The Tree of Life project per se will be much like the Human Genome project, merely the first step in a vastly broader research agenda. The complete nucleotide sequence of a human genome provided a foundation for investigating genomic operations more deeply, such as enabling geneticists to map and characterize the structures and functions of genes responsible for particular phenotypes. Analogously, a robust Tree of Life will provide a foundation for delving much deeper into nature’s evolutionary operations, such as enabling phylogeneticists to map the origins and evolutionary transitions among particular organismal phenotypes or document instances of reticulate evolution (Avise, 2006). All of these sentiments are simply conventional wisdom.
I suggest that either of two polar-opposite patterns (or more likely some mixture of the two) could emerge from the Tree of Life project and that either outcome would offer its own unprecedented grand opportunity for the field of evolutionary biology. One possibility is that the tree metaphor will apply well to many or most taxonomic groups, in which case an opportunity (described later) would arise for the field of systematics. Alternatively, the tree metaphor may prove to be inadequate, and an anastomose web or network of life would better describe the histories of descent of many taxa. Several recent authors have argued that genetic exchanges across lineages, via endosymbiotic mergers and lateral DNA transfers especially in microbes (Doolittle, 1999; Margulis and Sagan, 2002; Arnold, 2006) or via hybridization in metazoan plants and animals (Arnold, 1997), have played important evolutionary roles. For example, McCarthy (2008) builds a case that new species seldom arise from the standard population genetic processes of gradual divergence via mutation, drift, and selection in allopatry, but instead that novel life forms often originate via the genetic stabilization of recombinant lineages following hybridization events (Fig. 15.1). If this hypothesis is correct, the ramifications for many areas of evolutionary biology would be profound (as described later).
Deciding whether the tree metaphor or the network metaphor better explains the history of life is a stiff challenge requiring detailed and critical appraisals of empirical evidence for many taxonomic groups. But the two hypotheses do have several distinct predictions. In terms of genealogical expectations, for example, the Tree of Life model predicts that gene trees should be topologically concordant with one another and with the species tree they compose [barring potential complications such as insufficient resolution, hemiplasy (idiosyncratic lineage sorting across successive nodes in a species phylogeny) (Avise and Robinson, 2008), and homoplasy]. In contrast, the network of life model predicts that multiple gene trees in a given taxonomic group will often be qualitatively