies using phylogenomic datasets. Generally, incomplete lineage sorting and hybridization/introgression occurred in closely related species, whereas phylogenetic discrepancy at the family, order or above levels are usually the combined effects of gene duplication and/or loss, recombination, and genome duplication. Therefore, it is always important to understand the mechanisms causing the incongruence and explore approaches to better model the processes that generate the discordance. In recent decades, new models and methods in phylogenomic studies have been developed and have shed light on the species trees of some candidate groups. Thus, we still look to a bright future for the Tree of Life and its applications in related biological sciences despite the fact that we cannot achieve a completely congruent tree of the living world.