Defining Genealogical Species Is Problematic

Two genes in different species can by chance be more closely related than two genes from within the same species (see Fig. 22.10). Over time, all genes will come to have the same genealogy, and it has been suggested that when this happens, a genealogical species can be defined:

A group of individuals forms a genealogical species when the genealogies at all loci in the genome are reciprocally monophyletic. In other words, for each locus, all genes within the group must be more closely related to each other than they are to any organisms outside the group. For neutral genes, this will happen after about 10N_e generations of separation. However, this definition is hardly satisfactory. Populations that have been separated by accidents of geography for long enough would eventually be deemed different genealogical species, and conversely two new species could not be classified as such until enough time had elapsed for their genealogies to agree (Hudson and Coyne 2002). Moreover, we have seen that balancing selection can maintain polymorphisms for extremely long times. For example, incompatibility alleles are shared between plant species (Figs. 19.22 and 19.23), and humans share alleles at the major histocompatibility complex (MHC) with other primates. Any groups that share such polymorphisms must be deemed to be members of the same genealogical species.

We can see the history of an entire sexual population more accurately by looking at the relationships among many genes. If we examined enough genes, and if there was enough recombination, we would recover the pedigree: That is, we could trace back from each individual to its two parents, its four grandparents, and so on. As we saw on page 428, after we trace back 20 to 30 generations, the whole recent population would share the same set of ancestors, and we could think of this sharing of ancestors as defining the entire species (Fig. WN22.1). However, this hardly captures what we usually think of as species. Populations that were genetically identical, but that happened to be geographically isolated, would appear as separate branches of the pedigree (right of Fig. WN22.1B). Thus, even if we knew the whole pedigree, we still would not know how to draw lines around different branches so as to separate species from each other (Fig. WN22.1).