Several definitions of “species” have been proposed. Most evolutionary biologists use the biological species concept, which defines species as groups of actually or potentially interbreeding organisms that are reproductively isolated from other such groups, mean-ing they do not (or would not) exchange genes even if they encounter each other. Under this definition, speciation is the evolution of reproductive isolation. Some other biologists favor the phylogenetic species concept, according to which species are sets of popu-lations with character states that distinguish them.
Under any definition of species, the defining qualities (such as reproductive isolation) usually evolve gradually, so some popula-tions cannot be clearly classified as the same or different species.
The biological differences that constitute reproductive isolation include prezygotic barriers to gene exchange (e.g., ecological or sexual isolation) and postzygotic barriers (hybrid inviability or sterility). Several potential isolating barriers may be discovered be-tween two species. Some of them may have evolved before the others, and been the actual basis of speciation. Some barriers (e.g., postzygotic barriers) may not come into play because an earlier-acting difference already prevents gene exchange.
Speciation is rapid in some cases, requiring only a few thousand years or even less. Partial reproductive isolation has evolved even in laboratory populations. Occasionally, a new species is generated instantly by whole genome duplication. In other cases, it may take millions of years for populations to evolve reproductive isolation.
The causes of the evolution of prezygotic reproductive isolation include divergent natural selection arising from ecological factors (ecological speciation) and divergent sexual selection. When hybrids between two divergent populations have low fitness, there is selection for stronger prezygotic isolation, which may result in reinforcement of a prezygotic barrier.
The causes of evolution of postzygotic isolation are less well understood. Hybrid inviability and sterility are often based on incom-patible interactions among two or more genetic loci that diverged between populations by genetic conflict or divergent selection. Hybrid sterility can also be caused by differences in the numbers or arrangements of chromosomes. In some cases, these chromo-some differences may have been established by random genetic drift.
New species sometimes evolve from hybrids between parent species. In many cases, the hybrid species is polyploid.
Evolutionary biologists agree that allopatric speciation is common. Here a physical barrier separates populations of an ancestral species, and evolutionary changes in one or both populations result in biological barriers to gene flow if the populations come back into contact. One possible mode of allopatric speciation, peripatric or founder effect speciation, is thought to be initiated by genetic drift in a small local population of an ancestral species. This is generally thought to be rare.
In speciation with gene flow, a species evolves into two species because of strong divergent selection, without a physical barrier between populations. The evolution of reproductive isolation is hindered by ongoing interbreeding (which maintains gene ex-change) and recombination (which opposes the buildup of divergent sets of genes and characteristics).
Sympatric speciation is the evolution of reproductive isolation within an initially randomly mating population. It is the most ex-treme instance of speciation with gene flow, and requires very strong selection. It is made more likely if traits that are disruptively selected because of their ecological function also automatically reduce gene exchange (e.g., seasonal timing of reproduction).