Speciation

Abstract

In sexually reproducing organisms, speciation is lineage branching—the origin of reproductive isolation between sister populations descended from a single interbreeding parent population. Obviously, speciation is a process of fundamental importance in evolution. In sexually reproducing organisms, every persistent branching point of a phylogenetic tree, whether between very similar species or higher taxa, reflects a speciation event. Because complete reproductive isolation means the end of gene flow between populations, there is no doubt that it can facilitate genetic and phenotypic divergence. So speciation is a major cause of the diversification of living things. In nonsexual or uniparental populations, isolation between divergent populations may also be called speciation, but reduced gene flow can play no role. Such populations may become genetically distinctive and divergent due to differences in mutation, selection, and drift and thereby qualify as species under some definitions (see M. B. Williams, 1992, for a discussion of the species concept in asexual organisms). This chapter deals only with speciation in sexually reproducing organisms. By the usual view of speciation, some barrier to interbreeding comes first, followed or accompanied by genetic and phenotypic divergence. Reproductive isolation leads to divergence. Here I argue that the reverse may sometimes occur—that divergence, mediated by developmental plasticity and selection, may sometimes originate first and contribute to the evolution of reproductive isolation. As discussed in part III, evolution by disruptive and frequencydependent selection can produce a developmental switch between alternative phenotypes rather than loss of intermediate genotypes. This is particularly well documented in insects, often leading to misidentification of intraspecific morphs as species. Since polymorphic insects may have host-associated morphs, host shifts accompanied by distinctive morphology cannot be assumed to represent sympatric speciation or host-race formation, and sympatric speciation hypotheses need to decisively eliminate the possibility of a role for sympatric divergence in the form of polymorphism or behavioral and physiological plasticity. Such intraspecific host shifts may contribute to speciation, whether sympatric or allopatric, as discussed further below.