Edited 8 April 2011: Following some interesting comments from population geneticist Lou Jost, I’ve edited my characterization of the rule of thumb that one migrant per generation is generally enough to prevent populations from differentiating.
It’s the little differences. I mean, they got the same shit over there that we got here, but it’s just—it’s just there it’s a little different.
— Vincent, Pulp Fiction
Different places are different from each other. This is a truism bordering on tautology, but it also has real implications for the ways in which life evolves and diversifies. The specific differences between one environment and another shape how well living things can move between those environments, and what happens when they do—the evolutionary consequences of migration.
Individuals moving between populations can alter the evolution of those populations in a number of ways [$a]. Migration introduces traits from one population into another, making it a source of variation not unlike mutation. Migration can swamp out the effects of local natural selection, if enough migrants come from a population experiencing a different selective regime. Migration from a larger population to a smaller one can offset the loss of variation to genetic drift; but a small group of individuals migrating to an empty habitat can be strongly affected by drift. Depending on what species concept you prefer, migration between two populations is either what prevents them from becoming separate species, or what conclusively proves that they are the same species.
Migration mixes things up
Strictly speaking, I’m not talking simply about the movement of living things from place to place, but about gene flow, which also requires interbreeding between migrants and the populations to which they migrate. Individuals might be able to travel to another environment, and even do so frequently, but fail to survive when they get there [PDF], or fail to find a mate in the local population, or have offspring that are themselves less fit than the offspring of the locals. Because tracking each of these steps directly is daunting at best, most population biology studies estimate the rate of successful migration by proxy, using some measure of the genetic similarity of populations—the more migrants successfully move between populations, the more similar the traits and gene frequencies of those populations will be.
The rate of gene flow between two populations is essentially a measure of how much those populations evolve as a single unit—if there’s no gene flow, selection or drift can eventually make the two populations into completely different things. An effective migration rate of one migrant per generation
is has been generally understood to be enough to prevent drift from causing populations to diverge [$a]. (But there are significant objections to the one-migrant-per-generation rule of thumb, including the question of how we determine that populations have differentiated! See the discussion of this in the comments.) Populations linked by migration along several intervening populations may be isolated by distance [PDF] if the line of connecting populations is long enough.
If natural selection is operating in different directions in the two populations, more migration is necessary to prevent them evolving different traits. If individual genes experience different selection in each population, then those genes may still evolve differently even as migration continues to mix in neutral genes [PDF]. This movement of genes across recognized population and species boundaries is called introgression.
Gene flow versus selection
One well-documented case of gene flow between apparently “good” species is that of the Louisiana irises Iris fulva and I. brevicaulis. These two species fill fairly distinct ecological niches: red-orange I. fulva is mainly pollinated by hummingbirds, and grows in very wet conditions; blue I. brevicaulis is pollinated by bees, and grows best in drier habitats.
Yet when the two species co-occur, they sometimes do cross-pollinate. What prevents the two species from merging into a single iris? (Perhaps it would have purple flowers.) Natural selection, in this case, overwhelms migration. Experimentally created fulva–brevicaulis hybrids grown in wet conditions are more likely to survive if they carry specific genes from wet-tolerant fulva; and pollinators tend to favor hybrids that look more like their preferred parent species. I. fulva and I. brevicaulis share some genes that don’t affect wet tolerance or pollinator attraction, but generally remain separate evolutionary entities.
This kind of partial reproductive isolation is most widely documented in plants, but it has been found in all sorts of organisms—even the chipmunks Tamias ruficaudus and T. amoenus [PDF]. In an evolving world, this shouldn’t be surprising—it makes sense to find many cases of partial reproductive isolation as populations evolve toward the point of being separate species. Sometimes, of course, divergent selection weakens, or previous barriers to migration are removed, and populations re-merge into single evolutionary entities. But sometimes, the balance of the selection, mutation, drift, and migration is just right for just long enough to create a new, independently evolving form of life. And thus, as Darwin wrote, “endless forms most beautiful have been, and are being, evolved.”
Arnold, M., Hamrick, J., & Bennett, B. (1990). Allozyme variation in Louisiana irises: a test for introgression and hybrid speciation. Heredity, 65 (3), 297-306 DOI: 10.1038/hdy.1990.99
Good J.M., Hird S., Reid N., Demboski J.R., Steppan S.J., Martin-Nims T.R., & Sullivan J. (2008). Ancient hybridization and mitochondrial capture between two species of chipmunks. Molecular ecology, 17 (5), 1313-27 PMID: 18302691
Hedrick, P.W. (2005). Genetics of Populations. Boston: Jones and Bartlett Publishers. Google Books.
Martin, N.H., Bouck, A.C., & Arnold, M.L. (2005). Detecting adaptive trait introgression between Iris fulva and I. brevicaulis in highly selective field conditions. Genetics, 172 (4), 2481-9 DOI: 10.1534/genetics.105.053538
Martin, N., Sapir, Y., & Arnold, M. (2008). The genetic architecture of reproductive isolation in Louisiana irises: Pollination syndromes and pollinator preferences. Evolution, 62 (4), 740-52 DOI: 10.1111/j.1558-5646.2008.00342.x
Nosil, P., Egan, S., & Funk, D. (2008). Heterogeneous genomic differentiation between walking-stick ecotypes: “Isolation by adaptation” and multiple roles for divergent selection. Evolution, 62 (2), 316-36 DOI: 10.1111/j.1558-5646.2007.00299.x
Nosil, P., Vines, T., & Funk, D. (2005). Reproductive isolation caused by natural selection against immigrants from divergent habitats. Evolution, 59 (4), 705-19 DOI: 10.1554/04-428
Slatkin, M. (1987). Gene flow and the geographic structure of natural populations. Science, 236 (4803), 787-92 DOI: 10.1126/science.3576198
Wang, J. (2004). Application of the one-migrant-per-generation rule to conservation and management. Conservation Biology, 18 (2), 332-43 DOI: 10.1111/j.1523-1739.2004.00440.x
Wright, S.J. (1943). Isolation by distance. Genetics, 28, 139-56 PMCID: PMC1209196